The role of arginine and tryptophan
metabolism in sepsis
by
Christabelle Jean Darcy
BSc (Hons)
A thesis submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy
Menzies School of Health Research
Institute of Advanced Studies
Charles Darwin University
October 2010
I hereby declare that the work herein now submitted as a thesis for the degree of
Doctor of Philosophy of the Charles Darwin University is the result of my own
investigations, and all references to ideas and work of other researchers have been
specifically acknowledged. I hereby certify that the work embodied in this thesis has
not already been accepted in substance for any degree, and is not being currently
submitted in candidature for any other degree.
Christabelle Jean Darcy
i
Table of contents
Table of contents ........................................................................................................................................i
Table of figures........................................................................................................................................ iii
Table of tables............................................................................................................................................ v
Acknowledgments.....................................................................................................................................vi
Abstract ...................................................................................................................................................viii
Declaration of author’s contribution........................................................................................................x
Publications forming the basis of this thesis ........................................................................................ xiii
Abbreviations ........................................................................................................................................... xv
1. Introduction: Aim and scope. ..........................................................................................................1
2. Background: Sepsis .........................................................................................................................3
2.1. Introduction.............................................................................................................................3
2.2. What is sepsis? ........................................................................................................................3
2.3. Endothelial dysfunction in sepsis ............................................................................................6
2.4. Immune dysfunction in sepsis ..................................................................................................7
2.5. Amino acids in sepsis ............................................................................................................10
2.6. Conclusion.............................................................................................................................11
3. Background: Arginine and tryptophan bioavailability.................................................................12
3.1. Introduction...........................................................................................................................12
3.2. Arginine metabolism..............................................................................................................12
3.3. Tryptophan metabolism.........................................................................................................18
3.4. Conclusion.............................................................................................................................21
4. Background: Regulation of microvascular reactivity ...................................................................22
4.1. Introduction...........................................................................................................................22
4.2. Endothelial regulation of microvascular reactivity...............................................................22
4.3. Nitric oxide and microvascular reactivity .............................................................................23
4.4. Kynurenine and microvascular reactivity .............................................................................24
4.5. Conclusion.............................................................................................................................24
5. Background: Regulation of T cell function ..................................................................................25
5.1. Introduction...........................................................................................................................25
5.2. Overview of T cell regulation................................................................................................25
5.3. T cell zeta-chain expression ..................................................................................................26
5.4. Arginine and T cells ..............................................................................................................27
5.5. Tryptophan and T cells..........................................................................................................29
5.6. Myeloid derived suppressor cells ..........................................................................................29
5.7. Conclusion.............................................................................................................................31
6. Experimental design and hypotheses ............................................................................................32
6.1. Introduction...........................................................................................................................32
6.2. Clinical studies forming the basis of this project ..................................................................32
ii
6.3. Earlier results........................................................................................................................33
6.4. Generation of hypotheses......................................................................................................53
6.5. Conclusion ............................................................................................................................55
7. Methods: Measuring amino acids in plasma ................................................................................56
7.1. Introduction...........................................................................................................................56
7.2. High performance liquid chromatography (HPLC)..............................................................56
7.3. General amino acids assay ...................................................................................................57
7.4. ADMA assay..........................................................................................................................75
7.5. Effect of processing time on amino acid concentration ........................................................92
7.6. Conclusion ..........................................................................................................................106
8. Results: Arginine bioavailability in sepsis ..................................................................................108
8.1. Introduction.........................................................................................................................108
8.2. Arginase activity in sepsis ...................................................................................................108
8.3. Arginine/ADMA ratio in sepsis ...........................................................................................117
8.4. Conclusion ..........................................................................................................................134
9. Results: Tryptophan bioavailability in sepsis..............................................................................135
9.1. Introduction.........................................................................................................................135
9.2. Tryptophan bioavailability in sepsis ...................................................................................135
9.3. Conclusion ..........................................................................................................................154
10. Results: Inflammation and T cell suppression in sepsis........................................................155
10.1. Introduction.........................................................................................................................155
10.2. Arginine, tryptophan and T cell suppression in sepsis........................................................156
10.3. Myeloid derived suppressor cells in sepsis .........................................................................159
10.4. Conclusion ..........................................................................................................................180
11. Discussion and conclusion .....................................................................................................181
11.1. Introduction.........................................................................................................................181
11.2. Sepsis decreases amino acid bioavailability .......................................................................181
11.3. Amino acid bioavailability contributes to the pathophysiology of sepsis............................184
11.4. Future directions.................................................................................................................190
11.5. Conclusion ..........................................................................................................................193
References .............................................................................................................................................194
Appendix: Published papers from this thesis .......................................................................................222
iii
Table of figures
Figure 2.1 The relationship between sepsis and the systemic inflammatory immune response............ 4
Figure 3.1 The two reactions of nitric oxide synthesis as catalysed by nitric oxide synthase. ............ 15
Figure 3.2 Molecular structures of arginine, monomethylarginine (MMA), asymmetric
dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA). ................................................ 16
Figure 3.3 The enzyme indoleamine 2,3-dioxygenase (IDO) oxidises tryptophan to kynurenine........ 18
Figure 3.4 The reciprocal relationship between indoleamine 2,3 dioxygenase (IDO) and nitric oxide
synthase (NOS) mediated by heme oxygenase and nitric oxide. .......................................................... 20
Figure 5.1 T cell receptor structure showing showing the arrangement of the αβγδε and ζ chains. ... 26
Figure 5.2 Potential pathway of zeta down-regulation in response to arginase activity..................... 28
Figure 5.3 Relationship between inflammatory mediators and myeloid derived suppressor cell
(MDSC) induction. ............................................................................................................................... 31
Figure 6.1 Representative normal and abnormal peripheral arterial tonometry traces. .................... 37
Figure 6.2 Baseline microvascular reactivity is impaired in sepsis, in proportion to disease severity45
Figure 6.3 (a) Longitudinal change in microvascular reactivity in sepsis subjects, (b) Longitudinal
change in plasma arginine in sepsis subjects....................................................................................... 47
Figure 7.1 Photo of a high performance liquid chromatography (HPLC) unit................................... 57
Figure 7.2 Chromatogram of quality control plasma using the Nova-Pak method............................. 68
Figure 7.3 Chromatogram of quality control plasma using the Shim-pack method. ........................... 69
Figure 7.4 Chromatograms from dimethylarginine assay................................................................... 85
Figure 7.5 Plasma arginine time profile at room temperature and on ice. .........................................99
Figure 7.6 Time profile of median plasma arginine and ornithine concentrations in blood stored at
room temperature............................................................................................................................... 100
Figure 7.7 Time profile of median plasma taurine concentrations in blood stored at room temperature
and on ice. .......................................................................................................................................... 103
Figure 8.1 Relationship between circulating neutrophil count and plasma argininase activity (a) and
plasma arginine concentration (b). .................................................................................................... 114
Figure 8.2 Group comparison of arginase activity and arginine. ..................................................... 115
Figure 8.3 (a) Arginine to asymmetric dimethylarginine ratio and microvascular reactivity according
to disease severity. ............................................................................................................................. 127
Figure 8.4 Baseline plasma concentration of asymmetric dimethylarginine according to disease
category.............................................................................................................................................. 128
Figure 9.1 Plasma assessment of tryptophan catabolism.................................................................. 147
Figure 9.2 Proposed model of tryptophan catabolism in sepsis....................................................... 150
Figure 10.1 Ex vivo T cell zeta-chain expression in sepsis patients compared to controls (a) and the
association of T cell zeta-chain expression with plasma concentrations of arginine (b) and tryptophan
(c) in sepsis patients. .......................................................................................................................... 157
Figure 10.2 Change in T cell zeta-chain expression (a), plasma arginine concentration (b) and
plasma tryptophan concentration (c) between day 0 and day 2 of the study...................................... 157
iv
Figure 10.3 Recovery of T cell zeta-chain expression in unstimulated cells in media with
physiological concentrations of amino acids (a) and comparison of recovery with or without arginine
(b)........................................................................................................................................................158
Figure 10.4 Representative forward side scatter plots. ......................................................................169
Figure 10.5 Percentage of CD66b+ granulocytes in PBMC from septic shock, sepsis without shock
and control patients on day 0 (a) and day 2 – 4 (b) of the study.........................................................170
Figure 10.6 The relationship between the baseline percentage of CD66b+ granulocytes in the PBMC
and plasma inteluekin-6 in sepsis patients. .........................................................................................170
Figure 10.7 Staining comparison of CD66b+ granulocytes in PBMC, monocytes and PMN from a
single sepsis patient. ...........................................................................................................................172
Figure 10.8 Combination staining of CD66b and CD16....................................................................173
Figure 10.9 Comparison of T cell proliferation with and without CD66b+ cells in two sepsis patients.
............................................................................................................................................................173
Figure 10.10 T cell zeta-chain expression in septic shock patients, sepsis patients without shock and
hospital controls on day 0 (a) and day 2 (b) of the study....................................................................174
Figure 10.11 The longitudinal relationship between T cell zeta-chain expression and percentage of
CD66b+ cells in PBMC in 3 individual patients (a, b and c). .............................................................175
Figure 10.12 Percentage of CD66b+ cells in PBMC, T cell zeta-chain expression and plasma
arginase activity..................................................................................................................................176
Figure 10.13 Representative association between T cell zeta-chain expression and the percentage of
sepsis PBMC CD66b+ cells added back to the CD66b+ depleted cell culture, and arginine used from
the supernatant....................................................................................................................................177
Figure 11.1 Proposed relationship between amino acid metabolism and the pathophysiology of sepsis
without shock ......................................................................................................................................188
Figure 11.2 Proposed relationship between amino acid metabolism and the pathophysiology of septic
shock ...................................................................................................................................................189
v
Table of tables
Table 6.1 Baseline characteristics of patients ..................................................................................... 43
Table 6.2 RH-PAT index and related variables ................................................................................... 44
Table 7.1 Column gradient regime for the Nova-Pak column method.................................................64
Table 7.2 Column gradient regime for the Shim-pack column method................................................ 65
Table 7.3 Comparison of Nova-Pak and Shim-pack method. .............................................................. 72
Table 7.4 Comparison of amino acid concentrations measured in the quality control plasma by three
independent, NATA-certified laboratories with the two MSHR methods. ............................................ 73
Table 7.5 Mobile phase delivery program ........................................................................................... 81
Table 7.6 Average absolute and relative recovery of analytes. ........................................................... 86
Table 7.7 Intra-assay and inter-assay precision calculated from pooled quality control plasma ....... 88
Table 7.8 Assay accuracy calculated from spiked plasma samples ..................................................... 89
Table 7.9 Healthy plasma arginine, homoarginine and methylated arginine values........................... 90
Table 7.10 Characteristics of study subjects........................................................................................ 98
Table 7.11 Median (IQR) arginine and ornithine plasma concentrations over time from blood stored
at room temperature compared to blood stored on ice...................................................................... 100
Table 7.12 Changes in amino acid concentrations in whole blood after 24 hours at room temperature
and on ice........................................................................................................................................... 102
Table 8.1 Cohort information ............................................................................................................ 113
Table 8.2 Baseline characteristics .................................................................................................... 124
Table 8.3 Baseline plasma asymmetric dimtheylarginine and related variables.............................. 125
Table 8.4 Longitudinal results in subjects with sepsis ...................................................................... 129
Table 9.1 Baseline clinical characteristics of participants................................................................ 144
Table 9.2 Immunological characteristics of participants .................................................................. 145
Table 10.1 Patient details for the three cryopreserved PBMC groups. .............................................167
vi
Acknowledgments
This thesis only exists because many people have helped along the way, either
directly or indirectly, and I would like to take this opportunity to thank them.
Firstly, thank you to my supervisors: Dr Tonia Woodberry, for having an
unshakeable faith in my ability (despite glaring evidence to the contrary) and who
was my inspiration in the first place; Professor Nicholas Anstey (my Dumbledore)
for giving kind, thoughtful advice even when pressed with much weightier problems
than I; and Dr Yvette McNeil for introducing me to the joys and horrors of HPLC
and who read parts of this thesis after she had left Menzies and was sailing around
the world.
I sincerely appreciate the work of all the people who helped with the logistics of this
project, particularly Dr Joshua Davis who coordinated the sepsis clinical trials at
Royal Darwin Hospital, contributed to the design of this project, patiently answered
all my questions about sepsis and, on top of everything, helped move all our samples
when the freezer broke down. Thanks also to Mark McMillan, Jane Thomas and the
staff at ICU for recruiting and consenting patients and collecting blood.
Thanks to everyone who shared the load of the lab work, especially Kim Piera who
did an amazing job coordinating the lab side of things: efficiently running ELISAs,
processing blood, staining cells and keeping freezers organized. Thank you, Kim,
for putting up with me. Thank you Gabriela Minigo for eagerly helping with the
immunology part of my project when Tonia was away and for teaching me how to
separate cells with magnetic beads. Thanks also to Catherine Jones, who took over
vii
the ADMA assay, and Barbara MacHunter, who helped with blood processing, for
your help in the lab and your friendship.
Thanks to the support team at Menzies including research admin, IT and especially
operations (Jo Bex, Sue Hutton and Bronwyn Kennard) for ensuring everything ran
as smoothly possible. Thanks to the staff and students from the laboratory and Global
Health division for random bits of advice, tearoom troubleshooting and great
company. There are too many to name – but you know who you are! And thanks to
my fellow students in the Beach Club (Annette Dougall, Paul Burgess, Steve Tong,
Megan Lawrence, Jaqui Hughes, Leisa McCarthy and Anna Stephens) for sharing the
‘music that got me through my PhD’, desk plants and the tea club.
Sometimes there are repetitive jobs that need doing and a spoonful of sugar really
does help the medicine go down. Thanks to Radio National for their thought-
provoking podcasts and Territory FM (the only radio station I could get in the lab)
for their daggy late night music.
Of course, the biggest thank you goes to my family. Thanks to Mum and Dad for all
their help and support. Thanks to my little brother Damian for reminding me that
there is more to life than a PhD. And thanks to my husband Norman (my Mr Darcy)
for forgiving my long hours and grumpiness, and, in times of emergency, for being
the world’s most charming cleaner, chef, comedian and personal trainer.
viii
Abstract
A better understanding of the pathophysiology of sepsis is required to improve
current treatments. The bioavailability of two amino acids, arginine and tryptophan,
can help regulate both microvascular and immune function. The aim of this thesis
was to investigate the role of inflammation and amino acid bioavailability in the
pathophysiology of sepsis.
The methodology presented in this thesis shows that HPLC is an effective way to
accurately measure amino acids and their metabolites in plasma from sepsis patients.
Furthermore, results from a time-course experiment demonstrated that plasma needs
to be promptly separated from blood after collection in order to obtain reliable
measurements of arginine.
We found that increased circulating neutrophil counts were associated with increased
plasma arginase activity, and decreased plasma arginine concentrations in sepsis.
Furthermore we found that arginine bioavailability to nitric oxide synthase was
further reduced by increased concentrations of asymmetric dimethylarginine in
patients with septic shock. The ratio of arginine to asymmetric dimethylarginine was
associated with decreased microvascular reactivity and increased inflammation in
sepsis.
We found that sepsis patients have decreased plasma tryptophan and increased
plasma kynurenine concentrations, suggesting increased indoleamine 2,3-
dioxygenase activity. An increased ratio of kynurenine to tryptophan was associated
ix
with decreased microvascular reactivity, increased inflammation and circulating T
cell lymphopenia.
Finally, we identified circulating activated granulocytes with a myeloid derived
suppressor cell phenotype in septic shock patients which impair T cell signalling,
partly via arginine depletion. The percentage of myeloid derived suppressor cells in
sepsis was directly associated with plasma interleukin-6 concentrations.
In summary, these findings demonstrate that systemic inflammation in sepsis is
associated with both decreased amino acid bioavailability and increased circulating
myeloid derived suppressor cells. Decreased amino acid bioavailability may
contribute to endothelial and immune dysfunction in sepsis. Therapies which
improve amino acid bioavailability may be potential adjunctive treatments in sepsis.
x
Declaration of author’s contribution
This thesis is my own work with the following clarifications regarding the multi-
author papers and manuscripts in this thesis:
This project developed from a study by clinician and infectious diseases specialist,
Dr Joshua Davis, investigating the epidemiology of sepsis in the NT and endothelial
dysfunction in sepsis. My project investigated the role of amino acid metabolism in
sepsis and its effects on endothelial and immune dysfunction in sepsis. Our projects
relied on the same cohorts but were very different conceptually and we had different
responsibilities. Dr Davis, Ms Thomas or Mr McMillan usually recruited and
consented patients, measured endothelial function and collected blood. Myself, Dr
Woodberry, Ms Piera or Ms MacHunter usually collected the blood from the
hospital, removed the plasma and separated the peripheral blood mononuclear cells.
Dr Davis applied for ethics and coordinated the clinical side of the project while I, Dr
Woodberry and Ms Piera coordinated the laboratory side of the project.
Chapter 6 Experimental design and hypotheses
The Critical Care paper was written by Dr Joshua Davis. I helped process the blood
samples in the laboratory, did most of the extraction and derivatisation of the plasma
for HPLC analysis in the general amino acid assay, made up the buffers for the
HPLC machines and helped integrate the chromatograms. I provided intellectual
input into the whole paper.
xi
Chapter 7 Measuring amino acids in plasma
General amino acids assay: All methods of the general amino acid assay were
developed by Dr Yvette McNeil. I prepared buffers, extracted and derivatised
samples, integrated results, assisted with troubleshooting, HPLC machine
maintenance and method validation for both the Nova-Pak and Shim-pack methods.
Dr McNeil performed the validation for the Gemini method while I wrote the draft
version of the manuscript with her input at all stages.
ADMA assay: I developed the original extraction and HPLC methods for the
ADMA assay. I trained Catherine Jones and she further optimised the extraction and
HPLC methods. Ms Jones validated the assay with my input and the resulting
published paper was written together with joint first authorship. I was the
corresponding author for the paper.
STOPWATCH: The BMC Clinical Pathology paper was written by Dr Joshua Davis.
The experiment was designed by Dr Davis, myself, Dr Woodberry and Dr McNeil. I
helped coordinate the logistics of the experiment and helped run the samples on the
HPLC. Dr Davis analysed the results and drafted the first version of the paper while
I had intellectual input into the paper at all stages.
Chapter 8 Arginine bioavailability in sepsis
Arginase: I developed the hypotheses of the experiment, analysed the results and
wrote the paper. The arginase activity assay was performed in the lab of Professor
J.B. Weinberg, Duke University, Durham, NC, USA.
ADMA results: ADMA concentrations were determined by HPLC by Ms Jones.
Both Dr Davis and I analysed the results. The paper was written together with joint
first authorship.
xii
Chapter 9 Tryptophan bioavailability in sepsis
The draft manuscript of the KT paper was written by myself and Dr Davis as joint
first authorship. I ran the samples on the HPLC, assisted integrating the
chromatograms, analysed the results and wrote the first version of the draft.
Chapter 10 Inflammation and T cell suppression in sepsis
I developed the MDSC hypothesis and designed the cell staining and cell function
experiments with input from Dr Woodberry and Dr Minigo. I performed the
experiments with the assistance of Ms Kim Piera. I analysed the results with input
from Dr Woodberry and wrote the first version of the paper. Dr McNeil measured
amino acid concentrations in the media and cell supernatants on the HPLC unit and
either Dr McNeil or myself integrated the chromatograms. Dr Woodberry organised
the arginine and tryptophan free RPMI with input from myself, Dr McNeil and Ms
Piera.
xiii
Publications forming the basis of this thesis
Publications
Peer-reviewed journal articles
Davis, J. S., C. J. Darcy, K. Piera, Y. R. McNeil, T. Woodberry and N. M. Anstey (2009). "Ex-vivo changes in amino acid concentrations from blood stored at room temperature or on ice: implications for arginine and taurine measurements." BMC Clin Pathol 9: 10.
Davis, J. S., T. W. Yeo, J. H. Thomas, M. McMillan, C. J. Darcy, Y. R. McNeil, A. C. Cheng, D. S. Celermajer, D. P. Stephens and N. M. Anstey (2009). "Sepsis-associated microvascular dysfunction measured by peripheral arterial tonometry: an observational study." Crit Care 13(5): R155.
Jones, C. E., C. J. Darcy, T. Woodberry, N. M. Anstey and Y. R. McNeil (2010). "HPLC analysis of asymmetric dimethylarginine, symmetric dimethylarginine, homoarginine and arginine in small plasma volumes using a Gemini-NX column at high pH." J Chromatogr B Analyt Technol Biomed Life Sci 878(1): 8-12.
Davis, J. S., C. J. Darcy, T. W. Yeo, C. Jones, Y. R. McNeil, D. P. Stephens, D. S. Celermajer and N. M. Anstey (2011). "The arginine: asymmetric dimethylarginine ratio, microvascular reactivity and organ failure in sepsis." PLoS One 6(2): e17260
Darcy, C. J., J. S. Davis, T. Woodberry, Y. R. McNeil, D. P. Stephens, T. W. Yeo and N. M. Anstey (2011). " An observational cohort study of the kynurenine to tryptophan ratio in sepsis: association with impaired immune and microvascular function." PLoS One 6(6): e21185
Under review
Darcy, C. J., T. Woodberry, J. S. Davis, K. Piera, Y. R. McNeil, D. P. Stephens, T. W. Yeo, J. B. Weinberg and N. M. Anstey "Increased plasma arginase activity in sepsis is associated with increased circulating neutrophils."
In preparation
Darcy, C. J., K. Piera, G. Minigo, J. S. Davis, Y. R. McNeil, J. B. Weinberg, N. M. Anstey and T. Woodberry "Myeloid derived suppressor cells impair T cell signalling in septic shock patients."
Darcy, C. J., N. M. Anstey and Y. R. McNeil "Routine analysis of plasma amino acids using HPLC and AccQ-Fluor™ derivatives: a comparison of two different HPLC methods."
xiv
Abstracts
Darcy C. J, T. Woodberry, J. S. Davis, Y. R. McNeil, C. Jones, D. P. Stephens and N. M. Anstey. Tryptophan metabolism is associated with lymphopenia and disease severity in sepsis. Australasian Society for Immunology 38th Annual Scientific Meeting, Canberra, December 2008.
Darcy C .J., K. A. Piera, G. Minigo, J. S. Davis, Y. R. McNeil, N. M. Anstey, and T. Woodberry. Decreased T cell zeta-chain expression and low plasma arginine in human sepsis. Australasian Society for Immunology 39th Annual Scientific Meeting, Gold Coast, December 2009.
Darcy, C. J., K. A. Piera, G. Minigo, J. S. Davis, Y. R. McNeil, N. M. Anstey and T. Woodberry. Myeloid derived suppressor cells and arginase contribute to human T cell suppression in sepsis. Molecular and Cellular Biology of Immune Escape in Cancer, Keystone Symposia, Keystone, Colorado, USA, February 2010.
Darcy, C. J., K. A. Piera, G. Minigo, J. S. Davis, Y. R. McNeil, D. P. Stephens, N. M. Anstey and T. Woodberry. Septic shock patients have increased numbers of circulating myeloid derived suppressor cells. Australiasian Society for Infectious Diseases Annual Scientific Meeting, Darwin, May 2010
xv
Abbreviations
ACCP/SCCM American College of Chest Physicians / Society of Critical Care Medicine
ADMA Asymmetric Dimethylarginine AMQ Aminoquinoline Ang-2 Angiopoietin-2 ANOVA Analysis of variance BLING Beta-Lactam InfusioN Group (study name) CAT Cationic Amino acid Transporter cGMP Cyclic Guanosine Monophosphate CRRT Continuous Renal Replacement Therapy DDAH Dimethylarginine dimethylaminohydrolyase EDTA Ethylenediaminetetraacetic acid EIF2 Eukaryotic translation Initial Factor 2 ELISA Enzyme Linked Immunosorbent Assay eNOS Endothelial Nitric Oxide Synthase (NOS3) FRESH Finger Reactive hyperaemia to measure Endothelial Function in
Sepsis and in Health (study name) GCN2 General Control Non-depressible 2 GTP Guanosine Triphosphate HCl Hydrochloric acid HPLC High Performance Liquid Chromatography ICAM-1 Intracellular Adhesion Molecule-1 ICU Intensive Care Unit IDO Indoleamine-2,3-dioxygenase IFNγ Interferon-γ IL10 Interleukin-10 IL6 Interelukin-6 IL8 Interleukin-8 iNOS Inducible Nitric Oxide Synthase (NOS2) IQR Interquartile Range ITAM Immunoreceptor Tyrosine-based Activation Motif KT ratio Kynurenine to Tryptophan ratio LLD Lower Limit of Detection LOD Limit of Detection LOQ Limit of Quantification LPS Lipopolysaccharide MAP Mean Arterial Pressure MDSC Myeloid Derived Suppressor Cells MISTICS Myeloid Immune Suppression of T cells in Sepsis (study name) MMA Monomethylarginine MQ Milli-Q water mRNA Messager Ribonucleic Acid MSHR Menzies School of Health Research NIRS Near Infrared Spectroscopy
xvi
nNOS Neuronal Nitric Oxide Synthase (NOS1) NO Nitric Oxide NOS Nitric Oxide Synthase NPLA n- Propyl L-arginine OPA o-Phthaldialdehyde PBMC Peripheral Blood Mononuclear Cells RBC Red Blood Cell RH-PAT Reactive Hyperaemia Peripheral Arterial Tonometry RP-HPLC Reversed Phase – High Performance Liquid Chromatography RSD Relative Standard Deviation SD Standard Deviation SDMA Symmetric Dimethylarginine sGC Soluble Guanylyl Cyclase SIRS Systemic Inflammatory Response Syndrome SOFA score Sequential Organ Failure Assessment score SPE Solid Phase Extraction SSA Sulphosalicylic acid STATINS STudy of Atovastatin Therapy in Sepsis (study name) STOPWATCH Separation Time of Plasma – Whether Arginine is Time and
Temperature Critical (study name) STREAMS Statins to Reduce Endothelial dysfunction Adjuvant therapy Study
(study name) TCR T Cell Receptor TDO Tryptophan-2,3-pyrrolase TEA Triethylamine Th1 T helper 1 Th2 T helper 2 TNFα Tumor Necrosis Factor α tRNA Transfer Ribonucleic Acid UV Ultra-violet
1
1. Introduction: Aim and scope.
Globally, it is estimated that 18 million people per year develop sepsis, an excessive
inflammatory response to an infection (Slade, Tamber et al. 2003). Severe sepsis can
be fatal, particularly in developing countries where mortality rates may be over 50%
(Tanriover, Guven et al. 2006; Cheng, West et al. 2008; Becker, Theodosis et al.
2009). Even in developed countries, with antibiotic treatment and appropriate
intensive care, severe sepsis is still the leading cause of death in critically ill patients
(Hotchkiss and Karl 2003).
Despite years of research, the pathogenesis of sepsis is still incompletely understood.
Over 30 anti-inflammatory drugs have been trialled in sepsis and all have either
failed or shown minimal benefit (Hotchkiss and Karl 2003). Current treatment for
sepsis is limited to correcting its immediate manifestations and consequences and has
changed little for decades. In order to reduce the burden and mortality rate of sepsis,
we need a better understanding of how it develops.
One of the many effects of inflammation is a disturbance of amino acid
bioavailability. Arginine and tryptophan are two amino acids which help regulate
endothelial and immune responses. As sepsis patients have signs of both endothelial
and immune dysfunction, understanding the role that amino acid bioavailability plays
in sepsis may lead to new treatment targets and improve patient survival. Thus the
aim of this project was to investigate the relationship between inflammation, amino
acid metabolism and the pathology of sepsis.
2
To achieve this aim, chapters 2 to 5 review the literature on sepsis and how arginine
and tryptophan bioavailability affect endothelial and immune function. Chapter 6
sets out the hypotheses of this project and contains a published paper of earlier work
which helped with the experimental design of this project. Chapter 7 includes two
published papers and a draft manuscript describing the methods used to measure
amino acids in sepsis patients. Chapter 8 contains one manuscript describing
arginase activity in sepsis and a published paper describing arginine metabolites in
sepsis. Chapter 9 contains a published paper describing the role of tryptophan
bioavailability in sepsis. Chapter 10 consists of a draft manuscript describing the
relationship between inflammation and immune suppression in sepsis and the role of
amino acid metabolism. The final chapter discusses how the results of this project
contribute to the understanding of sepsis.
3
2. Background: Sepsis
2.1. Introduction
The definition and diagnosis of sepsis has changed over time as our understanding of
sepsis has improved. The aim of this chapter is to give an overview of sepsis. This
chapter outlines the current definitions of sepsis; explains two key contributors to the
pathogenesis of sepsis, endothelial and immune dysfunction; and describes previous
studies of amino acids in sepsis.
2.2. What is sepsis?
Sepsis is difficult to define because it is a syndrome, a collection of clinical signs and
symptoms, rather than a disease in itself. Until recently, definitions of sepsis varied
widely (Bone, Sprung et al. 1992). In August 1991, the American College of Chest
Physicians / Society of Critical Care Medicine (ACCP/SCCM) Consensus
Conference agreed on a set of definitions that could be applied to patients with
sepsis. As a result of this conference, sepsis was defined as a severe inflammatory
response to an infection (Bone, Balk et al. 1992).
Sepsis is diagnosed as an infection, or suspected infection, accompanied by two or
more signs of the systemic inflammatory response syndrome (SIRS) (Bone, Balk et
al. 1992). The manifestations of SIRS are diverse, including: a temperature of more
than 38°C or less than 36°C; a heart rate of more than 90 beats per minute; a
respiratory rate of more than 20 breaths per minute; and a white blood cell count of
more than 12 000 cells/µL, less than 4000 cells/ µL, or more than 10% band forms
4
(Bone, Balk et al. 1992). Figure 2.1 from Bone 1992, illustrates the relationship
between sepsis and SIRS.
Figure 2.1 The relationship between sepsis and the systemic inflammatory immune response. Reproduced from Bone 1992.
Despite the clear guidelines set out by the ACCP/SCCM conference, sepsis is still
difficult to diagnose. There are no SIRS criteria specific to sepsis and there is no
sensitive, specific biomarker of sepsis (Wheeler 2007). Sepsis can develop from a
broad range of infections, usually bacterial, but sometimes viral or fungal. Often, the
causative organism in sepsis cannot be detected by culturing (Brun-Buisson, Doyon
et al. 1995; Martin, Mannino et al. 2003). Furthermore, the underlying infections of
sepsis can be acquired either in the community or in the hospital as a complication of
trauma (including severe injury, burns and surgery). Thus, sepsis is difficult to
5
diagnose and treat because it represents a highly heterogeneous group of patients
(Marshall and Reinhart 2009).
If sepsis is uncontrolled, it can progress to severe sepsis and septic shock. Severe
sepsis is sepsis with organ failure (Bone, Balk et al. 1992). Septic shock is severe
sepsis with sepsis-induced hypotension. This is defined as systolic blood pressure
less than 90 mmHg (or a drop of at least 40 mmHg from baseline) despite adequate
fluid resuscitation, or the need for drugs to maintain higher pressure (Bone, Balk et
al. 1992).
The severity of sepsis is estimated using hospital scoring systems. After admission
into an intensive care unit, all patients are rated with an Acute Physiology and
Chronic Health Evaluation II (APACHE II) score (Knaus, Zimmerman et al. 1981).
The score is determined by physiological measurements such as blood pressure, body
temperature and heart rate within 24 hours of admission. This scoring system
identifies patients with a poor prognosis. In clinical trials of sepsis, an additional
scoring system is often used called the Sequential Organ Failure Assessment (SOFA)
score (Vincent, de Mendonca et al. 1998). This system estimates the severity of
organ failure using markers of respiration, coagulation, liver function, renal function
and cardiovascular function. A higher APACHE II or SOFA score is associated with
a higher risk of mortality.
Although sepsis is defined by signs of a disturbed physiology, the pathogenesis of
these disturbances is unclear. Therefore, the present treatment of sepsis is directed at
the clinical signs of sepsis rather than the underlying mechanisms. Early goal-
6
directed therapy recommends rapid identification of the causative organism,
appropriate antibiotic treatment, blood pressure stabilisation using fluid resuscitation
and vasopressor therapy and mechanical support of organ function, if required
(Dellinger, Carlet et al. 2004). Even with the best available treatment, the mortality
rate for severe sepsis is still 20 - 40% (Martin, Mannino et al. 2003; Finfer, Bellomo
et al. 2004).
Both endothelial and the immune dysfunction appear to contribute to sepsis
pathophysiology. There is significant cross talk and feedback between the
endothelium and the immune response (Rosemblatt and Bono 2004). Endothelial
cells mediate leukocyte adhesion and also influence immune cell function via toll-
like receptors and major histocompatibility complexes I and II (Danese, Dejana et al.
2007) and signals from T cells can prevent apoptosis of endothelial cells
(Stromberg, Woolsey et al. 2009). Therefore my project investigated the dysfunction
of both the endothelium and the immune response.
2.3. Endothelial dysfunction in sepsis
2.3.1. The endothelium
The endothelium is a thin layer of cells lining the inside of blood vessels. The
endothelium regulates vascular tone, cellular adhesion, vessel wall inflammation and
coagulation. A healthy endothelium is constantly sensing and responding to changes
in the local environment. For example, when pathogens invade host tissues
endothelial cells release inflammatory mediators locally, recruit leukocytes and
promote clotting to limit the infection (Aird 2003).
7
Most endothelial cells are in the microcirculation, lining the microvessels where
oxygen transfer to the tissues occurs (Ince 2005). Microvascular reactivity is the
ability of these small blood vessels to dilate in response to shear stress. The
endothelial cells help maintain microvascular tone by sensing blood flow and
releasing molecules which dilate or constrict the vessel, as required (Deanfield,
Halcox et al. 2007). Chapter 4 examines the regulation of microvascular reactivity in
more detail.
2.3.2. Microvascular reactivity and organ failure
As a result of the disturbed signalling pathways in sepsis, endothelial cells are no
longer able to perform their regulatory functions. The lack of regulation results in
excessive and systemic activation of the endothelium (Aird 2003). Sepsis is
characterised by uncontrolled dilation of the larger blood vessels while the
microcirculation remains constricted. As the microcirculation is the source of
oxygen and nutrients to tissues, lack of microvessel blood flow can soon lead to
organ failure (Ince 2005). Furthermore, migrating leukocytes can compromise
endothelial cell integrity causing a ‘leaky endothelium’. This loss of normal barriers
allows bacteria to escape from the gut into the bloodstream and leads to loss of
proteins and macromolecules (McGown and Brookes 2007).
2.4. Immune dysfunction in sepsis
2.4.1. Inflammation in sepsis
Inflammation is a complex response to infection involving the immune system, blood
vessels, liver and brain (Grivennikov, Greten et al. 2010; Medzhitov 2010). An
appropriate inflammatory response eliminates the invading pathogen without causing
8
harm to the host (Remick 2007). Sepsis patients have excessive, systemic
inflammation including high levels of C-reactive protein (CRP), interleukin-6 (IL-6),
interleukin-8 (IL-8) and pro-calcitonin (Herzum and Renz 2008). There is
substantial evidence that the excessive inflammation in sepsis is harmful to the host.
Increased concentrations of tumour necrosis factor (TNF) (Waage, Halstensen et al.
1987; Damas, Reuter et al. 1989) and IL-6 predict mortality in sepsis (Oberholzer,
Souza et al. 2005) and increased inflammation can predict organ failure (Takala,
Jousela et al. 1999).
The association between excess inflammation and mortality in sepsis motivated the
search for suitable agents to suppress the immune system. Agents trialled included
corticosteroids (Bone, Fisher et al. 1987), anti-endotoxin antibodies (Ziegler, Fisher
et al. 1991), tumor necrosis factor antagonists (Abraham, Wunderink et al. 1995;
Fisher, Agosti et al. 1996), cyclooxygenase inhibitors (Bernard, Wheeler et al. 1997),
interleukin-1 receptor antagonists (Fisher, Slotman et al. 1994) and activated protein
C (Bernard, Vincent et al. 2001). Unfortunately, none of these interventions reduced
the mortality rate of sepsis (Hotchkiss and Karl 2003), except for activated protein C
under limited circumstances (Marti-Carvajal, Salanti et al. 2007).
The failure of so many anti-inflammatory agents to reduce the mortality of sepsis led
researchers to question whether sepsis really was simply an uncontrolled
inflammatory response (Bone 1996; Warren 1997). Clearly, inflammation was
present in sepsis patients, but whether it was the cause of death was less certain.
Therefore, investigators turned their attention to immune suppression in sepsis.
9
2.4.2. Immune suppression in sepsis
Despite obvious inflammation, sepsis patients also have signs consistent with
immune suppression, including failed delayed type hyper-sensitivity response; viral
reactivation; lymphocyte apoptosis and impaired T cell function.
Delayed type hypersensitivity is an in vivo test of immune suppression. Patients are
injected with antigens that they should be immune to and tested for a skin response.
Immune suppressed patients are unable to mount a response to the antigen. Delayed
type hypersensitivity is impaired in sepsis and failure corresponds to disease severity
and mortality (Meakins, Pietsch et al. 1977; Christou, Meakins et al. 1995).
Another in vivo sign of impaired immunity is viral reactivity. Cytomegalovirus
reactivity is common in critically ill patients and is associated with longer episodes
of bacteraemia (Curtsinger, Cheadle et al. 1989) and increased mortality (Limaye,
Kirby et al. 2008). Herpes simplex virus reactivation is also common in critically ill
patients and is associated with poorer outcomes (Luyt, Combes et al. 2007).
Both viral reactivation and failed delayed type hypersensitivity tests suggest a defect
in T cell responses. In vivo, sepsis patients have decreased T cells in circulating
blood (Holub, Kluckova et al. 2000) and in the spleen (Hotchkiss, Swanson et al.
1999) as a result of apoptosis and lymphocyte apoptosis is associated with increased
mortality (Le Tulzo, Pangault et al. 2002). In vitro, impaired T cell proliferation in
response to mitogen is associated with mortality in sepsis (Heidecke, Hensler et al.
1999). Similarly, burns patients with impaired T cell response to mitogens are more
10
likely to develop sepsis (Baker, Miller et al. 1979). Thus, sepsis patients have
decreased T cells and impaired T cell function.
Therefore, although sepsis patients have inflammation, there is also suppression of
the adaptive immune response. Indeed sepsis patients unable to clear the original
infection are highly susceptible to secondary infections (Hotchkiss, Coopersmith et
al. 2009). These observations have led to the current theory that the pathogenesis of
sepsis changes with time. Initially sepsis patients tend to have increased
inflammation but over time there is gradual decline into an immunosuppressive state
(Hotchkiss and Karl 2003).
2.5. Amino acids in sepsis
In addition to endothelial and immune dysfunction, sepsis patients also have a
disturbed metabolism. Free amino acids circulate in the blood and can be transported
into cells for protein synthesis. Previous studies measuring circulating amino acid
concentrations in sepsis have had conflicting results. Freund et al. reported that most
amino acids were either normal or high in sepsis (Freund, Ryan et al. 1978), whereas
Druml et al. reported that most amino acids were low or normal in sepsis (Druml,
Heinzel et al. 2001). Arginine has been reported as low (Freund, Ryan et al. 1978) or
increased in sepsis (Chiarla, Giovannini et al. 2006). Similarly tryptophan has been
reported as normal (Freund, Ryan et al. 1978), low (Moyer, McMenamy et al. 1981;
Pellegrin, Neurauter et al. 2005) or high in sepsis (Sprung, Cerra et al. 1991).
Possible reasons for these conflicting reports include inaccurate quantification in
early methods, heterogeneous patient groups (including sepsis with and without
trauma) and small sample sizes.
11
2.6. Conclusion
Sepsis is a complication that can develop in response to a broad range of infections.
Both endothelial and immune dysfunction contribute to the pathogenesis of sepsis.
Endothelial dysfunction impairs the flow of blood to tissue and can lead to organ
failure. Sepsis patients have signs of excessive inflammation and impaired adaptive
immune responses. This means that the immune system in sepsis can harm the host
yet be unable to efficiently clear the infection. Amino acid metabolism appears to be
disturbed in sepsis, but the results so far are conflicting. The next three chapters will
outline the relationships between amino acid metabolism, endothelial function and
the immune response.
12
3. Background: Arginine and tryptophan bioavailability
3.1. Introduction
As discussed in section 2.5 above, early evidence suggested that sepsis patients have
disturbed amino acid metabolism. Many cells are sensitive to extra-cellular
concentrations of amino acids and the metabolism of some amino acids can have
important regulatory effects. Arginine and tryptophan are two amino acids that can
regulate both endothelial and immune function. Amino acid bioavailability is
influenced by many factors including altered catabolism, absorption, synthesis and
recycling. This chapter will introduce arginine and tryptophan and explain the
enzymes that affect their bioavailability.
3.2. Arginine metabolism
Arginine is a semi-essential amino acid found in plant and animal protein. It is a
precursor for many important biological molecules including nitric oxide (NO), urea,
and polyamines (Wu and Morris 1998). Endogenously synthesized levels of arginine
are sufficient for healthy adults, however external sources are required for growth
(Rose 1937), wound healing (Popovic, Zeh et al. 2007) and an effective immune
response (Bronte and Zanovello 2005). Many factors determine arginine
bioavailability including nutrition, muscle breakdown and enzymatic activity. Three
enzymes in particular are relevant to this project: arginase, nitric oxide synthase and
dimethylarginine dimethylaminohydrolyase (DDAH).
13
3.2.1. Arginase
Arginase is an enzyme which converts arginine to ornithine and urea. There are two
isoforms of arginase, arginase I and arginase II which are encoded by different genes.
Arginase I is a cytosolic enzyme expressed in hepatocytes (Husson, Bouazza et al.
1984), neutrophils (Munder, Mollinedo et al. 2005), red blood cells (Bernard, Kasten
et al. 2008), myeloid derived suppressor cells (MDSC) (Rodriguez, Ernstoff et al.
2009) and endothelial cells and smooth muscle cells (Morris 2009). Arginase II is a
mitochondrial enzyme expressed in renal cells, neurons and macrophages (Bronte
and Zanovello 2005). Arginase is usually an intra-cellular enzyme that is not
released until cell death (Morris 2007); however both human neutrophils and
myeloid derived suppressor cells secrete arginase I into the extra-cellular
environment (Rodriguez, Ernstoff et al. 2009).
Arginase I is differentially regulated in mice and humans. In mice, resting
leukocytes do not express arginase I, but arginase I expression is induced in
macrophages and dendritic cells in response to T helper 2 (Th2) cytokines or
lipopolysaccharide (Corraliza, Soler et al. 1995; Sonoki, Nagasaki et al. 1997;
Munder, Eichmann et al. 1999). In contrast, arginase I is constitutively expressed in
human neutrophils and is upregulated in activated neutrophils (Rodriguez, Ernstoff et
al. 2009), but apparently not inducible in human peripheral blood mononuclear cells
(PBMC) (Munder, Mollinedo et al. 2005). In addition, although both mouse and
human MDSC have been reported to express arginase, mouse MDSC arginase is
intracellular, whereas human MDSC arginase is secreted into the micro-environment
(Rodriguez, Ernstoff et al. 2009).
14
3.2.2. Nitric oxide synthase
Arginine is the primary substrate of nitric oxide synthase and thus essential for nitric
oxide (NO) synthesis (Figure 3.1). There are 3 isoforms of nitric oxide synthase,
neuronal (nNOS or NOS1), inducible (iNOS or NOS2) and endothelial (eNOS or
NOS3). Both nNOS and eNOS are constitutively expressed and usually produce
small amounts of NO whereas iNOS is induced by interferon-γ (IFNγ) and
lipopolysaccharide (LPS) and can produce 20 times more NO than constitutive
enzymes (Bruckdorfer 2005).
The role of nitric oxide in sepsis is controversial. Excess iNOS expression and
activity in sepsis is well established and is associated with hypotension (Jia, Pan et
al. 2006). However, a clinical trial of a NOS inhibitor increased mortality in septic
shock (Lopez, Lorente et al. 2004). The inhibitor in this trial did not distinguish
between iNOS and eNOS and may have inhibited both isoforms. eNOS is important
for microvessel dilation (Yamashita, Kawashima et al. 2001) and maintaining tissue
perfusion. Increasing data suggest that intravascular dysfunction in severe sepsis is a
state of NO deficiency (Trzeciak, Cinel et al. 2008). Thus, potential adjunctive
treatments for sepsis should aim to decrease iNOS and increase eNOS (McGown and
Brookes 2007).
This project will mainly focus on eNOS as it is the most important isoform for
regulating microvasular reactivity and tissue blood flow. Both eNOS expression and
activity affect endothelial NO production. Although constitutively expressed,
various conditions can modulate eNOS transcription and mRNA stability
(Hammerman, Klings et al. 1999; Searles 2006). For example, shear stress increases
15
eNOS mRNA transcription (Harrison, Kurz et al. 1992) while LPS decreases eNOS
mRNA stability (Lu, Schmiege et al. 1996). The activity of eNOS is dependent on
the availability of substrate and co-factors. Arginine bioavailability to eNOS is
reflected by plasma arginine concentrations, rather than intracellular arginine
concentrations. Endothelial NOS is closely associated with the cationic amino acid
(CAT) transporter within the caveolae of the cell membrane (McDonald, Zharikov et
al. 1997) and the rate of endothelial NO synthesis is dependent on cellular uptake of
arginine through the CAT transporter (Zani and Bohlen 2005). Healthy plasma
arginine concentrations are reported to be between 30 µM and 100 µM (Bode-Boger,
Scalera et al. 2007), however as delayed blood processing can deplete arginine
concentrations (see Chapter 6) the lower limit is more likely to be about 60 µM (see
Chapter 8).
Figure 3.1 The two reactions of nitric oxide synthesis as catalysed by nitric oxide synthase. Reproduced from Ashina 2004 (Ashina 2004)
16
3.2.3. DDAH
Methylation of arginine residues in proteins is a common post-translational
modification. Monomethylarginine (MMA) is arginine with one additional methyl
group and dimethylarginine has two additional methyl groups. There are two
dimethylarginines, asymmetric dimethylarginine (ADMA) and symmetric
dimethylarginine (SDMA) (Figure 3.2). Free methylarginines are released during
proteolysis and are not incorporated back into proteins. Healthy ADMA and SDMA
plasma concentrations are estimated to be between 0.4 µM and 0.6 µM (Teerlink
2007). Healthy MMA concentrations are about one tenth of this concentration.
Figure 3.2 Molecular structures of arginine, monomethylarginine (MMA), asymmetric dimethylarginine (ADMA) and symmetric dimethylargin ine (SDMA). Reproduced from Dweik 2007 (Dweik 2007)
ADMA, SDMA and MMA all have a higher affinity for the CAT transporter than
arginine (Bode-Boger, Scalera et al. 2006). ADMA and MMA also competitively
inhibit arginine binding to NOS (Vallance, Leone et al. 1992). Thus the
arginine/ADMA ratio is an estimate of arginine availability for NOS. The
17
arginine/ADMA ratio in healthy people ranges between 132 and 227, with a median
of 211 (Bode-Boger, Scalera et al. 2007).
All three methylarginines are eliminated via renal excretion. ADMA and MMA are
also metabolised by dimethylarginine dimethylaminohydrolyase 1 and 2 (DDAH1
and DDAH2). About 80% of ADMA is degraded by DDAH (Achan, Broadhead et
al. 2003). Both the kidney (Ogawa, Kimoto et al. 1989) and the liver (Nijveldt,
Teerlink et al. 2003) are important sources of DDAH. Leiper et al. demonstrated that
loss of DDAH activity led to an accumulation of ADMA, a reduction in NO,
endothelial dysfunction, increased systemic vascular resistance and elevated systemic
and pulmonary blood pressure (Leiper, Nandi et al. 2007). Inflammation inhibits
DDAH activity, leading to increased ADMA and decreased arginine bioavailability
to NOS (Ito, Tsao et al. 1999; Puchau, Hermsdorff et al. 2009).
High levels of ADMA have been associated with a range of disease states. A strong
link has been established between ADMA levels and cardiovascular events in end-
stage renal disease (Zoccali, Mallamaci et al. 2006). ADMA has been described as
an independent risk factor for coronary heart disease (Schulze, Lenzen et al. 2006)
and increased ADMA in patients with type 2 diabetes has been linked to subsequent
cardiovascular disease (Krzyzanowska, Mittermayer et al. 2007).
The precise mechanism by which ADMA inhibits endothelial nitric oxide synthase is
not completely understood. Recent studies suggest that ADMA contributes to eNOS
uncoupling by causing mitochondrial dysfunction and reducing heat shock protein-90
activity (Sud, Wells et al. 2008).
18
3.3. Tryptophan metabolism
Tryptophan cannot be synthesised by humans and is an essential amino acid. It is
used for protein synthesis and can be metabolised into the important co-enzyme,
nicotinamide adenine dinucleotide and the neurotransmitter, serotonin. There are
two enzymes that oxidise tryptophan to kynurenine, tryptophan-2,3-pyrrolase (TDO)
and indoleamine-2,3-dioxygenase (IDO). TDO is expressed in the liver in response
to excess tryptophan. IDO is expressed in the brain, lung, heart, kidney, intestines,
endothelium and leukocytes and is up-regulated in response to inflammation and
infection (Takikawa, Yoshida et al. 1986; Carlin, Borden et al. 1989; Beutelspacher,
Tan et al. 2006). Enhanced IDO activity in pathological conditions suppresses TDO
activity (Takikawa, Yoshida et al. 1986).
Figure 3.3 The enzyme indoleamine 2,3-dioxygenase (IDO) oxidises tryptophan to kynurenine. Reproduced from Barth 2009 (Barth, Ahluwalia et al. 2009)
19
3.3.1. Indoleamine 2,3-dioxygenase
There are two types of IDO, IDO-1 and IDO-2. In response to inflammation, IDO-1
increases expression, whereas IDO-2 tends to decrease expression (Ball, Sanchez-
Perez et al. 2007). The two types of IDO use a similar range of substrates but differ
in their sensitivity to inhibitors (Ball, Yuasa et al. 2009).
Increased IDO activity is associated with increased inflammation in cancer (Muller,
Sharma et al. 2008), chronic kidney disease (Schefold, Zeden et al. 2009) and human
immunodeficiency virus infection (Fuchs, Forsman et al. 1990). IFN-γ can increase
IDO expression in a range of cell types including endothelial cells, monocytes, renal
tubular epithelial cells and hepatocytes (Carlin, Borden et al. 1989; Larrea, Riezu-
Boj et al. 2007; Mohib, Guan et al. 2007; Iwamoto, Ito et al. 2009; Wang, Liu et al.
2010). Lipopolysaccharide (LPS) can induce IDO expression in the lung and brain
in an IFN-γ independent manner, via TNF-α (Fujigaki, Saito et al. 2001). IDO
activity is estimated in vivo by the ratio of kynurenine to tryptophan (the KT ratio).
3.3.2. Relationship between NOS and IDO
There are several mechanisms of negative feedback between IDO and iNOS and
these may be applicable to eNOS as well. Increased IDO activity decreases iNOS
expression and activity. Significantly less iNOS is expressed in response to IFN-γ in
a low tryptophan environment and iNOS expression increases more than 10 fold after
the addition of tryptophan but not arginine or nicotinamide (Chiarugi, Rovida et al.
2003). The mechanism of IDO suppression of iNOS appears to be via up-regulation
of the regulatory enzyme, heme oxygenase (Oh, Pae et al. 2004). As heme
oxygenase also regulates eNOS activity (Seki, Naruse et al. 1997), it is likely that
20
this mechanisms also applies to eNOS. Increased NOS activity decreases IDO
activity and expression. The presence of NO inhibits IDO activity in a dose-
dependent manner (Thomas, Mohr et al. 1994) and accelerates degradation of IDO
in the proteosome (Hucke, MacKenzie et al. 2004). Figure 3.4 summarises the
feedback between IDO and NOS, via heme oxygenase.
Figure 3.4 The reciprocal relationship between indoleamine 2,3 dioxygenase (IDO) and nitric oxide synthase (NOS) mediated by heme oxygenase and nitric oxide. Green arrows represent up-regulation or production and red arrows represent down-regulation or inhibition.
21
3.4. Conclusion
Inflammation can decrease the bioavailability of both arginine and tryptophan.
Arginine and tryptophan concentrations depend on nutrition, muscle breakdown,
protein synthesis and enzymatic activity. Arginine bioavailability is influenced by
the interaction of arginase, NOS and DDAH. In addition, increased arginase,
decreased DDAH activity and increased IDO activity all limit NOS activity.
Tryptophan bioavailability depends on nutrition, muscle breakdown, protein
synthesis, IDO activity and, to a lesser extent, TDO activity. Increased NOS activity
limits IDO activity and increased IDO activity limits TDO activity. Thus
inflammation and the negative feedback between IDO and NOS link arginine and
tryptophan metabolism. The next two chapters consider how arginine and
tryptophan bioavailability influence the regulation of microvascular reactivity and T
cell function.
22
4. Background: Regulation of microvascular reactivity
4.1. Introduction
As mentioned in the previous chapter, arginine is the primary substrate for nitric
oxide synthesis and tryptophan is the pre-cursor of kynurenine. Both nitric oxide and
kynurenine play an important role in regulating vascular tone, via the endothelium.
This chapter will discuss how microvascular reactivity is regulated with a particular
focus on nitric oxide and kynurenine.
4.2. Endothelial regulation of microvascular reactivity
As mentioned in section 2.3, microvascular reactivity is the ability of microvessels to
dilate in response to shear stress. The endothelium regulates microvascular reactivity
by releasing vasoactive molecules in response to shear stress and signalling
molecules such as bradykinin, adenosine, vascular endothelial growth factor and
serotonin (Deanfield, Halcox et al. 2007). Vasoactive molecules include
vasodilation and vasoconstriction molecules. Nitric oxide is a key endothelial
vasodilation molecule; however nitric oxide-independent pathways also exist.
Endothelium-Derived Hyperpolarizing Factor (EDHF) increases potassium
conductance and relaxes the blood vessel via hyperpolarisation of the surrounding
smooth muscle cells (Busse, Edwards et al. 2002). The cyoclooxygenase molecule
prostacylin also seems to have a minor role in maintaining vascular tone.
Endothelium-derived vasoconstrictor molecules include endothelin, vasoconstrictor
prostanoids and angiotensin II (Deanfield, Halcox et al. 2007).
23
4.3. Nitric oxide and microvascular reactivity
In 1980, Furchgott and Zawadski observed that blood vessels treated with
acetylcholine relaxed if the endothelial lining was present but contracted if the
endothelium had been removed (Furchgott and Zawadzki 1980). The mysterious
substance released by the endothelium to relax blood vessels was called
endothelium-derived relaxing factor. It was not until the late 1980s that two groups
established that endothelium-derived relaxing factor was actually nitric oxide (NO)
(Ignarro, Buga et al. 1987; Palmer, Ferrige et al. 1987). Nitric oxide is a gas and free
radical molecule. It was partly because NO was already an established
environmental pollutant that it took so long to confirm it that also plays an important
role in maintaining vascular stability (Bruckdorfer 2005).
Nitric oxide causes vasodilation by signalling to the smooth muscle cells surrounding
the blood vessel (Bruckdorfer 2005). NO diffuses through the lipid layer of the cell
membrane into the cytosol of the smooth muscle cells, where it interacts with soluble
guanylyl cyclase (sGC). Soluble guanylyl cyclase is an enzyme which converts the
nucleotide guanosine triphosphate (GTP) into the signalling molecule, cyclic
guanosine monophosphate (cGMP). By binding to the haem moity of sGC, NO
increases the activity of the enzyme by up to 200 fold. Through a series of
intermediates, cGMP reduces phosphorylation of the myosin light chain, thereby
inhibiting contraction of the smooth muscle cells. Thus, NO increases the production
of cGMP, which inhibits muscle contraction, allowing the blood vessel to dilate. The
lifetime of the NO-haem complex in sGC is about 0.2 seconds, therefore small
amounts of NO need to be constantly produced to maintain vascular stability. As
24
mentioned in chapter 2, most endothelial cells are in the microcirculation. Therefore
it is eNOS that is most important for the regulation of microvascular reactivity.
4.4. Kynurenine and microvascular reactivity
Kynurenine was recently identified as an endogenous vasodilator (Wang, Liu et al.
2010). Experiments by Wang et al. (Wang, Liu et al. 2010) demonstrated that the
endothelium was an important source of IDO activity and, thus, kynurenine. Similar
to NO, kynurenine activates the cGMP-pathways and increases tissue concentrations
of cGMP.
4.5. Conclusion
The endothelium produces several vasoactive molecules which regulate vascular
tone. Two of these molecules are products of arginine and tryptophan metabolism,
nitric oxide and kynurenine. The next chapter will consider how the bioavailability
of these two amino acids can also regulate T cell function.
25
5. Background: Regulation of T cell function
5.1. Introduction
As well as having important effects on endothelial function, both arginine and
tryptophan are immunoregulatory amino acids. This chapter will focus on T cell
suppression via down-regulation of the T cell receptor zeta-chain. Both arginine and
tryptophan bioavailability can regulate T cell zeta-chain expression. This chapter
will also give an overview of myeloid derived suppressor cells which can impair T
cell zeta-chain expression, partly by depleting amino acids.
5.2. Overview of T cell regulation
T cells are an important part of the adaptive immune response. T cells, so called
because they mature in the thymus, recognise specific, presented antigen via the T
cell receptor (TCR). After a naïve T cell is activated by its cognate antigen, it
proliferates and differentiates into an effector T cell. Cytotoxic T cells kill cells that
are infected with viruses or intracellular pathogens. Helper T cells help activate B
cells and macrophages (Murphy, Travers et al. 2008). T cells are important
regulators of the entire immune response, however T cells can harm the host if
uncontrolled (Romagnani 2006).
T cell function is regulated by many mechanisms including cytokines, suppressive
cells and amino acid availability. Cytokines can either suppress or stimulate T cells.
For example, interleukin-10 suppresses T cell function (Akdis and Blaser 1999)
while interleukin-2 stimulates T cell proliferation (Lan, Selmi et al. 2008). Cells
which suppress T cell function include regulatory T cells (Vignali, Collison et al.
2008) and myeloid derived suppressor cells (Gabrilovich and Nagaraj 2009). There
26
is also a growing appreciation of the role of amino acid availability in the regulation
of T cells (Li, Yin et al. 2007), which will be the focus of this chapter.
5.3. T cell zeta-chain expression
The T cell receptor is a trans-membrane molecule that is composed of six CD3
subunits (αβγδεζ) (Figure 5.1). The TCR is expressed on the surface of all T cells
and allows them to distinguish between self and non-self. After recognition and
binding of foreign antigen, the TCR initiates signaling cascades that can cause cell
activation, proliferation and cytokine secretion.
Figure 5.1 T cell receptor structure showing showing the arrangement of the αβγγγγδε and ζ chains. Reproduced from Baniyash 2004 (Baniyash 2004)
The CD3 zeta-chain (ζ) is the primary signal transduction unit of the TCR. When the
T cell receptor recognizes its cognate antigen, it is the zeta-chain which
communicates this to the rest of the cell. The zeta-chain contains three
immunoreceptor tyrosine-based activation motifs (ITAMs) which are phosphorylated
at the tyrosine residues after TCR engagement (Baniyash, Garcia-Morales et al.
1988). The phosphorylated ITAMs in the zeta-chain recruit the cell signaling
27
molecule ZAP70, which initiates a signaling cascade throughout the cell, leading to
T cell activation. In some circumstances, the zeta-chain can become dissociated
from the rest of the TCR, leaving the TCR on the cell membrane without the zeta-
chain. When this occurs, the TCR is unable to effectively signal to the rest of the
cell. It follows that T cells with impaired zeta-chain expression respond poorly to
TCR-mediated stimulation with impaired proliferation (Rodriguez, Zea et al. 2003)
and cytokine production (Ochoa, Zea et al. 2007). Thus zeta-chain expression is
essential for an effective T cell response.
There are several factors that can impair T cell zeta-chain expression including
amino acid starvation, reactive oxygen species and chronic inflammation (Bronstein-
Sitton, Cohen-Daniel et al. 2003; Ezernitchi, Vaknin et al. 2006). Possible
mechanisms for zeta-chain down-regulation include a shorter half-life of zeta-chain
mRNA (Rodriguez, Zea et al. 2002), decreased zeta-chain synthesis (Zea, Rodriguez
et al. 2004) and increased lysosomal degradation (Bronstein-Sitton, Cohen-Daniel et
al. 2003). Dysfunctional T cells with impaired zeta-chain expression have been
described in patients with HIV, leprosy, autoimmune diseases and cancer (Finke, Zea
et al. 1993; Gunji, Hori et al. 1994; Zea, Curti et al. 1995; Liossis, Ding et al. 1998;
Matsuda, Ulfgren et al. 1998).
5.4. Arginine and T cells
In vitro studies have demonstrated that T cells have impaired T cell zeta-chain
expression when stimulated in the absence of arginine and recover zeta-chain
expression when arginine is added back into the culture (Taheri, Ochoa et al. 2001;
Rodriguez, Zea et al. 2002; Zea, Rodriguez et al. 2004; Rodriguez, Quiceno et al.
28
2007). T cell zeta-chain expression is also impaired when T cells are cultured in the
presence of arginase or arginase-producing cells and zeta-chain expression recovers
when excess arginine or an arginase inhibitor such as N-hydroxy-nor-L-arginine is
added to the culture (Rodriguez, Zea et al. 2003; Munder, Schneider et al. 2006). It
has also been noted that arginase-producing cells can deplete the extra-cellular
environment of arginine, and hence suppress zeta-chain expression more efficiently
than NO-producing cells (Rodriguez, Zea et al. 2003).
The mechanism by which arginine starvation leads to T cell zeta-chain expression is
not completely understood. One mechanism appears to be via the general control
non-depressible 2 (GCN2) kinase pathway. GCN2 is a stress-response kinase that is
activated by amino acid starvation. When extra-cellular arginine concentrations are
low, some transfer RNA molecules (tRNAs) specific for arginine are left empty
within the cell. These uncharged tRNAs bind GCN2 which phosphorylates the
signalling molecule eukaryotic translation initial factor 2 alpha (EIF2α) (Figure 5.2).
Phosphorylated EIF2α leads to altered mRNA translation of a range of proteins
including decreased T cell zeta-chain translation (Fallarino, Grohmann et al. 2006).
Figure 5.2 Potential pathway of zeta down-regulation in response to arginase activity. Reproduced from Bronte 2005, (Bronte and Zanovello 2005)
29
5.5. Tryptophan and T cells
Low extra-cellular concentrations of tryptophan as a result of IDO activity also leads
to impaired T cell zeta-chain expression and decreased T cell proliferation via the
GCN2 pathway (Munn, Sharma et al. 2005; Fallarino, Grohmann et al. 2006).
In addition, low tryptophan and high kynurenine concentrations can induce or
increase T cell apoptosis (Fallarino, Grohmann et al. 2002; Lee, Park et al. 2002;
Fallarino, Grohmann et al. 2003) and T cells cultured in the presence of IDO-
producing cells show increased apoptosis (Fallarino, Vacca et al. 2002). The
mechanism of IDO-dependent apoptosis of T cells appears to be mediated by
caspase-8 (Fallarino, Grohmann et al. 2002). Thus, decreased tryptophan availability
can lead to impaired T cell function and increased T cell apoptosis.
5.6. Myeloid derived suppressor cells
Amino acid deprivation is one of the suppressive mechanisms used by myeloid
derived suppressor cells (MDSC). MDSC, previously called myeloid suppressor
cells or natural suppressor cells, are defined as myeloid derived cells with the ability
to suppress T cell function (Gabrilovich, Bronte et al. 2007). MDSC can impair T
cell zeta-chain expression, T cell proliferation and cytokine production (Bronte and
Zanovello 2005).
MDSC are a heterogeneous group of cells that can express arginase (Bronte, Serafini
et al. 2003), IDO (Jia, Jackson-Cook et al. 2010), NOS (Mazzoni, Bronte et al. 2002)
and produce hydrogen peroxide (Schmielau and Finn 2001), peroxynitrite (Gallina,
Dolcetti et al. 2006) and interleukin-10, depending on the stimulus. MDSC can also
30
reduce T cell access to the amino acid cysteine (Srivastava, Sinha et al. 2009).
MDSC have specifically been shown to down-regulate T cell zeta-chain expression
via arginase (Zea, Rodriguez et al. 2005) and hydrogen peroxide (Schmielau and
Finn 2001).
MDSC link inflammation and adaptive immune suppression (Ostrand-Rosenberg and
Sinha 2009). MDSC accumulate in response to inflammatory signals such as IL6
(Bunt, Yang et al. 2007), prostaglandin E2 and/or cyclooxygenase-2 (Rodriguez,
Hernandez et al. 2005; Sinha, Clements et al. 2007), molecules which stimulate toll-
like receptor 4 (Bunt, Clements et al. 2009) and the pro-inflammatory proteins
S100A8/A9 (Sinha, Okoro et al. 2008), see Figure 5.3 MDSC are a major
mechanism of T cell suppression in patients with inflammation such as cancer (Zea,
Rodriguez et al. 2005; Filipazzi, Valenti et al. 2007; Diaz-Montero, Salem et al.
2009) and inflammatory bowel disease (Haile, von Wasielewski et al. 2008; Nagaraj,
Collazo et al. 2009).
31
Figure 5.3 Relationship between inflammatory mediators and myeloid derived suppressor cell (MDSC) induction. Reproduced from Ostrand-Rosenberg 2009 (Ostrand-Rosenberg and Sinha 2009)
5.7. Conclusion
T cells down-regulate zeta-chain expression in response to arginine starvation,
tryptophan starvation and reactive oxygen species. MDSC suppress T cell zeta-chain
expression, proliferation and cytokine production via arginase and reactive oxygen
species production. MDSC develop in response to inflammation and link
inflammation and immune suppression in cancer.
The next chapter will outline how the literature in this review, combined with our
preliminary results, led to the hypotheses of this project.
32
6. Experimental design and hypotheses
6.1. Introduction
This project builds on earlier sepsis research by the Global Health Division starting
in 2006. As decreased amino acid bioavailability can regulate both immune function
and endothelial function, the aim of this project was to investigate the role of amino
acid bioavailability in sepsis. This chapter describes the three sepsis studies that
form the basis of this project, presents the earlier results that influenced the
experimental design of this project and sets out the hypotheses of this project.
6.2. Clinical studies forming the basis of this project
This project was a sub-study of a series of sepsis studies at Royal Darwin Hospital
between 2007 and 2010. All studies were approved by the Menzies School of Health
Research Human Research Ethics Committee. Sepsis patients or hospital controls
were either participants in the “Finger reactive hyperaemia to measure endothelial
function in sepsis and in health” (FRESH) study, “Statins to reduce endothelial
dysfunction adjuvant therapy study” (STREAMS) or “Beta-lactam infusion group”
(BLING) study. All participants gave informed consent to participate in the sub-
study.
The FRESH study was a local observational study based at Royal Darwin Hospital
which ran from March 2006 until November 2007, investigating endothelial function
in sepsis using reactive hyperaemia peripheral arterial tonometry (RH-PAT).
Endothelial function was measured and blood was collected from 85 sepsis patients
on day 0, 2 and 7 of the study until discharge from the hospital or death. Endothelial
33
measurements and blood was also collected from 45 hospital controls on enrolment
only.
STREAMS is a local sub-study of a national, multi-centre trial called STATINS
(study of atorvastatin therapy in sepsis) which was a phase I clinical trial
investigating the use of statins in sepsis. STREAMS had enrolled 37 patients up to
September 2010. Patients enrolled in both STATINS and STREAMS have
endothelial function measured using RH-PAT and blood collected on day 0, 1, 2, 3,
5, 7, 9 and 10 until discharge from the hospital or death.
BLING is a national clinical trial comparing administration methods of beta-lactam
antibiotics in sepsis patients. Our local sub-study is collecting extra blood for
analysis of immune cell function on day 0, 3 and 4 (MISTICS – Myeloid Immune
Suppression of T cells in Sepsis). Three patients have been enrolled in MISTICS up
to September 2010. Endothelial function is not being measured in these patients.
6.3. Earlier results
In March 2006 (one year before this project began) Global Health Division team
members started the enrolment of patients for the FRESH study. The results of the
FRESH study reproduced in this section helped form the hypotheses of this project,
detailed in section 6.4.
34
6.3.1. Published paper: Sepsis-associated microvascular dysfunction
measured by peripheral arterial tonometry: an observational study.
Authors: Joshua S. Davis1, 2, Tsin W. Yeo1, Jane H. Thomas3, Mark McMillan1,
Christabelle J. Darcy1, Yvette R. McNeil1, Allen C. Cheng1, 2, David S. Celermajer4,
Dianne P. Stephens3, Nicholas M. Anstey1, 2
Authors’ affiliations: 1 – International Health Division, Menzies School of Health
Research and Charles Darwin University, Darwin, NT 0810, Australia. 2 – Division
of Medicine, Royal Darwin Hospital, Darwin, NT, 0810, Australia. 3- Intensive Care
Unit, Royal Darwin Hospital, Darwin, NT, 0810, Australia 4 – Department of
Medicine, University of Sydney and Department of Cardiology, Royal Prince Alfred
Hospital, Sydney, NSW 2006, Australia
Abstract
Introduction
Sepsis has a high mortality despite advances in management. Microcirculatory and
endothelial dysfunction contribute to organ failure, and better tools are needed to
assess microcirculatory responses to adjunctive therapies. We hypothesised that: i)
Peripheral arterial tonometry, a novel user-independent measure of endothelium-
dependent microvascular reactivity, would be feasible in septic subjects; ii) that
microvascular reactivity would be impaired, in proportion to sepsis severity, plasma
arginine concentration and circulating markers of endothelial activation.
Methods
Observational cohort study in a 350-bed teaching hospital in tropical Australia.
Microvascular reactivity was measured at the bedside in 85 adults with sepsis and 45
35
controls at baseline and 2-4 days later by peripheral arterial tonometry.
Microvascular reactivity was related to measures of disease severity, plasma
concentrations of L-arginine (the substrate for nitric oxide synthase), and biomarkers
of endothelial activation.
Results
Baseline reactive hyperaemia index (RH-PAT index), measuring endothelium-
dependent microvascular reactivity; mean [95% CI]) was lowest in severe sepsis
(1.57 [1.43-1.70]), intermediate in sepsis without organ failure (1.85 [1.67-2.03]) and
highest in controls (2.05 [1.91-2.19]); p<0.00001. Independent predictors of baseline
RH-PAT index in sepsis were APACHE II score and mean arterial pressure, but not
plasma L-arginine or markers of endothelial activation. Low baseline RH-PAT index
was significantly correlated with an increase in SOFA score over the first 2-4 days
(r=-0.37, p=0.02).
Conclusions
Endothelium-dependent microvascular reactivity is impaired in proportion to sepsis
severity. Peripheral arterial tonometry has potential as a method of monitoring
responses to novel adjunctive therapies targeting endothelial dysfunction in sepsis.
Introduction
Mortality from severe sepsis remains high, despite advances in its management
(Angus, Pereira et al. 2006). Organ failure commonly occurs despite the achievement
of normal haemodynamics in response to fluid resuscitation, vasopressors and the
treatment of infection. This may be due to impaired vasomotor regulation of the
microcirculation (Ince and Sinaasappel 1999). In sepsis, the endothelium has key
36
roles in regulating vascular tone and permeability and its activation is pivotal in
initiating both the inflammatory and coagulation cascades (Aird 2003).
Endothelial function is assessed clinically by the ability of blood vessels to
vasodilate in response to pharmacological stimuli or to shear stress, and is primarily
dependent on endothelial nitric oxide (NO) production (Deanfield, Halcox et al.
2007). Currently, measurement of endothelial function using techniques such as
laser Doppler, plethysmography and flow-mediated dilatation of the brachial artery
requires skilled operators and is technically difficult to perform at the bedside. As a
result, many clinical studies investigating the endothelium in sepsis have measured
circulating endothelial activation markers, as a surrogate for endothelial function.
Some studies have assessed endothelial function by measuring reactive hyperaemia
in human sepsis using these operator-dependant techniques (Hartl, Gunther et al.
1988; Astiz, DeGent et al. 1995; Young and Cameron 1995; Neviere, Mathieu et al.
1996; Kubli, Boegli et al. 2003; Vaudo, Marchesi et al. 2007). These studies have
generally shown normal baseline blood flow and impaired reactive hyperaemic
responses in sepsis, but have been small (n= 8-45) and have not correlated reactive
hyperaemia with L-arginine or circulating markers of endothelial activation. More
recently, investigators using dynamic near-infrared spectroscopy (NIRS) have found
impaired microvascular responses in sepsis, however the nature of the relationship
between NIRS and endothelial NO activity is unclear (Creteur, Carollo et al. 2007).
Reactive hyperaemia peripheral arterial tonometry (RH-PAT) is a novel, simple and
user-independent bedside technique used to measure microvascular endothelial
function (Celermajer 2008) (Figure 6.1). It is increasingly being used to measure
37
endothelial function as a cardiovascular risk assessment tool in ambulatory patients
(Chenzbraun, Levin et al. 2001; Bonetti, Pumper et al. 2004; Haller, Stein et al.
2007; Kuvin, Mammen et al. 2007; Celermajer 2008), including in the third
generation Framingham cohort (Hamburg, Keyes et al. 2008). RH-PAT has been
shown to be at least 50% dependent on endothelial NO activity (Nohria, Gerhard-
Herman et al. 2006). RH-PAT uses finger probes to measure digital pulse wave
amplitude detected by a pressure transducer, and has been validated against the
operator-dependent flow-mediated dilatation method (Kuvin, Patel et al. 2003;
Dhindsa, Sommerlad et al. 2008) and with endothelial function in other vascular
beds, including the coronary arteries (Bonetti, Pumper et al. 2004). Using RH-PAT,
we have demonstrated endothelial dysfunction in subjects with severe malaria (Yeo,
Lampah et al. 2007) but it has not previously been evaluated in subjects with sepsis.
Figure 6.1 Representative normal and abnormal peripheral arterial tonometry traces. The tracings represent the pulse wave amplitude from a fingertip over a15-minute period. The y axis is pulse wave amplitude in arbitrary units (derived from millivolts). The top trace was taken from a control subject whose reactive hyperaemia peripheral arterial tonometry; (RH-PAT) index was 1.98, and the bottom from a severe sepsis subject whose RH-PAT index was 1.16. The horizontal axis is time. The first shaded section is averaged as a baseline signal. The middle section is arterial occlusion, with consequent loss of the pulse wave signal. The final section is the pulse wave amplitude following release of the cuff. The random vertical spikes are movement artefacts. In the top trace there is reactive hyperaemia, with an increase in average pulse wave amplitude. The shaded post-occlusion section is compared with the shaded baseline section to give a ratio -- the RH-PAT index.
38
Vasodilatory shock in sepsis has been hypothesized to reflect a state of NO excess.
However, several recent isotope studies have shown no net increase in NO synthesis
in humans with sepsis (Villalpando, Gopal et al. 2006; Kao, Bandi et al. 2008;
Luiking, Poeze et al. 2009). To explain this, it has been proposed that sepsis may be
a state of imbalance between the NOS isoforms inducible NOS (iNOS) and
endothelial NOS (eNOS) in the microvasculature (McGown and Brookes 2007). This
could lead to a relative deficiency of endothelial NO, which is required to maintain
the microvascular endothelium in a healthy, quiescent state.
Another possible reason for endothelial NO deficiency is a decreased availability of
L-arginine, the substrate for NOS and the precursor for NO (Hecker, Sessa et al.
1990). Sepsis has been hypothesised to be an arginine deficient state (Luiking, Poeze
et al. 2004), although plasma L-arginine levels in humans with sepsis have been
variably reported to be high (Chiarla, Giovannini et al. 2000), normal (Askanazi,
Carpentier et al. 1980; Ochoa, Udekwu et al. 1991) or low (Sprung, Cerra et al. 1991;
Druml, Heinzel et al. 2001; Luiking, Poeze et al. 2009). Decreased plasma L-
arginine has been linked with decreased NO production in animal and in vitro models
(Hallemeesch, Lamers et al. 2002).
We hypothesised that RH-PAT would be a feasible technique to measure
microvascular reactivity in sepsis and that microvascular reactivity would be
impaired in subjects with sepsis in proportion to disease severity. Our secondary
hypotheses were that microvascular reactivity would correlate with plasma L-
arginine and measures of endothelial activation, and that plasma L-arginine would be
decreased in sepsis.
39
Materials and methods
Study design and setting
We performed a prospective observational cohort study in a 350-bed teaching
hospital in tropical northern Australia, with an 18-bed mixed intensive care unit
(ICU). Approval was obtained from Human Research Ethics Committee of the
Menzies School of Health Research and the Department of Health and Community
Services. Written informed consent was obtained from all participants or next of kin.
Participants
Between March 2006 and November 2007, all adult subjects (≥ 18 years) admitted to
the hospital were screened regarding eligibility for the study. Inclusion criteria for
sepsis subjects were: suspected or proven infection; presence of 2 or more criteria for
the systemic inflammatory response syndrome (SIRS) within the last 4 hours (Bone,
Balk et al. 1992); and admission to ICU within the preceding 24 hours or to the
wards within the preceding 36 hours. Exclusion criteria were coagulopathy (Platelets
≤ 20x109/L, APTT≥70 seconds, INR≥2.0); smoking of tobacco within the preceding
4 hours; and current administration of intravenous nitrates. Control subjects were
adults recruited from hospital subjects with no clinical or laboratory evidence of
inflammation or infection, and who had not met SIRS criteria within the last 30 days.
Severe sepsis was defined as sepsis with organ dysfunction or shock at the time of
enrolment according to American College of Chest Physicians/Society of Critical
Care Medicine consensus criteria (Bone, Balk et al. 1992; Stephens, Thomas et al.
2008).
40
Measurement of Microvascular Reactivity
Sepsis subjects underwent standardised demographic and clinical data collection,
bedside RH-PAT measurement (Endopat 2000, Itamar Medical), and blood
collection at days 0 and 2-4. All studies were performed after resuscitation and at
least an hour of hemodynamic stability (defined as no change in vasopressor dose or
need for fluid boluses) in a quiet room at 25°C, with the patient recumbent. Control
subjects had the same assessment at a single time point.
In this study, probes were placed on the index fingers of both hands, or on other
fingers if the index fingers were not suitable. Digital pulse wave amplitude was
recorded from both hands for a resting baseline period of 10 minutes and then a
blood pressure cuff was rapidly inflated on the study arm up to 200 mm Hg, or 50
mmHg above systolic blood pressure, whichever was greater. After 5 minutes +/- 10
seconds, the cuff was deflated. Pulse wave amplitude was then recorded for a further
5 minutes. An automated computerised algorithm provided by the manufacturer
(Endo-PAT 2000 software version 3.1.2) was used to calculate a post-pre occlusion
ratio (RH-PAT index), thus making the measurements user-independent. The
software also normalises the RH-PAT index to the control arm to correct for changes
in systemic vascular tone (Figure 6.1).
There was no systematic difference between RH-PAT indices generated by different
observers. We have previously examined the reproducibility of RH-PAT
measurements by repeating them after 0.5 – 0.75 h in 37 healthy adults (Yeo,
Lampah et al. 2007). Reproducibility was acceptable according to the method of
Bland and Altman (Bland and Altman 1986), and was comparable with previous
41
reproducibility results for RH-PAT (Bonetti, Barsness et al. 2003) and with those
obtained with the flow-mediated dilatation method (Jarvisalo, Jartti et al. 2006).
Laboratory assays
Blood was collected in lithium heparin tubes at each time point and the plasma was
frozen. Plasma arginine concentrations were determined using high-performance
liquid chromatography, with a method modified from van Wandelen and Cohen (van
Wandelen and Cohen 1997). To assess circulating measures of endothelial activation,
ICAM1 and E-selectin were measured by ELISA (R&D Systems). Plasma
Interleukin 6 (IL6) was measured by flow cytometry using a cytokine bead array (BD
Biosciences, CA, USA). Ex-vivo plasma arginase activity causes significant
degradation of L-arginine at room temperature (Nuttall, Patton et al. 1998), thus only
L-arginine levels derived from blood frozen within 30 minutes of collection were
included in the analysis.
Statistical methods
Predefined groups for analysis were sepsis without organ failure, severe sepsis and
controls. Continuous variables were compared using Student’s t-test/ANOVA or
Mann Whitney U test for parametric and non-parametric variables respectively.
Categorical variables were compared using Fisher’s exact test. Correlates with
baseline RH-PAT index were determined using Pearson’s (parametric) or
Spearman’s (non-parametric) coefficient for univariate analysis. For multivariate
analysis, linear regression with backward selection was used. To examine
longitudinal correlations, linear mixed-effects models were used. A 2-sided p-value
42
of <0.05 was considered significant. All analyses were performed using Stata version
10 (Stata Corp).
Results
Participants
Over the 19-month study period, 85 subjects with sepsis and 45 control subjects were
enrolled. Of the sepsis subjects, 54 had organ failure due to sepsis at baseline (severe
sepsis group) and 31 did not (sepsis without organ failure). The three groups were
well matched in terms of risk factors for endothelial dysfunction and other baseline
characteristics (Table 6.1). Of the 85 sepsis subjects, 92% had community-acquired
sepsis, with no preceding trauma or surgery, and pneumonia was the most common
focus of infection.
Baseline microvascular reactivity
Baseline microvascular reactivity was impaired in sepsis subjects compared with
controls (p<0.0001, Table 6.2). Mean RH-PAT index was lowest in the severe sepsis
group (1.57 [95% CI: 1.43-1.70]), intermediate in the sepsis without organ failure
group (1.85 [1.67-2.03]), and highest in the control group (2.05 [1.91-2.19]);
p<0.00001, Figure 6.2. Subjects with severe sepsis were more likely to have
endothelial dysfunction than control subjects (odds ratio [OR] 9.4 [95% CI 3.5-
25.0]). This relationship persisted after controlling for known associations with and
risk factors for endothelial dysfunction (diabetes, smoking, ischaemic heart disease,
chronic renal disease, hypercholesterolemia, hypertension, statin use and age).
(Adjusted OR 17.0 [95% CI 5.0-58.0]). Within the severe sepsis group, mean RH-
PAT index was not significantly different in the 27 subjects requiring vasopressors
43
(1.48 [1.30-1.66]) than in those not requiring vasopressors (1.64 [1.39-1.89]), p=NS.
In those receiving noradrenaline (n =25), there was no correlation between RH-PAT
index and noadrenaline dose (r=0.19, p=NS). There was also no relationship between
body temperature and RH-PAT index.
Table 6.1 Baseline characteristics of patients
Severe sepsis 54
Sepsis without organ failure 31
Control 45
p valuea
Ageb 52.4 (48.3-56.5) 50.8 (46.5-55.2) 47.2 (43.1-51.4) NS
Male n (%) 33 (61) 21 (68) 30 (67) NS
Diabetic n (%) 18 (33) 7 (23) 14 (31) NS
Smoker n (%) 28 (57) 12 (39) 18 (41) NS
IHD c n (%) 9 (17) 6 (19) 6 (13) NS
On statin n (%) 13 (24) 9 (29) 13 (29) NS
APACHE II d,e 19.0 (15-23) 7.5 (5-11) <0.0001
SOFA scored,f 6 (3-9) 1 (0-2) <0.0001
Focus of Infection – n (%)
Pleuropulmonary n (%) 26 (48) 16 (52)
Skin/Soft tissue n (%) 9 (17) 9 (29)
Intra-abdominal n (%) 6 (11) 1 (3)
Urinary n (%) 4 (7) 3 (10)
Other n (%) 9 (17) 2(6)
Causative Organism
None Cultured n (%) 25 (46) 20 (65)
Gram Positive n (%) 15 (28) 5 (16)
Gram Negative n (%) 14 (26) 6 (19)
Origin of Sepsis
Community-acquired n (%) 47 (87) 30 (97)
Nosocomial n (%) 7 (13) 1 (3)
a. For difference between all 3 groups by one way analysis of variance b. Mean (95% Confidence Interval) c. IHD=Ischaemic Heart disease d. Median (Interquartile range) e. APACHE II=Acute Physiology and Chronic Health Evaluation 2 score f. SOFA=Sequential Organ Failure Assessment score
44
Table 6.2 RH-PAT index and related variables Severe sepsis
54
Sepsis without organ failure 31
Control 45
p value pooled sepsis v control
p value severe sepsis vs SWOFa
RH-PAT indexb 1.57 (1.43-1.70) 1.85 (1.67-2.03) 2.05 (1.91-2.19) <0.00001 0.01
Plasma L-arginine (µmol/L)b 35.8 (30.2-41.4) 40.9 (33.5-48.3) 80.4 (72.3-88.6) <0.00001 NS
MAP (mmHg)b,c 77 (74-81)
89 (83-95) 83 (79-87) NS 0.0006
Receiving vasopressors n(%) 27 (50) 0
Noradrenaline dose (µg/kg/min) d, e 0.08 (0.03-0.42)
Receiving assisted ventilation n(%) 20 (37) 0
CVP (cm H20)b,f 12.2 (10.3-14.1)
Plasma ICAM-1 (ng/ml)g 811 (500-1502) 507 (368-673) 323 (252-397) <0.00001 0.003
Plasma E-selectin (ng/ml)g 329 (138-502) 90 (51-164) 38 (26-63) <0.00001 0.0003
Plasma Interleukin 6 (pg/ml)g 385 (124-996) 148 (46-315) 5 (2-8) <0.00001 0.009
White blood cell countb 16.7 (14.2-19.2) 15.5 (13.3-17.7) 8.4 (6.9-9.8) <0.00001 NS
C-reactive proteing 190 (131-255) 102 (84-234) 7 (3-24) <0.00001 NS
a. SWOF= Sepsis without organ failure b. mean (95% CI) c. MAP=Mean arterial pressure d. median (range) e. Of 27 patients receiving vasopressors, 25 were receiving Noradrenaline f. CVP=Central venous pressure g. median (interquartile range) h. Severe sepsis n=30 , sepsis without organ failure n=26 , control n=27
45
Figure 6.2 Baseline microvascular reactivity is impaired in sepsis, in proportion to disease severity. Solid circles represent mean values, with error bars representing 95% confidence intervals. P values indicate pairwise comparisons between groups.
RH-PAT was well tolerated by all subjects. In 18 of 227 measurements (8%), a result
was not obtainable. This occurred in 15/182 (8%) of measurements in sepsis
subjects, and 3/45 (7%) in controls and was due either to inability to obtain a
baseline pulse wave reading, or failure to completely occlude forearm blood flow due
to oedema.
Plasma markers of endothelial activation (ICAM-1 and E-selectin) were both
significantly raised in sepsis subjects compared with controls (Table 6.2), however
they did not correlate with RH-PAT index. Blood lactate levels were routinely
measured only in subjects with severe sepsis, in whom the baseline median lactate
was 1.6 mmol/L (Range 0.5-12.7; IQR 1.0-2.3). Among severe sepsis subjects,
lactate correlated inversely with RH-PAT index, but this was not statistically
significant (r= -0.28, p = 0.06).
46
Among all sepsis subjects, baseline RH-PAT index correlated with mean arterial
pressure (MAP) (r=0.55, p<0.0001) and serum albumin (r=0.27, p=0.03), and was
inversely related to APACHE II score (r=-0.36, p=0.002), C-reactive protein (r=-
0.30, p=0.02) and the cardiovascular component of the SOFA score (r=-0.29,
p=0.01), but not with total SOFA score. Independent predictors of baseline RH-PAT
index on multivariate analysis were APACHE II score (β=-0.014, p=0.03) and MAP
(β=0.012, p<0.0001).
Baseline plasma L-arginine
In the subjects whose blood samples were processed within 30 minutes of collection,
baseline mean plasma L-arginine concentration was significantly lower in sepsis
subjects (38.6 µmol/L [34.2-43.1] n=56) than in controls (80.3 µmol/L [72.5-88.1]
n=27), p<0.0001. There was no significant difference in L-arginine levels between
severe sepsis and sepsis without organ failure (Table 6.2). When all subjects
including controls were considered, baseline plasma L-arginine correlated with
baseline RH-PAT index (r=0.32, p=0.007), however, this association was no longer
significant when stratified by disease severity.
47
Figure 6.3 (a) Longitudinal change in microvascular reactivity in sepsis subjects, (b) Longitudinal change in plasma arginine in sepsis subjects. Solid circles represent mean values, with error bars representing 95% confidence intervals.
Longitudinal changes in RH-PAT and L-arginine
Longitudinal RH-PAT readings were only available in 70% of subjects. There was
no difference in disease severity, as measured by APACHE II score in those with
(median [IQR] 14 [8-23]) and without (15.5 [8.5-20.5], p=NS) longitudinal data. In
sepsis subjects, there was no statistically significant change in mean RH-PAT index
from baseline to day 2-4 (1.67 to 1.85, p=NS; Figure 6.3). The same was true in the
severe sepsis subgroup (1.57 to 1.76, p=NS). In contrast, mean plasma L-arginine
48
concentrations significantly increased from baseline to day 2-4 (38.2 to 49.9 µmol/L,
p=0.01). In a mixed-effects linear regression model, change in microvascular
reactivity over the first 2-4 days of treatment correlated significantly with increasing
MAP and decreasing C-reactive protein, but not with change in plasma L-arginine.
Subject outcomes
Low baseline RH-PAT index was significantly correlated with an increase in SOFA
score over the first 2-4 days (r=-0.37, p=0.02). In subjects whose SOFA score
worsened over the first 2-4 days, the median RH-PAT index was 1.54, compared
with 1.74 in those whose SOFA score improved or did not change (p=0.01). At both
hospital discharge and 28-day follow-up, 8 of 85 (9%) subjects with sepsis had died.
Among those with septic shock at baseline, 6 of 29 (21%) had died at 28-day follow-
up. The mean baseline RH-PAT index was 1.67 among survivors and 1.60 among
non-survivors (p=NS). The strongest baseline predictors of death on univariate
analysis were APACHE II score (p=0.008), SOFA score (p=0.002) and IL-6 level
(p=0.004).
Discussion
This is the largest published study to date assessing reactive hyperaemia in human
sepsis and the first to use peripheral arterial tonometry. We have found that
endothelium-dependent microvascular reactivity is impaired in sepsis, in proportion
to disease severity, even after controlling for known associations with endothelial
dysfunction, suggesting that sepsis itself is the explanation for the observed
impairment in microvascular reactivity, rather than traditional cardiovascular risk
factors. RH-PAT proved to be a practical and feasible method of measuring
49
microvascular reactivity at the bedside in critically ill septic subjects, with a low
proportion of technical failures, which were no more common in sepsis subjects than
in controls, and no relationship with noradrenaline dose.
The findings of this study are generally consistent with those of the previous small
studies of reactive hyperaemia in adult subjects with sepsis using other methods.
Plethysmographic measures of forearm blood flow in sepsis have found a post-pre
occlusion ratio of 1.6 (Astiz, DeGent et al. 1995) and forearm skin laser Doppler
studies have found a ratio of 1.4 (Young and Cameron 1995). These results are very
similar to our observed ratio of 1.57, suggesting that the finding of impaired reactive
hyperaemia in adults with sepsis is a true phenomenon, which is independent of the
method used to measure it.
Because RH-PAT is at least 50% NO-dependent (Nohria, Gerhard-Herman et al.
2006), impaired RH-PAT responses in sepsis suggest reduced endothelial NO
bioavailability. Our results are in accord with increasing data from radiolabelled
arginine flux studies suggesting that NO synthesis is decreased in sepsis
(Villalpando, Gopal et al. 2006; Kao, Bandi et al. 2008; Luiking, Poeze et al. 2009).
Impaired RH-PAT has been demonstrated to be reversible with L-arginine infusion
in falciparum malaria, providing direct evidence for NO-dependence in acute
inflammatory states (Yeo, Lampah et al. 2007). However, we cannot exclude
contributions by other mechanisms, including impaired production of prostacyclin
and endothelium-derived hyperpolarizing factor (Bellien, Thuillez et al. 2008;
Mitchell, Ali et al. 2008).
50
There was a significant correlation between plasma L-arginine and microvascular
reactivity when all subjects were considered together, but this was not significant
within groups. Furthermore, the improvement of plasma L-arginine over the first 2-4
days was not significantly correlated with change in microvascular reactivity. These
findings suggest that NO production and endothelial function in sepsis are influenced
by other factors in addition to circulating L-arginine. Such factors may include an
increase in competitive inhibitors of NOS, such as asymmetric dimethylarginine
(O'Dwyer, Dempsey et al. 2006); deficiency of NOS cofactors such as
tetrahydrobiopterin; NO quenching by microvascular reactive oxygen intermediates
(Xia, Roman et al. 1998); and the enhanced local expression and activity of
endothelial cell arginase (Argaman, Young et al. 2003).
The marked hypoargininaemia which we found in subjects with sepsis supports the
hypothesis that L-arginine is decreased in sepsis, independent of trauma (Luiking,
Poeze et al. 2004). This finding is strengthened by the fact that we only included
subjects within 24-36 hours of admission, with standardised sepsis criteria and with
over 90% having community-acquired sepsis.
Our study has several potential limitations. Baseline blood flow measurements were
not available, and it is possible that the apparent decrease in reactive hyperaemia in
sepsis is an artefact of marked baseline vasodilatation. This could limit the subjects’
ability to respond to ischaemia by increased blood flow, because they already have
near-maximal vasodilatation. This is unlikely to be the case because baseline forearm
blood flow in septic subjects has been found to be normal or decreased by multiple
investigators (Astiz, Tilly et al. 1991; Neviere, Mathieu et al. 1996; Kubli, Boegli et
51
al. 2003; Vaudo, Marchesi et al. 2007). Furthermore, skeletal muscle has the capacity
to increase blood flow by up to ten-fold (Hudlicka 1985), which greatly exceeds the
increase seen in both healthy and septic subjects in this and other studies.
Due to variations in sample processing time, we were unable to determine accurate
plasma arginine values for all subjects. Thus the reported arginine values may not be
representative of the groups as a whole. Limitations of the longitudinal data include
incomplete follow up. Of the subjects who had an initial measurement of RH-PAT
index, 70% had a repeat measurement 2-4 days later. Although those who were not
followed up had a similar baseline APACHE II score to those who were followed up,
this may not have been a representative population, as subjects who rapidly improved
and were discharged home did not have repeat measurements. Thus the observed
degree of recovery in microvascular reactivity is likely to be an underestimate.
Finally, the mortality rate in this cohort was low (hospital and 28 day mortality 9%
overall and 21% among those with septic shock). Although this is consistent with the
relatively low mortality rate in severe sepsis previously documented in our ICU
(Stephens, Thomas et al. 2008), it does mean that the study may have been
underpowered to detect associations of measured variables with mortality.
Conclusions
In summary, we have found that peripheral arterial tonometry is a feasible tool for
measuring microvascular reactivity in sepsis, and that it is impaired in sepsis in
proportion to disease severity, suggesting reduced endothelial function and decreased
endothelial NO bioavailability. Given the growing interest in HMG CoA reductase
52
inhibitors (Terblanche, Almog et al. 2006) and other potential adjunctive therapies
targeting the endothelium in sepsis (Aird 2007), better tools for monitoring the
response of the endothelium in clinical trials are needed. RH-PAT is an attractive
option for such studies, as other current methods are user-dependent and have limited
availability.
53
6.4. Generation of hypotheses
In the FRESH study (Davis, Yeo et al. 2009) we found that sepsis patients have
impaired NO-dependent microvascular reactivity proportional to the severity of
disease. We identified that arginine concentrations were profoundly low in sepsis,
however, they were not related to disease severity or microvascular reactivity. This
led to questions regarding the role of arginine bioavailability in sepsis. Furthermore,
in the amino acid chromatograms, we noted that sepsis patients had particularly low
tryptophan levels. The median tryptophan concentration in sepsis patients was 24µM
compared to 49µM in controls, however we were unable to measure kynurenine at
this stage. This led us to question the importance of tryptophan bioavailability in
sepsis. As both arginine and tryptophan bioavailability can influence immune
responsiveness, we also sought to investigate associations between these two amino
acids and T cell viability and function.
The overall aim of this thesis was to investigate the relationship between
inflammation, amino acid bioavailability and the pathology of sepsis. Our
preliminary data combined with the literature reviewed in chapters 2 to 5 led to the
formation of the following hypotheses:
1. As neutrophils constitutively express arginase, we hypothesised that sepsis
patients with increased circulating neutrophil counts would have increased
plasma arginase activity and decreased plasma arginine concentrations.
54
2. As ADMA competes with arginine for binding to NOS, we hypothesised that
a decreased plasma arginine/ADMA ratio would be associated with decreased
microvascular reactivity in sepsis patients.
3. As IDO inhibits NOS, we hypothesised that an increased plasma
kynurenine/tryptophan (KT) ratio would be associated with decreased
microvascular reactivity in sepsis patients.
4. As decreased extra-cellular concentrations of tryptophan and increased extra-
cellular concentrations of kynurenine can induce T cell apoptosis, we
hypothesised that the KT ratio would be related to circulating T cell numbers
in sepsis.
5. As low extra-cellular concentrations of either arginine or tryptophan can
impair T cell zeta-chain expression, we hypothesised that T cell zeta-chain
expression would be decreased in sepsis patients and related to plasma
arginine and tryptophan concentrations.
6. As sepsis patients have increased inflammation and myeloid derived
suppressor cells are induced by inflammation, we hypothesised that sepsis
patients would have increased circulating myeloid derived suppressor cells
which would impair T cell zeta-chain expression via amino acid metabolism.
55
6.5. Conclusion
This project developed as a result of earlier research by our group. The FRESH
study found that sepsis patients had decreased microvascular reactivity and decreased
plasma arginine but that there was not a significant association between the two. We
hypothesised that the bioavailability of arginine and tryptophan, which takes into
account competing amino acids and metabolites, would be more informative then
arginine or tryptophan concentrations alone. The next chapter will describe the
methods used to develop and validate assays to measure amino acids and their
metabolites in plasma.
56
7. Methods: Measuring amino acids in plasma
7.1. Introduction
As this project relied on accurate arginine and tryptophan measurements, a major
focus of my PhD was measuring amino acids in plasma. To investigate the question
of tryptophan metabolism, it was necessary to modify the existing general amino
acids assay so that it could measure an important tryptophan metabolite, kynurenine.
To investigate the role of arginine metabolism in sepsis and particularly the effect of
methylated arginines in sepsis, it was necessary to develop an entirely different
assay. This chapter considers the development and optimisation of the methods for
measuring amino acids. The chapter consists of a draft manuscript describing the
general amino acids assay, a published paper describing the ADMA assay and a
published paper describing the effect of delayed blood processing on amino acid
concentration.
7.2. High performance liquid chromatography (HPLC)
In this project, plasma amino acids were measured using high performance liquid
chromatography (HPLC), see Figure 7.1. The HPLC process, including extraction
and derivatisation, separates analytes of interest (such as amino acids) out of
complex matrices (such as plasma) according to biochemical properties. As it is the
concentration of free amino acids which can regulate endothelial and immune
function, we measured the concentration of free plasma amino acids, not those
making up proteins. Free amino acids were extracted by precipitating the proteins
from the plasma. The amino acids were labelled with a fluorescent tag to enhance
detection. The extracts were then loaded onto a HPLC column and gradually eluted
57
off one by one. The key to HPLC is to make sure that only one compound of interest
elutes at a time by carefully controlling the conditions of pH and hydrophobicity.
Figure 7.1 Photo of a high performance liquid chromatography (HPLC) unit
7.3. General amino acids assay
7.3.1. Introduction to the general amino acids assay
The aim of this assay was to measure a broad range of common amino acids from a
small amount of patient plasma. HPLC assays are constantly being upgraded and
optimised as technology improves. The stable extracts from the ethanol extraction
process meant that extracts could be re-run if necessary without more plasma being
used.
This project relies on results from the general amino acids assay run between 2006
and 2009. During this time, two different brands of reversed-phase columns were
used, each with a different HPLC method – although the method of extraction and
derivatisation remained the same. The original method was run on a Nova-Pak
column, could separate 28 amino acids and was used until March 2008.
The Nova-Pak method, although reliable, had a relatively long run time and
58
eventually had difficulty measuring tryptophan. The next method used two Shim-
pack columns, connected in series, and gave much sharper peaks, allowing the
separation of 48 amino acids. One of the additional amino acids that the Shim-pack
method could measure was kynurenine, an important amino acid for understanding
tryptophan metabolism.
The plasma samples for this study were first run in 2006 using the Nova-Pak method
and then extracts were re-run in 2008 using the Shim-pack method to quantify
additional amino acids including kynurenine. Cell cultures and cell supernatants
were analysed in 2008-2010 using either the Shim-pack method or the Gemini
method. The Gemini method is a very recent method that is currently undergoing
validation and is not described in this chapter.
7.3.2. Draft manuscript: Routine analysis of plasma amino acids
using HPLC and AccQ-Fluor™ derivatives: a comparison of 2
different HPLC methods
Authors: Christabelle J. Darcy, Nicholas M. Anstey, Yvette R. McNeil*
Affiliations: Global Health Division, Menzies School of Health Research and
Charles Darwin University, Darwin, NT, Australia.
59
Abstract
Accurate analysis of free amino acids in plasma depends on the extraction process,
derivatisation agent and HPLC method. This report describes an extraction and
derivatisation procedure that was analysed using two different HPLC methods over a
period of six years. The ethanol extraction required only 50µL of plasma and the
extract was stable for at least 2 years at -80°C. The non-endogenous internal
standard, norleucine, was used. AccQ-Fluor derivatisation gave stable adducts that
reacted with a broad range of amino acids. Here, we compare the two different
HPLC methods that were used to quantify the extracted and derivatised amino acids.
The first method used a Nova-Pak column and separated 28 amino acids. The
second method used a Shim-pack column and separated 48 amino acids. The two
different HPLC methods gave very similar concentrations of amino acids in the
pooled, quality control plasma. Both HPLC methods achieved recovery rates of 97 -
101% and inter-assay coefficients of variation of less than 6% for most of the
detected amino acids. This extraction and derivatisation procedure gives good
recovery and reproducibility with either HPLC method on a non-dedicated machine.
Introduction
The reference method for amino acid analysis in biological fluids is ion-exchange
chromatography using post-column ninhydrin derivatisation (Cynober 2004).
However, reversed-phase high performance liquid chromatography (RP-HPLC) is a
simple alternative to ion-exchange that does not require expensive, dedicated
equipment. Furthermore, RP-HPLC often requires less sample volume while
offering greater detection sensitivity (Cohen and Michaud 1993).
60
The move towards RP-HPLC analysis has led to the development of several
derivatisation reagents. The choice of reagent greatly affects recovery,
reproducibility, length of the chromatographic run and equipment required. The
most common pre-column derivatisation reagent used for RP-HPLC analysis of
amino acids in physiological fluids is o-phthaldialdehyde (OPA). Although good
separation can be achieved in a short run with OPA, this reagent does not react with
secondary amino acids (such as proline or hydroxyproline) (Diaz, Lliberia et al.
1996). Furthermore, OPA produces unstable derivatives with glycine, alanine, lysine
and ornithine (Reverter, Lundh et al. 1997). These highly unstable derivatives may
result in higher coefficients of variation and lower recovery unless expensive
automated equipment is used. Similarly, dimethylaminoaphthalensulphonyl chloride
(Dansyl-Cl) derivatives may be unstable and furthermore this reagent requires a long
derivatization time (Bosch, Alegria et al. 2006). Phenylisothiocyanate (PITC) also
needs a long derivatization time and the excess reagent must be removed in several
stages of drying under vacuum (Reverter, Lundh et al. 1997). 9-Fluorenylmethyl-
chloroformate (FMOC) may yield multiple derivatives and the reagent needs to be
extracted with pentane (Diaz, Lliberia et al. 1996). This extraction process may
result in poor recovery as some amino acids are also extracted by pentane (Reverter,
Lundh et al. 1997).
The use of 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AccQ-Fluor) as a
pre-column derivatization agent was first described by Cohen and Michaud (Cohen
and Michaud 1993). AccQ-Fluor reacts rapidly with both primary and secondary
amines, yielding stable and highly fluorescent compounds. The excess reagent
hydrolyses to 6-aminoquinoline (AMQ) within minutes. The derivatization process
61
is quick and simple with excellent recovery. Results using this reagent agree well
with the reference method (Bosch, Alegria et al. 2006). The disadvantage of using
AccQ-Fluor is that the chromatographic run time is relatively long.
Over a six year period, two different HPLC methods were used to analyse amino
acids in the plasma of sepsis patients and malaria patients participating in clinical
trials. An important consideration during method development was that the amino
acid profiles of severely ill patients may contain extraneous peaks. A further
consideration was that patient plasma was limited and we wanted to accurately
measure a broad range of amino acids using a small amount of plasma. The aim of
this method was to resolve most amino acids down to baseline with particular interest
in arginine, ornithine, phenylalanine and tryptophan. Here, we show that the choice
of column and HPLC method greatly influence that number of amino acids that can
be detected from an AccQ-Fluor derivative.
Material and methods
Chemicals
All reagents were analytical or HPLC grade. Ethanol was purchased from Merck
(Darmstadt, Germany), acetic acid from BDH, acetonitrile from Burdick and Jackson
(Muskego, MI, USA) and sodium acetate trihydrate from Riedel-de Haën (Germany).
High purity grade amino acids were from Sigma-Aldrich (St. Louis, MO, USA) and
Calbiochem (La Jolla, CA, USA). The derivatizing agent AccQ-Fluor was
purchased from Waters (Milford, MA, USA) in kit form. The kit contained borate
buffer, AccQ-Fluor reagent powder and acetonitrile as reagent diluent. All aqueous
62
solutions were prepared with purified water (>18 MΩ/cm, Milli-Q plus, QPAK®1,
Millipore, Billerica, MA, USA).
Standards
Calibration concentrates, containing 32 - 48 amino acids (0.25 mM- 2.5 mM final
concentration dissolved in 0.1 M HCl) were prepared in the laboratory, aliquoted and
stored at –80°C. Tryptophan, glutamine, asparagine and ethanolamine were prepared
freshly in MQ and added to the amino acid stock solution when preparing calibrators.
Amino acid concentrates were prepared at relative physiological concentrations.
Calibrators were mixed with internal standard and ethanol-extracted prior to
derivatisation. A five level calibration was run on a monthly basis. Calibrators were
either prepared fresh, as described above, or derivatised from frozen ethanol extracts.
Quantitation was based on the peak areas and amount ratios of the amino acids to
those of the internal standard, norleucine.
Plasma
Heparinised plasma, from healthy donors, was pooled, aliquoted and stored at –80°C
as quality control plasma. Aliquots of pooled plasma were sent to three independent,
NATA-certified laboratories to measure amino acid concentrations. An aliquot of
quality control plasma was extracted, derivatised and run for every extraction batch
of samples. Immediately prior to analysis plasma was thawed, centrifuged (1 min at
9472 x g), extracted and derivatised as per procedures below.
63
Extraction and derivatisation
50µL of standard or plasma was added to 50µL of internal standard (0.2 mM
norleucine in 0.4 M aqueous hydrogen chloride), mixed, 200 µL cold (-20C) ethanol
was added and vortexed. The mixture was centrifuged at 9472 x g for 3 mins, the
supernatant removed and derivatized. AccQ-Fluor reagent, prepared to kit
instructions, was used to derivatise samples and calibrators. 65µL of Waters borate
buffer was added to 250µL vial inserts, 15µL of extracted sample or calibrator was
added to the buffer, the insert vortexed and then 20µL AccQ-Fluor reagent added
with immediate mixing. Typically 6 µL of derivatised mixture was injected onto the
columns. Derivatised samples were stable for at least 10 days at room temperature.
Equipment and eluents
A Shimadzu HPLC (Class VP series, Shimadzu Corporation, Kyoto, Japan) was
used for amino acid determinations. The HPLC unit consisted of an auto sampler,
quaternary pumping system, heated column oven and degasser. Column effluent was
monitored simultaneously using UV (250nm) and fluorescence detectors (ex 250 nm,
em 395nm) connected in series. The analytical columns were either a Nova-pack C18
column (3.9 mm x 300 mm, 4 µm) or two Shimadzu Shim-Pak C18 analytical
columns (4.6 x100 mm, 2 µm) were connected in series. A C18 4.0 x 3.0 mm
Security Guard cartridge (Phenomenex, Torrance, CA) was used to protect both the
Nova-Pak and Shim-Pak columns.
For the Nova-Pak column, flow rate was 1.1 mL/min at 33 °C, protected by a C18
4.0 x 3.0 mm Security Guard cartridge (Phenomenex, Torrance, CA). Two sodium
acetate buffer concentrates (x10) were prepared by dissolving 191.04g sodium
64
acetate in 1L of MQ, mixing and taking to pH 5.6 and 5.0 with concentrated
phosphoric acid. 10mL of Na EDTA and 4.2mL of TEA were added, mixed and the
final pH titrated to 5.5 and 4.9 respectively. Buffer concentrates were stored at 4°C.
Upon use buffers were diluted to 1:10 with MQ and a final concentration of 2%
methanol added to the pH 5.5 buffer and 3% methanol to the pH 4.9 buffer. Buffers
were filtered through 0.45 µm filters and sonicated under vacuum. An example
gradient regime for the Nova-Pak column is set out in Table 7.1. Note that the
gradient regime was usually slightly modified for each new column to ensure
consistent separation and accurate quality control concentrations.
Table 7.1 Column gradient regime for the Nova-Pak column method Time %A %B %C %D
(min) Buffer pH 5.5 100% Acetonitrile MQ Buffer pH 4.9
0 98 2 64 97 3 64.01 0 3 97 80 3 97 107 9 91 109 11 89 117 12 88 127 13 87 132 0 156 86 14 159 86 14 159.01 0 60 40 174 60 40 174.01 98 2 187 98 2
For the shim-pack column, the flow rate was 0.65 mL/min, 37°C, protected by a C18
4.0 x 3.0 mm Security Guard cartridge (Phenomenex, Torrance, CA). For the shim-
pack column method, two sodium acetate buffer concentrates (1M) were titrated to
pH 4.9 and 5.8 with glacial acetic acid and stored at 4°C. The pH 4.9 buffer had
65
1.36% absolute ethanol (wt/v) and 0.025% 1mg/mL EDTA (v/v) while the pH 5.8
buffer had 2.88% absolute ethanol (wt/v) and 0.05% 1mg/mL EDTA (v/v). Buffer
concentrates were diluted 1:20 with water and filtered through 0.45 µm membrane
(Millipore Inc., Milford, MA) prior to use. Table 7.2 shows an example gradient
regime for the Shim-pack method.
Table 7.2 Column gradient regime for the Shim-pack column method Time %A %B %C %D
(min) Buffer pH 5.8 100% Acetonitrile MQ Buffer pH 4.9
0 97 3 0 10.00 97 gradient 0 12.00 96.5 3.5 0 45.00 96 4 0 50.00 96 4 0 50.01 26 gradient 70 66 24.5 5.5 70 68 22 8 70 70 62 gradient 30 80 60.5 9.5 30 94 58.5 11.5 30 101 54 16 30 106.99 54 16 30 107.00 54 16 30 112 54 16 30 114 52 18 30 118 52 18 30 122 12 18 70 142 12 18 0 70 142.01 0 60 40 0 158 0 60 40 0 158.01 97 3 0 0 172 97 3 0 0
Results and Discussion
Extraction and derivatisation
The extraction and derivatisation produced an extract in which a broad range of
amino acids could be detected from only 50 µL of plasma. This is a distinct
66
advantage in a clinical setting where plasma is limiting and may need to be used for
many other biochemical assays. The ethanol extracts were very stable at -80 °C,
allowing extracts to be thawed and re-derivatised at a later date, as necessary. The
AccQ-Fluor derivatisation produced a stable adduct that lasted about 10 days,
allowing large batches to be processed at once.
Initially plasma samples were extracted using sulphosalicylic acid (SSA) as
described in other studies using AccQ-Fluor derivatisation (Teerlink, van Leeuwen et
al. 1994; Reverter, Lundh et al. 1997; Badiou, Lehmann et al. 2004). However SSA
was found to elute on the column as a peak at approximately 2 minutes. The method
wash conditions were inadequate for the complete removal of SSA and peak height
and width increased with each injection until the peak width interfered with the
resolution of early eluting amino acids. The SSA peak could be removed gradually
with extensive column washing. As a result this obviated the use of SSA for routine
analysis. This problem was eliminated with the use of ethanol to precipitate plasma
protein.
Once extracted with ethanol, samples and calibrators were found to be stable when
frozen at –80 °C, unlike acid-extracted amino acids. Both ethanol-extracted samples
and calibrators could be frozen and derivatized much later. Small volumes of plasma
could be analysed for amino acids using ethanol extraction. Accurate analyses could
be performed with volumes as low as 10 µL of plasma.
AccQ-Fluor reacts rapidly with both primary and secondary amino acids; ammonia
and AMQ elute as the two most prominent, clearly separated, by-products of the
67
reaction. Smaller peaks, attributable to the AccQ-Fluor derivitisation reaction, appear
near threonine, cystine, ornithine and lysine. These additional derivatizing reagent
peaks result from the exposure of the reagent to atmospheric water. AccQ-Fluor
reagent was stored in a dessicator when not in use, however short interval of
exposure to the atmosphere, when derivatising samples resulted in more AMQ
formation. The peak heights of the AccQ-Fluor reagent-associated peaks changed as
the reagent aged after solubilisation. As reported (Cohen and Michaud 1993;
Strydom and Cohen 1994), the AccQ-Fluor derivatives were extremely stable with
time at room temperature and control plasma amino acid concentrations were within
2 standard deviations of mean values for more than 10 days after derivatisation.
HPLC amino acid separation
Figure 7.2 illustrates a typical chromatogram of quality control plasma detected by
fluorescence using the Nova-Pak method while Figure 7.3 shows a chromatogram of
the same quality control plasma using the Shim-pack method. In both methods most
of the amino acids were baseline separated. The Shim-pack method had a better
separation capacity with taller, sharper peaks compared to the Nova-Pak method. In
addition, the Shim-pack method separated more amino acids in a shorter amount of
time than the Nova-Pak method. The majority of amino acids could be quantified by
both UV and fluorescence detection. Although tryptophan and kynurenine were
below the quantifiable limit in the fluorescence chromatogram, both had
considerably stronger responses in the UV and could be quantified using UV
detection. The UV trace was also used as an integration guide to confirm values
obtained for select amino acids integrated in the fluorescence trace.
68
Figure 7.2 Chromatogram of quality control plasma using the Nova-Pak method.
69
Figure 7.3 Chromatogram of quality control plasma using the Shim-pack method.
70
Method validation
The Nova-Pak column gave good accuracy and recoveries however the Shim-pack
column separated more amino acids in a shorter amount of time without
compromising accuracy or recovery.
71
Table 7.3 compares amino acid concentrations, inter-assay CVs and recoveries for
the quality control plasma using the Nova-Pak and Shim-pack method. The mean
inter-assay were 2.7% for the Nova-Pak method and 3.5 % for the Shim-pack
method. Amino acid recovery, as determined from spiked samples, was 97-101 % for
the majority of amino acids in both methods.
Amino acids present in lower concentrations in the quality control plasma tended to
have higher inter-assay CV. The higher CVs in aspartate and glutamate were also
due to deamidation. Although plasma free amino acids were not acid-extracted
glutamine and asparagine deamidation was observed in a few extractions, which
elevated CV values for aspartate and glutamate. Deamidation may have resulted
from elevated temperature during centrifugation, as a result of numerous plasma
freeze-thaw cycles or both.
72
Table 7.3 Comparison of Nova-Pak and Shim-pack method. Summary of mean amino acid concentrations and co-efficients of variation and inter-assay relative standard deviations (RSD) for the Nova-Pak method (n=161) and Shim-pack method (n=50) of quality control plasma. Recovery of amino acids from spiked quality control plasma was calculated from 10 x cal 3 spikes for the Nova-pak method and 5 x cal 1-5 spikes for the Shim-pack method. NM = not measureable. QC values Inter-assay RSD Recovery
Nova-Pak
Shim-pack
Nova-Pak
Shim-pack
Nova-Pak
Shim-pack
hydroxyproline 10.7 10.4 2.7 5.7 99.1 98.0 aspartate 3.2 3.2 3.4 11.5 96.2 99.4 asparagine 50.3 48.8 3.0 4.0 99.0 100.8 serine 108.6 109.1 2.6 3.3 99.5 98.2 glutamate 36.1 41.9 2.8 32.5 100.4 83.9 glutamine 556.6 554.1 3.0 5.1 99.2 98.5 glycine 238.7 232.3 2.5 3.0 99.4 96.8 histidine 79.3 80.0 2.6 2.7 99.2 97.6 1-MHis* NM 9.7 NM 6.7 NM 95.6 taurine 53.9 54.4 3.1 3.5 98.5 98.1 citrulline 30.6 29.6 1.1 3.6 99.8 101.9 threonine 118.1 113.7 2.9 100.5 96.6 arginine 76.6 76.4 2.8 3.1 101.5 98.6 alanine 381.9 358.5 2.4 2.7 97.9 98.2 ethanolamine 7.5 6.8 5.2 4.8 100.8 104.9 proline 208.1 200.7 1.6 2.9 97.0 95.5 BAIB* 2.0 1.9 31.6 13.3 112.7 96.0 aab* 17.6 16.8 2.2 4.0 99.2 89.5 cystine NM 31.3 NM 22.5 NM tyrosine 62.7 60.7 2.3 6.0 99.1 97.9 valine 215.4 204.4 2.0 2.4 98.5 97.7 methionine 25.5 24.2 2.9 2.3 98.6 98.5 ornithine 62.5 61.6 4.4 3.4 99.4 99.8 lysine 193.1 188.5 2.2 2.9 102.6 100.8 isoleucine 71.7 67.5 2.7 2.9 99.3 97.4 leucine 127.5 126.2 2.2 3.5 98.8 95.7 kynurenine NM 1.5 NM 5.6 NM 96.4 phenylalanine 63.5 60.9 2.4 2.9 99.5 97.0 tryptophan 60.7 59.3 2.8 2.0 99.4 97.6 Average 2.7 3.5 99.8 97.4
*1-MHis = 1 methylhistidine, BAIB = β aminoisobuyrate, aab = α aminoisobutyrate
73
Table 7.4 Comparison of amino acid concentrations measured in the quality control plasma by three independent, NATA-certified laboratories with the two MSHR methods. Source of data PMH WCH Lab 3 MSHR MSHR Method of analysis IE IE RP NovaPak Shimpack Precipitation SSA SSA acetonitrile ethanol ethanol Samples analysed 2 1 2 161 50 1-Methyl Histidine 12.6 16.1 NM NM 9.7 3-Methyl Histidine 3.3 5.5 NM NM NM AAB 17.4 21.5 NM 17.6 16.8 Alanine 366.5 430.5 305 381.9 358.5 Amino adipic acid <1 6.2 NM BQL BQL Arginine 77.8 93.9 61.5 76.6 76.4 Asparagine 53.9 NM 43.5 50.3 48.8 Aspartate 4.4 NM 4 3.2 3.2 BAIB <1 NM NM 2 1.9 Carnosine <1 27.6 NM NM NM Citrulline 32.8 39.7 NM 30.6 29.6 Cystine 20.5 16.5 NM NM 31.3 Ethanolamine NM NM NM 7.5 6.8 Glutamate 28.3 56.2 29 36.1 41.9 Glutamine 580.8 430.5 504 556.6 554.1 Glycine 238.1 244.8 205.5 238.7 232.3 Histidine 83 90.8 53 79.3 80 Homocystine <1 NM NM BDL BDL Hydroxyproline 9.9 NM NM 10.7 10.4 Isoleucine 66.7 78.8 58.5 71.7 67.5 Leucine 123 139.6 118.5 127.5 126.2 Lysine 176.1 205.7 134 193.1 188.5 Methionine 26.7 27.9 28.5 25.5 24.2 Ornithine 65.6 76.9 43.5 62.5 61.6 Phenylalanine 57.6 67.3 54.5 63.5 60.9 Phosphoserine 4.4 NM NM BDL BDL Proline 214.1 195.7 228 208.1 200.7 Serine 106.1 NM 87 108.6 109.1 Taurine 53.3 73.8 48 53.9 54.4 Threonine 119.4 87.1 76 118.1 113.7 Tryptophan NM NM 61 60.7 59.3 Tyrosine 63.1 69.4 117 62.7 60.7 Valine 221.3 244.2 165.5 215.4 204.4
PMH=Princess Margaret Hospital, WA; WCH= Metabolic Laboratory, Dept. of Genetic Medicine, Women and Children’s Hospital, S.A.; Laboratory 3 (requested not be identified), Vic; MSHR= Menzies School of Health Research, NT. Results for MSHR controls (n=161). NM = not measured, BDL = below detectable limit, BQL = below quantifiable limit. PMH, WCH and Lab 3 all participants in the quality control program, ERNDIM (European Research Network for evaluation and improvement of screening, Diagnosis and treatment of Inherited disorders of Metabolism). Control samples were analysed by PMH & WCH using cation exchange gradient chromatography, SSA extraction and post-column detection with ninhydrin. Lab 3 analysed samples using reversed phase gradient chromatography, acetonitrile protein precipitation, derivatisation with PITC and UV detection at 540 nm.
74
External method comparisons
The pooled quality control plasma was analysed for free amino acid concentrations at
three independent commercial laboratories (Table 7.4); extraction, derivatization and
chromatographic methods varied between the laboratories. Two laboratories used
ion-exchange chromatography to separate plasma amino acids, the 3rd laboratory
used RP-HPLC and PITC to separate and detect amino acids. Concentration results
obtained by RP-HPLC using AccQ-Fluor compared favourably with those from other
laboratories, as did the number of amino acids measured. Lab 3, using RP-HPLC and
PITC derivatisation, trended towards reporting lower amino acid concentrations for
the plasma samples but was the only laboratory to report tryptophan concentration.
Tryptophan concentrations were similar for both methods of derivatisation.
Conclusion
In summary this reversed phase HPLC method using AccQ-Fluor derivatised amino
acids provides clear resolution for most of the amino acids analysed. The precision
and recovery was comparable to that of ion-exchange chromatography. The method
requires small plasma volumes, minimal processing, the plasma and standard extracts
are stable for months when stored –80 °C and the derivatised extracts are stable.
75
7.4. ADMA assay
7.4.1. Introduction to the ADMA assay
As described in chapter 2, ADMA is a methylated arginine which competes with
arginine for binding to eNOS. The plasma arginine/ADMA ratio is used to estimate
the bioavailability of arginine to eNOS. Because methylated arginines are present in
such small quantities in plasma, we were unable to accurately measure them with the
general amino acids assay. To measure the methylated arginines, it was necessary to
use a different extraction method which removed many other amino acids and
concentrated the methylated arginines. One of the challenges in developing this
assay was accurately measuring the methylated arginines from a small volumne of
plasma. The following paper describing the ADMA assay was published in the
Journal of Chromatography B in 2010.
7.4.2. Published paper: HPLC analysis of asymmetric
dimethylarginine, symmetric dimethylarginine, homoarginine and
arginine in small plasma volumes using a Gemini-NX column at high
pH
Catherine E. Jonesa#, Christabelle J. Darcya#, Tonia Woodberrya, Nicholas M.
Ansteya, Yvette R. McNeila
# These authors contributed equally
Authors’ affiliations: aMenzies School of Health Research, Rocklands Drive, Tiwi,
Darwin, N.T., Australia
76
Abstract
There is increasing recognition of the clinical importance of endogenous nitric oxide
synthase inhibitors in critical illness. This has highlighted the need for an accurate
high performance liquid chromatography (HPLC) method for detection of
asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA) in
small volumes of blood. Here, the validation of an accurate, precise HPLC method
for the determination of ADMA, SDMA, homoarginine and arginine concentrations
in plasma is described. Solid phase extraction is followed by derivatisation with
AccQ-Fluor™ and reversed phase separation on a Gemini-NX column at pH 9.
Simultaneous detection by both UV-visible and fluorescence detectors affords extra
validation. This solid phase extraction method gives absolute recoveries of more than
85% for ADMA and SDMA and relative recoveries of 102% for ADMA and 101%
for SDMA. The intra-assay relative standard deviations are 2.1% and 2.3% for
ADMA and SDMA, respectively, with inter-assay relative standard deviations of
2.7% and 3.1% respectively. Advantages of this method include improved recovery
of all analytes using propan-2-ol in the solid phase extraction; sharp, well-resolved
chromatographic peaks using a high pH mobile phase; a non-endogenous internal
standard, n- propyl L-arginine; and accurate and precise determination of methylated
arginine concentrations from only 100µL of plasma.
77
Introduction
The clinical importance of endogenous nitric oxide synthase (NOS) inhibitors has
long been recognised in chronic disease (Vallance, Leone et al. 1992). Nitric oxide
(NO) is important in the maintenance of normal endothelial function (Vallance,
Collier et al. 1989) and the prevention of platelet aggregation (Mellion, Ignarro et al.
1981). NO synthesis from L-arginine is reduced in the presence of asymmetric
dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA), which are
products of methylated protein degradation.
ADMA and homoarginine compete with arginine for specific binding sites on NOS.
Homoarginine is an alternative but less efficient substrate for NOS (Moali, Boucher
et al. 1998) whereas ADMA directly inhibits nitric oxide synthases. ADMA, SDMA
and homoarginine each compete with arginine for transport into the cell (McDonald,
Zharikov et al. 1997) and may, therefore, also limit the amount of arginine available
to NOS (Closs, Basha et al. 1997; Kakoki, Kim et al. 2006). High concentrations of
methylated arginines have been associated with a broad range of chronic diseases,
including hypertension (Goonasekera, Rees et al. 1997), renal failure (Vallance,
Leone et al. 1992), hypercholesterolemia (Boger, Tsikas et al. 2004) and diabetes
(Fard, Tuck et al. 2000). Indeed, elevated ADMA is an independent risk factor for
both cardiovascular disease (Schulze, Lenzen et al. 2006) and all-cause mortality
(Boger, Sullivan et al. 2009).
In addition to the importance of ADMA in chronic disease, there is increasing
recognition of its important role in acute critical illness (Nijveldt, Teerlink et al.
2003; Nijveldt, Teerlink et al. 2003) and acute inflammatory conditions such septic
78
shock (O'Dwyer, Dempsey et al. 2006). As limited blood is available from critically
ill patients, there is a need for an accurate high performance liquid chromatography
(HPLC) method for detection of ADMA and SDMA in small volumes of blood.
This paper describes a reversed phase HPLC method for the measurement of
arginine, ADMA, SDMA and homoarginine from 100 µL of plasma. The
chromatography utilised a Gemini-NX column with a novel, high pH borate buffer-
acetonitrile gradient, and the non-endogenous internal standard n- propyl L-arginine
(NPLA). Sample preparation utilised solid phase extraction (SPE) and fluorescent
derivatisation. The extraction procedure and HPLC method give accurate and
precise results from a small volume of plasma.
Experimental
Materials
L-arginine HCl, L-homo-arginine-HCl, NG,NG di-methyl-L-arginine and NG,NG’ di-
methyl-L-arginine were purchased from Calbiochem (La Jolla, CA, USA). N-propyl-
L-arginine was a product of Cayman Chemicals (Ann Arbor, MI, USA). Sodium
tetra borate decahydrate and boric acid were obtained from Sigma-Aldrich (St. Louis,
MO, USA). Oasis Mixed Mode Cation Exchange (MCX) cartridges (1mL, 30cc)
were purchased from Waters (Milford, MA USA). Isopropanol and ammonia
solution 28-30% were purchased from Merck (Darmstadt, Germany). HPLC-grade
acetonitrile was obtained from Burdick and Jackson (Muskego, MI, USA). High
purity water was used to prepare all aqueous solutions (Milli-Q water system, Milli-
Pore, Billerica, MA, USA). The AccQ-Fluor™ kit from Waters (Milford, MA,
USA) contained the fluorescent reagent 6-aminoquinolyl-N-hydroxysuccinimidyl, a
79
vial of acetonitrile diluent, and a vial of aqueous borate buffer (0.2 M, pH 8.8) for the
derivatisation reaction.
Plasma samples
Venous blood from healthy volunteers or patients was collected into lithium heparin
tubes, centrifuged (492 x g for 8 min) within 120 minutes of collection and the
plasma was frozen at –80 °C until analysis. A pool of plasma from Australian Red
Cross blood donors was used as quality control plasma.
Plasma from 30 apparently healthy volunteers was used to determine healthy
concentrations of ADMA and SDMA. 8 of these volunteers were laboratory staff
(blood collected as above) and 22 were blood bank donors (blood collected
according to standard Australian Red Cross blood bank procedures). Blood from
blood bank donors was usually separated the day after collection. The age range of
the healthy volunteers was 16-61; 18 were female and 12 were male. The use of this
plasma was approved by the Ethics Committees of the Australian Red Cross and the
Menzies School of Health Research.
Extraction
Oasis MCX cartridges were affixed to a vacuum manifold and pre-equilibrated with
1 mL of isopropanol, followed by 1 mL of 50 mM borate buffer (pH 9). 100 µL of
plasma or calibrator was mixed with 100 µL 15 µM NPLA and diluted with 800 µL
50 mM borate buffer (pH 9) and then loaded onto the cartridge. Cartridges were then
washed with 1 mL of water and then 1 mL of isopropanol. Extracts were eluted from
the cartridges into glass collection tubes with 1 mL of eluting solvent
80
(isopropanol:water:28-30% ammonia solution (5:4:1)). Flow rates were controlled by
vacuum adjustment. The vacuum manifold pressure was less than 254 mm Hg for
the pre-equilibration and wash steps, and less than 127 mm Hg for the loading and
eluting steps. Extracts were dried under nitrogen at 75°C (for approximately 1 hour).
Dried eluates were reconstituted in 0.2 mL water and transferred to glass storage
vials.
Derivatisation
Extracts were derivatised with Waters AccQ-Fluor™ kit prior to chromatography. In
a 250 µL HPLC vial insert; 20 µL of extract, diluted with 70 µL of Waters’ borate
buffer, was reacted with 10 µL AccQ-Fluor™ reagent by immediate vortexing for 10
seconds.
Chromatography
The Shimadzu VP series HPLC system consisted of a gradient pump, degasser,
column oven (42 oC) and UV and fluorescence detectors. The detectors were
connected in series for simultaneous detection of UV (absorption wavelength = 250
nm) and fluorescence (excitation wavelength = 250 nm, emission wavelength = 395
nm). Extracts were separated on a C18 Gemini-NX analytical column (150 x 4.6
mm, 3 µm) protected by a C18 Gemini NX security guard cartridge (4.0 x 3.0 mm),
both from Phenomenex (Lane Cove, NSW, Australia). Mobile phase flow rate was 1
mL min-1.
A 100 mM stock solution of sodium tetra borate/ boric acid was prepared and filtered
(0.2 µm) into a sterile container. The stock was kept at room temperature. Eluent A
81
was a 1:5 dilution of the borate buffer stock solution. The mobile phase delivery
program of 20 mM borate buffer pH 9 (A), acetonitrile (B) and water (C) is shown in
Table 7.5. All eluents were filtered through 0.45 µm filters before use.
Table 7.5 Mobile phase delivery program
Time (minutes)
0.00-18.00
18.01-21.00
21.01-29.00
29.01-40.00
40.01-52.00
Eluenta
A:B
A:B
A:B
A:B
B:C
Value (%)
93:7
93:7>>92:8
92:8
87:13
65:35
Event
Isocratic
Gradient 7-8% over 3 minutes
Isocratic
Isocratic
Wash a Eluents: 20 mM borate buffer pH 9 (A), acetonitrile (B) and water (C)
Calibration and validation
Stock solutions of arginine (2.5 mM), homoarginine (500 µM), ADMA (100 µM),
SDMA (100 µM) and NPLA (2.5 mM) were prepared, aliquoted and stored at -80
°C. Seven calibration standards were made to encompass physiological and disease-
associated concentration ranges. Arginine covered the range of 7.5-200 µM,
homoarginine 0.5-12 µM, ADMA 0.25-6 µM and SDMA 0.25-6 µM. The calibration
standards were extracted and derivatised in the same manner as plasma samples.
Identification of analytes within plasma samples was based on the retention time of
the corresponding standard. A seven level calibration curve for each analyte, using
peak area/amount ratios of the analytes to internal standard was constructed from
integrated chromatograms.
82
Analyte recovery during the extraction process was determined by calculating the
relative recovery and absolute concentrations recovered after calibration standards
were subjected to SPE compared with un-extracted calibrator concentrations. Seven
standards were run without undergoing SPE in parallel with aliquots of the same
standards subjected to SPE. Absolute recovery was calculated by comparing the area
of the extracted peaks to the area of the un-extracted peaks. This ensured that no
particular analyte was preferentially lost through extraction. Relative recovery was
calculated by plotting the extracted calibrators onto the curve of the un-extracted
calibrators. The percent recovery was calculated by dividing the measured
concentration by the theoretical concentration from the un-extracted curve.
The HPLC method was validated by calculating the intra-assay and inter-assay
precision of pooled quality control plasma and by determining the spike recovery of
analyte added to control plasma. The intra-assay precision of the HPLC method was
determined by running a single extract of control plasma 10 times consecutively and
calculating the concentration of the analytes of interest. Inter-assay precision was
calculated by extracting and running 30 separate control plasmas over 2 months. In
order to determine the accuracy of the HPLC method, the pooled quality control
plasma was spiked with known concentrations of arginine, homoarginine, ADMA
and SDMA. The percent spike recovery was expressed as the recovery of added
analyte from spiked plasma samples. This process was repeated 3 times in 6 months.
Limit of detection (LOD) was determined by a signal to noise ratio of 2:1 and the
limit of quantification (LOQ) was determined by a signal to noise ratio of 10:1.
83
Results and discussion
Chromatography
Homoarginine, ADMA and SDMA were detected simultaneously using UV and
fluorescence detection. Arginine was out of range of fluorescence detection once
above 30 µM and was therefore primarily detected by UV. There was less than 5%
deviation between ADMA and SDMA values measured by either fluorescence or
UV. Validation data presented in this paper was from the fluorescent detection of
ADMA, SDMA and homoarginine and the UV detection of arginine.
This method provided excellent separation of arginine, homoarginine, ADMA,
SDMA and NPLA. Figure 7.4 shows the separation of analytes in a standard, the
pooled quality control plasma and plasma from a malaria patient. Blank samples of
water also underwent the extraction and derivitisation processes and were
chromatographed to ensure there were no co-eluting peaks originating from the SPE
method or the derivatising agent. The pooled quality control plasma and plasma from
2 patients with bacterial sepsis and 2 patients with falciparum malaria were subjected
to SPE without the addition of internal standard, to ensure there was a flat baseline
under NPLA (see Figure 7.4B).
The coefficient of determination (r2) for each analyte was >0.999. Limit of detection
was 0.04 µM for arginine, 0.06 µM for homoarginine, 0.04 µM for ADMA and 0.03
µM for SDMA. The limit of quantification was 0.20 µM for arginine, 0.30 µM for
homoarginine, 0.20 µM for ADMA and 0.15 µM for SDMA.
84
Borate was chosen as the mobile phase buffer in this method as it is also the matrix
of the derivatised samples and greatest retention time reproducibility is obtained
when samples are dissolved in a similar solution to the mobile phase. The borate
buffer was prepared to pH 9 as the pKa of borate buffer is 9.2 and buffers are most
effective within 0.5 pH units of their pKa. The combination of high pH and
acetonitrile resulted in sharp, well resolved chromatographic peaks. The Gemini-NX
column was selected for this method as it has a large pH stability range of 1-12.
85
Figure 7.4 Chromatograms from dimethylarginine assay. Fluorescence detection of a calibration standard (A) with 30 µM arginine, 2 µM homoarginine, 1 µM ADMA and 1 µM SDMA; and (B) the pooled quality control plasma (black) with 23.68 µM arginine, 1.82 µM homoarginine, 0.48 µM ADMA and 0.39 µM SDMA, overlaid with a chromatogram from a patient with falciparum malaria (red) without internal standard added. Peak identity: (1) arginine; (2) homoarginine, (3) ADMA, (4) SDMA, (5) NPLA. Inset B: region 27-31 min magnified 40x.
86
Extraction and derivatisation
A number of different extraction solvents and procedures were trialled, including the
procedures recommended in the Oasis MCX cartridge literature. Most published
methods use methanol in the final eluting solution and/or during the pre-equilibration
and wash stages. However, optimal recovery of all analytes, especially NPLA, was
obtained by substituting methanol with the slightly less polar alcohol, isopropanol.
The cleanest extracts were produced when the cartridges were pre-equilibrated with
the sample matrix (50mM borate pH 9). Water was added to the eluting mixture to
increase arginine recovery (Teerlink, Nijveldt et al. 2002). The absolute and relative
recoveries of the SPE method are shown in Table 7.6.
Table 7.6 Average absolute and relative recovery of analytes. Calculated from 7 level calibration standards after solid phase extraction (n = 4) Analyte (conc. range)
Arginine (7.5-200 µM)
Homoarginine (0.5-12 µM)
ADMA (0.25-6 µM)
SDMA (0.25-6 µM)
NPLA (15 µM)
Absolute recovery
mean ±±±± SD %
80.9 ± 5.6
78.1 ± 5.6
85.1 ± 6.5
86.3 ± 5.2
83.4 ± 5.5
Relative recovery
mean ±±±± SD %
98.9 ± 2.5
94.9 ± 3.2
101.6 ± 1.3
101.4 ± 2.4
100.0
As the fluorescent adducts of AccQ-Fluor™ are stable for at least 7 days (Heresztyn,
Worthley et al. 2004), large batches of samples can be efficiently extracted and
derivatised
87
Method validation
Method precision was evaluated using the pooled quality control plasma. The inter-
assay percent relative standard deviations (RSDs) (n=10) were less than 2.3% for all
analytes. The inter-assay RSDs for ADMA (2.7%) and SDMA (3.1%) compare very
well to other HPLC assays using fluorescence detection (Teerlink, Nijveldt et al.
2002; Heresztyn, Worthley et al. 2004; Nonaka, Tsunoda et al. 2005; Tsunoda,
Nonaka et al. 2005; Blackwell, O'Reilly et al. 2009) and to HPLC or gas
chromatography mass spectrometry methods (Albsmeier, Schwedhelm et al. 2004;
Huang, Guo et al. 2004). As ADMA and SDMA have a very narrow concentration
range in the general population, high analytical precision is required to produce
clinically useful results (Teerlink 2005). Blackwell et al. (Blackwell, O'Reilly D et
al. 2007) recently determined the intra-individual variability for ADMA and SDMA
to be 7.4% and 5.8% respectively in healthy European volunteers. The minimum
required precision of an assay is defined as 0.75 times the intra-individual variability
(Petersen, Fraser et al. 2002; Blackwell, O'Reilly D et al. 2007). This definition
requires that inter-assay RSDs be ≤5.6% for ADMA and ≤4.4% for SDMA.
Desirable imprecision goals are defined as 0.5 times the intra-individual variability
(Petersen, Fraser et al. 2002) which is ≤3.7% for ADMA and ≤2.9% for SDMA
(Blackwell, O'Reilly D et al. 2007). The inter-assay RSDs for ADMA with this
method are within the desirable imprecision goals. The inter-assay RSDs for SDMA
come close to the desirable imprecision goals and are well within the minimum
requirements. As Blackwell et al. note, few published methods for measuring
ADMA and SDMA meet these desirable precision goals. Data on the precision of
this method are presented in Table 7.7.
88
An aliquot of pooled quality control plasma was analysed by HPLC at an
independent research laboratory with an established, validated method (Heresztyn,
Worthley et al. 2004). This laboratory reported mean values of 0.48 µM ADMA and
0.35 µM SDMA, which concurred with the results obtained using this method. Data
on accuracy, expressed as recovery of added analyte from spiked quality control
plasma (n=3), are presented in Table 7.8.
Table 7.7 Intra-assay and inter-assay precision calculated from pooled quality control plasma Analyte
Arginine
Homoarginine
ADMA
SDMA
Intra-assay
n=10
mean ±SD
21.06 ± 0.2
1.87 ± 0.02
0.49 ± 0.01
0.39 ± 0.01
Intra-assay
n=10
RSD (%)
0.93
1.22
2.06
2.26
Inter-assay
n=30
mean ± SD
23.68 ± 1.86
1.88 ± 0.09
0.48 ± 0.01
0.38 ± 0.01
Inter-assay
n=30
RSD (%)
7.88
4.57
2.69
3.07
This assay has since been used successfully to measure plasma dimethylarginines in
over 194 patients with critical illness. It is important to note that of these patients,
only 15 had ADMA more than 1 µM (unpublished data). Hence this assay was
optimised to be accurate and precise at low concentrations of ADMA and SDMA.
89
Table 7.8 Assay accuracy calculated from spiked plasma samples (n = 3)
Concentration (µM)
Analyte Mean
unspiked
plasma
Spike
added
Mean
spiked
plasma
SD RSD (%) Mean
spike
recovered
(µM)
Accuracy/
Spike
recoverya
(%)
Arginine
Homo
arginine
ADMA
SDMA
11.70
0.94
0.25
0.20
3.78
7.55
12.60
15.10
25.20
50.50
0.50
0.75
1.00
1.50
3.00
0.13
0.25
0.38
0.50
0.75
1.50
0.13
0.25
0.38
0.50
0.75
1.50
15.58
19.86
25.47
26.91
37.50
64.14
1.35
1.63
1.86
2.42
4.03
0.36
0.51
0.62
0.75
1.00
1.78
0.32
0.46
0.58
0.70
0.96
1.72
0.41
0.91
1.01
0.82
0.48
1.81
0.26
0.27
0.29
0.31
0.43
0.02
0.03
0.02
0.02
0.06
0.10
0.03
0.05
0.03
0.04
0.03
0.07
2.63
4.61
3.95
3.05
1.27
2.82
18.94
16.34
15.83
12.97
10.65
4.81
5.70
2.45
2.05
6.09
5.55
7.78
9.96
5.51
5.71
3.13
4.08
3.88
8.16
13.78
15.22
25.80
52.44
0.41
0.69
0.92
1.48
3.09
0.12
0.26
0.38
0.50
0.76
1.53
0.13
0.27
0.39
0.51
0.77
1.53
102.8
108.1
109.4
100.8
102.4
103.8
81.3
92.0
91.7
98.4
103.0
92.0
104.7
100.9
100.3
101.1
102.1
102.7
106.0
103.6
101.0
102.0
101.9
a Calculated as a percentage of spike recovered from spiked plasma after subtraction of the unspiked
plasma concentration.
90
Healthy plasma levels
Thirty apparently healthy volunteers provided plasma samples. The mean and
standard deviation of each analyte of interest are shown in Table 7.9. These values
were within the healthy range reported by others (Blackwell, O'Reilly D et al. 2007;
Horowitz and Heresztyn 2007), with the exception of L-arginine concentration,
which was lower than expected due to the delay in processing blood from blood bank
donors (Nuttall, Chen et al. 1998).
Table 7.9 Healthy plasma arginine, homoarginine and methylated arginine values (n = 30)
Min
Max
Mean
SD
Arginine (µM)
23.40
152.92
66.91
33.46
Homoarginine (µM)
0.86
3.95
2.15
0.75
ADMA (µM)
0.30
0.58
0.45
0.07
SDMA (µM)
0.20
0.54
0.40
0.09
Limitations and strengths of the assay
A limitation of this assay is the need to condition new HPLC columns before
retention times stabilise, a requirement noted in other methods (Takenaga, Ishii et al.
1995; Skotty, Lee et al. 1996; Rustum and Estrada 1998; Morrison and Dolan 2005).
After conditioning the new column with repeated injections of either standards or the
quality control plasma, retention times stabilised and excellent retention times were
then obtained for the duration of the column life. This method has been used with
three Gemini-NX columns, each lasting approximately 900 injections.
91
This method is not as short as a number of other published methods because it uses
AccQ-Fluor™ derivatisation and a non-endogenous internal standard. AccQ-
Fluor™ derivatisation leads to longer chromatography (Ahmed, Argirov et al. 2002;
Oreiro-Garcia, Vazquez-Illanes et al. 2005), however the stable adducts produced by
AccQ-Fluor™ give accurate results without requiring on-line derivatisation.
Furthermore, the shorter published methods tend to use either monomethylarginine
(MMA) or homoarginine as internal standards, concentrations of which may be
altered in disease states (Martens-Lobenhoffer and Bode-Boger 2007; Blackwell,
O'Reilly et al. 2009). Using a non-endogenous internal standard gives more accurate
results and also allows all analytes to be quantitated in plasma.
This method has several strengths. Firstly, the substitution of methanol with
propanol in the SPE method gives improved recovery of all analytes. Secondly, a
combination of the acetonitrile gradient and borate buffer at pH 9 on the Gemini NX
column produced clearly defined chromatographic peaks. Thirdly, the average
accuracy of ADMA was 100.2% + 4.3% while for SDMA it was 102.9% + 1.8%.
Finally, the inter-assay RSDs for ADMA are within the desirable precision goals set
out by Blackwell et al. (Blackwell, O'Reilly D et al. 2007) while SDMA
measurements easily meet the minimum standards and come close to achieving the
desirable precision goals. Importantly, as this method achieves accurate and precise
results from small volumes of plasma it is particularly useful for research into critical
illness.
92
7.5. Effect of processing time on amino acid concentration
7.5.1. Introduction to the STOPWATCH experiment
The HPLC methods outlined above described how plasma amino acids were
measured throughout this study. As has been discussed there are many factors which
affect the accuracy of measurements including extraction and derivatisation
procedures and separation achieved in HPLC methods. Another factor which
influences accuracy is the delay between collecting blood and removing plasma. As
we sought to determine in vivo free amino acid concentrations, it was important to
minimise any ex vivo effects. There were some reports that after blood is collected,
lysed red blood cells release arginase which rapidly degrades plasma arginine.
However, we could not find any detailed guidelines to help us determine how
quickly blood had to be processed to achieve accurate plasma arginine
concentrations. Furthermore, we wanted to know how other free amino acid
concentrations might change if there is a delay in blood processing.
Therefore, we designed the following experiment called the STOPWATCH study
(Separation Time of Plasma – Whether Arginine is Time and Temperature Critical).
This study involved taking blood from 6 volunteers simultaneously and then
comparing the plasma amino acid concentrations as the blood was processed
immediately and at various time points over 24 hours. The results showed that
plasma needed to be separated within 30 minutes of blood collection for plasma
arginine concentrations to be accurate. Furthermore, it also showed that ex vivo
arginine concentrations were much more stable if blood was kept on ice. This
experiment helped develop standard operating procedures to ensure a smooth and
93
rapid transfer of blood from the hospital to the laboratory. The following paper was
published in BMC Clinical Pathology in 2009.
After this paper was published, the same plasma samples were also used to test the
effect of blood processing time on methylated arginines. It was found that blood
needs to be processed within 2 hours for ADMA to be reliable.
7.5.2. Published paper: Ex-vivo changes in amino acid concentrations
from blood stored at room temperature or on ice: implications for
arginine and taurine measurements.
Authors: Joshua S Davis 1, 2, Christabelle J Darcy1, Kim Piera1, Yvette R McNeil1,
Tonia Woodberry1, Nicholas M Anstey1, 2
Authors’ affiliations: 1 Global Health Division, Menzies School of Health Research
and Charles Darwin University, Darwin, NT 0810, Australia. 2 Division of
Medicine, Royal Darwin Hospital, Darwin, NT, 0810, Australia.
94
Abstract
Determination of the plasma concentrations of arginine and other amino acids is
important for understanding pathophysiology, immunopathology and nutritional
supplementation in human disease. Delays in processing of blood samples cause a
change in amino acid concentrations, but this has not been precisely quantified. We
aimed to describe the concentration time profile of twenty-two amino acids in blood
from healthy volunteers, stored at room temperature or on ice. Venous blood was
taken from six healthy volunteers and stored at room temperature or in an ice slurry.
Plasma was separated at six time points over 24 hours and amino acid levels were
determined by high-performance liquid chromatography. Median plasma arginine
concentrations decreased rapidly at room temperature, with a 6% decrease at 30
minutes, 25% decrease at 2 hours and 43% decrease at 24 hours. Plasma ornithine
increased exponentially over the same period. Plasma arginine was stable in blood
stored on ice, with a <10% change over 24 hours. Plasma taurine increased by 100%
over 24 hours, and this change was not prevented by ice. Most other amino acids
increased over time at room temperature but not on ice. Plasma arginine
concentrations in stored blood fall rapidly at room temperature, but remain stable on
ice for at least 24 hours. Blood samples taken for the determination of plasma amino
acid concentrations either should be placed immediately on ice or processed within
30 minutes of collection.
95
Introduction
Quantification of plasma amino acids is not routinely offered by clinical laboratories
and thus plasma often needs to be transported to research or reference laboratories
for testing. In order to accurately assess the concentration of plasma amino acids, it is
important to know their stability in human blood which has been stored or
transported prior to testing. Previous studies addressing this question have been small
and the rate of degradation has not been well quantified.
Arginine, the precursor of nitric oxide (NO) (Boger 2007), is important for
endothelial (Ganz and Vita 2003) and immunological (Bogdan 2001) function and is
acutely decreased in sepsis (Luiking, Poeze et al. 2004) and trauma (Ochoa, Udekwu
et al. 1991), and was thus the focus of this study. The major routes for arginine
metabolism in humans are metabolism by arginase to urea and ornithine; use for
creatine synthesis; and metabolism by nitric oxide synthase to NO and citrulline
(Boger and Bode-Boger 2001). Both red blood cells (RBCs) (Bernard, Meier et al.
2007) and macrophages (Mori and Gotoh 2004) are rich in arginase. In stored
packed RBCs, arginase is released and the resulting degradation of plasma arginine is
thought to be a mechanism of transfusion-associated immunosuppression (Prins,
Houdijk et al. 2001; Bernard, Meier et al. 2007). Other amino acids which are
commonly added to supplementary nutrition for critically ill patients may also play
an important role in immune function including tryptophan (Munn, Zhou et al. 1998)
glutamine (Ardawi and Newsholme 1983) and taurine (Stapleton, Mahon et al. 1994;
Muhling, Campos et al. 2002).
96
Hainque and colleagues studied eight healthy volunteers and found a “significant
degradation” of plasma arginine following 4 hours at room temperature but this was
not quantified and no other time points were reported (Hainque, Gerbet et al. 1985).
Schaefer et al. studied one volunteer and found a 50% decrease in plasma arginine
after 6 hours at room temperature compared with a 10% decrease after 6 hours at 4
degrees centigrade, with earlier time points not reported (Schaefer, Piquard et al.
1987). Nutall and colleagues reported time profile data from one volunteer, which
showed an approximate 33% decrease in plasma arginine by 2 hours at room
temperature (Nuttall, Chen et al. 1998).
To determine the impact of delayed processing we undertook a study to estimate the
rate of arginine degradation in human plasma at room temperature and on ice. We
hypothesised that this degradation would be primarily due to plasma arginase activity
and that there would be less than 10% degradation at 2 hours in samples placed
immediately on ice. We also sought to determine the effect of delayed separation and
freezing of plasma on the concentration of other amino acids.
Methods
The study was considered by the chair of the human research ethics committee of the
Menzies School of Health Research and Northern Territory Department of Health
and Families, and was approved as a quality assurance activity which did not require
full ethical review. Following written informed consent, six healthy normotensive
fasting volunteers had venous blood collected into 12 x 2mL lithium heparin tubes
(Vacutainer, Becton Dickinson, Franklin Lakes, New Jersey) using a 21 gauge
needle and vacutainer system. For each subject, the first six tubes were immediately
97
placed into an ice slurry and the second six were left at room temperature (25°
Celsius (C)) in an air conditioned laboratory. After intervals of 0 minutes, 30
minutes, 2 hours, 4 hours, 8 hours and 24 hours from the time of venepuncture, the
tubes were centrifuged at 3000rpm for 10 minutes (either at 4°C or at room
temperature as appropriate) and the plasma immediately separated and stored at -80
°C.
Subsequently, following thawing, plasma amino acids were extracted with ethanol,
then derivatized with AccQ-Fluor (Waters, Milford, MA). Amino acid
concentrations were then determined by reverse-phase high performance liquid
chromatography (HPLC; Shimadzu corporation, Kyoto, Japan) with UV (250 nm)
and fluorescence (excitation 250 nm, emission 395 nm) detection, using a method
modified from van Wandelen and Cohen (van Wandelen and Cohen 1997).
The data were analysed using Stata 10 (Statacorp, College Station, Texas) and
GraphPad Prism 5 (Graphpad software, San Diego, California). Due to the small
number of subjects, data were summarized using median and interquartile range.
Median amino acid concentrations over time were compared using a paired
Wilcoxon test, with a p-value of <0.05 considered significant. The arginine
degradation curve was fitted using a one-phase exponential decay model. The sample
size was determined using data from an earlier experiment (unpublished data), which
found that there was 31.8% (std dev=14%) degradation of arginine at room
temperature by 2 hours. Using a power of 80% and a significance level of 5%, five
subjects in each group would be needed to detect a difference of 22% degradation at
98
2 hours, meaning less than 10% degradation in the ice group. To allow for sample
wastage and errors, we recruited six subjects.
Results
Of the six study subjects, half were male, and the median age was 37.5 years, with a
range of 19-47 years (Table 7.10). All were healthy, of normal weight and
normotensive, and none had cardiovascular disease or diabetes mellitus. The median
baseline plasma arginine concentration was 74.9 µmol/L, similar to previously
reported mean plasma arginine concentrations from healthy volunteers, the majority
of which are between 60 and 80 µmol/L (Martens-Lobenhoffer and Bode-Boger
2006).
Table 7.10 Characteristics of study subjects Subject Age (years) Gender Ethnicity 1 36 F Caucasian 2 39 M Caucasian 3 47 F Caucasian 4 27 F Caucasian 5 19 M Caucasian 6 44 M Caucasian
Arginine and ornithine time profiles at room temperature
Plasma arginine concentration decreased rapidly at room temperature (Figure 7.5,
Figure 7.6, Table 7.11) with 6% degradation within 30 minutes, 25% degradation
within 2 hours and 43% degradation within 24 hours. A non-linear model of the
plasma arginine profile over time was defined by the equation Y=((Y0-P)*e-kt )+P,
where t=time in hours, P=the plateau value, Y0=initial value. The parameters of the
model were Y0=81.3, P=37.8, and k=0.6273. This model fitted the data well, with an
R2 of 0.73. Plasma ornithine concentration increased exponentially at room
temperature (Table 7.11, Figure 7.6) , with a 4% increase at 30 minutes, a 62%
99
increase at 2 hours, and a 183% increase at 24 hours. If the fall in arginine
concentrations had been paralleled exactly by the rise in ornithine concentrations,
one would expect a proportional increase in plasma ornithine over 24 hours of
1/0.43=2.32. The fact that the observed increase was 1.83-fold suggests that a small
proportion of arginine degradation occurred by pathways other than metabolism by
arginase to ornithine.
Figure 7.5 Plasma arginine time profile at room temperature and on ice. Each curve represents an individual subject. (a) whole blood stored at room temperature (25° C), (b) aliquots of the same blood samples stored in an ice slurry. Arginine time profile on ice compared with room temperature
Plasma arginine was extremely stable on ice, with a less than 10% change over a 24
hour period. At 2 hours, the median plasma arginine concentration had decreased by
6% in the ice specimens compared with 25% in the room temperature specimens
100
(p<0.001) (Figure 7.5). At 24 hours, the change in arginine was negligible for the ice
specimens compared with a 43% decrease at room temperature (p<0.001). Ornithine
was also very stable on ice, with a 24% increase over the 24 hour period, compared
with a 183% increase at room temperature.
Table 7.11 Median (IQR) arginine and ornithine plasma concentrations over time from blood stored at room temperature compared to blood stored on ice Baseline 30 minutes 2hours 4hours 8hours 24hours Arginine RT a 74.9 70.3 49.6 40.4 37.3 42.6 73.2-87.8 63.4-75.5 46.0-53.6 35.8-45.8 32.1-42.6 25.5-42.8 Arginine Ice 79.6 77.1 74.8 78.6 80.4 81.0 76.8-93.0 74.6-90.8 73.4-86.9 74.6-86.1 79.4-86.7 79.9-83.0 Ornithine RT 44.7 45.6 72.6 87.4 101.6 114.1 32.9-60.8 39.5-69.8 58.7-94.2 69.1-112.3 79.4-125.9 100.9-153.6 Ornithine Ice 38.6 31.6 39.2 40.5 36.3 43.1 29.4-57.3 38.3-59.2 30.2-60.0 30.5-61.9 32.8-61.5 36.2-68.1 a. RT=Room Temperature
Figure 7.6 Time profile of median plasma arginine and ornithine concentrations in blood stored at room temperature. Each point represents the median value for that time, and the error bars rpresent the interquartile range. Median plasma arginine is indicated by triangles, and ornithine by solid circles.
101
Time profile of other amino acids
For the majority of other amino acids, concentrations increased by >10% over 24
hours at room temperature (Table 7.12). The majority of these changes were largely
or completely prevented in the blood that was placed on ice. The most notable room
temperature concentration increases at 24 hours were seen with taurine (which
doubled) and glutamate (which increased more than five fold). The change in taurine
was unusual in that it was more marked in the blood placed on ice (a 126% increase)
than the room temperature specimens (a 100% increase), suggesting that the increase
in taurine may be due to release from lysed cells rather than to an enzymatic process
(Figure 7.7). Tryptophan was very stable both at room temperature and on ice.
102
Table 7.12 Changes in amino acid concentrations in whole blood after 24 hours at room temperature and on ice %Change at 24h
at RTa,b
%Change at 24h
on ice b
Group 1 - ≤10% change at RTa over 24 h
Citrulline -4 (-6, 4) -7 (-9, -4)
Glutamine -10 (-13, -10) -5 (-5, -4)
Hydroxyproline 8 (7,9) -3 (-4,-3)
Methionine 0 (-2, 1) 1 (1, 6)
Tryptophan 7 (5, 8) 4 (4, 6)
Tyrosine 8 (5, 12) -2 (-3, -1)
Valine 8 (4, 11) 1 (0, 1)
Group 2 - >10% increase at RTa over 24h
Alanine 18 (16,20) 0 (-1, 0)
Asparagine 17 (12, 21) 0 (-1, +3)
Glutamate 593 (563, 612) 38 (92, 186)
Glycine 26 (24, 34) 3 (2, 4)
Histidine 23 (17, 27) 1 (0, 1)
Isoleucine 16 (10, 21) 0 (-1, 2)
Leucine 23 (17, 34) 2 (1, 5)
Lysine 19 (18, 19) 2 (1, 5)
Ornithine 183 (180, 224) 24 (23,25)
Phenylalanine 15 (14,22) 1 (1, 3)
Proline 11 (6,13) 1 (-1,2)
Serine 18 (17,28) 6 (2,6)
Taurine 100 (94, 102) 126 (120, 147)
Threonine 11 (10, 14) -2 (-5, 0)
Group 3 - >10% decrease at RTa over 24h
Arginine -43 (-65, -43) -1 (-5, 4)
a. RT – Room temperature
103
Figure 7.7 Time profile of median plasma taurine concentrations in blood stored at room temperature and on ice. Each point represents the median value for that time and the error bars represent the interquartile range. Median plasma taurine at room temperature is represented by solid circles, and median plasma taurine on ice is represented by triangles.
Discussion
Plasma arginine concentration decreases rapidly in whole blood held at room
temperature, and this decrease is greatly attenuated by placing the blood on ice.
Ornithine, the metabolic product of arginine metabolism by arginase, rises
exponentially at room temperature, and this rise does not occur on ice, suggesting
that it is due to an enzymatic process. Thus, it is likely that arginase is the primary
mechanism of arginine degradation in ex-vivo blood samples. This arginase could
come from either lysed RBCs or lysed macrophages, but we did not evaluate the
source of arginase, and thus cannot determine which of these was more important. In
vitro hemolysis is difficult to measure, as the released cell-free haemoglobin is
immediately bound by haptoglobin. Nonetheless, our observations suggest that the
decrease in arginine in ex-vivo blood is due to arginase activity.
104
Most other amino acids increase at room temperature but not on ice, which also
implies an enzymatic reaction. Tryptophan is very stable both at room temperature
and on ice. Taurine and glutamine are unusual, in that they increase markedly both at
room temperature and on ice; this may be due to their release from peripheral blood
mononuclear cells (PBMC).
The rate of decrease of plasma arginine which we found in blood held at room
temperature is similar to that found by Nuttall and colleagues in the only published
paper to have reported plasma arginine concentrations at room temperature at more
than two time points (Nuttall, Patton et al. 1998). The lack of early time points in
other papers make it difficult to estimate the rate of decline and whether it is linear or
exponential. Nuttall et al. reported data in graphical form, from a single subject up to
2.5 hours post venepuncture. They found a fall from 89 µmol/L to approximately
60µmol/L at 2 hours (a 33% drop), similar to our reported decrease of 25% at 2
hours.
The large increases seen in taurine and glutamate in our study have not previously
been reported. Sahai et al. measured amino acid levels in whole blood from twenty-
two volunteers, stored on ice for 1 hour or 2 hours, and found a less than 10%
decrease in plasma taurine and glutamate at 1 and 2 hours (Sahai and Uhlhaas 1985).
Shaeffer et al. reported a <10% decrease in plasma taurine and glutamate at 6 hours
in blood held at room temperature from one healthy volunteer (Schaefer, Piquard et
al. 1987). The reason for this discrepancy is unclear. These papers used different
methods of amino acid quantification. Sahai et al. did not measure time points
beyond 2 hours, and most of the increase in both taurine and glutamine in our study
105
occurred beyond 2 hours. However, until this finding is reproduced by other
investigators, it should be regarded with caution.
The primary limitations of this study are the relatively small number of subjects and
the lack of subjects suffering from sepsis, trauma or other conditions of interest. A
larger number of subjects would allow a more accurate estimate of the time profile of
arginine degradation over time. Considering arginase activity is increased in severe
sepsis (Argaman, Young et al. 2003) and trauma (Bernard, Mistry et al. 2001), it is
unclear if blood from patients with these conditions would yield the same results as
we observed. We did not directly measure arginase activity in blood or plasma, and
thus our inference that plasma arginase is primarily responsible for the observed ex-
vivo arginine degradation is based on indirect evidence and may be incorrect.
However, the only other significant mechanism for arginine degradation likely to
occur ex-vivo is the breakdown of arginine to NO and citrulline by nitric oxide
synthase, which accounts for less than 5% of arginine metabolism in healthy humans
(Castillo, Beaumier et al. 1996).
One potential implication of these data is that whole blood stored for the purpose of
transfusion is likely to contain non-physiological concentrations of amino acids,
which may have unintended immunosuppressive effects. These data also reinforce
the importance of accurate methodological descriptions in papers reporting plasma
amino acid levels. In a hospital setting, it is not always possible to process samples
within 30 minutes of collection. It is therefore essential to note the time between
collection and freezing when reporting concentrations of plasma amino acids. This
is particularly important if the sample cannot be kept on ice - for example, if the
106
blood is to be used for both peripheral blood mononuclear cell (PBMC) collection
and amino acid analysis. As PBMC lyse on ice, these samples must be kept at room
temperature and processed as soon as possible to allow accurate analysis of both
PBMC function and amino acid concentrations. Furthermore, where plasma amino
acids are being measured for clinical applications, our data emphasise the importance
of timely separation and freezing of plasma; if this is not done, there could be
important clinical consequences such as incorrect diagnosis of metabolic disorders.
In conclusion, arginine undergoes rapid ex-vivo degradation at room temperature but
this does not occur on ice; plasma tryptophan is stable for at least 24 hours both at
room temperature and on ice; plasma taurine concentrations show large increases
both at room temperature and on ice. Blood collected for the purposes of plasma
amino acid determination should be placed immediately on ice; if this is not possible,
plasma should be frozen with 30 minutes of collection.
7.6. Conclusion
HPLC is a simple and accurate method for measuring amino acids. HPLC assays
constantly undergo development as technology improves. The general amino acids
assay has had the same extraction and derivatisation process for six years but there
have been two different HPLC methods. The second method was developed as
column technology improved, allowing more amino acids to be separated in a shorter
amount of time. The ADMA assay was novel because it used a new type of column
and high pH buffer which gave accurate results from 100 µL of plasma. The
STOPWATCH experiment demonstrated that to accurately measure most amino
acids, plasma needs to be separated from blood within 2 hours. To accurately
107
measure arginine, blood must be processed within 30 minutes or placed on ice. In
the following results chapters, amino acids were only reported if the plasma had been
processed as required, according to the STOPWATCH experiment.
Together, the general amino acids assay and the ADMA assay allowed us to measure
the free amino acids in plasma required to investigate arginine and tryptophan
metabolism in sepsis. The STOPWATCH experiment helped us plan the logistics of
getting the blood from the hospital to the laboratory and ensured that the results that
we obtained were accurate. Occasionally, it was not possible to process blood
quickly enough, in which case it was not used for amino acid analysis.
The next two chapters present the results obtained with these methods. The first
discusses arginine metabolism in sepsis and the second discusses tryptophan
metabolism in sepsis.
108
8. Results: Arginine bioavailability in sepsis
8.1. Introduction
The preliminary results in chapter 6 demonstrated that sepsis patients have low
plasma arginine concentrations. This chapter consists of two draft manuscripts
which further investigate the role of arginine metabolism in sepsis. The first
manuscript investigates why arginine is low in sepsis and examines plasma arginase
activity in sepsis, which has not previously been reported. The second manuscript
considers the availability of arginine to nitric oxide, which is measured by the ratio
of arginine to ADMA in plasma. These two manuscripts add to our understanding of
the role of arginase and nitric oxide in the pathology of sepsis.
8.2. Arginase activity in sepsis
The following arginase results have been prepared in short report format.
8.2.1. Draft manuscript: Increased plasma arginase activity in sepsis is
associated with increased circulating neutrophils
Authors: C. J. Darcy 1, T. Woodberry 1, J.S. Davis 1,2, K. Piera 1 , Y. R. McNeil 1,
D.P. Stephens 4, T. W. Yeo 1,2, J. B. Weinberg3, N.M. Anstey 1,2
Authors’ affiliations: 1 – Global Health Division, Menzies School of Health
Research and Charles Darwin University, Darwin, NT 0810, Australia. 2 – Division
of Medicine, Royal Darwin Hospital, Darwin, NT, 0810, Australia. 3 –Division of
Hematology-Oncology, Duke University and Veterans’ Affairs Medical Centers,
109
Durham, NC 27710 USA. 4 - Intensive Care Unit, Royal Darwin Hospital, Darwin,
NT, 0810, Australia
Abstract
Sepsis patients have low plasma arginine concentrations and abnormal neutrophil
counts. As human neutrophils constitutively express arginase I, we hypothesised that
the circulating neutrophil count would be related to the plasma arginase activity and
plasma arginine concentrations in sepsis. We measured plasma arginase activity in
18 sepsis patients and 12 hospital controls and found that plasma arginase activity in
sepsis correlates with the number of circulating neutrophils. Sepsis patients with the
highest circulating neutrophil count had the highest plasma arginase activity and the
lowest plasma arginine concentration. These results suggest that neutrophil-derived
arginase contributes to the low plasma arginine concentrations in sepsis.
110
Introduction
Sepsis patients have decreased plasma arginine concentrations compared to healthy
controls but the reasons for this are incompletely understood (Davis and Anstey
2010). As plasma arginine concentration is essential for both endothelial (Moncada
and Higgs 2006) and immune (Bronte and Zanovello 2005) function, it is important
to understand what factors might contribute to the low arginine in sepsis.
In humans, neutrophils constitutively express arginase I (Munder, Mollinedo et al.
2005) and this is upregulated in response to activation (Rodriguez, Ernstoff et al.
2009). Although murine macrophage, monocytes and dendritic cells express
arginase in response to Th2 cytokines, this does not appear to be the case in humans
(Munder, Mollinedo et al. 2005). As sepsis patients have increased numbers of
circulating neutrophils, we hypothesised that sepsis patient with increased numbers
of circulating neutrophils would have increased plasma arginase activity and
decreased plasma arginine compared to controls.
Methods
We investigated plasma arginase activity in 18 sepsis patients and 12 hospital
controls with detailed peripheral blood mononuclear cell phenotyping (see Chapter
10), plasma processed within 30 minutes and no signs of haemolysis. Haemolysis
was defined as visibly haemolysed or with cell free haemoglobin over 400 µg/mL
and 4 of the original 22 samples were excluded to avoid artefactual measurement of
arginase released from red blood cells during sample collection or processing. These
patients were part of a previously reported study of endothelial function in sepsis
111
(Davis, Yeo et al. 2009) and were representative of the rest of the cohort in terms of
age, ethnicity, gender and sepsis severity.
Sepsis patients had suspected or proven infection and the presence of two or more
criteria for the systemic inflammatory response syndrome (SIRS) within the last 4
hours (Bone, Balk et al. 1992). Sepsis severity was estimated using the Acute
Physiology and Chronic Health Evaluation (APACHE) II score. Patients were
enrolled within 24 hours of ICU admission or within 36 hours of ward admission.
Control subjects were recruited from hospital patients who had not met SIRS criteria
within the last 30 days and who had no clinical or laboratory evidence of
inflammation or infection. Written informed consent was obtained from all
participants or next of kin. The study was approved by the Human Research Ethics
Committee of Menzies School of Health Research and the Department of Health and
Community Services.
Venous blood was collected in lithium heparin tubes and plasma was separated
within 30 minutes and stored at -80 °C. Plasma arginine concentrations were
measured by High Pressure Liquid Chromatography (HPLC; Shimadzu, Kyoto,
Japan) with UV (250 nm) and fluorescence (excitation 250 nm, emission 395 nm)
detection, using a method modified from van Wandelen and Cohen (van Wandelen
and Cohen 1997). Plasma arginase activity was measured using a radiometric assay,
as previously described, and reported as micromole/millliter/hour (Morris, Kato et al.
2005). Plasma concentrations of cell free hemoglobin were measured by ELISA,
according to the manufacturer’s instructions (Bethyl Laboratories). White blood cell
counts were measured by an automated counter (T890; Beckman Coulter) and high
112
circulating neutrophil count in sepsis was defined as over 14 x 103/µL (median
circulating neutrophil count in this cohort).
Continuous parametric variables were compared using Student’s t-test, continuous
non-parametric variables were compared using Mann-Whitney, Kruskal-Wallis or
Wilcoxon tests as appropriate. Correlations were examined using Pearson’s or
Spearman’s tests for parametric and non-parametric data respectively. A 2-sided p-
value of <0.05 was considered significant. Analyses were performed using Prism
version 5.01 (GraphPad Software, CA, USA).
Results
Plasma arginase activity was investigated in 18 sepsis patients and 12 hospital
control (Table 8.1). Sepsis patients had a significantly higher circulating white blood
cell count, mostly due to a high circulating neutrophil count.
In sepsis patients, increased circulating neutrophils correlated with increased plasma
arginase activity (r2 = 0.42, p = 0.003) and decreased plasma arginine concentration
(r2 = 0.27, p = 0.027; Figure 8.1). As neutrophils made up most of the white blood
cell count, the white blood cell count was also associated with plasma arginase
activity (r2 = 0.37, p = 0.007) and plasma arginine concentration (r2 = 0.26, p =
0.032), but not to same extent. There was no association with any other circulating
cells including monocytes, immature granulocytes and lymphocytes (data not
shown).
113
Table 8.1 Cohort information All
sepsis
Sepsis
Nphil>14
Sepsis
Nphil≤14
Control All sepsis
vs control
Sepsis
nphil>14
vs control
sepsis
nphil≤14
vs
control
Sepsis
nphil>14 vs
nphil≤14
Subjects, n 44 22 22 25
Age * 50 (46 - 55) 50 (44 - 55) 51 (44 – 58) 45 (40 - 50) ns ns ns ns
Male, n (%) 26 (59%) 11 (50%) 15 (68%) 17 (68%) ns ns ns ns
ATSI (%) 26 (59%) 14 (64%) 12 (55%) 13 (52%) ns ns ns ns
SOFA score 3 (1 – 8) 2 (1 – 7) 5 (1 – 9) N/A N/A N/A N/A ns
SOFA hepatic component 0 (0 – 0.75) 0 (0 – 0.25) 0 (0 – 1) N/A N/A N/A N/A ns
SOFA renal component 0 (0 – 1) 0 (0 – 1.25) 0.5 (0 – 1.25) N/A N/A N/A N/A ns
APACHE II score‡ 15 (8 -20) 13.5 (8 – 20.5) 16 (5.5 – 18.5) N/A N/A N/A N/A ns
White blood cell x 103/µL ‡ 15 (10 – 18) 18 (17 – 25) 11 (7 – 14) 8 (6 -10 )# 0.001 <0.0001 ns <0.0001
Neutrophil x 103/µL‡ 14 (9 – 16) 16 (14 – 21) 9 (4 – 10) 5 (3 – 6) # 0.0002 <0.0001 ns <0.0001
Monocyte x 103/µL ‡ 0.6 (0.4 – 1.1) 0.95 (0.5 – 1.2) 0.5 (0.3 – 0.6) 0.5 (0.5 – 0.6) # ns 0.02 ns 0.006
Lymphocyte x 103/µL ‡ 0.9 (0.50 – 1.3) 1.1 (0.8 – 1.7) 0.8 (0.5 -1.1) 2.2 (1.9 – 2.2) # 0.002 0.04 0.004 0.07
Immature granulocyte x 103/µL‡ 0 [0 – 6.7] 0 [0 – 6.7] 0 [0 – 3.1] 0 [0 – 0] # ns ns ns ns
Plasma interleukin 6 (pg/mL) 267 (76 – 563) 277 (105 – 832) 267 (63 – 428) 5 (5 - 5) <0.0001 <0.0001 <0.0001 ns
Plasma cell free hemoglobin (µM) ‡ 0.74 (0.56 – 1.2) 0.88 (0.52 – 1.32) 0.68 (0.56 – 1.1) 0.66 (0.43 – 1.2) ns ns ns ns
Plasma arginase activity
µmol/mL/hr ‡
0.17 (0.09 –0.23) 0.21 (0.14 – 0.26) 0.10 (0.05 – 0.18) 0.13 (0.05 –0.16) 0.07 0.004 ns 0.0009
Plasma L-arginine (µM) ‡ 33 (27 – 47) 30 (20 – 41) 39 (30 – 53) 81 (69 – 91) <0.0001 <0.0001 <0.0001 0.02
ATSI = Aboriginal or Torres Strait Islander, * Mean (95% confidence interval), ‡Median (interquartile range) or [range], #n = 12
114
Figure 8.1 Relationship between circulating neutrophil count and plasma argininase activity (a) and plasma arginine concentration (b).
Sepsis patients with high circulating neutrophil counts (more than 14 x 103/µL) had
significantly higher plasma arginase activity (median 0.21, IQR [0.18 – 0.25])
compared to hospital controls (0.14 [0.09 – 0.16]; p = 0.002), whereas sepsis patients
with lower circulating neutrophil counts did not (0.1 [0.05 – 0.18]; p = ns) (Figure
8.2a). Furthermore, sepsis patients with high circulating neutrophil counts had lower
plasma arginine concentrations (median 27 µM, IQR [17 – 40]) than sepsis patients
with lower circulating neutrophils counts (42 µM [35 – 63]) and hospital controls (74
µM [65 – 88]) (Figure 8.2 b).
There was no association between plasma arginase activity and disease severity (as
measured by APACHE II score).
115
(a) (b)
Figure 8.2 Group comparison of arginase activity and arginine. Comparison of plasma arginine activity (a) and plasma arginine concentration (b) in sepsis patients with high circulating neutrophil counts (>14 x 103/µL), sepsis patients with lower circulating neutrophil counts (<14 x 103/µl) and hospital controls. Bars represent the median and inter-quartile range.
Discussion
It is unclear why sepsis patients have low plasma concentrations of arginine.
Potential mechanisms include decreased intestinal absorption, increased protein
synthesis and increased arginase activity, as reviewed in (Davis and Anstey 2010).
Plasma arginase activity represents extra-cellular arginase which is circulating in the
plasma. Human hepatocytes, erythrocytes, endothelial cells and smooth muscle cells
express intra-cellular arginase, which is only released from these cells if they are
damaged or when they die (Morris 2007; Morris 2009). In contrast, activated
neutrophils release arginase into the extra-cellular environment via de-granulation
(Rodriguez, Ernstoff et al. 2009). Our results demonstrate that the number of
circulating neutrophils in sepsis is positively correlated with plasma arginase activity
and inversely correlated with plasma arginine concentration.
116
This study has several limitations. We did not directly measure arginase activity or
arginase protein in neutrophils from sepsis patients. Therefore, it is unclear whether
the neutrophils are directly responsible for the increased plasma arginase or whether
the neutrophils simply represent an inflammatory milieu in which other cell types or
tissues release arginase, possibly during cell death. However, patients with increased
inflammation or organ failure did not have increased plasma arginase activity, as
there was no association between plasma arginase activity and APACHE II score.
Similarly, plasma arginase activity was not associated with any circulating leukocyte
measured by coulter counter other than neutrophils. Thus, although more work
needs to be done to confirm these results, our preliminary data suggest that
circulating neutrophils do contribute to plasma arginase activity in sepsis.
The relationship between circulating neutrophil count, plasma arginase activity and
plasma arginine concentration suggests that increased plasma arginase activity
contributes to the low arginine concentrations in sepsis. This is consistent with a
recent stable-isotope infusion study which demonstrated that sepsis patients have
increased arginase activity, as sepsis patients convert a higher percentage of whole-
body arginine production to urea compared to controls (Luiking, Poeze et al. 2009).
In the context of the existing literature, our results suggest that neutrophil-derived
arginase may contribute to the low plasma arginine concentrations in sepsis.
117
8.3. Arginine/ADMA ratio in sepsis
The FRESH paper reproduced in chapter 6 demonstrated that sepsis patients have
low plasma arginine concentrations and impaired microvascular reactivity. As the
ratio of plasma arginine to asymmetric dimethylarginine represents arginine
bioavailability to nitric oxide synthase, we hypothesised that this would better reflect
nitric oxide mediated microvascular reactivity, than plasma arginine concentration
alone.
8.3.1. Published paper: The arginine/asymmetric dimethylarginine
ratio, microvascular reactivity and organ failure in sepsis
Joshua S Davis*1,2, Christabelle J Darcy *1, Tsin W Yeo1,2 , Catherine Jones1, Yvette
R McNeil1, Dianne P Stephens4, David S Celermajer3 , Nicholas M Anstey1,2
*These authors contributed equally to this work
Authors’ affiliations: 1 Global Health Division, Menzies School of Health Research
and Charles Darwin University, Darwin, NT 0810, Australia. 2 – Division of
Medicine, Royal Darwin Hospital, Darwin, NT, 0810, Australia. 3 – Department of
Medicine, University of Sydney and Department of Cardiology, Royal Prince Alfred
Hospital, Sydney, NSW 2006, Australia 4- Intensive Care Unit, Royal Darwin
Hospital, Darwin, NT, 0810, Australia
118
Abstract
Objective
Arginine bioavailability to nitric oxide synthase is estimated by the ratio of L-
arginine to asymmetric dimethylarginine (ADMA). We hypothesised that plasma
arginine/ADMA ratio would be decreased in sepsis, in proportion to disease severity,
and would correlate with microvascular reactivity.
Methods and Results
In a prospective longitudinal study of 67 patients with sepsis and 31 hospital
controls, blood was collected and microvascular reactivity was measured at baseline
and 2-4 days later. Digital microvascular reactivity was measured by peripheral
arterial tonometry and plasma arginine and ADMA concentrations were determined
by high performance liquid chromatography (HPLC). Baseline plasma
arginine/ADMA ratio was significantly lower in sepsis patients (median [IQR] 63
[45-103] than in hospital controls (143 [123-166], p<0.0001). The plasma
arginine/ADMA ratio correlated with microvascular reactivity (rs=3.4, p=0.02) and
inversely correlated with severity of illness (rs=-4.0, p=0.003) and organ failure (rs= -
5.0 p=0.0001) in sepsis. Baseline plasma ADMA was independently associated with
28-day mortality (Odds ratio [95% CI] for death in those in the highest quartile
(≥0.66 µM ) = 20.8 [2.2-195.0], p=0.008).
Conclusions
Plasma arginine/ADMA ratio is decreased in sepsis, in proportion to disease severity.
Reduced endothelial nitric oxide bioavailability may impair microvascular reactivity
and lead to organ failure in sepsis.
119
Introduction
Asymmetric dimethylarginine (ADMA) is an endogenous non-specific nitric oxide
synthase (NOS) inhibitor that is associated with chronic endothelial dysfunction
(Juonala, Viikari et al. 2007) and increased cardiovascular risk (De Gennaro
Colonna, Bianchi et al. 2009). Symmetric dimethylarginine (SDMA) is the
stereoisomer of ADMA which competes with arginine for transport into the cell but
does not inhibit NOS. The role of ADMA and SDMA in endothelial dysfunction in
acute infections has not been well characterised.
Severe sepsis is the leading cause of death in intensive care units in the USA (Angus,
Linde-Zwirble et al. 2001), and is increasing in incidence globally (Martin, Mannino
et al. 2003). Microvascular and endothelial dysfunction are key contributors to organ
failure and death in sepsis but the mechanisms linking sepsis with vascular
dysfunction remain incompletely understood (Aird 2003). Microvascular reactivity
is the ability of the microvessels to dilate in response to shear stress. Endothelial
nitric oxide synthase (eNOS) produces nitric oxide (NO) in response to shear stress.
A relative deficiency of constitutively expressed endothelial NO may underlie sepsis-
associated endothelial and microvascular dysfunction (Trzeciak, Cinel et al. 2008;
Davis, Yeo et al. 2009). NO is produced by NOS from its primary substrate, L-
arginine. ADMA competitively inhibits the production of NO by NOS; hence the
arginine/ADMA ratio is considered a better indicator of the availability of arginine to
NOS than plasma arginine concentration alone (Bode-Boger, Scalera et al. 2007).
Infusion of ADMA in both rats (De Gennaro Colonna, Bonomo et al. 2007) and
humans (Vallance, Leone et al. 1992) acutely decreases NO production, resulting in
120
endothelial dysfunction. Plasma ADMA concentrations are increased in patients with
chronic renal disease (Kielstein, Boger et al. 2002), hypertension (Surdacki, Nowicki
et al. 1999), diabetes mellitus (Abbasi, Asagmi et al. 2001) and peripheral vascular
disease (Boger, Bode-Boger et al. 1997). Furthermore, ADMA has been shown to be
an independent predictor of cardiovascular events in patients with existing coronary
artery disease (Valkonen, Paiva et al. 2001) and end-stage renal disease (Zoccali,
Bode-Boger et al. 2001).
In contrast the few clinical studies that have reported plasma ADMA concentrations
during acute infection have had conflicting results (O'Dwyer, Dempsey et al. 2006;
Zoccali, Maas et al. 2007; Nakamura, Sato et al. 2009). A study of experimental
human endotoxemia (Mittermayer, Namiranian et al. 2004) found a decreased
arginine/ADMA ratio however no studies have reported arginine/ADMA ratios in
sepsis patients or examined microvascular reactivity in this context. Using
peripheral arterial tonometry, we have previously shown that digital microvascular
reactivity, a measure of endothelial NO bioavailability (Nohria, Gerhard-Herman et
al. 2006), is decreased in patients with sepsis (Davis, Yeo et al. 2009). However, we
did not find a correlation between concentrations of plasma arginine and
microvascular reactivity. We also found that despite an increase in plasma arginine
concentrations over time, there was no corresponding improvement in microvascular
reactivity. A potential explanation for these findings in sepsis is competitive
inhibition of NOS by ADMA.
We hypothesised that plasma arginine/ADMA ratio would be decreased in sepsis, in
proportion to disease severity, and would correlate with reactive hyperaemia
121
peripheral arterial tonometry (RH-PAT) index, an in vivo measure of endothelial NO
bioavailability. Furthermore, we hypothesised that increased plasma ADMA would
be associated with mortality.
Methods
Study design and setting
We performed a prospective observational study at a 350-bed Australian teaching
hospital, with an 18-bed mixed intensive care unit (ICU). Approval was obtained
from the Human Research Ethics Committee of the Menzies School of Health
Research and the Department of Health and Community Services. Written informed
consent was obtained from all participants or next of kin where necessary.
Subjects
The study subjects were adults (≥18 years) hospitalised with sepsis, who were
enrolled in a previously-reported study of microvascular reactivity; more detail of
subject recruitment and study procedures are provided in this paper (Davis, Yeo et al.
2009). Sepsis was defined as a proven or suspected infection plus at least 2 criteria
for the systemic inflammatory response syndrome (SIRS) present within the last 4
hours (Bone, Balk et al. 1992). Septic patients were eligible for enrolment within 24
hours of their admission to the ICU, or within 36 hours of admission to the ward.
Control subjects were adults recruited from hospitalised patients with no clinical or
laboratory evidence of inflammation or infection, and who had not met SIRS criteria
within the last 30 days. Septic patients were classified as septic shock, or sepsis
without shock. Septic shock was defined at the time of enrolment as systolic blood
pressure <90mmHg or a reduction of ≥ 40mmHg from baseline despite adequate
122
fluid resuscitation, or the need for vasopressors to maintain these targets (Bone, Balk
et al. 1992). Disease severity was assessed by the Acute Physiology and Chronic
Health Evaluation (APACHE) II score and organ failure was determined using the
Sequential Organ Failure Assessment (SOFA) score.
Laboratory assays
Blood from arterial lines if present, or venepuncture if not, was collected in lithium
heparin tubes at baseline and 2-4 days later, and plasma was separated and stored at -
70⁰C within 2 hours of blood collection. Control patients had blood collected at
baseline only.
ADMA and SDMA were measured by reversed phase HPLC with simultaneous
fluorescence and UV-visible detection, as previously described (Jones, Darcy et al.
2010). Arginine was measured using a method modified from van Wandelen and
Cohen (van Wandelen and Cohen 1997), if plasma was separated within 30 minutes
of collection. IL-6 and TNFα were measured by flow cytometry using a cytokine
bead array (BD Biosciences, CA, USA). Angiopoietin-2 (Ang-2) and intracellular
adhesion molecule-1 (ICAM-1) were measured by ELISA (R&D systems).
Measurement of microvascular reactivity
Microvascular reactivity was measured at the bedside by RH-PAT (Itamar Medical,
Caesarea, Israel), a non-invasive method of assessing endothelial function(Kuvin,
Patel et al. 2003; Hamburg and Benjamin 2009) that is at least 50% dependent on
endothelial NO production (Nohria, Gerhard-Herman et al. 2006). Peripheral arterial
tonometry (PAT) was measured in a fingertip before and after a 5-minute ischemic
123
stress at the forearm, generating an RH-PAT index, normalized to the control arm, as
previously reported (Davis, Yeo et al. 2009).
Statistical methods
Continuous variables were compared using Mann Whitney U test, and categorical
variables using Fisher’s exact test. Correlates with baseline ADMA were determined
using Spearman’s coefficient for univariate analysis. Day 2 values were compared
with baseline values using paired Wilcoxon signed-rank test. The relationship
between baseline ADMA and mortality among sepsis patients was examined using
logistic regression with ADMA divided into quartiles, as previously described
(Nijveldt, Teerlink et al. 2003). To examine longitudinal correlations, linear mixed-
effects models were used. A 2-sided p-value of <0.05 was considered significant. All
analyses were performed using Intercooled Stata 10 (Statacorp, Texas).
Results
There were 20 subjects with septic shock, 47 with sepsis without shock and 31
controls. The three groups were well-matched in terms of age, sex and known
associations with chronically raised ADMA (Table 8.2). Arginine measurements
were available in 19 patients with septic shock, 37 patients in sepsis without shock
and 27 controls. Follow-up measurements 2 to 4 days after study enrolment were
available in 47 out of 67 sepsis patients. Six of 67 sepsis patients (9%) had died by
day 28 of follow-up, 5 of whom were in the septic shock subgroup.
124
Disease severity and outcome
Baseline plasma arginine/ADMA ratio was significantly lower in sepsis patients
(median [IQR] 63 [45-103] than in hospital controls (143 [123-166], p<0.0001)
(Table 8.3, Figure 8.3). Furthermore, septic shock patients had significantly lower
arginine/ADMA ratio (median [IQR] 43 [34-73]) than sepsis patients without shock
(91 [56-108], p<0.0001). The plasma arginine/ADMA ratio inversely correlated with
severity of illness as measured by APACHE II score (rs=-4.0, p=0.003) and organ
failure as measured by SOFA score (rs= -5.0 p=0.0001). The arginine/ADMA ratio
negatively correlated with IL-6 (rs=-0.37, p=0.005) but was not significantly
associated with lactate or C-reactive protein (CRP) (data not shown).
Table 8.2 Baseline characteristics Septic Shock Sepsis without
shock
Controls p valuea
n 20 47 31
Ageb 51.5(12.0) 52.5 (14.4) 45.4 (12.7) NS
Malec 11 (55) 30 (63) 24 (75) NS
Diabeticc 6 (30) 13 (27) 10 (31) NS
Smokerc 8 (40) 22 (46) 14 (44) NS
IHD c 4 (20) 8 (17) 4 (13) NS
Hypertensionc 5 (25) 17 (35) 9 (28) NS
Hyperlipidemia c 4 (20) 11 (22) 11 (34) NS
Chronic renal diseasec 4 (20) 4 (8) 3 (10) NS
APACHE II scored 20.0 (16-23) 10.0 (6-16) <0.0001
SOFA scored 6 (3-9) 2.0 (0.5-4.0) <0.0001
a – by Chi2 test for difference between all 3 groups
b – Mean (sd)
c – n (%)
d – Median (Interquartile range)
125
Table 8.3 Baseline plasma asymmetric dimtheylarginine and related variables All sepsis Septic shock Sepsis without
shock
Control p value
pooled
sepsis v
control
p value
septic
shock vs
control
n 67 20 47 31
Plasma ADMA(µmol/L)a 0.52 (0.39-0.65) 0.64 (0.54-0.85) 0.47 (0.38-0.57) 0.57 (0.50-0.62) 0.10 0.09
Plasma arginine (µmol/L)a,b 35.5 (27.3-51.2) 31.0 (23.7-40.4) 38.1 (29.4-51.7) 81.8 (68.9-91.3) <0.001 <0.001
Plasma arginine/ADMA ratioa,b 63.2 (45.3-103.4) 43.4(33.6-73.3) 91.4 (55.5-108.3) 142.9 (123.0-165.7) <0.001 <0.001
Plasma SDMA (µmol/L)a 0.66 (0.50-1.29) 1.05 (0.77-1.45) 0.56 (0.45-0.80) 0.47 (0.43-0.65) 0.002 <0.001
Plasma lysine (µmol/L)a 128 (100-171) 129 (90-190) 128 (104-162) 184 (157-216) <0.001 0.006
Receiving mechanical ventilationc 14 (21) 9 (47) 5 (26) - - -
RH-PAT indexd 1.70 (0.47) 1.47 (0.40) 1.78 (0.47) 2.05 (0.46) 0.001 <0.001
Plasma Interleukin 6 (pg/ml)a 223 (76.6-563) 885 (298-2412) 148 (46.0-322) 4.7 (2.2-9.5) <0.001 <0.001
White blood cell counta 15.2 (10.1-20.2) 17.5 (11.0-27.8) 15.2 (9.1-17.8) 7.7 (5.7-9.0) <0.001 <0.001
C-reactive proteina 180 (87.3-259) 202 (126-297) 143(84-259) 7 (4-22) <0.001 <0.001
a . median (interquartile range); b . n=septic shock 19, sepsis without shock 37, controls 27; c. n (%) d. mean (sd)
126
The median [IQR] plasma concentration of ADMA was significantly higher in septic
shock patients (0.64 [0.54-0.85] µM) than sepsis patients without shock (0.47 [0.38-
0.57] µM) (p=0.008) (Table 8.3, Figure 8.4) and correlated with SOFA score (r=0.45,
p<0.001). Median [IQR] baseline ADMA was approximately twice as high in those who
died (1.07 [0.75-1.31] µM) as in survivors (0.51 [0.39-0.61] µM), p=0.001. Sepsis
patients with a baseline plasma ADMA concentration in the highest quartile (≥0.66 µM)
had an odds ratio for death of 20.8 (95% CI 2.2-195.0, p=0.008). In a multivariate model
incorporating SOFA score, age, gender and IL-6 concentration, baseline ADMA was the
only significant predictor of death (p=0.04).
SDMA was highest in septic shock, intermediate in sepsis without shock and lowest in
controls (Table 8.3). SDMA, which is predominantly renally excreted (Kielstein,
Salpeter et al. 2006), correlated strongly with serum creatinine (r=0.70, p<0.001),
whereas ADMA did not (r=0.16, p=NS). On univariate analysis, sepsis patients with a
plasma SDMA concentration in the highest quartile (≥1.30 µmol/L) had an odds ratio for
death of 8.12 (95% CI 1.33-50.0), however this became insignificant on controlling for
disease severity (using either IL-6 or SOFA score).
Microvascular reactivity and endothelial activation
Baseline arginine/ADMA ratio was associated with microvascular reactivity as
measured by RH-PAT (rs=0.34, p=0.02) and systolic blood pressure (rs=0.32, p=0.02)
but not diastolic blood pressure. The arginine/ADMA ratio was significantly lower in
sepsis patients who required vasopressors (median [IQR] 42 [32-55] compared to those
who did not (74 [54-108], p=0.002). Baseline plasma ADMA concentration correlated
127
with markers of endothelial activation including Ang-2 (rs=0.45, p=0.0002) and ICAM
(rs=0.47, p=0.0001). This relationship persisted after controlling for disease severity in a
multivariate analysis.
Figure 8.3 (a) Arginine to asymmetric dimethylarginine ratio and microvascular reactivity according to disease severity. (a)Ratio of arginine to asymmetric dimethylarginine in baseline plasma samples according to disease category. Solid circles represent individual sepsis subjects and solid triangles represent individual control subjects. Horizontal lines represent median group values, and error bars represent the inter-quartile range. (b) Baseline microvascular reactivity according to disease category. P values represent comparisons between groups. Solid circles represent mean group values for sepsis subjects and the solid triangle for control subjects. Error bars represent standard error of the mean.
128
Figure 8.4 Baseline plasma concentration of asymmetric dimethylarginine according to disease category. P values represent comparisons between groups. Solid circles represent individual sepsis subjects and solid triangles represent individual control subjects. Open circles represent subjects with a fatal outcome at 28 day follow-up. Horizontal lines represent median group values, and error bars represent the inter-quartile range.
129
Table 8.4 Longitudinal results in subjects with sepsis Day 0 Day 2-4 P Day 0 to 2-4
n 67 47
ADMA 0.53 (0.39-0.66) 0.64 (0.51-0.78) 0.002
Arginine 35.5 (27.3-51.2) 47.2 (30.8-58.1) 0.03
Arginine: ADMA ratio 63.2 (45.3-103.4) 63.0 (41.7-108.0) NS
RH-PAT index 1.70 (1.57-1.82) 1.81 (1.65-1.96) NS
SDMA 0.66 (0.50-1.30) 0.71 (0.47-1.36) NS
IL-6 223 (78.2-530) 54.5 (16.1-201) <0.001
SOFA score 3 (1-7) 2 (1-7) 0.04
Note: ADMA=Asymmetric dimethylarginine. RH-PAT index=Reactive hyperaemia peripheral arterial
tonometry index. SDMA=Symmetric dimethylarginine. IL-6=Interleukin 6. SOFA score=Sequential
Organ Failure Assessment Score.
Longitudinal changes
Over the first 2-4 days of follow up, ADMA increased in the sepsis patients (0.53 to
0.64, p=0.002) (Table 8.4). Plasma arginine concentrations also increased, but due to the
increase in ADMA, there was no significant change in the arginine/ADMA ratio.
Similarly, there is no significant improvement in RH-PAT index between day 0 and day
2-4. In a mixed effects linear regression model examining change from baseline to day
2-4, increase in ADMA over time significantly correlated with increase in SOFA score
(p<0.001) and decrease in RH-PAT index (p=0.03), but not with change in IL-6 or CRP.
130
It also correlated with increase in the liver (p<0.001) but not the renal (p=0.09)
components of the SOFA score.
Discussion
This study found that the arginine/ADMA ratio is significantly reduced in sepsis, in
proportion to disease severity. This is consistent with findings in malaria patients, where
the arginine/ADMA ratio also correlates with disease severity (Yeo, Lampah et al.
2010). Reduced blood flow to organs as a result of endothelial dysfunction is one of the
key contributors to organ failure in sepsis (Vallet 2003). In rats, decreased plasma
arginine/ADMA ratio reduced blood flow through the kidney, liver and spleen (Richir,
van Lambalgen et al. 2009). We found that decreased arginine/ADMA ratio was
associated with organ failure and disease severity in sepsis patients.
Decreased arginine/ADMA ratio may contribute to organ failure in sepsis by reducing
microvascular reactivity. The arginine/ADMA ratio is a marker of arginine availability
to NOS (Bode-Boger, Scalera et al. 2007). In this study we found that baseline
arginine/ADMA ratio, but not arginine or ADMA alone, correlated with endothelial NO-
dependent microvascular reactivity. Furthermore, plasma ADMA concentrations
correlated with increased plasma concentrations of Ang-2 and ICAM-1, both of which
are associated with reduced endothelial nitric oxide (Yeo, Lampah et al. 2008). In
malaria, plasma ADMA concentrations are associated with reduced exhaled nitric oxide
and endothelial function (Yeo, Lampah et al. 2010). Together, these findings suggest
that a low arginine/ADMA ratio reduces endothelial nitric oxide and impairs
microvascular reactivity in sepsis.
131
Microvascular reactivity did improve by day 2 - 4 of the study, along with a significant
increase in plasma arginine concentrations. ADMA concentrations paralleled the
increase in arginine during this time; therefore the availability of arginine to NOS
remained the same. Thus the lack of significant improvement in microvascular
reactivity within the first few days of sepsis may partly be explained by the
arginine/ADMA ratio.
Unlike the arginine/ADMA ratio, ADMA has been previously reported in sepsis -
however there have been inconsistent findings. O’Dwyer and colleagues enrolled 47
patients with severe sepsis and found that baseline ADMA was increased compared with
controls, but that it did not correlate with mortality (O'Dwyer, Dempsey et al. 2006).
Zoccali studied 17 patients with bacterial infections and raised C-reactive protein (CRP),
but no organ failure, and found that ADMA was not raised compared with controls, but
that it significantly increased after resolution of fever (Zoccali, Maas et al. 2007).
Finally, Nakamura studied 10 patients with septic shock and found that ADMA
concentrations were increased and correlated with mortality (Nakamura, Sato et al.
2009). We found that ADMA was significantly higher in septic shock than in sepsis
without shock. ADMA tended to be increased in septic shock compared to hospital
controls; however this did not reach significance. The two previous studies which found
that ADMA was significantly higher in septic shock compared to controls had a higher
mortality and greater illness severity than our study (O'Dwyer, Dempsey et al. 2006;
Nakamura, Sato et al. 2009). In contrast to septic shock, we found that ADMA was
significantly reduced in sepsis without shock compared to hospital controls. This
finding has not been reported before in septic humans, however it agrees with animal
132
models of sepsis where ADMA is significantly decreased after LPS injection (Nijveldt,
Teerlink et al. 2003). As patients with septic shock had increased ADMA and patients
without shock had decreased ADMA, there was no significant difference between the
ADMA concentrations in pooled sepsis and hospital controls – a potentially misleading
finding unless patients are stratified by severity. The only other published study to enrol
non-severe sepsis patients also found no difference in plasma ADMA concentrations
between sepsis and control patients; however, they did not consider non-severe sepsis
and septic shock separately (Zoccali, Maas et al. 2007).
The disparity between ADMA concentrations in shock and without shock may be due to
different mechanisms within these two states. Early sepsis is a hyperdynamic state, with
increased cardiac output and liver and kidney blood flow (Lang, Bagby et al. 1984; Di
Giantomasso, May et al. 2003). This may lead to increased degradation of ADMA in the
liver by dimethylarginine dimethylaminohydrolase (DDAH) and, to a lesser extent,
increased renal excretion. This hypothesis is supported by a study which found that the
liver fractional extraction rate for ADMA is significantly higher and circulating ADMA
is significantly lower in endotoxemic rats compared to controls (Nijveldt, Teerlink et al.
2003). Patients with septic shock have generally developed multiple organ failure and
down-regulation of cellular functions (Singer 2008) and thus hepatic metabolism and
renal excretion of ADMA may drop back to baseline concentrations. This hypothesis is
supported by our finding that ADMA concentrations inversely correlate with liver
function, both at baseline and longitudinally.
133
The final finding of this study was that increased plasma ADMA predicted mortality in
sepsis. Raised plasma ADMA predicts short-term mortality in critically ill surgical
patients with multiple organ failure (Nijveldt, Teerlink et al. 2003) and predicts
mortality in cardiovascular patients (Schnabel, Blankenberg et al. 2005). Our results
suggest that impaired microvascular reactivity may be a mechanistic link between
plasma ADMA concentrations and death in this and other studies (Nijveldt, Teerlink et
al. 2003; Nakamura, Sato et al. 2009).
It is unlikely that the decreased arginine/ADMA ratio in sepsis is a protective response
to excess inducible NO. Inducible nitric oxide synthase (iNOS) produces NO in
response to inflammatory makers whereas eNOS maintains vascular reactivity by
producing NO in response to shear stress. Sepsis patients have excessive iNOS,
resulting in uncontrolled vasodilation of major blood vessels, but insufficient eNOS,
resulting in impaired microvascular reactivity and decreased blood flow to organs
(McGown and Brookes 2007). We found that sepsis patients requiring vasopressors to
maintain vascular stability have the lowest arginine/ADMA ratio. This suggests that a
decreased arginine/ADMA ratio is not protective against excess iNOS.
This study has several limitations. Although it is at least 50% dependant on endothelial
NO (Nohria, Gerhard-Herman et al. 2006), peripheral arterial tonometry is not a direct
measure of NO activity. Other factors may contribute to endothelial NO bioavailability
besides the arginine/ADMA ratio, including CAT transport inhibitors and oxidative
stress resulting in NO-quenching. Plasma nitrate was not used as a marker for NO
because it is not specific to the endothelium and is confounded by renal failure (Lopez,
134
Lorente et al. 2004). The 67 sepsis patients were not all followed up on day 2 - 4,
largely because of hospital discharge; thus the longitudinal results may underestimate
the degree of improvement in microvascular and organ function.
Plasma arginine/ADMA ratio is decreased in sepsis, in proportion to disease severity.
Decreased arginine/ADMA ratio is associated with impaired microvascular reactivity
and increased organ failure. Plasma ADMA is significantly higher in septic shock
compared to sepsis without shock and predicts mortality in sepsis, thus it may be a
useful prognostic marker. Reduced endothelial nitric oxide bioavailability is a potential
mechanism linking decreased plasma arginine/ADMA ratio with endothelial dysfunction
and organ failure in sepsis. An improved understanding of the role of arginine
metabolism in sepsis may lead to potential therapeutic targets (Wang, Liu et al. 2010).
8.4. Conclusion
This chapter outlines two different mechanisms that contribute to the disturbed arginine
metabolism in sepsis. Sepsis patients have increased circulating neutrophils, increased
plasma arginase activity and decreased plasma arginine concentrations. The decreased
plasma arginine in sepsis results in an overall decreased arginine/ADMA ratio, which
reduces the amount of arginine available to NOS. In addition, septic shock patients have
higher ADMA concentrations than non-shock patients, further limiting arginine
bioavailability to NOS. As the arginine/ADMA ratio is associated with impaired
microvascular reactivity and increased disease severity, our data suggest that arginine
bioavailability has an important role in the pathophysiology of sepsis.
135
9. Results: Tryptophan bioavailability in sepsis
9.1. Introduction
The previous chapter demonstrated that decreased arginine bioavailability is associated
with increased disease severity in sepsis. Similarly, this chapter demonstrates the
importance of tryptophan bioavailability in sepsis. This published paper shows that
tryptophan bioavailability is associated with a dysfunctional immune response and
impaired microvascular reactivity in sepsis.
9.2. Tryptophan bioavailability in sepsis
9.2.1. Published paper: An observational cohort study of the kynurenine
to tryptophan ratio in sepsis: association with impaired immune and
microvascular function
Authors: C. J Darcy 1*, J.S Davis 1,2*, T. Woodberry 1, Y. R. McNeil 1, D.P. Stephens 3,
T. W. Yeo 1,2, N.M. Anstey 1,2
* These authors contributed equally to this work
Authors’ affiliations: 1 – Global Health Division, Menzies School of Health Research
and Charles Darwin University, Darwin, NT 0810, Australia. 2 – Division of Medicine,
Royal Darwin Hospital, Darwin, NT, 0810, Australia. 3 – Intensive Care Unit, Royal
Darwin Hospital, Darwin, NT, 0810, Australia
136
Abstract
Both endothelial and immune dysfunction contribute to the high mortality rate in human
sepsis, but the underlying mechanisms are unclear. In response to infection, interferon-γ
activates indoleamine 2,3-dioxygenase (IDO) which metabolizes the essential amino
acid tryptophan to the toxic metabolite kynurenine. IDO can be expressed in endothelial
cells, hepatocytes and mononuclear leukocytes, all of which contribute to sepsis
pathophysiology. Increased IDO activity (measured by the kynurenine to tryptophan
[KT] ratio in plasma) causes T-cell apoptosis, vasodilation and NO synthase inhibition.
We hypothesized that IDO activity in sepsis would be related to plasma interferon-γ,
interleukin-10, T-lymphocytopenia and impairment of microvascular reactivity, a
measure of endothelial NO bioavailability. In a longitudinal study of 80 sepsis patients
and 40 controls, we determined the relationship between IDO activity and selected
plasma cytokines, microvascular reactivity and lymphocyte subsets in sepsis. The
plasma KT ratio was increased in sepsis (median 141 [IQR 64-235]) compared to
controls (36 [28-52]); p<0.0001), and correlated with plasma interferon-γ and
interleukin-10, and inversely with total lymphocyte count, CD8+ and CD4+ T-
lymphocytes, systolic blood pressure and microvascular reactivity. In response to
treatment of severe sepsis, the median KT ratio decreased from 162 [IQR 100-286] on
day 0 to 89 [65-139] by day 7; p=0.0006) and this decrease in KT ratio correlated with a
decrease in the Sequential Organ Failure Assessment score (p<0.0001). IDO-mediated
tryptophan catabolism is associated with dysregulated immune responses and impaired
microvascular reactivity in sepsis and may link these two fundamental processes in
sepsis pathophysiology.
137
Introduction
Sepsis is a systemic inflammatory response to infection (Bone, Balk et al. 1992).
Despite advances in its management, severe sepsis still has a mortality rate of 30-50%
(Angus, Linde-Zwirble et al. 2001; Finfer, Bellomo et al. 2004; Blanco, Muriel-Bombin
et al. 2008). Both immune and endothelial dysfunction are thought to contribute to the
high mortality rate in sepsis (Aird 2003; Hotchkiss and Karl 2003); however, the
underlying mechanisms are not completely understood.
Tryptophan is an essential amino acid that is central to cellular respiration (Ellinger and
Abdel Kader 1947) and neurotransmission (Fernstrom and Wurtman 1971), and is a key
immune mediator. During inflammation, tryptophan is metabolised by indoleamine 2,3-
dioxygenase (IDO) to the toxic metabolite kynurenine. IDO activity is measured by the
ratio of kynurenine to tryptophan (the KT ratio). Endothelial cells, monocytes, renal
tubular epithelial cells and hepatocytes express IDO in response to interferon-γ (Carlin,
Borden et al. 1989; Larrea, Riezu-Boj et al. 2007; Mohib, Guan et al. 2007; Iwamoto, Ito
et al. 2009; Wang, Liu et al. 2010) and IL10 stabilises IDO expression (Munn, Sharma et
al. 2002).
IDO activity regulates a number of immune responses. Increased IDO activity inhibits T
cell function (Fallarino, Grohmann et al. 2006) and proliferation (Munn, Shafizadeh et
al. 1999; Munn, Sharma et al. 2002; Boasso, Herbeuval et al. 2007) and contributes to T
cell apoptosis (Fallarino, Grohmann et al. 2002). Furthermore, elevated IDO activity
inhibits nitric oxide synthase and vice versa (Sekkai, Guittet et al. 1997; Chiarugi,
Rovida et al. 2003; Samelson-Jones and Yeh 2006). Recent isotope studies have shown
138
that systemic NO production is either reduced or unchanged in human sepsis compared
with healthy controls (Villalpando, Gopal et al. 2006; Kao, Bandi et al. 2009; Luiking,
Poeze et al. 2009).
In addition to regulating the immune response, IDO activity may also regulate
endothelial function. Kynurenine, a metabolite of IDO, has recently been described as
an endogenous vasorelaxing factor (Wang, Liu et al. 2010). Increased IDO activity
would therefore be expected to directly decrease systemic vascular resistance.
Additionally, as IDO inhibits NOS, IDO may indirectly affect endothelial function by
impairing NO-dependent microvascular reactivity. NO is essential for normal
endothelial function and NO-dependent microvascular reactivity has been previously
shown to be impaired in patients with sepsis, in proportion to disease severity (Vaudo,
Marchesi et al. 2008; Davis, Yeo et al. 2009).
IDO activity correlates with disease severity in patients with chronic inflammatory
diseases such as human immunodeficiency virus (Huengsberg, Winer et al. 1998),
systemic lupus erythematosus (Widner, Sepp et al. 2000) and malignancy (Huang, Fuchs
et al. 2002), but little is known about IDO activity in acute inflammatory states. A
raised KT ratio has recently been reported in patients with bacteremia (Huttunen,
Syrjanen et al. 2009).
We investigated the relationship between the KT ratio and disease severity in sepsis.
We hypothesised that the KT ratio would be related to IFN-γ and IL10 concentrations,
139
and inversely related to both T cell lymphopenia and microvascular reactivity, a measure
of endothelial NO bioavailability.
Materials and methods
Participants
We evaluated patients with sepsis and hospital controls who were part of a previously
reported study of endothelial function in sepsis (Davis, Yeo et al. 2009). Sepsis patients
had suspected or proven infection and the presence of two or more criteria for the
systemic inflammatory response syndrome (SIRS) within the last 4 hours (Bone, Balk et
al. 1992). Severe sepsis patients had organ dysfunction or shock at the time of
enrolment according to the American College of Chest Physicians/Society of Critical
Care Medicine criteria (Bone, Balk et al. 1992; Stephens, Thomas et al. 2008). Sepsis
severity was estimated using the Acute Physiology and Chronic Health Evaluation
(APACHE) II score from the first 24 hours of admission and daily modified Sequential
Organ Failure Assessment (SOFA) score (Vincent, de Mendonca et al. 1998). Patients
were enrolled within 24 hours of ICU admission or within 36 hours of ward admission.
Control subjects were recruited from hospital patients who had not met SIRS criteria
within the last 30 days and who had no clinical or laboratory evidence of inflammation
or infection. Written informed consent was obtained from all participants or next of kin.
All sepsis patients had undergone resuscitation and were haemodynamically stable at the
time of study enrolment. The study was approved by the Human Research Ethics
Committee of Menzies School of Health Research and the Department of Health and
Community Services.
140
Blood collection and lymphocyte counts
Venous blood was collected in lithium heparin tubes at enrolment, day 2 - 4, and day 7
until discharge from the hospital or death. Whole blood differential white cell counts
were measured by Coulter Counter. Lymphopenia was defined as an absolute
lymphocyte count less than 1.2 x103/µL (Hotchkiss, Swanson et al. 1999). Plasma was
separated and stored at -80°C.
Lymphocytes were analysed in more detail in a subset of patients from whom samples
were processed within 30 minutes of collection, matched for age and gender. Peripheral
blood mononuclear cells were separated using Ficoll-Paque™ Plus (GE Healthcare
Biosciences, Uppsala, Sweden) and cryopreserved in fetal calf serum and dimethyl
sulfoxide. Cells were thawed and stained with appropriate antibodies and analysed on a
FACSCalibur flow cytometer (Becton Dickinson Immunocytometry Systems, MA,
USA). Antibodies were sourced from Biolegend, California, USA (CD3, CD16 and
CD56) or BD Biosciences Pharmingen, California, USA (CD4 and CD8). Results were
analysed using Flow Jo software (Tree Star, Oregon, USA). T cells were defined as
CD3+ lymphocytes and natural killer cells were defined as CD3-CD16+CD56+
lymphocytes.
Tryptophan and kynurenine measurements
Plasma tryptophan and kynurenine concentrations were measured by High Pressure
Liquid Chromatography (HPLC; Shimadzu, Kyoto, Japan) with UV (250 nm) and
fluorescence (excitation 250 nm, emission 395 nm) detection, using a method modified
from van Wandelen and Cohen (van Wandelen and Cohen 1997). The kynurenine to
141
tryptophan (KT) ratio was calculated by dividing the kynurenine concentration (µmol/L)
by the tryptophan concentration (µmol/L) and multiplying the quotient by 1000
(Huengsberg, Winer et al. 1998; Zangerle, Widner et al. 2002; Pellegrin, Neurauter et al.
2005).
Plasma cytokine measurements
Concentrations of plasma IFN-γ, IL6 and IL10 were determined using a cytometric bead
array (Human Th1/Th2 Cytokine Kit II, BD Biosciences Pharmingen, CA, USA) and a
FACSCalibur flow cytometer (Becton Dickinson Immunocytometry Systems, MA,
USA). Results were analysed using FCAP array version 1.0.1 (Soft Flow Hungary for
Becton Dickinson Biosciences). The lower limits of detection (LLD) of the assay were
2.5 pg/mL for IFN-γ and 10 pg/mL for IL6 and IL10. Values below the LLD were
assigned a value halfway between zero and the LLD for statistical analysis. Cytokines
were only measured if plasma had been frozen within 2 hours of collection.
Measurement of endothelial function
Sepsis patients underwent serial bedside reactive hyperemia peripheral arterial
tonometry (RH-PAT) measurements at enrolment, day 2 - 4, and day 7 (Davis, Yeo et al.
2009). Control patients had the same assessment at a single time point. RH-PAT (Itamar
Medical, Caesarea, Israel) is a non-invasive operator-independent method of assessing
endothelial function. Endothelial function is defined by the ability of blood vessels to
vasodilate in response to an ischemic stress, which invasive studies have demonstrated
to be dependent on endothelial cell NO production (Deanfield, Halcox et al. 2007). RH-
PATis at least 50% NO-dependent (Kuvin, Mammen et al. 2007). RH-PAT uses finger
142
probes to measure digital pulse wave amplitude detected by a pressure transducer
(Celermajer 2008), and has been validated against the more operator-dependent flow-
mediated dilatation method (Kuvin, Patel et al. 2003) and with endothelial function in
other vascular beds (Bonetti, Pumper et al. 2004).
Statistical methods
Predefined groups for analysis were severe sepsis, non-severe sepsis (meaning sepsis
without evidence of organ dysfunction or shock at enrolment), and hospital controls.
Continuous parametric variables were compared using Student’s t-test or ANOVA while
continuous non-parametric variables were compared using Mann-Whitney, Kruskal-
Wallis or Wilcoxon tests as appropriate. Correlations were examined using Pearson’s or
Spearman’s tests for parametric and non-parametric data respectively. As SOFA score
was highly right-skewed and no transformation gave a normal distribution, Kendall’s tau
coefficient for partial correlation was used for multivariate analysis involving SOFA
(Gibbons and Chakraborti 2003). Linear mixed-effects models were used to examine
longitudinal correlations. A 2-sided p-value of <0.05 was considered significant.
Analyses were performed using Stata version 10.0 (Stata Corp TX, USA) and Prism
version 5.01 (GraphPad Software, CA, USA).
143
Results
Patients
The study included 50 patients with severe sepsis, 30 with non-severe sepsis and 40
hospital controls. The three groups did not differ significantly in age or gender (Table
9.1). Ninety percent of severe sepsis patients and all non-severe sepsis patients were
either orally or enterally fed at the time of enrolment; none were receiving parenteral
nutrition.
IDO activity and sepsis severity
Plasma tryptophan concentrations were significantly reduced in patients with sepsis (p <
0.0001, Figure 9.1 and Table 9.2). In all sepsis patients, plasma tryptophan was
inversely related to SOFA score (r = -0.45, p < 0.0001). There was no difference in the
baseline plasma tryptophan concentrations among severe sepsis patients who were orally
fed (n = 29), enterally fed (n = 16) or who were nil by mouth (n = 5). Conversely,
plasma kynurenine concentrations were elevated in sepsis patients compared to hospital
controls (p < 0.0001, Figure 9.1 and Table 9.2), and correlated with SOFA score (r =
0.34, p = 0.005). As kynurenine is renally excreted and accumulates in renal failure,
(Pawlak, Tankiewicz et al. 2003; Schefold, Zeden et al. 2009) kynurenine concentrations
were tested for relationships with renal impairment. Kynurenine concentrations were
significantly higher in patients requiring continuous renal replacement therapy (CRRT)
(median 4.5 µmol [IQR 4-5.3]) than in patients not receiving CRRT (2.8 µmol [2.1-4.4];
p = 0.03). In all sepsis patients, kynurenine concentration correlated with plasma
creatinine (r = 0.41, p = 0.0002). Nevertheless, the association between plasma
144
kynurenine concentration and SOFA score remained significant even after controlling
for creatinine (ktau = 0.24, p < 0.01).
Table 9.1 Baseline clinical characteristics of participants
Severe sepsis Non-severe
sepsis
Controls p value*
Subjects (n) 50 30 40
Age† 52 (48-57) 50 (46-55) 48 (44-52) NS
Male – n (%) 29 (58%) 20 (67%) 27 (68%) NS
Diabetic – n (%) 16 (32%) 7 (23%) 13 (33%) NS
Mean Arterial Pressure‡ 74 (70-82)
n=50
88 (77-104)
n=30
80 (73-93)
n=37
0.001
Systolic Blood Pressure‡ 113 (105-132)
n=49
123 (110-140)
n=24
115 (110-128)
n=37
NS
Diastolic Blood Pressure‡ 60 (54-68)
n=49
70 (60-90)
n=24
60 (60-75)
n=37
0.002
APACHE II ‡ 19 (15-23) 7 (5-12) <0.0001
SOFA score (day 0)‡ 6 (3-9) 1 (0-2) <0.0001
RH-PAT index† 1.59
(1.45-1.73)
n=45
1.86
(1.67-2.05)
n=26
2.04
(1.91-2.18)
n=36
<0.0001
Causative Organism – n (%)
None Cultured 23 (46%) 20 (67%)
Gram Positive Bacterium 14 (28%) 4 (13%)
Gram Negative Bacterium 13 (26%) 6 (20%)
Nutrition – n (%)
Oral feeding 29 (58%) 29 (97%)
Enteral feeding 16 (32%) 1 (3%)
Nil By Mouth 5 (10%)
*For difference between all 3 groups by one way analysis of variance
† Mean (95% confidence interval)
‡Median (interquartile range)
145
Table 9.2 Immunological characteristics of participants
Severe sepsis
50
Non-severe sepsis
30
Combined sepsis
80
Controls
40
Sepsis vs
Control*
Plasma tryptophan µmol/L 21 (13 - 29) 31 (23 - 37) 24 (14 – 35) 49 (40 - 55) <0.0001
Plasma kynurenine µmol/L 3.5 (2.4 - 5.2) 2.3 (1.9 - 3.9) 3.1 (2.1 – 4.7) 1.9 (1.5 - 2.3) <0.0001
KT ratio 162 (100 - 286) 82 (55 - 159) 141 (64 – 235) 36 (28 - 52) <0.0001
Plasma IFN-γ pg/mL 8 (1.3 - 20.1) n = 38 27 (3 - 84) n = 29 9 (3 – 48) n = 67 1.3 (1.3 - 7) n = 37 <0.0001
Plasma IL6 pg/mL 380 (121 - 979) n = 38 136 (44 - 320) n = 29 222 (75 – 596) n = 67 5 (5-5) n = 37 <0.0001
Plasma IL10 pg/mL 23 (13 - 64) n = 38 5 (4 - 25) n = 29 16 (5 – 41) n = 67 5 (5 - 5) n = 37 <0.0001
Neutrophils x103/µL 13.5 (8.7 - 20.4) n = 49 14.1 (9.2 - 16.3) 14 (8.8 – 16.6) n = 79 5.1 (3.2 - 6.5) n=20 0.049
Lymphocytes x103/µL 0.9 (0.5 - 1.2) n = 49 1.0 (0.7 - 1.3) 0.9 (0.5 – 1.2) n = 79 2.1 (1.2 - 2.2) n = 20 <0.0001
Lymphocyte subsets n = 11 n = 12 n = 23 n = 4
T cells x103/µL 0.65 (0.34 - 1.8) 0.67 (0.34 - 1.0) 0.65 (0.34 – 1.1) 1.49 (1.0 - 1.7) NS
CD4+ T cells x103/µL 0.35 (0.17 - 0.85) 0.35 (0.17 - 0.59) 0.35 (0.18 – 0.67) 0.89 (0.52 - 1.2) NS
CD8+ T cells x103/µL 0.18 (0.07 - 0.72) 0.16 (0.10 - 0.33) 0.18 (0.1 – 0.34) 0.46 (0.31 - 0.61) NS
NK cells x103/µL 0.07 (0.03-0.12) 0.06 (0.03-0.17) 0.06 (0.03 – 0.11) 0.08 (0.04-0.20) NS
* All sepsis vs controls, Mann Whitney test † severe sepsis = 11, non-severe sepsis = 12, controls = 4
146
IDO activity was significantly increased in sepsis patients (median KT ratio 141 [IQR
64-235]) compared to controls (36 [28-52]) (p < 0.0001) and in severe sepsis compared
to non-severe sepsis (p = 0.0006, Table 9.2). The baseline KT ratio correlated with
APACHE II (rs = 0.51, p < 0.0001) and total SOFA scores (rs = 0.54, p < 0.0001) in
sepsis patients. The KT ratio positively correlated with the hepatic (rs = 0.28, p = 0.01),
renal (rs = 0.53, p < 0.0001), cardiovascular (rs = 0.42, p < 0.0001) and respiratory (rs =
0.36, p = 0.0009) components of the SOFA score but not the coagulation component (rs
= 0.13, p = ns). The baseline KT ratio was higher in non-survivors than survivors
(median 270 [IQR 102 - 431] versus 138 [63-232]) but this difference was not
statistically significant.
In longitudinal analysis of severe sepsis, the KT ratio significantly decreased between
day 0 (median 162 [IQR 100-286]) and day 7 (89 [65-139]), p = 0.0006); Figure 9.1 D.
Among all sepsis patients, decrease in KT ratio correlated with decrease in SOFA score
over time (p < 0.0001).
IDO activity and plasma cytokines
Plasma IFN-γ, IL6 and IL10 were all significantly increased in patients with sepsis
(Table 9.2). Plasma concentrations of interluekin-1, interleukin-2, interleukin-4 and
tumour necrosis factor-α were not significantly increased in this cohort and were not
analysed further. Both IL6 and IL10 positively correlated with SOFA score (rs = 0.55, p
< 0.0001 and rs =0.55, p < 0.0001 respectively) but there was no association between
IFN-γ and SOFA score.
147
Figure 9.1 Plasma assessment of tryptophan catabolism. The concentration of plasma tryptophan (Fig 1A), kynurenine (Fig 1B) and the KT ratio (Fig 1C) in 50 severe sepsis patients, 30 non-severe sepsis patients and 40 hospital controls. Fig 1D shows the KT ratio in severe sepsis patients on admission (n=50), day 2 (n=34) and day 7 (n=16). The KT ratio is determined by dividing the plasma kynurenine concentration (µmol/L) by the plasma tryptophan concentration (µmol/L) and multiplying the quotient by 1000. Horizontal lines represent median values for the group. P value analysis in Figs 1A-C used a Mann Whitney test, and in Fig 1D, a paired Wilcoxon test.
In sepsis patients, the KT ratio correlated with plasma IFN-γ (rs = 0.44, p = 0.0002), IL6
(r s= 0.49, p < 0.0001) and IL10 (rs = 0.62, p < 0.0001). The associations between KT
ratio and IL6 and IL10 remained significant after controlling for SOFA score (ktau =
0.30, p < 0.003 and ktau = 0.45, p < 0.0001 respectively).
Severe Non-severe Control0
20
40
60
80
100
p = 0.002
p < 0.0001
p < 0.0001P
lasm
a tr
ypto
ph
an (
µµ µµ mo
l/L)
Severe Non-severe Control0
5
10
15
20
25 p < 0.0001
p = 0.01
p = 0.02
Pla
sma
kyn
ure
nin
e ( µµ µµ
mo
l/L)
Severe Non-severe Control0
500
1000
1500
p = 0.0006
p < 0.0001
p < 0.0001
KT
rat
io
Day 0 Day 2 - 4 Day 70
500
1000
1500p=0.0006
p=0.02
p=0.04KT
rat
io
A B
C D
148
In a univariate mixed effects model, the decrease in KT ratio over time correlated with
the decrease in IL6 (p < 0.0001) and IL10 (p < 0.0001) between day 0 and day 7. In a
multivariate model, these relationships remained significant after controlling for change
in SOFA score (IL6 p = 0.009; IL10 p = 0.02).
IDO activity and lymphocyte counts
Sepsis patients had significantly higher total white blood cell (p < 0.0001) and
neutrophil (p < 0.05) counts than hospital controls (Table 9.2). Conversely, sepsis
patients had significantly lower total lymphocyte counts compared with hospital controls
(p < 0.0001, Table 9.2). In all sepsis patients the baseline KT ratio was weakly
associated with absolute lymphocyte count (rp = 0.26, p = 0.02). In a linear mixed
effects model, absolute lymphocyte count increased as the KT ratio decreased over time
(p = 0.001). This relationship persisted after controlling for SOFA score (p = 0.008).
When all subjects were grouped according to lymphopenia, lymphopenic patients (n =
63) had a median KT ratio of 128 [IQR 63-236], compared with 59 [33-86] in non-
lymphopenic patients (n = 57) (p < 0.0001).
As IDO activity contributes to T cell apoptosis (Fallarino, Grohmann et al. 2002), we
examined the relationship between KT ratio and lymphocyte subsets. Peripheral blood
mononuclear cells were analysed from 23 of the 80 sepsis patients whose blood had
been processed within 30 minutes of collection. This subset of patients was
representative of the cohort in terms of age, gender distribution, total lymphocyte count
and KT ratio. In this subset of patients, the KT ratio negatively correlated with absolute
numbers of lymphocytes (rp = -0.54, p = 0.007), T cells (rp =- 0.53, p = 0.01), CD4+ T
149
cells (rp = -0.50, p = 0.01), CD8+ T cells (rp= -0.49, p = 0.02) and natural killer cells (rp=
-0.46, p = 0.03) (Table 9.2).
IDO activity and endothelial function
In sepsis, the KT ratio at baseline correlated inversely with NO-dependent microvascular
reactivity (rs = -0.45, p = 0.001) even after controlling for disease severity (using SOFA
score; p = 0.001). In a multivariate mixed effects model controlling for SOFA score,
improvement in KT ratio between day 0 and day 7 correlated with improvement in
microvascular reactivity (p = 0.001). In all sepsis patients, there was an inverse
association between the baseline KT ratio and mean arterial pressure (rs = -0.29, p =
0.009) and diastolic blood pressure (rs = -0.29, p = 0.01) but no association with systolic
blood pressure.
Discussion
IDO activity is increased in sepsis, in proportion to disease severity. IDO-mediated
tryptophan catabolism is associated with dysregulated immune responses and impaired
microvascular reactivity in sepsis. IFN-γ and IL10 are associated with, and may
contribute to, increased IDO activity in sepsis. The independent inverse longitudinal
association with total lymphocyte counts suggests a potential role in sepsis-associated
lymphopenia. Similarly, the independent inverse association between the KT ratio and
NO-dependent microvascular reactivity suggests that IDO activity may also contribute
to impaired endothelial function in sepsis. Based on these associations we propose a
model of interpretation outlined in Figure 9.2.
150
IncreasedIFN-γ +/- IL10
Increased KT ratioin plasma
Stabilised IL6 mRNA andincreased plasma IL6
Increased IDO activity
Decreased microvascular
reactivity
Decreased endothelialnitric oxide
Increased lymphocyteapoptosis
Decreased plasma tryptophan and increased plasma kynurenine
Figure 9.2 Proposed model of tryptophan catabolism in sepsis IDO = Indoleamine 2,3-dioxygenase, IL6 = interleukin-6, IL10 = interleukin-10, IFN-γ = interferon gamma and NO = nitric oxide.
Increased expression of IFN-γ (Yoshida, Imanishi et al. 1981), IL6 (Maes, Meltzer et al.
1993; Bonaccorso, Lin et al. 1998) and IL10 (Munn, Sharma et al. 2002) have each been
associated with increased tryptophan catabolism by IDO in other disease states. In
sepsis patients in our study, IFN-γ concentration correlated with the KT ratio only at
baseline, whereas IL6 and IL10 correlated with the KT ratio both at baseline and
longitudinally. Our findings agree with the in vitro literature, where IFN-γ induces IDO
(Yoshida, Imanishi et al. 1981; Carlin, Borden et al. 1989). Although under certain
151
conditions, IL-10 has been reported to suppress IDO activity (MacKenzie, Gonzalez et
al. 1999), our findings support the majority of in vitro studies which have shown that IL-
10 induces or stabilises IDO (Munn, Sharma et al. 2002; van der Sluijs, Nijhuis et al.
2006; Maneechotesuwan, Supawita et al. 2008; Yanagawa, Iwabuchi et al. 2009). The
high IFN-γ associated with early sepsis (Hunsicker, Kullich et al. 1997) may lead to
increased IDO activity while high IL10 may sustain or potentially enhance IDO activity
(Yanagawa, Iwabuchi et al. 2009) throughout the course of the disease. The role of IL6
in IDO expression is unclear. Orabona et al suggest that IL6 inhibits IDO activity by
increasing murine dendritic cell SOCS3 expression, which drives IDO
breakdown(Orabona, Pallotta et al. 2008). On the other hand, a low tryptophan
environment created by IDO activity stabilises IL6 mRNA and increases IL6
responses(van Wissen, Snoek et al. 2002). Given the conflicting evidence in these and
other studies regarding IL6 and IDO, we investigated the relationship between the KT
ratio and IL6 in sepsis patients. The strong positive correlation between plasma KT ratio
and IL6 concentration is consistent with findings in murine models of sepsis where IDO-
/- mice or mice treated with IDO inhibitors have lower levels of plasma IL6
concentrations (Ulloa, Ochani et al. 2002; Jung, Lee et al. 2009).
We report that the high KT ratio in sepsis is associated with a decreased lymphocyte
count, independent of disease severity, a finding similar to that found in patients with
trauma (Pellegrin, Neurauter et al. 2005), human immunodeficiency virus (Huengsberg,
Winer et al. 1998) and cancer (Ino, Yamamoto et al. 2008). Previous studies in sepsis
have associated lymphopenia with disease severity (Le Tulzo, Pangault et al. 2002),
duration of ICU stay (Le Tulzo, Pangault et al. 2002) and mortality (Felmet, Hall et al.
152
2005) and prevention of lymphocyte apoptosis improves survival in animal models of
sepsis (Hotchkiss, Tinsley et al. 1999; Hotchkiss, Chang et al. 2000; Bommhardt, Chang
et al. 2004; Schwulst, Muenzer et al. 2008). T cells co-cultured with IDO-producing
cells have reduced proliferation and increased death (Fallarino, Vacca et al. 2002;
Odemuyiwa, Ghahary et al. 2004). Both high kynurenine concentrations and low
tryptophan concentrations appear to contribute to T cell death. In vivo, kynurenine
treatment in mice depletes overall thymocyte counts and, in vitro, thymocytes die of
apoptosis when cultured in media with kynurenines (Fallarino, Grohmann et al. 2002).
Furthermore, T cells cultured in low tryptophan media have reduced proliferation and
increased apoptosis via activated GCN2 kinase (Lee, Park et al. 2002; Forouzandeh,
Jalili et al. 2008). These in vitro studies suggest a potential mechanism through which
increased IDO activity may contribute to lymphopenia and its deleterious consequences
in sepsis.
IDO activity regulates vascular tone in sepsis. In this study IDO activity in sepsis
patients correlated with diastolic blood pressure but not systolic blood pressure. This is
in keeping with the recent finding that kynurenine is a vascular relaxation factor (Wang,
Liu et al. 2010). Another important regulator of endothelial function in sepsis is NO.
There is significant cross-talk between IDO and NOS, with IDO activity inhibiting both
expression and activity of NOS (Sekkai, Guittet et al. 1997; Chiarugi, Rovida et al.
2003; Samelson-Jones and Yeh 2006) and vice versa. We found the KT ratio in sepsis is
inversely associated with microvascular reactivity as measured by RH-PAT, which is at
least 50% dependent on endothelial NO production (Nohria, Gerhard-Herman et al.
2006). Increased IDO activity in sepsis may regulate vascular tone directly, via the
153
vasorelaxing effects of kynurenine, and indirectly, by impairing NO-dependent
microvascular reactivity. Increased plasma kynurenine concentrations may further
impede endothelial function in sepsis by mediating adhesion of monocytes and
neutrophils to the vascular endothelium (Barth, Ahluwalia et al. 2009).
A limitation of this study is that we did not directly measure IDO expression.
Alternative enzymes which can affect the KT ratio include tryptophan-2,3-pyrrolase
(TDO) and IDO-2 - however neither of these enzymes increase in response to
inflammation (Takikawa, Yoshida et al. 1986; Ball, Sanchez-Perez et al. 2007). The KT
ratio is an established measure of systemic IDO activity (Huengsberg, Winer et al. 1998;
Suzuki, Suda et al. 2010) with tissue IDO expression and activity directly correlated
with plasma KT ratio in multiple human disease states, including celiac disease (Torres,
Lopez-Casado et al. 2007), hepatitis C (Larrea, Riezu-Boj et al. 2007) and pre-eclampsia
(Kudo, Boyd et al. 2003). There are several possible sources of IDO activity in sepsis
patients including the endothelium, kidney, liver, lungs and leukocytes (Carlin, Borden
et al. 1989; Larrea, Riezu-Boj et al. 2007; Mohib, Guan et al. 2007; Iwamoto, Ito et al.
2009; Yanagawa, Iwabuchi et al. 2009; Wang, Liu et al. 2010), although a recent study
was unable to detect spontaneous IDO expression in PBMC from sepsis patients
(Tattevin, Monnier et al. 2010). Importantly, the effects of the high KT ratio in sepsis on
immune function and endothelial function would be the same whether the high KT ratio
was the result of increased IDO activity alone or in combination with decreased feeding
and impaired renal excretion of kynurenine. Furthermore, it is unlikely that nutritional
deficiency and renal impairment accounted for the differences we found, because
controlling for these factors made no difference to our results.
154
The generation of a low tryptophan environment may be a maladaptive host response to
infection. In murine models of sepsis, IDO-/- mice have significantly increased survival
compared to wild type mice (Jung, Lee et al. 2009) and treatment of wild-type mice with
IDO inhibitors such as 1-methyl-tryptophan (Jung, Lee et al. 2009) or ethyl pyruvate
also significantly increase survival (Ulloa, Ochani et al. 2002). While growth of some
bacterial species is inhibited by low tryptophan (MacKenzie, Hadding et al. 1998), most
can synthesize tryptophan (Merino, Jensen et al. 2008) and others have specialized
tryptophan transport systems (Yanofsky, Horn et al. 1991). The KT ratio is significantly
higher in bacteremic patients with a fatal outcome (Huttunen, Syrjanen et al. 2009) and
we and others have demonstrated that the KT ratio is associated with disease severity in
sepsis (Huttunen, Syrjanen et al. 2009; Schefold, Zeden et al. 2010; Tattevin, Monnier et
al. 2010). Together, this evidence supports the hypothesis that increased IDO activity is
a deleterious host response in human sepsis. IDO inhibitors are being considered as
potential adjunctive cancer treatments (Lob, Konigsrainer et al. 2009) and these
treatments may also have therapeutic potential in sepsis.
9.3. Conclusion
IDO activity is elevated in sepsis and associated with disease severity, T cell
lymphopenia and microvascular dysfunction. Because excessive IDO activity is
associated with both immune and endothelial dysfunction, the increased tryptophan
catabolism we have described may link these two key aspects of sepsis pathophysiology.
Modulation of IDO activity warrants investigation as a therapeutic strategy in sepsis.
155
10. Results: Inflammation and T cell suppression in sepsis
10.1. Introduction
The two previous chapters have demonstrated that sepsis patients have increased
arginine and tryptophan metabolites, in proportion to disease severity. Both a decreased
arg/ADMA ratio and increased KT ratio were associated with endothelial dysfunction
and increased plasma cytokine concentrations.
As outlined in chapter 5, the inflammatory environment that is associated with disturbed
amino acid metabolism is also associated with the accumulation of MDSC. In humans,
the unusual density of MDSC results in them co-eluting with PBMC in the Ficoll-
Paque™ layer (Schmielau and Finn 2001). As it had been previously reported that
PBMC from sepsis patients were ‘contaminated’ with granulocytes (van den Akker,
Baan et al. 2008), we hypothesised that those granulocytes would be MDSC.
MDSC suppress T cell function by impairing the expression of the T cell receptor zeta-
chain (Ezernitchi, Vaknin et al. 2006). Similarly, low concentrations of arginine and
tryptophan impair T cell zeta chain expression (Zea, Rodriguez et al. 2004; Fallarino,
Grohmann et al. 2006). Therefore, we hypothesised that sepsis patients would have
impaired T cell zeta-chain expression and that zeta-chain expression would be related to
MDSC and/or plasma arginine and tryptophan concentrations. Section 10.2 outlines the
relationship between T cell zeta-chain expression and plasma arginine and tryptophan
concentrations in sepsis. The draft manuscript in section 10.3 reports that sepsis patients
156
do have impaired T cell receptor zeta-chain expression and that expression is related to
the percentage of the MDSC co-eluting with the PBMC.
10.2. Arginine, tryptophan and T cell suppression in
sepsis
As sepsis patients have low plasma concentrations of arginine and tryptophan, we
hypothesised that T cells from sepsis patients would have low zeta-chain expression.
We found that sepsis patients do have low T cell zeta-chain expression but that it is not
associated with plasma concentrations of arginine or tryptophan (Figure 10.1 a and b). In
the immunology subset of sepsis patients, T cell zeta-chain expression recovered
between day 0 and day 2, however plasma arginine and tryptophan concentrations did
not significantly improve over the same time (Figure 10.2) and there was no longitudinal
association between recovery of T cell zeta-chain expression and plasma arginine or
tryptophan. We found that unstimulated T cells from sepsis patients significantly
improved T cell zeta-chain expression when cultured in media with physiological
concentrations of amino acids (Figure 10.3a), whereas zeta-chain expression on T cells
from hospital controls stayed the same (data not shown). Furthermore, T cells from
sepsis patients recovered zeta-chain expression in culture, regardless of whether arginine
was present (Figure 10.3b).
157
(a) (b) (c)
Figure 10.1 Ex vivo T cell zeta-chain expression in sepsis patients compared to controls (a) and the association of T cell zeta-chain expression with plasma concentrations of arginine (b) and tryptophan (c) in sepsis patients.
(a) (b) (c)
Figure 10.2 Change in T cell zeta-chain expression (a), plasma arginine concentration (b) and plasma tryptophan concentration (c) between day 0 and day 2 of the study. P values determined by a paired T test.
158
(a) (b) Figure 10.3 Recovery of T cell zeta-chain expression in unstimulated cells in media with physiological concentrations of amino acids (a) and comparison of recovery with or without arginine (b)
As we were performing these experiments, we noted that, in shock patients, zeta-chain
expression tended to be lower when granulocytes were present in the PBMC.
Granulocytes co-eluting with PBMC from sepsis patients had previously been noted
(van den Akker, Baan et al. 2008) but the significance of this had not been investigated.
In cancer patients, granulocytes co-eluting with PBMC have been found to suppress T
cells and are called myeloid derived suppressor cells. This lead to a new hypothesis, that
T cell zeta-chain expression was suppressed by MDSC.
159
10.3. Myeloid derived suppressor cells in sepsis
10.3.1. Draft manuscript: Myeloid derived suppressor cells impair T
cell signalling in septic shock patients
Authors: C.J. Darcy1, K. A. Piera1, G. Minigo1, J.S. Davis1, 2, Y. R. McNeil1, J. B.
Weinberg3, N. M. Anstey1, 2, T. Woodberry1
Authors’ affiliations: 1 – Global Health Division, Menzies School of Health Research
and Charles Darwin University, Darwin, NT 0810, Australia. 2 – Division of Medicine,
Royal Darwin Hospital, Darwin, NT, 0810, Australia. 3 –Division of Hematology-
Oncology, Duke University and Veterans’ Affairs Medical Centers, Durham, NC 27710
USA.
Abstract
Septic shock is a systemic inflammatory response to an infection with hypotension and
organ failure. Impaired T cell function in septic shock is associated with poor outcome,
but the mechanism of this dysfunction is not well understood. Myeloid derived
suppressor cells (MDSC) are myeloid derived cells that can suppress T cell function. In
a longitudinal case-control study of sepsis, MDSC were increased in septic shock and
associated with plasma interleukin-6 concentrations. MDSC from sepsis patients were
mature granulocytes which co-eluted with the peripheral blood mononuclear cells during
density gradient separation and phenotypically and functionally different to
polymorphonuclear neutrophils. There was an ex vivo association between the
160
percentage of MDSC and T cell zeta-chain expression in septic shock patients both at
baseline and longitudinally. In vitro depletion of MDSC restored T cell-zeta chain
expression and capacity for T cell proliferation. We describe a population of circulating
MDSC in septic shock patients which impair T cell receptor zeta-chain expression and T
cell function. The identification of MDSC in septic shock patients links inflammation
and T cell dysfunction in sepsis.
Introduction
Sepsis is a systemic inflammatory response to infection (Bone, Balk et al. 1992).
Despite improvements in its management, septic shock still has a mortality rate of 30-
50% (Angus, Linde-Zwirble et al. 2001; Finfer, Bellomo et al. 2004; Blanco, Muriel-
Bombin et al. 2008) and is a leading cause of death in intensive care units (Angus,
Linde-Zwirble et al. 2001).
Although sepsis patients have high concentrations of inflammatory mediators,
components of their immune system are suppressed (Lyn-Kew and Standiford 2008;
Hotchkiss, Coopersmith et al. 2009). Sepsis patients have widespread apoptosis of
lymphocytes leading to lymphopenia (Hotchkiss, Swanson et al. 1999). In vivo evidence
of T cell dysfunction in sepsis is demonstrated by cytomegalovirus and herpes simplex
virus re-activation (Kutza, Muhl et al. 1998; von Muller, Klemm et al. 2006) and
impaired delayed type hypersensitivity (MacLean, Meakins et al. 1975). This is
confirmed ex vivo by impaired T cell proliferation and cytokine production in response
to stimulation (Heidecke, Hensler et al. 1999). Viral reactivation (Limaye, Kirby et al.
2008), lymphocyte apoptosis (Le Tulzo, Pangault et al. 2002) and impaired T cell
161
function (Heidecke, Hensler et al. 1999) are all associated with increased mortality in
critically ill or septic patients. However, the mechanism of T cell suppression in sepsis
is not well understood.
A possible link between inflammation and T cell dysfunction are myeloid derived
suppressor cells (MDSC). MDSC are a heterogeneous group of myeloid derived cells
that can suppress T cell function. MDSC are induced or activated by multiple pro-
inflammatory mediators including interleukin-1ß, interleukin-6 and vascular endothelial
growth factor (Ostrand-Rosenberg and Sinha 2009). MDSC suppress T cell activation
and proliferation (Mazzoni, Bronte et al. 2002; Zea, Rodriguez et al. 2005; Movahedi,
Guilliams et al. 2008). Human MDSC have been reported to co-elute with peripheral
blood mononuclear cells (PBMC) during density gradient separation (Schmielau and
Finn 2001; Zea, Rodriguez et al. 2005).
One mechanism that MDSC use to suppress T cells is by down-regulating T cell
receptor zeta-chain expression (Ezernitchi, Vaknin et al. 2006). The zeta-chain is the
principal signal transduction component of the T cell receptor. When zeta-chain
expression is impaired, T cells proliferate less and produce less cytokines in response to
stimulation. Low T cell zeta-chain expression and T cell dysfunction has been reported
in many human cancers, autoimmune disease, HIV and leprosy, as reviewed in
(Baniyash 2004).
162
MDSC have been described in mouse models of sepsis but have not yet been
investigated in human sepsis (Delano, Scumpia et al. 2007; Sander, Sackett et al. 2010).
As sepsis patients have high concentrations of inflammatory mediators, we hypothesised
that the granulocytes co-eluting with PBMC (van den Akker, Baan et al. 2008) would be
MDSC. Furthermore, as T cell dysfunction is well-described in sepsis, we hypothesised
that the MDSC would suppress T cells by down-regulating zeta-chain expression. Here
we report that septic shock patients have significantly more MDSC compared to controls
and sepsis patients without shock. Furthermore, these MDSC impair T cell zeta-chain
expression in septic shock patients.
Materials and methods
Participants
We evaluated 24 patients with sepsis and 12 hospital controls who were part of a
previously reported study of endothelial function in sepsis (Davis, Yeo et al. 2009). The
subset of patients had blood processed within 30 minutes of collection and were
representative of the entire cohort in terms of age, gender, ethnicity and disease severity.
Sepsis patients had suspected or proven infection and the presence of two or more
criteria for the systemic inflammatory response syndrome (SIRS) within the last 4 hours
(Bone, Balk et al. 1992). Septic patients were classified as septic shock, or sepsis
without shock. Septic shock was defined at the time of enrolment as systolic blood
pressure <90mmHg or a reduction of ≥ 40mmHg from baseline despite adequate fluid
resuscitation, or the need for vasopressors to maintain these targets (Bone, Balk et al.
1992). Sepsis severity was estimated using the Acute Physiology and Chronic Health
Evaluation (APACHE) II score from the first 24 hours of admission and daily modified
163
Sequential Organ Failure Assessment (SOFA) score (Vincent, de Mendonca et al. 1998).
Patients were enrolled within 24 hours of ICU admission or within 36 hours of ward
admission. Control subjects were recruited from hospital patients who had not met SIRS
criteria within the last 30 days and who had no clinical or laboratory evidence of
inflammation or infection. Written informed consent was obtained from all participants
or next of kin. The study was approved by the Human Research Ethics Committee of
Menzies School of Health Research and the Department of Health and Community
Services.
Blood collection and lymphocyte counts
Venous blood was collected in lithium heparin tubes at enrolment, day 2 - 4, and day 7
until discharge from the hospital or death. Whole blood differential white cell counts
were measured by Coulter Counter. Lymphopenia was defined as an absolute
lymphocyte count less than 1.2 x109/µL (Hotchkiss, Swanson et al. 1999). Plasma was
separated within 30 minutes of collection and stored at -80 °C. Peripheral blood
mononuclear cells were separated within 2 hours by density gradient using Ficoll-
Hypaque™ Plus (GE Healthcare Biosciences, Uppsala, Sweden) and either stained fresh
or cryopreserved in liquid nitrogen in fetal calf serum and dimethyl sulfoxide.
Evaluation of cell phenotype
All thawed, cryopreserved samples from each patient were analysed simultaneously. In
order to estimate the percentage of granulocytes from cryopreserved sepsis PBMC, cells
were thawed in media with 50 units/mL benzonase nuclease to reduce cell clumping and
stained immediately after thawing. We obtained reasonable estimates of the percentage
164
of granulocytes in PBMC in 23 out of 24 sepsis patients and all control patients.
Freshly isolated cells from 5 additional patients were used to confirm quantification of
granulocytes in fresh PBMC versus thawed and to undertake detailed staining profiles
and functional analysis. Freshly processed blood was also used to collect PMN from
underneath the Ficoll-Hypaque™ Plus layer during density gradient separation.
Antibodies were sourced from Biolegend, California, USA (CD3, CD16, CD56, CD11b,
CD15, CD33), BD Biosciences Pharmingen, California, USA (CD4, CD8, CD66b,
CD14) or eBioscience (CD115). Matched isotype controls were used. T cells were
stained for intra-cellular zeta-chain expression (Beckman Coulter, Immunotech) after
surface staining and permeabilising with digitonin using CD3 zeta-chain antibody.
Results were read on a FACSCalibur flow cytometer (Becton Dickinson
Immunocytometry Systems, MA, USA) and analysed using Flow Jo software (Tree Star,
Oregon, USA).
Zeta-chain mean fluorescence intensity was normalized to an internal control. The
internal control consisted of multiple aliquots of PBMC from a blood bank donor,
cryopreserved from a single donation. A standardized zeta-chain value for the internal
control was established by calculating the mean fluorescence of three aliquots, thawed
and stained in separate experiments using a single vial of zeta-chain antibody. A single
blood bank aliquot was then thawed for each experiment and used as an internal control
for zeta-chain fluorescence. The zeta-chain mean fluorescence of the internal control in
each experiment was compared to the standardized value to obtain a normalization
factor. All zeta-chain values in an experiment were then multiplied by the same
165
normalization factor. Un-normalized zeta-chain values were significantly higher each
time a new vial of zeta-chain antibody was opened and reduced over time, even though
care was taken to minimize light exposure and all antibody vials were the same lot
number. The normalization factor proved important to minimize variation caused by
the age of the antibody and counter small discrepancies in incubation times.
Isolation and depletion of granulocytes from PBMC
Granulocytes were isolated from freshly separated sepsis PBMC by labeling with
CD66b+ followed by anti-FITC magnetic bead selection (MACS, Miltenyi Biotech),
according to the manufacturer’s instructions. Proliferation assays were set up with
PBMC from sepsis patients either with ex vivo percentages of MDSC or after MDSC
depletion. Proliferation was determined using carboxyfluorescein diacetate succinimidyl
ester (CFSE, Invitrogen)-labelled PBMC stimulated with immobilized anti-CD3
(Biolegend) and anti-CD28 (Biolegend). In some cultures the arginase inhibitor nor-
NOHA was added at 50ug/mL.
Plasma arginine and arginase activity
Plasma arginine concentrations were measured by High Pressure Liquid
Chromatography (HPLC; Shimadzu, Kyoto, Japan) with UV (250 nm) and fluorescence
(excitation 250 nm, emission 395 nm) detection, using a method modified from van
Wandelen and Cohen (van Wandelen and Cohen 1997). Plasma arginase activity was
measured using a radiometric assay, as previously described, and reported as
micromole/milliliter/hour (Morris, Kato et al. 2005).
166
Plasma cytokine measurements
Concentrations of plasma IFN-γ, IL6 and IL10 were determined using a cytometric bead
array (Human Th1/Th2 Cytokine Kit II, BD Biosciences Pharmingen, CA, USA) and a
FACSCalibur flow cytometer (Becton Dickinson Immunocytometry Systems, MA,
USA). Results were analysed using FCAP array version 1.0.1 (Soft Flow Hungary for
Becton Dickinson Biosciences). The lower limits of detection (LLD) of the assay were
2.5 pg/mL for IFN-γ and 10 pg/mL for IL6 and IL10. Values below the LLD were
assigned a value halfway between zero and the LLD for statistical analysis.
Statistical methods
Groups for analysis were septic shock, sepsis without shock and hospital controls.
Continuous parametric variables were compared using Student’s t-test or ANOVA while
continuous non-parametric variables were compared using Mann-Whitney, Kruskal-
Wallis or Wilcoxon tests as appropriate. Correlations were examined using Pearson’s or
Spearman’s tests for parametric and non-parametric data respectively. Linear mixed-
effects models were used to examine longitudinal correlations. A 2-sided p-value of
<0.05 was considered significant. Analyses were performed using Stata version 10.0
(Stata Corp TX, USA) and Prism version 5.01 (GraphPad Software).
167
Results
Patients
Longitudinal detailed phenotyping or functional analyses were done on 18 samples of
freshly isolated PBMC from five sepsis patients. To compare T cell zeta-chain
expression and MDSC percentages between groups, cryopreserved PBMC were
analysed from an additional 12 patients with septic shock, 12 patients with non-shock
sepsis and 12 hospital controls. The three cryopreserved PBMC groups did not differ
significantly in age or gender (
Table 10.1).
Table 10.1 Patient details for the three cryopreserved PBMC groups. Septic
shock
Sepsis
without shock
Hospital
controls
p value*
Subjects (n) 12 12 12
Age‡ 52 (45 - 57) 45 (39 - 55) 49 (4 0- 56) NS
Male – n (%) 7 (58%) 6 (50%) 8 (67%) NS
APACHE II 20 (29-23) 8 (4-14) <0.0001
SOFA score (day 0) 10 (4 - 10) 1 (0 - 2) <0.0001
IL-6 (pg/mL) 1433 (400 -4290) 82 (42 - 302) 5 (5 - 5) <0.0001
% CD66b+ in PBMC 19.2 (4.4 – 29.5) 2.7 (1.5 – 6.1) 1.5 (0 – 2.0) 0.001
Neutrophil count 13.1 (7.2 – 19.4) 14.2 (11.4 – 16.6) 6 (4.0-9.6)
Imm. granulocyte count 0.4 (0 - 2.6) 0 (0 - 0) 0 (0 - 0)
Monocyte count 0.45 (0.1 – 1.2) 0.65 (0.35 – 1) 0.55 (0.5 – 0.68)
Lymphocyte count 1.2 (0.5 – 2.1) 1.2 (0.8 – 1.6) 2.2 (1.5 – 2.4)
Causative Organism
n (%)
None Cultured 5 (42%) 9 (75%)
Gram Positive Bacterium 4 (33%) 2 (17%)
Gram Negative Bacterium 3 (25%) 1 (8%)
168
Septic shock patients had more granulocytes co-eluting with PBMC than non-shock
patients
Sepsis patients had granulocytes which co-elute with PBMC. These PBMC
granulocytes sat in a similar position in the forward side scatter as PMN (Figure 10.4).
Septic shock patients had more granulocytes co-eluting with PBMC compared to sepsis
patients without shock both at day 0 and day 2 (Figure 10.5 a and b and Table 10.1).
The percentage of granulocytes co-eluting with PBMC in all sepsis patients was related
to plasma IL-6 concentrations (Figure 10.6).
169
(a) (b) (c)
Figure 10.4 Representative forward side scatter plots. PBMC from hospital control patient (a), PBMC from a septic shock patient (b) and polymorphonuclear neutrophils (collected from underneath the ficoll layer) from a sepsis patient (c).
170
a) Day 0 (b) Day 2 - 4
Figure 10.5 Percentage of CD66b+ granulocytes in PBMC from septic shock, sepsis without shock and control patients on day 0 (a) and day 2 – 4 (b) of the study. Symbols indicate the median and the interquartile rangeand p values were calculated with a Mann-Whitney test.
Figure 10.6 The relationship between the baseline percentage of CD66b+ granulocytes in the PBMC and plasma inteluekin-6 in sepsis patients. Correlated with a Spearman’s test.
171
Granulocytes which co-eluted with PBMC from sepsis patients were MDSC
The granulocytes which co-eluted with PBMC during density gradient separation were
phenotypically and functionally different to the PMN from the same patient. The PBMC
granulocytes from sepsis patients were CD66b+CD11b+CD15+CD45RO+,
CD16lowCD14low and negative for CD33, CD115 and HLA-DR (Figure 10.7). The
CD66b+ cells in the PBMC were consistently higher in CD66b and CD15 and lower in
CD16 compared to PMN and monocytes from the same patient. The combination of
CD66b and CD16 had the potential to separate granulocytes which co-eluted with
PBMC from PMN in whole blood from humans (Figure 10.8).
As MDSC are described as myeloid derived cells that suppress T cells, we investigated
whether the CD66b+ granulocytes which co-eluted with PBMC could suppress T cell
proliferation. We compared T cell proliferation between PBMC with CD66b+ cells and
PBMC depleted of CD66b+ cells and found that T cell proliferation was suppressed in
the presence of CD66b+ cells at ex vivo concentrations (Figure 10.9a). This depletion
experiment was repeated on a second patient and PMN were added to match the original
MDSC:T cell ratio. The PMN did not suppress T cell proliferation or T cell zeta-chain
expression to the same extent as MDSC at the same ratio (Figure 10.9b). T cell
proliferation was still suppressed in the presence of CD66b+ cells even after the addition
of an arginase inhibitor (data not shown).
172
Figure 10.7 Staining comparison of CD66b+ granulocytes in PBMC, monocytes and PMN from a single sepsis patient. Grey = isotype control, black = stain. Stains include CD66b, CD11b, CD15, CD14, CD115, CD33, CD16, HLA-DR and CD45RO.
CD66b CD11b CD15 CD14 CD33 HLA-DR CD16 CD115 CD45RO
CD66b+ PBMC
Monocytes
PMN
173
Figure 10.8 Combination staining of CD66b and CD16. Differentiation between CD66b+ granulocytes in PBMC (black) and PMN (red).
(a) (b)
Figure 10.9 Comparison of T cell proliferation with and without CD66b+ cells in two sepsis patients. (a) T cell proliferation in sepsis PBMC with CD66b+ cells and without CD66b+ cells. Grey = unstimulated control, black = sepsis PBMC with ex vivo concentrations of CD66b+ cells (1:1 CD66b+ granulocyte:T cell ratio), orange = sepsis PBMC depleted of CD66b+ cells. (b): Comparison of T cell proliferation with either CD66b+ granulocytes from PBMC or PMN at a 1:2 ratio to T cells. Grey = unstimulated control, black = PBMC with CD66b+ granulocytes, orange = PBMC with PMN.
174
To determine whether the CD66b+ granulocytes in the PBMC could be detected by a
Coulter Counter, we compared the percentage of CD66b+ granulocytes in PBMC to
Coulter counts from 12 septic shock patients. We found that the percentage of CD66b+
granulocytes in PBMC from septic shock patients was associated with the circulating
neutrophil count (r= 0.6, p= 0.04) but not the immature granulocyte or monocyte count.
CD66b+ granulocytes in PBMC impaired T cell zeta-chain expression in sepsis patients
As we hypothesised that the CD66b+ granulocytes in the PBMC from sepsis patients
were MDSC, we investigated whether sepsis patients have impaired T cell zeta-chain
expression. We found both that septic shock and non-shock patients had low T cell zeta-
chain expression at baseline. Furthermore, by day 2 of the study non-shock patients had
recovered T cell zeta-chain expression, whereas septic shock patients had not (Figure
10.10).
(a) Day 0 (b) Day 2
Figure 10.10 T cell zeta-chain expression in septic shock patients, sepsis patients without shock and hospital controls on day 0 (a) and day 2 (b) of the study. Red symbols indicate patients who died.
175
In an additional 3 sepsis patients whose cells were freshly isolated and stained
immediately ex vivo, we investigated the longitudinal association between zeta-chain
expression and the percentage of CD66b+ cells in the PBMC. We found an inverse
longitudinal association between the ex vivo percentage of CD66b+ cells in the PBMC
and T cell zeta-chain expression in sepsis, with zeta expression lower with increasing
%MDSC (Figure 10.11). This longitudinal inverse association was significant in a
mixed effects model (p=0.01).
(a) (b) (c)
Figure 10.11 The longitudinal relationship between T cell zeta-chain expression and percentage of CD66b+ cells in PBMC in 3 individual patients (a, b and c). Green = T cell zeta-chain expression. Red = % of CD66b+ cells co-eluting with PBMC.
In shock patients only, the ex vivo percentage of CD66b+ cells in the PBMC was directly
related to zeta-chain expression both at day 0 and day 2 of the study (Figure 10.12 a and
b). There was a positive trend between the percentage of CD66b+ cells in the PBMC
and plasma arginase activity in shock patients, but this was not significant (Figure 10.12
c).
176
(a) Day 0 (b) Day 2 (c) Day 0
Figure 10.12 Percentage of CD66b+ cells in PBMC, T cell zeta-chain expression and plasma arginase activity. Association between the percentage of CD66b+ cells in PBMC from septic shock patients and T cell zeta-chain expression on day 0 (n = 11) and day 2 - 4 of the study (n = 9) (b). Relationship between the percentage CD66b+ cells in PBMC and plasma arginase activity on day 0 (n = 9).
To confirm these associations, a series of add-back experiments were set up where
sepsis PBMC were depleted of CD66b+ cells, then the CD66b+ cells were added back at
different percentages. In vitro, the percentage of CD66b+ cells in culture correlated with
T cell zeta-chain expression and arginine consumed from media (Figure 10.13).
177
Figure 10.13 Representative association between T cell zeta-chain expression and the percentage of sepsis PBMC CD66b+ cells added back to the CD66b+ depleted cell culture, and arginine used from the supernatant. Green = T cell zeta-chain expression. Brown = Arginine used from the cell supernatant (µM). A control experiment showed no association between the percentage of PMN (separating underneath the ficoll layer) and T cell zeta-chain expression. Discussion
Septic shock patients have increased circulating MDSC which suppress T cell signaling.
Although in a mouse model of sepsis, MDSC appeared to be beneficial to survival
(Sander, Sackett et al. 2010), our results from humans with sepsis suggest that excessive
numbers of circulating MDSC may be harmful. Patients with more circulating MDSC
had more severe disease, higher concentrations of plasma IL-6 and slower recovery of T
cell zeta-chain expression. Similarly, increased circulating MDSC in cancer patients is
associated with more aggressive disease (Diaz-Montero, Salem et al. 2009). The slower
recovery of T cell zeta-chain expression in septic shock patients is consistent with
reports that patients with more severe sepsis have T cell dysfunction for longer. The
slower recovery of T cell function puts these patients at risk of secondary, nosocomial
infections.
178
The MDSC we describe in sepsis are phenotypically as well as functionally different to
PMN from the same patient. While the MDSC have a similar flow cytometry
forward/side scatter as PMN, they have a notably different staining phenotype, with
consistently higher CD66b and lower CD16 expression when compared to PMN.
Furthermore, MDSC were more effective at suppressing T cell proliferation and zeta-
chain expression than PMN from the same patient, at the same concentration.
As analysing MDSC from whole blood may be more informative in the future
(Mandruzzato, Solito et al. 2009), it would be helpful to design a staining regimen that
could differentiate MDSC from PMN in whole blood. Our results show that a
combination of CD66b and CD16 has the potential to separate most MDSC from PMN,
although there is still some overlap in both populations.
Although some human MDSC have staining profiles consistent with immature
granulocytes (Gabrilovich and Nagaraj 2009), the MDSC co-eluting with PBMC from
sepsis patients appear to be mature granulocytes. In particular, high CD66b, high
CD45RO and low CD33 expression are all consistent with mature granulocyte staining
(Elghetany 2002). MDSC co-eluting with PBMC with a mature granulocyte phenotype
have also been described in renal cell carcinoma patients (Rodriguez, Ernstoff et al.
2009) and patients with other advanced cancers (Schmielau and Finn 2001). Schmeilau
et al demonstrated that when PMN from a healthy donor are activated with N-formyl-L-
methionyl-L-leucyl-L-phenylalanine, the PMN will co-elute with PBMC and suppress T
cells in a dose-dependent manner (Schmielau and Finn 2001). Furthermore, we found
that the percentage of MDSC in septic shock patients correlated with the mature
179
neutrophil count, rather than the immature granulocyte count, from the Coulter Counter.
The sum of all our evidence suggests that the MDSC co-eluting with PBMC from sepsis
patients are hyper-activated mature granulocytes. Thus, the overlap between the MDSC
and PMN populations in combined CD66b and CD16 staining may represent a
transformation in progress.
As wells as having more MDSC, septic shock patients also have significantly higher
plasma concentrations of IL-6, a marker of prognosis in sepsis. Moreover, in all sepsis
patients, the percentage of MDSC correlated with plasma IL-6. As IL-6 can induce
MDSC (Lechner, Liebertz et al. 2010), the inflammatory environment of sepsis may
contribute to the accumulation of MDSC. The role of neutrophils in sepsis has long
been controversial. Potentially, the neutrophil count from the Coulter Counter may
include MDSC. Thus, even though septic shock and non-shock patients have similar
numbers of circulating neutrophils, septic shock patients have significantly more MDSC.
This suggests that the neutrophil count in shock represents cells which are
phenotypically and functionally different to the neutrophils in non-shock patients.
Our results suggest that arginase may be one mode of action of MDSC suppression.
There was a positive trend, although not statistically significant, between the percentage
of MDSC in septic shock patients and plasma arginase activity. Furthermore, cultures
with MDSC consumed more arginine compared to cultures without MDSC. Curiously,
however, MDSC still suppressed T cell proliferation even in the presence of arginase
inhibitors suggesting more than one mechanism of suppression and likely redundancy.
180
Multiple mechanisms of suppression have been reported in other human MDSC and
seem to be dependent on the mode of induction (Lechner, Liebertz et al. 2010).
Conclusion
We have demonstrated that the granulocytes which co-elute with PBMC are circulating
MDSC which suppress T cell proliferation by impairing T cell zeta-chain expression.
The percentage of MDSC in sepsis patients is proportional to disease severity and
correlates with plasma IL-6 concentrations. Together, these results demonstrate that the
inflammatory milieu of sepsis increases circulating MDSC which suppress T cell
function. Thus, as in cancer, MDSC appear to be a major link between inflammation
and T cell suppression in sepsis.
10.4. Conclusion
Increased inflammation in sepsis patients is with associated both with increased arginine
and tryptophan metabolism and increased circulating MDSC. Sepsis patients have low
plasma concentrations of arginine and tryptophan and increased circulating MDSC, all
of which can potentially suppress T cell activation and proliferation by impairing T cell
zeta-chain expression. We found that sepsis patients have low T cell zeta-chain
expression compared to hospital controls and that T cell zeta-chain expression is related
to the percentage of MDSC, rather than plasma concentrations of arginine or tryptophan.
The percentage of MDSC was related to plasma IL-6 concentrations and was highest in
septic shock patients. Therefore, MDSC may link inflammation and T cell suppression
in sepsis.
181
11. Discussion and conclusion
11.1. Introduction
We need a better understanding of the pathophysiology of sepsis to be able to design
better adjunctive treatments. Sepsis patients have dysfunctional immune responses and
impaired microvascular reactivity, but the mechanisms behind these disturbances are not
well understood. As the bioavailability of both arginine and tryptophan can influence the
immune response and endothelial function, the aim of this project was to investigate the
relationship between inflammation, amino acid bioavailability and the pathophysiology
of sepsis. This chapter will discuss the results from this project in the context of the
existing literature and will consider two questions. Firstly, what factors contribute to the
decreased amino acid bioavailability in sepsis? And secondly, how does the decreased
amino acid bioavailability contribute to the pathophysiology of sepsis? Flow diagrams
are used to summarise the proposed model of amino acid bioavailability in sepsis with
and without shock. Finally, the potential future directions of this research are discussed.
11.2. Sepsis decreases amino acid bioavailability
There are several factors likely to decrease amino acid bioavailability in sepsis
including, altered catabolism, absorption, synthesis and recycling, all of which are likely
influenced by the balance of pro-inflammatory and anti-inflammatory cytokines,
bacterial endotoxin, leukocyte arginase and organ failure.
182
Sepsis patients have decreased plasma concentrations of tryptophan and increased
plasma concentrations of kynurenine, indicating increased IDO activity. By referring to
the extensive in vitro work on IDO activity and comparing those findings to what we
know about sepsis, three aspects of sepsis in particular seem likely to contribute to the
increased KT ratio. Firstly, plasma IFN-γ concentrations increase early in sepsis
(Hunsicker, Kullich et al. 1997) and it is well established that IFN-γ increases IDO
expression in a range of cell types including endothelial cells, monocytes, renal tubular
epithelial cells and hepatocytes (Carlin, Borden et al. 1989; Larrea, Riezu-Boj et al.
2007; Mohib, Guan et al. 2007; Iwamoto, Ito et al. 2009; Wang, Liu et al. 2010).
Secondly, IL10 stabilises IDO expression (Munn, Sharma et al. 2002) and may increase
IFN-γ-dependent IDO expression (Yanagawa, Iwabuchi et al. 2009). In our cohort, both
severe and non-severe sepsis patients had elevated plasma IFN-γ but only severe patients
had significantly increased IL10. Therefore, the combined elevation of IFN-γ and IL10
in severe sepsis may contribute to higher IDO expression and a higher plasma KT ratio
compared to non-shock patients. Finally, LPS which is often present in bacterial sepsis
(Dunn 1990), can induce or enhance IDO expression in a broad range of tissues
(Takikawa, Yoshida et al. 1986) in an IFN-γ-independent manner (Fujigaki, Saito et al.
2001).
Arginine bioavailability to NOS (arginine/ADMA ratio) is decreased in sepsis, in
proportion to disease severity. In non-severe sepsis, this is mostly due to the decreased
plasma arginine concentrations. Septic shock patients have low plasma arginine
concentrations and increased plasma ADMA concentrations, resulting in a markedly
183
decreased arginine/ADMA ratio. Both increased arginase activity and decreased DDAH
activity appear to contribute to the decreased arginine/ADMA ratio in sepsis.
Several aspects of sepsis are likely to contribute to decreased plasma arginine
concentrations including decreased absorption, increased protein synthesis and increased
enzymatic activity. Our results and a recent study by Luiking et al. (Luiking, Poeze et
al. 2009) suggest that arginase activity contributes to the low plasma arginine
concentrations in sepsis. Luiking et al. used stable isotope infusion to investigate
arginine metabolism in sepsis. They found that sepsis patients have increased whole
body arginase activity (urea synthesis from arginine). Human neutrophils constitutively
express arginase I (Munder, Mollinedo et al. 2005) and we found a strong association
between the circulating neutrophil count and plasma arginase activity and plasma
arginine concentrations, which suggests that neutrophil-derived arginase activity may
contribute to the low arginine concentrations in sepsis. This observation appears to be
unrelated to disease severity as both severe and non-severe sepsis patients had similarly
elevated neutrophil counts and plasma arginase activity and similarly reduced plasma
arginine concentrations.
There are two mechanisms which are likely to increase plasma ADMA concentrations in
septic shock. Firstly, both the liver and kidney are important sources of DDAH activity
and impairment of these organs leads to decreased DDAH activity and increased ADMA
(Nijveldt, Teerlink et al. 2003; Nijveldt, Siroen et al. 2004; Mookerjee, Malaki et al.
2007). Secondly, septic shock patients have significantly more IL6 than non-shock
patients. The positive association between plasma IL6 and plasma ADMA
184
concentrations is consistent with reports that increased inflammation inhibits DDAH
activity, leading to increased ADMA (Ito, Tsao et al. 1999; Puchau, Hermsdorff et al.
2009). Therefore, decreased DDAH activity as a result of organ failure and
inflammation likely leads to the severely decreased arginine/ADMA ratio in shock.
11.3. Amino acid bioavailability contributes to the
pathophysiology of sepsis
The results presented in this thesis suggest that decreased amino acid bioavailability may
contribute to the pathophysiology of sepsis by impairing both microvascular reactivity
and cellular immune responses.
We found that both the decreased arginine/ADMA ratio and increased KT ratio were
associated with impaired microvascular reactivity in sepsis. Microvascular reactivity is
the ability of microvessels to dilate in response to shear stress and is at least 50%-
dependent on eNOS (nitric oxide produced by endothelial cells). Both an increased KT
ratio and decreased arginine/ADMA ratio can contribute to impaired microvascular
reactivity (Boger, Bode-Boger et al. 1998; Wang, Liu et al. 2010). Furthermore, the
negative feedback between IDO and NOS (Thomas, Mohr et al. 1994; Sekkai, Guittet et
al. 1997; Samelson-Jones and Yeh 2006) means that a decreased arginine/ADMA ratio
also increases the KT ratio and vice versa. An important consideration here is the
location of IDO and NOS expression. As both IDO and NOS are intra-cellular enzymes,
the negative feedback between these enzymes would be most evident within cell types
185
that express both IDO and NOS, such as endothelial cells. A recent study found that
IDO expression is greatly increased in endothelial cells in sepsis (Wang, Liu et al. 2010),
thus we would anticipate that high IDO expression in endothelial cells would result in
decreased eNOS. Although this study was unable to measure IDO expression within
endothelial cells in sepsis patients, since both the KT ratio and arginine/ADMA ratio
were associated with impaired microvascular reactivity, it is likely that the IDO/NOS
balance is disturbed in endothelial cells lining the microvessels in sepsis. When
microvascular reactivity is severely impaired it limits tissue blood flow, leading to organ
failure. This is consistent with the association between organ failure and both the KT
ratio and arginine/ADMA ratio in sepsis.
An increased KT ratio can increase both pro-inflammatory and anti-inflammatory
cytokines. Increased KT stabilises IL6 expression (van Wissen, Snoek et al. 2002) and
may also stabilise IL10 expression (van der Sluijs, Nijhuis et al. 2006). IDO inhibitors
reduce plasma IL6 concentrations (Ulloa, Ochani et al. 2002; Jung, Lee et al. 2009) and
plasma IL10 concentrations (van der Sluijs, Nijhuis et al. 2006). Thus the increased KT
ratio may help sustain both IL6 and IL10 expression in sepsis.
Decreased amino acid bioavailability in sepsis also impairs the adaptive immune
response. Firstly, an increased KT ratio cause lymphocyte apoptosis, particularly in T
cells (Fallarino, Grohmann et al. 2002; Fallarino, Grohmann et al. 2003). Lymphocyte
apoptosis is well described in sepsis and prevention of lymphocyte apoptosis improves
outcome in models of sepsis (Wang, Huang et al. 1994; Hotchkiss, Swanson et al. 1999;
186
Hotchkiss, Tinsley et al. 1999). Our results link the increased KT ratio in sepsis with
decreased circulating T cells, and are consistent with the in vitro data.
Both an increased KT ratio and decreased arginine have been linked to impaired T cell
zeta-chain expression (Rodriguez, Zea et al. 2002; Fallarino, Grohmann et al. 2006).
However, even though we found that sepsis patients do have low T cell zeta-chain
expression, it was not associated with the plasma KT ratio or plasma arginine
concentration in sepsis. Neither was there a longitudinal association, as T cell zeta-chain
expression significantly increased in vivo before either plasma arginine or tryptophan
concentrations improved. The reason for this lack of association despite the strong in
vitro evidence is unclear. One possibility is that the arginine and tryptophan
concentrations may differ within the microenvironment of the T cells, whereas we can
only measure plasma concentrations.
In shock patients, there was a clear association between increased circulating MDSC and
decreased T cell zeta-chain expression. This is an important finding as MDSC have not
previously been described in sepsis patients. The number of MDSC appeared to be
directly related to plasma IL6 concentrations and the KT ratio. IL6 creates or reflects an
inflammatory environment suitable for the generation and/or maintenance of MDSC.
The relationship between the KT ratio and MDSC is likely linked to IL6 as an increased
KT ratio can increase the IL6 response. A similar association between increased IDO,
increased inflammation and increased MDSC has been noted in cancer (Tinder,
Subramani et al. 2008). Non-shock sepsis patients also had impaired T cell zeta-chain
expression, but did not generally have increased circulating MDSC. The reason why
187
sepsis patients without shock also have decreased T cell zeta-chain expression remains
unclear and requires further investigation.
In septic shock patients, those with the most inflammation, including increased plasma
IL6 concentrations and increased MDSC, also had the lowest T cell zeta-chain
expression. These results and others (Tschaikowsky, Hedwig-Geissing et al. 2002)
suggest that characterisation of sepsis into an ‘inflammatory phase’ and
‘immunosuppressive phase’ (Hotchkiss, Coopersmith et al. 2009) may not be
straightforward. Our findings suggest that the dysfunction of the adaptive immune
response in septic shock is similar to that in cancer, where patients with the most
inflammation also have the most suppressed adaptive immune response (Ostrand-
Rosenberg and Sinha 2009).
Figure 11.1 and Figure 11.2 demonstrate the proposed relationships between amino acid
metabolism and the pathophysiology of sepsis with and without shock.
188
IncreasedIFN-γ and/or LPS*
Increased KT ratioin plasma
Increased plasma IL6
Increased IDO activity
Decreased microvascular
reactivity
Decreased endothelial nitric oxide
Increased T cell
apoptosis
DecreasedArg/ADMA ratio
Decreased circulating T cells
Decreased plasma arginineIncreased plasma arginase activity
Increased circulating neutrophils
Decreased plasma tryptophan andIncreased plasma kynurenine concentration
Sepsis without shock
Decreased T cell zeta-chain expression**
* LPS not measured** Mechanism still unclear
Figure 11.1 Proposed relationship between amino acid metabolism and the pathophysiology of sepsis without shock
189
++ IncreasedIFN-γ and/or LPS*
++ Increased KT ratioin plasma
Increased plasma IL6
++ Increased IDO activity
++ Decreased microvascular
reactivity
++ Decreased endothelial nitric oxide
Increased T cell
apoptosis
++ DecreasedArg/ADMA ratio
Decreased circulating T cells
Decreased plasma arginineIncreased plasma arginase activity
Increased circulating neutrophils
++ Decreased plasma tryptophan and++ Increased plasma kynurenine concentration
Septic shock
Decreased T cell zeta-chain expression
* LPS not measured++ more than non-shock patients
Increased plasma IL10
Organ failure
Decreased DDAHactivity in
liver and kidney
Increased plasma ADMA
Increased circulating MDSC
Increased plasma ADMA
Figure 11.2 Proposed relationship between amino acid metabolism and the pathophysiology of septic shock
190
11.4. Future directions
The clear links we have found between amino acid bioavailability and the
pathophysiology of sepsis generate several suggestions for possible future adjunctive
treatments for sepsis, including IDO inhibitors, arginase inhibitors and statins.
Our results suggest that IDO inhibitors may be an effective adjunctive treatment in
sepsis. As sepsis patients have low concentrations of tryptophan, exogenous tryptophan
could help restore tryptophan bioavailability. However, the kynurenine concentrations
suggest that tryptophan is low because of increased IDO activity. Therefore, giving
tryptophan to patients with increased IDO activity could be potentially dangerous as the
tryptophan would likely be metabolised to kynurenine, which may further exacerbate the
endothelial and immune dysfunction associated with high kynurenine concentrations.
Therefore, IDO inhibitors may be a more appropriate treatment option. IDO inhibition
has been shown to improve survival in mouse models of sepsis (Ulloa, Ochani et al.
2002; Jung, Lee et al. 2009). Likely mechanisms for increased survival include
decreased IL6 concentrations (Ulloa, Ochani et al. 2002; Jung, Lee et al. 2009),
decreased IL10 concentrations (van der Sluijs, Nijhuis et al. 2006), improved endothelial
function (Wang, Liu et al. 2010) and increased T cell survival (Fallarino, Grohmann et
al. 2002). Furthermore, prevention of T cell apoptosis is specifically associated with
increased survival in sepsis in murine models (Hotchkiss, Chang et al. 2000;
Bommhardt, Chang et al. 2004). A phase 1 clinical trial of 1-Methyl-D-Tryptophan, an
IDO inhibitor, is currently underway in metastatic cancer patients
191
(http://clinicaltrials.gov/ct2/show/NCT00567931). Such IDO inhibitors are potential
adjunctive treatments that are ready for evaluation in sepsis.
Exogenous arginine has a controversial history in sepsis. Increasing arginine in sepsis
could be harmful if it contributes to excess systemic nitric oxide and unresponsive
hypotension (Kalil and Danner 2006) or beneficial if it improves microcirculation and
immune responses (Luiking and Deutz 2007). Intravenous arginine has had
contradictory results in animal models, decreasing survival in a dog model of sepsis
(Kalil, Sevransky et al. 2006) but increasing survival in a rat model of sepsis (Madden,
Breslin et al. 1988). If sepsis is a state of excess nitric oxide then giving exogenous
arginine could be dangerous, but the role of nitric oxide is still not clear in sepsis with
animal models showing conflicting results with nitric oxide inhibitors (Teale and
Atkinson 1992; Satriano, Schwartz et al. 2001). Varying models of sepsis (cecal ligation
and puncture, lipopolysaccharide injection or live bacterial doses) and species
differences in amino acid metabolism may have contributed to the confusion in this area.
Furthermore, it is difficult for animal models to accurately reflect the complexity of
human sepsis (Zanotti-Cavazzoni and Goldfarb 2009).
Recent studies in human sepsis suggest that it is not a state of nitric oxide excess and
that arginine infusion is a potential treatment option. Stable-isotope studies suggest that
nitric oxide synthesis is not increased in human sepsis (Villalpando, Gopal et al. 2006;
Luiking, Poeze et al. 2009) and treatment with nitric oxide inhibitors increases mortality
in human sepsis (Lopez, Lorente et al. 2004). Intravenous arginine appears to be safe in
sepsis patients and continuous infusion avoids the transient hypotension associated with
192
bolus delivery (Lorente, Landin et al. 1993; Luiking, Poeze et al. 2006). Thus, arginine
is a potential target for treatment in sepsis.
As our results associate decreased arginine bioavailability with impaired microvascular
reactivity and increased organ failure, these findings support increasing arginine
concentrations, if it can be done safely. This could be achieved either by continuous
infusion of arginine or with an arginase inhibitor. In septic shock patients with increased
ADMA, treatments which target DDAH could also improve microvascular reactivity. A
recent study found that treatment with the selective beta1-adrenergic receptor agonist
nebivolol in hypertensive patients decreased plasma ADMA concentrations and
improved microvascular reactivity, apparently by increasing DDAH expression (Pasini,
Garbin et al. 2008).
An alternative treatment option currently under evaluation is the administration of HMG
CoA reductase inhibitors (“statins”). Statins are cardiovascular drugs that are widely
used to lower cholesterol in humans. Statins have pleiotropic effects including
upregulating eNOS, downregulating iNOS and reducing inflammation (McGown and
Brookes 2007). It remains unknown how statin treatment may affect, and be affected
by, amino acid bioavailability in sepsis. Some studies have found that statins can reduce
ADMA (Schroecksnadel, Weiss et al. 2007), however others have found that treatment
with statins does not change ADMA concentrations (Valkonen, Laakso et al. 2003).
Furthermore, the efficacy of statins may also be affected by amino acid concentrations,
as several studies have found that statins are less effective in the presence of high
ADMA concentrations (Boger, Rudolph et al. 2007; Vladimirova-Kitova and Deneva-
193
Koycheva 2010). Our group has recently taken part in a national, multi-centre trial of
atorvastatin in sepsis and we are investigating potential role of amino acid
bioavailability on the effectiveness of statins in sepsis.
11.5. Conclusion
This project has demonstrated that sepsis patients have decreased amino acid
bioavailability, which contributes to the pathophysiology of sepsis. In the context of the
existing literature, our results suggest that both the low arginine/ADMA ratio and
increased KT ratio contribute to impaired microvascular reactivity in sepsis and
contribute to organ failure in septic shock. The increased KT ratio may also increase T
cell apoptosis leading to decreased circulating T cells in both severe and non-severe
sepsis patients. Finally, the increased KT ratio may increase IL6 in sepsis, which in turn
may increase circulating MDSC in septic shock. Together, these results improve our
understanding of the pathogenesis of sepsis and support potential adjunctive treatments
targeting these pathways in human sepsis. With a case-fatality of 20 - 40 % (Martin,
Mannino et al. 2003; Finfer, Bellomo et al. 2004) with existing management strategies,
there remains a major need for adjunctive therapies to reduce the burden and mortality
of severe sepsis.
194
References
Abbasi, F., T. Asagmi, J. P. Cooke, C. Lamendola, T. McLaughlin, G. M. Reaven, M. Stuehlinger and P. S. Tsao (2001). "Plasma concentrations of asymmetric dimethylarginine are increased in patients with type 2 diabetes mellitus." Am J Cardiol 88(10): 1201-1203.
Abraham, E., R. Wunderink, H. Silverman, T. M. Perl, S. Nasraway, H. Levy, R. Bone, R. P. Wenzel, R. Balk, R. Allred and et al. (1995). "Efficacy and safety of monoclonal antibody to human tumor necrosis factor alpha in patients with sepsis syndrome. A randomized, controlled, double-blind, multicenter clinical trial. TNF-alpha MAb Sepsis Study Group." Jama 273(12): 934-941.
Achan, V., M. Broadhead, M. Malaki, G. Whitley, J. Leiper, R. MacAllister and P. Vallance (2003). "Asymmetric dimethylarginine causes hypertension and cardiac dysfunction in humans and is actively metabolized by dimethylarginine dimethylaminohydrolase." Arterioscler Thromb Vasc Biol 23(8): 1455-1459.
Ahmed, N., O. K. Argirov, H. S. Minhas, C. A. Cordeiro and P. J. Thornalley (2002). "Assay of advanced glycation endproducts (AGEs): surveying AGEs by chromatographic assay with derivatization by 6-aminoquinolyl-N-hydroxysuccinimidyl-carbamate and application to Nepsilon-carboxymethyl-lysine- and Nepsilon-(1-carboxyethyl)lysine-modified albumin." Biochem J 364(Pt 1): 1-14.
Aird, W. C. (2003). "The role of the endothelium in severe sepsis and multiple organ dysfunction syndrome." Blood 101(10): 3765-3777.
Aird, W. C. (2007). "Endothelium as a therapeutic target in sepsis." Curr Drug Targets 8(4): 501-507.
Akdis, C. A. and K. Blaser (1999). "IL-10-induced anergy in peripheral T cell and reactivation by microenvironmental cytokines: two key steps in specific immunotherapy." FASEB J 13(6): 603-609.
Albsmeier, J., E. Schwedhelm, F. Schulze, M. Kastner and R. H. Boger (2004). "Determination of NG,NG-dimethyl-L-arginine, an endogenous NO synthase inhibitor, by gas chromatography-mass spectrometry." J Chromatogr B Analyt Technol Biomed Life Sci 809(1): 59-65.
Angus, D. C., W. T. Linde-Zwirble, J. Lidicker, G. Clermont, J. Carcillo and M. R. Pinsky (2001). "Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care." Crit Care Med 29(7): 1303-1310.
Angus, D. C., C. A. Pereira and E. Silva (2006). "Epidemiology of severe sepsis around the world." Endocr Metab Immune Disord Drug Targets 6(2): 207-212.
Ardawi, M. S. and E. A. Newsholme (1983). "Glutamine metabolism in lymphocytes of the rat." Biochem J 212(3): 835-842.
Argaman, Z., V. R. Young, N. Noviski, L. Castillo-Rosas, X. M. Lu, D. Zurakowski, M. Cooper, C. Davison, J. F. Tharakan, A. Ajami and L. Castillo (2003). "Arginine and nitric oxide metabolism in critically ill septic pediatric patients." Crit Care Med 31(2): 591-597.
195
Ashina, M. (2004). "Neurobiology of chronic tension-type headache." Cephalalgia 24(3): 161-172.
Askanazi, J., Y. A. Carpentier, C. B. Michelsen, D. H. Elwyn, P. Furst, L. R. Kantrowitz, F. E. Gump and J. M. Kinney (1980). "Muscle and plasma amino acids following injury. Influence of intercurrent infection." Ann Surg 192(1): 78-85.
Astiz, M. E., G. E. DeGent, R. Y. Lin and E. C. Rackow (1995). "Microvascular function and rheologic changes in hyperdynamic sepsis." Crit Care Med 23(2): 265-271.
Astiz, M. E., E. Tilly, E. D. Rackow and M. H. Weil (1991). "Peripheral vascular tone in sepsis." Chest 99(5): 1072-1075.
Badiou, S., S. Lehmann, J. P. Cristol and H. Bellet (2004). "Determination of plasma amino acids by fluorescent derivatization and reversed-phase liquid chromatographic separation." Clin Lab 50(3-4): 153-158.
Baker, C. C., C. L. Miller and D. D. Trunkey (1979). "Predicting fatal sepsis in burn patients." J Trauma 19(9): 641-648.
Ball, H. J., A. Sanchez-Perez, S. Weiser, C. J. Austin, F. Astelbauer, J. Miu, J. A. McQuillan, R. Stocker, L. S. Jermiin and N. H. Hunt (2007). "Characterization of an indoleamine 2,3-dioxygenase-like protein found in humans and mice." Gene 396(1): 203-213.
Ball, H. J., H. J. Yuasa, C. J. Austin, S. Weiser and N. H. Hunt (2009). "Indoleamine 2,3-dioxygenase-2; a new enzyme in the kynurenine pathway." Int J Biochem Cell Biol 41(3): 467-471.
Baniyash, M. (2004). "TCR zeta-chain downregulation: curtailing an excessive inflammatory immune response." Nat Rev Immunol 4(9): 675-687.
Baniyash, M., P. Garcia-Morales, E. Luong, L. E. Samelson and R. D. Klausner (1988). "The T cell antigen receptor zeta chain is tyrosine phosphorylated upon activation." J Biol Chem 263(34): 18225-18230.
Barth, M. C., N. Ahluwalia, T. J. Anderson, G. J. Hardy, S. Sinha, J. A. Alvarez-Cardona, I. E. Pruitt, E. P. Rhee, R. A. Colvin and R. E. Gerszten (2009). "Kynurenic acid triggers firm arrest of leukocytes to vascular endothelium under flow conditions." J Biol Chem.
Becker, J. U., C. Theodosis, S. T. Jacob, C. R. Wira and N. E. Groce (2009). "Surviving sepsis in low-income and middle-income countries: new directions for care and research." Lancet Infect Dis 9(9): 577-582.
Bellien, J., C. Thuillez and R. Joannides (2008). "Contribution of endothelium-derived hyperpolarizing factors to the regulation of vascular tone in humans." Fundam Clin Pharmacol 22(4): 363-377.
Bernard, A., M. Kasten, C. Meier, E. Manning, S. Freeman, W. Adams, P. Chang, B. Boulanger and P. Kearney (2008). "Red blood cell arginase suppresses Jurkat (T cell) proliferation by depleting arginine." Surgery 143(2): 286-291.
Bernard, A., C. Meier, N. Lopez, J. May, P. Chang, B. Boulanger and P. Kearney (2007). "Packed red blood cell-associated arginine depletion is mediated by arginase." J Trauma 63(5): 1108-1112; discussion 1112.
Bernard, A. C., S. K. Mistry, S. M. Morris, Jr., W. E. O'Brien, B. J. Tsuei, M. E. Maley, L. A. Shirley, P. A. Kearney, B. R. Boulanger and J. B. Ochoa (2001). "Alterations in arginine metabolic enzymes in trauma." Shock 15(3): 215-219.
196
Bernard, G. R., J. L. Vincent, P. F. Laterre, S. P. LaRosa, J. F. Dhainaut, A. Lopez-Rodriguez, J. S. Steingrub, G. E. Garber, J. D. Helterbrand, E. W. Ely and C. J. Fisher, Jr. (2001). "Efficacy and safety of recombinant human activated protein C for severe sepsis." N Engl J Med 344(10): 699-709.
Bernard, G. R., A. P. Wheeler, J. A. Russell, R. Schein, W. R. Summer, K. P. Steinberg, W. J. Fulkerson, P. E. Wright, B. W. Christman, W. D. Dupont, S. B. Higgins and B. B. Swindell (1997). "The effects of ibuprofen on the physiology and survival of patients with sepsis. The Ibuprofen in Sepsis Study Group." N Engl J Med 336(13): 912-918.
Beutelspacher, S. C., P. H. Tan, M. O. McClure, D. F. Larkin, R. I. Lechler and A. J. George (2006). "Expression of indoleamine 2,3-dioxygenase (IDO) by endothelial cells: implications for the control of alloresponses." Am J Transplant 6(6): 1320-1330.
Blackwell, S., S. O'Reilly D and D. Talwar (2007). "Biological variation of asymmetric dimethylarginine and related arginine metabolites and analytical performance goals for their measurement in human plasma." Eur J Clin Invest 37(5): 364-371.
Blackwell, S., D. S. O'Reilly and D. K. Talwar (2009). "HPLC analysis of asymmetric dimethylarginine (ADMA) and related arginine metabolites in human plasma using a novel non-endogenous internal standard." Clin Chim Acta 401(1-2): 14-19.
Blanco, J., A. Muriel-Bombin, V. Sagredo, F. Taboada, F. Gandia, L. Tamayo, J. Collado, A. Garcia-Labattut, D. Carriedo, M. Valledor, M. De Frutos, M. J. Lopez, A. Caballero, J. Guerra, B. Alvarez, A. Mayo and J. Villar (2008). "Incidence, organ dysfunction and mortality in severe sepsis: a Spanish multicentre study." Crit Care 12(6): R158.
Bland, J. M. and D. G. Altman (1986). "Statistical methods for assessing agreement between two methods of clinical measurement." Lancet 1(8476): 307-310.
Boasso, A., J. P. Herbeuval, A. W. Hardy, S. A. Anderson, M. J. Dolan, D. Fuchs and G. M. Shearer (2007). "HIV inhibits CD4+ T-cell proliferation by inducing indoleamine 2,3-dioxygenase in plasmacytoid dendritic cells." Blood 109(8): 3351-3359.
Bode-Boger, S. M., F. Scalera and L. J. Ignarro (2007). "The l-arginine paradox: Importance of the l-arginine/asymmetrical dimethylarginine ratio." Pharmacol Ther 114.
Bode-Boger, S. M., F. Scalera and L. J. Ignarro (2007). "The L-arginine paradox: Importance of the L-arginine/asymmetrical dimethylarginine ratio." Pharmacol Ther 114(3): 295-306.
Bode-Boger, S. M., F. Scalera, J. T. Kielstein, J. Martens-Lobenhoffer, G. Breithardt, M. Fobker and H. Reinecke (2006). "Symmetrical dimethylarginine: a new combined parameter for renal function and extent of coronary artery disease." J Am Soc Nephrol 17(4): 1128-1134.
Bogdan, C. (2001). "Nitric oxide and the immune response." Nat Immunol 2(10): 907-916.
Boger, G. I., T. K. Rudolph, R. Maas, E. Schwedhelm, E. Dumbadze, A. Bierend, R. A. Benndorf and R. H. Boger (2007). "Asymmetric dimethylarginine determines the improvement of endothelium-dependent vasodilation by simvastatin effect of combination with oral L-arginine." J Am Coll Cardiol 49(23): 2274-2282.
197
Boger, R. H. (2007). "The pharmacodynamics of L-arginine." J Nutr 137(6 Suppl 2): 1650S-1655S.
Boger, R. H. and S. M. Bode-Boger (2001). "The clinical pharmacology of L-arginine." Annu Rev Pharmacol Toxicol 41: 79-99.
Boger, R. H., S. M. Bode-Boger, A. Szuba, P. S. Tsao, J. R. Chan, O. Tangphao, T. F. Blaschke and J. P. Cooke (1998). "Asymmetric dimethylarginine (ADMA): a novel risk factor for endothelial dysfunction: its role in hypercholesterolemia." Circulation 98(18): 1842-1847.
Boger, R. H., S. M. Bode-Boger, W. Thiele, W. Junker, K. Alexander and J. C. Frolich (1997). "Biochemical evidence for impaired nitric oxide synthesis in patients with peripheral arterial occlusive disease." Circulation 95(8): 2068-2074.
Boger, R. H., L. M. Sullivan, E. Schwedhelm, T. J. Wang, R. Maas, E. J. Benjamin, F. Schulze, V. Xanthakis, R. A. Benndorf and R. S. Vasan (2009). "Plasma asymmetric dimethylarginine and incidence of cardiovascular disease and death in the community." Circulation 119(12): 1592-1600.
Boger, R. H., D. Tsikas, S. M. Bode-Boger, L. Phivthong-Ngam, E. Schwedhelm and J. C. Frolich (2004). "Hypercholesterolemia impairs basal nitric oxide synthase turnover rate: a study investigating the conversion of L-[guanidino-(15)N(2)]-arginine to (15)N-labeled nitrate by gas chromatography--mass spectrometry." Nitric Oxide 11(1): 1-8.
Bommhardt, U., K. C. Chang, P. E. Swanson, T. H. Wagner, K. W. Tinsley, I. E. Karl and R. S. Hotchkiss (2004). "Akt decreases lymphocyte apoptosis and improves survival in sepsis." J Immunol 172(12): 7583-7591.
Bonaccorso, S., A. Lin, C. Song, R. Verkerk, G. Kenis, E. Bosmans, S. Scharpe, M. Vandewoude, A. Dossche and M. Maes (1998). "Serotonin-immune interactions in elderly volunteers and in patients with Alzheimer's disease (DAT): lower plasma tryptophan availability to the brain in the elderly and increased serum interleukin-6 in DAT." Aging (Milano) 10(4): 316-323.
Bone, R. C. (1996). "Sir Isaac Newton, sepsis, SIRS, and CARS." Crit Care Med 24(7): 1125-1128.
Bone, R. C., R. A. Balk, F. B. Cerra, R. P. Dellinger, A. M. Fein, W. A. Knaus, R. M. Schein and W. J. Sibbald (1992). "Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine." Chest 101(6): 1644-1655.
Bone, R. C., C. J. Fisher, Jr., T. P. Clemmer, G. J. Slotman, C. A. Metz and R. A. Balk (1987). "A controlled clinical trial of high-dose methylprednisolone in the treatment of severe sepsis and septic shock." N Engl J Med 317(11): 653-658.
Bone, R. C., C. L. Sprung and W. J. Sibbald (1992). "Definitions for sepsis and organ failure." Crit Care Med 20(6): 724-726.
Bonetti, P. O., G. W. Barsness, P. C. Keelan, T. I. Schnell, G. M. Pumper, J. T. Kuvin, R. P. Schnall, D. R. Holmes, S. T. Higano and A. Lerman (2003). "Enhanced external counterpulsation improves endothelial function in patients with symptomatic coronary artery disease." J Am Coll Cardiol 41(10): 1761-1768.
Bonetti, P. O., G. M. Pumper, S. T. Higano, D. R. Holmes, Jr., J. T. Kuvin and A. Lerman (2004). "Noninvasive identification of patients with early coronary
198
atherosclerosis by assessment of digital reactive hyperemia." J Am Coll Cardiol 44(11): 2137-2141.
Bosch, L., A. Alegria and R. Farre (2006). "Application of the 6-aminoquinolyl-N-hydroxysccinimidyl carbamate (AQC) reagent to the RP-HPLC determination of amino acids in infant foods." Journal of Chromatography B 831(1-2): 176-183.
Bronstein-Sitton, N., L. Cohen-Daniel, I. Vaknin, A. V. Ezernitchi, B. Leshem, A. Halabi, Y. Houri-Hadad, E. Greenbaum, Z. Zakay-Rones, L. Shapira and M. Baniyash (2003). "Sustained exposure to bacterial antigen induces interferon-gamma-dependent T cell receptor zeta down-regulation and impaired T cell function." Nat Immunol 4(10): 957-964.
Bronte, V., P. Serafini, A. Mazzoni, D. M. Segal and P. Zanovello (2003). "L-arginine metabolism in myeloid cells controls T-lymphocyte functions." Trends Immunol 24(6): 302-306.
Bronte, V. and P. Zanovello (2005). "Regulation of immune responses by L-arginine metabolism." Nat Rev Immunol 5(8): 641-654.
Bruckdorfer, R. (2005). "The basics about nitric oxide." Mol Aspects Med 26(1-2): 3-31. Brun-Buisson, C., F. Doyon, J. Carlet, P. Dellamonica, F. Gouin, A. Lepoutre, J. C.
Mercier, G. Offenstadt and B. Regnier (1995). "Incidence, risk factors, and outcome of severe sepsis and septic shock in adults. A multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis." JAMA 274(12): 968-974.
Bunt, S. K., V. K. Clements, E. M. Hanson, P. Sinha and S. Ostrand-Rosenberg (2009). "Inflammation enhances myeloid-derived suppressor cell cross-talk by signaling through Toll-like receptor 4." J Leukoc Biol 85(6): 996-1004.
Bunt, S. K., L. Yang, P. Sinha, V. K. Clements, J. Leips and S. Ostrand-Rosenberg (2007). "Reduced inflammation in the tumor microenvironment delays the accumulation of myeloid-derived suppressor cells and limits tumor progression." Cancer Res 67(20): 10019-10026.
Busse, R., G. Edwards, M. Feletou, I. Fleming, P. M. Vanhoutte and A. H. Weston (2002). "EDHF: bringing the concepts together." Trends Pharmacol Sci 23(8): 374-380.
Carlin, J. M., E. C. Borden, P. M. Sondel and G. I. Byrne (1989). "Interferon-induced indoleamine 2,3-dioxygenase activity in human mononuclear phagocytes." J Leukoc Biol 45(1): 29-34.
Castillo, L., L. Beaumier, A. M. Ajami and V. R. Young (1996). "Whole body nitric oxide synthesis in healthy men determined from [15N] arginine-to-[15N]citrulline labeling." Proc Natl Acad Sci U S A 93(21): 11460-11465.
Celermajer, D. S. (2008). "Reliable endothelial function testing: at our fingertips?" Circulation 117(19): 2428-2430.
Cheng, A. C., T. E. West, D. Limmathurotsakul and S. J. Peacock (2008). "Strategies to reduce mortality from bacterial sepsis in adults in developing countries." PLoS Med 5(8): e175.
Chenzbraun, A., G. Levin, J. Scheffy, A. Keren, S. Stern and D. Goor (2001). "The peripheral vascular response to exercise is impaired in patients with risk factors for coronary artery disease." Cardiology 95(3): 126-130.
Chiarla, C., I. Giovannini and J. H. Siegel (2006). "Plasma arginine correlations in trauma and sepsis." Amino Acids 30(1): 81-86.
199
Chiarla, C., I. Giovannini, J. H. Siegel, G. Boldrini and M. Castagneto (2000). "The relationship between plasma taurine and other amino acid levels in human sepsis." J Nutr 130(9): 2222-2227.
Chiarugi, A., E. Rovida, P. Dello Sbarba and F. Moroni (2003). "Tryptophan availability selectively limits NO-synthase induction in macrophages." J Leukoc Biol 73(1): 172-177.
Christou, N. V., J. L. Meakins, J. Gordon, J. Yee, M. Hassan-Zahraee, C. W. Nohr, H. M. Shizgal and L. D. MacLean (1995). "The delayed hypersensitivity response and host resistance in surgical patients. 20 years later." Ann Surg 222(4): 534-546; discussion 546-538.
Closs, E. I., F. Z. Basha, A. Habermeier and U. Forstermann (1997). "Interference of L-arginine analogues with L-arginine transport mediated by the y+ carrier hCAT-2B." Nitric Oxide 1(1): 65-73.
Cohen, S. A. and D. P. Michaud (1993). "Synthesis of a Fluorescent Derivatizing Reagent, 6-Aminoquinolyl-N-Hydroxysuccinimidyl Carbamate, and Its Application for the Analysis of Hydrolysate Amino Acids via High-Performance Liquid Chromatography." Analytical Biochemistry 211(2): 279-287.
Corraliza, I. M., G. Soler, K. Eichmann and M. Modolell (1995). "Arginase induction by suppressors of nitric oxide synthesis (IL-4, IL-10 and PGE2) in murine bone-marrow-derived macrophages." Biochem Biophys Res Commun 206(2): 667-673.
Creteur, J., T. Carollo, G. Soldati, G. Buchele, D. De Backer and J. L. Vincent (2007). "The prognostic value of muscle StO(2) in septic patients." Intensive Care Med 33(9): 1549-1556.
Curtsinger, L. J., W. G. Cheadle, M. J. Hershman, K. Cost and H. C. Polk, Jr. (1989). "Association of cytomegalovirus infection with increased morbidity is independent of transfusion." Am J Surg 158(6): 606-610; discussion 610-601.
Cynober, L., Ed. (2004). Metabolic and therapeutic aspects of amino acids in clinical nutrition. Boca Raton, CRC Press.
Damas, P., A. Reuter, P. Gysen, J. Demonty, M. Lamy and P. Franchimont (1989). "Tumor necrosis factor and interleukin-1 serum levels during severe sepsis in humans." Crit Care Med 17(10): 975-978.
Danese, S., E. Dejana and C. Fiocchi (2007). "Immune regulation by microvascular endothelial cells: directing innate and adaptive immunity, coagulation, and inflammation." J Immunol 178(10): 6017-6022.
Davis, J. S. and N. M. Anstey (2010). "Is plasma arginine concentration decreased in patients with sepsis? A systemic review and meta-analysis." Crit Care Med epub ahead of print.
Davis, J. S., T. W. Yeo, J. H. Thomas, M. McMillan, C. J. Darcy, Y. R. McNeil, A. C. Cheng, D. S. Celermajer, D. P. Stephens and N. M. Anstey (2009). "Sepsis-associated microvascular dysfunction measured by peripheral arterial tonometry: an observational study." Crit Care 13(5): R155.
De Gennaro Colonna, V., M. Bianchi, V. Pascale, P. Ferrario, F. Morelli, W. Pascale, L. Tomasoni and M. Turiel (2009). "Asymmetric dimethylarginine (ADMA): an endogenous inhibitor of nitric oxide synthase and a novel cardiovascular risk molecule." Med Sci Monit 15(4): RA91-101.
200
De Gennaro Colonna, V., S. Bonomo, P. Ferrario, M. Bianchi, M. Berti, M. Guazzi, B. Manfredi, E. E. Muller, F. Berti and G. Rossoni (2007). "Asymmetric dimethylarginine (ADMA) induces vascular endothelium impairment and aggravates post-ischemic ventricular dysfunction in rats." Eur J Pharmacol 557(2-3): 178-185.
Deanfield, J. E., J. P. Halcox and T. J. Rabelink (2007). "Endothelial function and dysfunction: testing and clinical relevance." Circulation 115(10): 1285-1295.
Delano, M. J., P. O. Scumpia, J. S. Weinstein, D. Coco, S. Nagaraj, K. M. Kelly-Scumpia, K. A. O'Malley, J. L. Wynn, S. Antonenko, S. Z. Al-Quran, R. Swan, C. S. Chung, M. A. Atkinson, R. Ramphal, D. I. Gabrilovich, W. H. Reeves, A. Ayala, J. Phillips, D. Laface, P. G. Heyworth, M. Clare-Salzler and L. L. Moldawer (2007). "MyD88-dependent expansion of an immature GR-1(+)CD11b(+) population induces T cell suppression and Th2 polarization in sepsis." J Exp Med 204(6): 1463-1474.
Dellinger, R. P., J. M. Carlet, H. Masur, H. Gerlach, T. Calandra, J. Cohen, J. Gea-Banacloche, D. Keh, J. C. Marshall, M. M. Parker, G. Ramsay, J. L. Zimmerman, J. L. Vincent and M. M. Levy (2004). "Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock." Crit Care Med 32(3): 858-873.
Dhindsa, M., S. M. Sommerlad, A. E. DeVan, J. N. Barnes, J. Sugawara, O. Ley and H. Tanaka (2008). "Interrelationships among noninvasive measures of postischemic macro- and microvascular reactivity." J Appl Physiol 105(2): 427-432.
Di Giantomasso, D., C. N. May and R. Bellomo (2003). "Vital organ blood flow during hyperdynamic sepsis." Chest 124(3): 1053-1059.
Diaz-Montero, C. M., M. L. Salem, M. I. Nishimura, E. Garrett-Mayer, D. J. Cole and A. J. Montero (2009). "Increased circulating myeloid-derived suppressor cells correlate with clinical cancer stage, metastatic tumor burden, and doxorubicin-cyclophosphamide chemotherapy." Cancer Immunol Immunother 58(1): 49-59.
Diaz, J., J. L. Lliberia, L. Comellas and F. Broto-Puig (1996). "Amino acid and amino sugar determination by derivatization with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate followed by high-performance liquid chromatography and fluorescence detection." Journal of Chromatography A 719(1): 171-179.
Druml, W., G. Heinzel and G. Kleinberger (2001). "Amino acid kinetics in patients with sepsis." Am J Clin Nutr 73(5): 908-913.
Dunn, D. L. (1990). "Development and potential use of antibody directed against lipopolysaccharide for the treatment of gram-negative bacterial sepsis." J Trauma 30(12 Suppl): S100-106.
Dweik, R. A. (2007). "The lung in the balance: arginine, methylated arginines, and nitric oxide." Am J Physiol Lung Cell Mol Physiol 292(1): L15-17.
Elghetany, M. T. (2002). "Surface antigen changes during normal neutrophilic development: a critical review." Blood Cells Mol Dis 28(2): 260-274.
Ellinger, P. and M. M. Abdel Kader (1947). "Tryptophane as precursor of nicotinamide in mammals." Nature 160(4072): 675.
Ezernitchi, A. V., I. Vaknin, L. Cohen-Daniel, O. Levy, E. Manaster, A. Halabi, E. Pikarsky, L. Shapira and M. Baniyash (2006). "TCR zeta down-regulation under
201
chronic inflammation is mediated by myeloid suppressor cells differentially distributed between various lymphatic organs." J Immunol 177(7): 4763-4772.
Fallarino, F., U. Grohmann, C. Vacca, R. Bianchi, C. Orabona, A. Spreca, M. C. Fioretti and P. Puccetti (2002). "T cell apoptosis by tryptophan catabolism." Cell Death Differ 9(10): 1069-1077.
Fallarino, F., U. Grohmann, C. Vacca, C. Orabona, A. Spreca, M. C. Fioretti and P. Puccetti (2003). "T cell apoptosis by kynurenines." Adv Exp Med Biol 527: 183-190.
Fallarino, F., U. Grohmann, S. You, B. C. McGrath, D. R. Cavener, C. Vacca, C. Orabona, R. Bianchi, M. L. Belladonna, C. Volpi, P. Santamaria, M. C. Fioretti and P. Puccetti (2006). "The combined effects of tryptophan starvation and tryptophan catabolites down-regulate T cell receptor zeta-chain and induce a regulatory phenotype in naive T cells." J Immunol 176(11): 6752-6761.
Fallarino, F., C. Vacca, C. Orabona, M. L. Belladonna, R. Bianchi, B. Marshall, D. B. Keskin, A. L. Mellor, M. C. Fioretti, U. Grohmann and P. Puccetti (2002). "Functional expression of indoleamine 2,3-dioxygenase by murine CD8 alpha(+) dendritic cells." Int Immunol 14(1): 65-68.
Fard, A., C. H. Tuck, J. A. Donis, R. Sciacca, M. R. Di Tullio, H. D. Wu, T. A. Bryant, N. T. Chen, M. Torres-Tamayo, R. Ramasamy, L. Berglund, H. N. Ginsberg, S. Homma and P. J. Cannon (2000). "Acute elevations of plasma asymmetric dimethylarginine and impaired endothelial function in response to a high-fat meal in patients with type 2 diabetes." Arterioscler Thromb Vasc Biol 20(9): 2039-2044.
Felmet, K. A., M. W. Hall, R. S. Clark, R. Jaffe and J. A. Carcillo (2005). "Prolonged lymphopenia, lymphoid depletion, and hypoprolactinemia in children with nosocomial sepsis and multiple organ failure." J Immunol 174(6): 3765-3772.
Fernstrom, J. D. and R. J. Wurtman (1971). "Brain serotonin content: physiological dependence on plasma tryptophan levels." Science 173(992): 149-152.
Filipazzi, P., R. Valenti, V. Huber, L. Pilla, P. Canese, M. Iero, C. Castelli, L. Mariani, G. Parmiani and L. Rivoltini (2007). "Identification of a new subset of myeloid suppressor cells in peripheral blood of melanoma patients with modulation by a granulocyte-macrophage colony-stimulation factor-based antitumor vaccine." J Clin Oncol 25(18): 2546-2553.
Finfer, S., R. Bellomo, J. Lipman, C. French, G. Dobb and J. Myburgh (2004). "Adult-population incidence of severe sepsis in Australian and New Zealand intensive care units." Intensive Care Med 30(4): 589-596.
Finke, J. H., A. H. Zea, J. Stanley, D. L. Longo, H. Mizoguchi, R. R. Tubbs, R. H. Wiltrout, J. J. O'Shea, S. Kudoh, E. Klein and et al. (1993). "Loss of T-cell receptor zeta chain and p56lck in T-cells infiltrating human renal cell carcinoma." Cancer Res 53(23): 5613-5616.
Fisher, C. J., Jr., J. M. Agosti, S. M. Opal, S. F. Lowry, R. A. Balk, J. C. Sadoff, E. Abraham, R. M. Schein and E. Benjamin (1996). "Treatment of septic shock with the tumor necrosis factor receptor:Fc fusion protein. The Soluble TNF Receptor Sepsis Study Group." N Engl J Med 334(26): 1697-1702.
Fisher, C. J., Jr., G. J. Slotman, S. M. Opal, J. P. Pribble, R. C. Bone, G. Emmanuel, D. Ng, D. C. Bloedow and M. A. Catalano (1994). "Initial evaluation of human recombinant interleukin-1 receptor antagonist in the treatment of sepsis
202
syndrome: a randomized, open-label, placebo-controlled multicenter trial." Crit Care Med 22(1): 12-21.
Forouzandeh, F., R. B. Jalili, M. Germain, V. Duronio and A. Ghahary (2008). "Skin cells, but not T cells, are resistant to indoleamine 2, 3-dioxygenase (IDO) expressed by allogeneic fibroblasts." Wound Repair Regen 16(3): 379-387.
Freund, H. R., J. A. Ryan, Jr. and J. E. Fischer (1978). "Amino acid derangements in patients with sepsis: treatment with branched chain amino acid rich infusions." Ann Surg 188(3): 423-430.
Fuchs, D., A. Forsman, L. Hagberg, M. Larsson, G. Norkrans, G. Reibnegger, E. R. Werner and H. Wachter (1990). "Immune activation and decreased tryptophan in patients with HIV-1 infection." J Interferon Res 10(6): 599-603.
Fujigaki, S., K. Saito, K. Sekikawa, S. Tone, O. Takikawa, H. Fujii, H. Wada, A. Noma and M. Seishima (2001). "Lipopolysaccharide induction of indoleamine 2,3-dioxygenase is mediated dominantly by an IFN-gamma-independent mechanism." Eur J Immunol 31(8): 2313-2318.
Furchgott, R. F. and J. V. Zawadzki (1980). "The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine." Nature 288(5789): 373-376.
Gabrilovich, D. I., V. Bronte, S. H. Chen, M. P. Colombo, A. Ochoa, S. Ostrand-Rosenberg and H. Schreiber (2007). "The terminology issue for myeloid-derived suppressor cells." Cancer Res 67(1): 425; author reply 426.
Gabrilovich, D. I. and S. Nagaraj (2009). "Myeloid-derived suppressor cells as regulators of the immune system." Nat Rev Immunol 9(3): 162-174.
Gallina, G., L. Dolcetti, P. Serafini, C. De Santo, I. Marigo, M. P. Colombo, G. Basso, F. Brombacher, I. Borrello, P. Zanovello, S. Bicciato and V. Bronte (2006). "Tumors induce a subset of inflammatory monocytes with immunosuppressive activity on CD8+ T cells." J Clin Invest 116(10): 2777-2790.
Ganz, P. and J. A. Vita (2003). "Testing endothelial vasomotor function: nitric oxide, a multipotent molecule." Circulation 108(17): 2049-2053.
Gibbons, J. and S. Chakraborti (2003). Nonparametric Statistical Inference, Marcel Dekker
Goonasekera, C. D., D. D. Rees, P. Woolard, A. Frend, V. Shah and M. J. Dillon (1997). "Nitric oxide synthase inhibitors and hypertension in children and adolescents." J Hypertens 15(8): 901-909.
Grivennikov, S. I., F. R. Greten and M. Karin (2010). "Immunity, inflammation, and cancer." Cell 140(6): 883-899.
Gunji, Y., S. Hori, T. Aoe, T. Asano, T. Ochiai, K. Isono and T. Saito (1994). "High frequency of cancer patients with abnormal assembly of the T cell receptor-CD3 complex in peripheral blood T lymphocytes." Jpn J Cancer Res 85(12): 1189-1192.
Haile, L. A., R. von Wasielewski, J. Gamrekelashvili, C. Kruger, O. Bachmann, A. M. Westendorf, J. Buer, R. Liblau, M. P. Manns, F. Korangy and T. F. Greten (2008). "Myeloid-derived suppressor cells in inflammatory bowel disease: a new immunoregulatory pathway." Gastroenterology 135(3): 871-881, 881 e871-875.
Hainque, B., D. Gerbet, J. P. Roisin, G. Le Moel, S. Troupel and A. Galli (1985). "[Course of the concentration of serum free amino acids as a function of time and the method of preservation]." Ann Biol Clin (Paris) 43(3): 221-226.
203
Hallemeesch, M. M., W. H. Lamers and N. E. Deutz (2002). "Reduced arginine availability and nitric oxide production." Clin Nutr 21(4): 273-279.
Haller, M. J., J. Stein, J. Shuster, D. Theriaque, J. Silverstein, D. A. Schatz, M. G. Earing, A. Lerman and F. H. Mahmud (2007). "Peripheral artery tonometry demonstrates altered endothelial function in children with type 1 diabetes." Pediatr Diabetes 8(4): 193-198.
Hamburg, N. M. and E. J. Benjamin (2009). "Assessment of endothelial function using digital pulse amplitude tonometry." Trends Cardiovasc Med 19(1): 6-11.
Hamburg, N. M., M. J. Keyes, M. G. Larson, R. S. Vasan, R. Schnabel, M. M. Pryde, G. F. Mitchell, J. Sheffy, J. A. Vita and E. J. Benjamin (2008). "Cross-sectional relations of digital vascular function to cardiovascular risk factors in the Framingham Heart Study." Circulation 117(19): 2467-2474.
Hammerman, S. I., E. S. Klings, K. P. Hendra, G. R. Upchurch, Jr., D. C. Rishikof, J. Loscalzo and H. W. Farber (1999). "Endothelial cell nitric oxide production in acute chest syndrome." Am J Physiol 277(4 Pt 2): H1579-1592.
Harrison, D. G., M. A. Kurz, J. E. Quillen, F. W. Sellke and A. Mugge (1992). "Normal and pathophysiologic considerations of endothelial regulation of vascular tone and their relevance to nitrate therapy." Am J Cardiol 70(8): 11B-17B.
Hartl, W. H., B. Gunther, D. Inthorn and G. Heberer (1988). "Reactive hyperemia in patients with septic conditions." Surgery 103(4): 440-444.
Hecker, M., W. C. Sessa, H. J. Harris, E. E. Anggard and J. R. Vane (1990). "The metabolism of L-arginine and its significance for the biosynthesis of endothelium-derived relaxing factor: cultured endothelial cells recycle L-citrulline to L-arginine." Proc Natl Acad Sci U S A 87(21): 8612-8616.
Heidecke, C. D., T. Hensler, H. Weighardt, N. Zantl, H. Wagner, J. R. Siewert and B. Holzmann (1999). "Selective defects of T lymphocyte function in patients with lethal intraabdominal infection." Am J Surg 178(4): 288-292.
Heresztyn, T., M. I. Worthley and J. D. Horowitz (2004). "Determination of l-arginine and NG, NG - and NG, NG' -dimethyl-L-arginine in plasma by liquid chromatography as AccQ-Fluor fluorescent derivatives." J Chromatogr B Analyt Technol Biomed Life Sci 805(2): 325-329.
Herzum, I. and H. Renz (2008). "Inflammatory markers in SIRS, sepsis and septic shock." Curr Med Chem 15(6): 581-587.
Holub, M., Z. Kluckova, B. Beneda, J. Hobstova, I. Huzicka, J. Prazak and A. Lobovska (2000). "Changes in lymphocyte subpopulations and CD3+/DR+ expression in sepsis." Clin Microbiol Infect 6(12): 657-660.
Horowitz, J. D. and T. Heresztyn (2007). "An overview of plasma concentrations of asymmetric dimethylarginine (ADMA) in health and disease and in clinical studies: methodological considerations." J Chromatogr B Analyt Technol Biomed Life Sci 851(1-2): 42-50.
Hotchkiss, R. S., K. C. Chang, P. E. Swanson, K. W. Tinsley, J. J. Hui, P. Klender, S. Xanthoudakis, S. Roy, C. Black, E. Grimm, R. Aspiotis, Y. Han, D. W. Nicholson and I. E. Karl (2000). "Caspase inhibitors improve survival in sepsis: a critical role of the lymphocyte." Nat Immunol 1(6): 496-501.
Hotchkiss, R. S., C. M. Coopersmith, J. E. McDunn and T. A. Ferguson (2009). "The sepsis seesaw: tilting toward immunosuppression." Nat Med 15(5): 496-497.
204
Hotchkiss, R. S. and I. E. Karl (2003). "The pathophysiology and treatment of sepsis." N Engl J Med 348(2): 138-150.
Hotchkiss, R. S., P. E. Swanson, B. D. Freeman, K. W. Tinsley, J. P. Cobb, G. M. Matuschak, T. G. Buchman and I. E. Karl (1999). "Apoptotic cell death in patients with sepsis, shock, and multiple organ dysfunction." Crit Care Med 27(7): 1230-1251.
Hotchkiss, R. S., K. W. Tinsley, P. E. Swanson, K. C. Chang, J. P. Cobb, T. G. Buchman, S. J. Korsmeyer and I. E. Karl (1999). "Prevention of lymphocyte cell death in sepsis improves survival in mice." Proc Natl Acad Sci U S A 96(25): 14541-14546.
Huang, A., D. Fuchs, B. Widner, C. Glover, D. C. Henderson and T. G. Allen-Mersh (2002). "Serum tryptophan decrease correlates with immune activation and impaired quality of life in colorectal cancer." Br J Cancer 86(11): 1691-1696.
Huang, L. F., F. Q. Guo, Y. Z. Liang, B. Y. Li and B. M. Cheng (2004). "Simultaneous determination of L-arginine and its mono- and dimethylated metabolites in human plasma by high-performance liquid chromatography-mass spectrometry." Anal Bioanal Chem 380(4): 643-649.
Hucke, C., C. R. MacKenzie, K. D. Adjogble, O. Takikawa and W. Daubener (2004). "Nitric oxide-mediated regulation of gamma interferon-induced bacteriostasis: inhibition and degradation of human indoleamine 2,3-dioxygenase." Infect Immun 72(5): 2723-2730.
Hudlicka, O. (1985). "Regulation of muscle blood flow." Clin Physiol 5(3): 201-229. Huengsberg, M., J. B. Winer, M. Gompels, R. Round, J. Ross and M. Shahmanesh
(1998). "Serum kynurenine-to-tryptophan ratio increases with progressive disease in HIV-infected patients." Clin Chem 44(4): 858-862.
Hunsicker, A., W. Kullich, W. Weissenhofer, D. Lorenz, J. Petermann, H. Rokos and G. Schwesinger (1997). "Correlations between endotoxin, interferon-gamma, biopterin and serum phospholipase A2-activities during lethal gram negative sepsis in rats." Eur J Surg 163(5): 379-385.
Husson, A., M. Bouazza, C. Buquet and R. Vaillant (1984). "Precocious induction of arginase in primary cultures of fetal rat hepatocytes." In Vitro 20(4): 314-320.
Huttunen, R., J. Syrjanen, J. Aittoniemi, S. S. Oja, A. Raitala, J. Laine, M. Pertovaara, R. Vuento, H. Huhtala and M. Hurme (2009). "High activity of indoleamine 2,3 dioxygenase enzyme predicts disease severity and case fatality in bacteremic patients." Shock.
Ignarro, L. J., G. M. Buga, K. S. Wood, R. E. Byrns and G. Chaudhuri (1987). "Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide." Proc Natl Acad Sci U S A 84(24): 9265-9269.
Ince, C. (2005). "The microcirculation is the motor of sepsis." Crit Care 9 Suppl 4: S13-19.
Ince, C. and M. Sinaasappel (1999). "Microcirculatory oxygenation and shunting in sepsis and shock." Crit Care Med 27: 1369 - 1377.
Ino, K., E. Yamamoto, K. Shibata, H. Kajiyama, N. Yoshida, M. Terauchi, A. Nawa, T. Nagasaka, O. Takikawa and F. Kikkawa (2008). "Inverse correlation between tumoral indoleamine 2,3-dioxygenase expression and tumor-infiltrating lymphocytes in endometrial cancer: its association with disease progression and survival." Clin Cancer Res 14(8): 2310-2317.
205
Ito, A., P. S. Tsao, S. Adimoolam, M. Kimoto, T. Ogawa and J. P. Cooke (1999). "Novel mechanism for endothelial dysfunction: dysregulation of dimethylarginine dimethylaminohydrolase." Circulation 99(24): 3092-3095.
Iwamoto, N., H. Ito, K. Ando, T. Ishikawa, A. Hara, A. Taguchi, K. Saito, M. Takemura, M. Imawari, H. Moriwaki and M. Seishima (2009). "Upregulation of indoleamine 2,3-dioxygenase in hepatocyte during acute hepatitis caused by hepatitis B virus-specific cytotoxic T lymphocytes in vivo." Liver Int 29(2): 277-283.
Jarvisalo, M. J., L. Jartti, J. Marniemi, T. Ronnemaa, J. S. Viikari, T. Lehtimaki and O. T. Raitakari (2006). "Determinants of short-term variation in arterial flow-mediated dilatation in healthy young men." Clin Sci (Lond) 110(4): 475-482.
Jia, W., C. Jackson-Cook and M. R. Graf (2010). "Tumor-infiltrating, myeloid-derived suppressor cells inhibit T cell activity by nitric oxide production in an intracranial rat glioma + vaccination model." J Neuroimmunol 223(1-2): 20-30.
Jia, Y. X., C. S. Pan, J. H. Yang, X. H. Liu, W. J. Yuan, J. Zhao, C. S. Tang and Y. F. Qi (2006). "Altered L-arginine/nitric oxide synthase/nitric oxide pathway in the vascular adventitia of rats with sepsis." Clin Exp Pharmacol Physiol 33(12): 1202-1208.
Jones, C. E., C. J. Darcy, T. Woodberry, N. M. Anstey and Y. R. McNeil (2010). "HPLC analysis of asymmetric dimethylarginine, symmetric dimethylarginine, homoarginine and arginine in small plasma volumes using a Gemini-NX column at high pH." J Chromatogr B Analyt Technol Biomed Life Sci 878(1): 8-12.
Jung, I. D., M. G. Lee, J. H. Chang, J. S. Lee, Y. I. Jeong, C. M. Lee, W. S. Park, J. Han, S. K. Seo, S. Y. Lee and Y. M. Park (2009). "Blockade of indoleamine 2,3-dioxygenase protects mice against lipopolysaccharide-induced endotoxin shock." J Immunol 182(5): 3146-3154.
Juonala, M., J. S. Viikari, G. Alfthan, J. Marniemi, M. Kahonen, L. Taittonen, T. Laitinen and O. T. Raitakari (2007). "Brachial artery flow-mediated dilation and asymmetrical dimethylarginine in the cardiovascular risk in young Finns study." Circulation 116(12): 1367-1373.
Kakoki, M., H. S. Kim, C. J. Edgell, N. Maeda, O. Smithies and D. L. Mattson (2006). "Amino acids as modulators of endothelium-derived nitric oxide." Am J Physiol Renal Physiol 291(2): F297-304.
Kalil, A. C. and R. L. Danner (2006). "L-Arginine supplementation in sepsis: beneficial or harmful?" Curr Opin Crit Care 12(4): 303-308.
Kalil, A. C., J. E. Sevransky, D. E. Myers, C. Esposito, R. W. Vandivier, P. Eichacker, G. M. Susla, S. B. Solomon, G. Csako, R. Costello, K. J. Sittler, S. Banks, C. Natanson and R. L. Danner (2006). "Preclinical trial of L-arginine monotherapy alone or with N-acetylcysteine in septic shock." Crit Care Med 34(11): 2719-2728.
Kao, C. C., V. Bandi, K. K. Guntupalli, M. Wu, L. Castillo and F. Jahoor (2008). "Arginine, citrulline, and nitric oxide metabolism in sepsis." Clin Sci (Lond).
Kao, C. C., V. Bandi, K. K. Guntupalli, M. Wu, L. Castillo and F. Jahoor (2009). "Arginine, citrulline and nitric oxide metabolism in sepsis." Clin Sci (Lond) 117(1): 23-30.
206
Kielstein, J. T., R. H. Boger, S. M. Bode-Boger, J. C. Frolich, H. Haller, E. Ritz and D. Fliser (2002). "Marked increase of asymmetric dimethylarginine in patients with incipient primary chronic renal disease." J Am Soc Nephrol 13(1): 170-176.
Kielstein, J. T., S. R. Salpeter, S. M. Bode-Boeger, J. P. Cooke and D. Fliser (2006). "Symmetric dimethylarginine (SDMA) as endogenous marker of renal function--a meta-analysis." Nephrol Dial Transplant 21(9): 2446-2451.
Knaus, W. A., J. E. Zimmerman, D. P. Wagner, E. A. Draper and D. E. Lawrence (1981). "APACHE-acute physiology and chronic health evaluation: a physiologically based classification system." Crit Care Med 9(8): 591-597.
Krzyzanowska, K., F. Mittermayer, N. Shnawa, M. Hofer, J. Schnabler, Y. Etmuller, S. Kapiotis, M. Wolzt and G. Schernthaner (2007). "Asymmetrical dimethylarginine is related to renal function, chronic inflammation and macroangiopathy in patients with Type 2 diabetes and albuminuria." Diabet Med 24(1): 81-86.
Kubli, S., Y. Boegli, A. D. Ave, L. Liaudet, J. P. Revelly, S. Golay, A. Broccard, B. Waeber, M. D. Schaller and F. Feihl (2003). "Endothelium-dependent vasodilation in the skin microcirculation of patients with septic shock." Shock 19(3): 274-280.
Kudo, Y., C. A. Boyd, I. L. Sargent and C. W. Redman (2003). "Decreased tryptophan catabolism by placental indoleamine 2,3-dioxygenase in preeclampsia." Am J Obstet Gynecol 188(3): 719-726.
Kutza, A. S., E. Muhl, H. Hackstein, H. Kirchner and G. Bein (1998). "High incidence of active cytomegalovirus infection among septic patients." Clin Infect Dis 26(5): 1076-1082.
Kuvin, J. T., A. Mammen, P. Mooney, A. A. Alsheikh-Ali and R. H. Karas (2007). "Assessment of peripheral vascular endothelial function in the ambulatory setting." Vasc Med 12(1): 13-16.
Kuvin, J. T., A. R. Patel, K. A. Sliney, N. G. Pandian, J. Sheffy, R. P. Schnall, R. H. Karas and J. E. Udelson (2003). "Assessment of peripheral vascular endothelial function with finger arterial pulse wave amplitude." Am Heart J 146(1): 168-174.
Lan, R. Y., C. Selmi and M. E. Gershwin (2008). "The regulatory, inflammatory, and T cell programming roles of interleukin-2 (IL-2)." J Autoimmun 31(1): 7-12.
Lang, C. H., G. J. Bagby, J. L. Ferguson and J. J. Spitzer (1984). "Cardiac output and redistribution of organ blood flow in hypermetabolic sepsis." Am J Physiol 246(3 Pt 2): R331-337.
Larrea, E., J. I. Riezu-Boj, L. Gil-Guerrero, N. Casares, R. Aldabe, P. Sarobe, M. P. Civeira, J. L. Heeney, C. Rollier, B. Verstrepen, T. Wakita, F. Borras-Cuesta, J. J. Lasarte and J. Prieto (2007). "Upregulation of indoleamine 2,3-dioxygenase in hepatitis C virus infection." J Virol 81(7): 3662-3666.
Le Tulzo, Y., C. Pangault, A. Gacouin, V. Guilloux, O. Tribut, L. Amiot, P. Tattevin, R. Thomas, R. Fauchet and B. Drenou (2002). "Early circulating lymphocyte apoptosis in human septic shock is associated with poor outcome." Shock 18(6): 487-494.
Lechner, M. G., D. J. Liebertz and A. L. Epstein (2010). "Characterization of cytokine-induced myeloid-derived suppressor cells from normal human peripheral blood mononuclear cells." J Immunol 185(4): 2273-2284.
207
Lee, G. K., H. J. Park, M. Macleod, P. Chandler, D. H. Munn and A. L. Mellor (2002). "Tryptophan deprivation sensitizes activated T cells to apoptosis prior to cell division." Immunology 107(4): 452-460.
Leiper, J., M. Nandi, B. Torondel, J. Murray-Rust, M. Malaki, B. O'Hara, S. Rossiter, S. Anthony, M. Madhani, D. Selwood, C. Smith, B. Wojciak-Stothard, A. Rudiger, R. Stidwill, N. Q. McDonald and P. Vallance (2007). "Disruption of methylarginine metabolism impairs vascular homeostasis." Nat Med 13(2): 198-203.
Li, P., Y. L. Yin, D. Li, S. Woo Kim and G. Wu (2007). "Amino acids and immune function." Br J Nutr: 1-16.
Limaye, A. P., K. A. Kirby, G. D. Rubenfeld, W. M. Leisenring, E. M. Bulger, M. J. Neff, N. S. Gibran, M. L. Huang, T. K. Santo Hayes, L. Corey and M. Boeckh (2008). "Cytomegalovirus reactivation in critically ill immunocompetent patients." Jama 300(4): 413-422.
Liossis, S. N., X. Z. Ding, G. J. Dennis and G. C. Tsokos (1998). "Altered pattern of TCR/CD3-mediated protein-tyrosyl phosphorylation in T cells from patients with systemic lupus erythematosus. Deficient expression of the T cell receptor zeta chain." J Clin Invest 101(7): 1448-1457.
Lob, S., A. Konigsrainer, H. G. Rammensee, G. Opelz and P. Terness (2009). "Inhibitors of indoleamine-2,3-dioxygenase for cancer therapy: can we see the wood for the trees?" Nat Rev Cancer 9(6): 445-452.
Lopez, A., J. A. Lorente, J. Steingrub, J. Bakker, A. McLuckie, S. Willatts, M. Brockway, A. Anzueto, L. Holzapfel, D. Breen, M. S. Silverman, J. Takala, J. Donaldson, C. Arneson, G. Grove, S. Grossman and R. Grover (2004). "Multiple-center, randomized, placebo-controlled, double-blind study of the nitric oxide synthase inhibitor 546C88: effect on survival in patients with septic shock." Crit Care Med 32(1): 21-30.
Lorente, J. A., L. Landin, R. De Pablo, E. Renes and D. Liste (1993). "L-arginine pathway in the sepsis syndrome." Crit Care Med 21(9): 1287-1295.
Lu, J. L., L. M. Schmiege, 3rd, L. Kuo and J. C. Liao (1996). "Downregulation of endothelial constitutive nitric oxide synthase expression by lipopolysaccharide." Biochem Biophys Res Commun 225(1): 1-5.
Luiking, Y., M. Poeze, M. Hendrikx, P. Breedveld, C. Dejong, P. de Feiter, F. Rubulotta, G. Ramsay and N. Deutz (2006). "Continuous L-arginine infusion does not deteriorate the haemodynamic condition in patients with severe sepsis." European Journal of Gastroenterology & Hepatology 18(1): A8.
Luiking, Y. C. and N. E. Deutz (2007). "Exogenous arginine in sepsis." Crit Care Med 35(9 Suppl): S557-563.
Luiking, Y. C., M. Poeze, C. H. Dejong, G. Ramsay and N. E. Deutz (2004). "Sepsis: an arginine deficiency state?" Crit Care Med 32(10): 2135-2145.
Luiking, Y. C., M. Poeze, G. Ramsay and N. E. Deutz (2009). "Reduced citrulline production in sepsis is related to diminished de novo arginine and nitric oxide production." Am J Clin Nutr 89(1): 142-152.
Luyt, C. E., A. Combes, C. Deback, M. H. Aubriot-Lorton, A. Nieszkowska, J. L. Trouillet, F. Capron, H. Agut, C. Gibert and J. Chastre (2007). "Herpes simplex virus lung infection in patients undergoing prolonged mechanical ventilation." Am J Respir Crit Care Med 175(9): 935-942.
208
Lyn-Kew, K. and T. J. Standiford (2008). "Immunosuppression in sepsis." Curr Pharm Des 14(19): 1870-1881.
MacKenzie, C. R., R. G. Gonzalez, E. Kniep, S. Roch and W. Daubener (1999). "Cytokine mediated regulation of interferon-gamma-induced IDO activation." Adv Exp Med Biol 467: 533-539.
MacKenzie, C. R., U. Hadding and W. Daubener (1998). "Interferon-gamma-induced activation of indoleamine 2,3-dioxygenase in cord blood monocyte-derived macrophages inhibits the growth of group B streptococci." J Infect Dis 178(3): 875-878.
MacLean, L. D., J. L. Meakins, K. Taguchi, J. P. Duignan, K. S. Dhillon and J. Gordon (1975). "Host resistance in sepsis and trauma." Ann Surg 182(3): 207-217.
Madden, H. P., R. J. Breslin, H. L. Wasserkrug, G. Efron and A. Barbul (1988). "Stimulation of T cell immunity by arginine enhances survival in peritonitis." J Surg Res 44(6): 658-663.
Maes, M., H. Y. Meltzer, S. Scharpe, E. Bosmans, E. Suy, I. De Meester, J. Calabrese and P. Cosyns (1993). "Relationships between lower plasma L-tryptophan levels and immune-inflammatory variables in depression." Psychiatry Res 49(2): 151-165.
Mandruzzato, S., S. Solito, E. Falisi, S. Francescato, V. Chiarion-Sileni, S. Mocellin, A. Zanon, C. R. Rossi, D. Nitti, V. Bronte and P. Zanovello (2009). "IL4Ralpha+ myeloid-derived suppressor cell expansion in cancer patients." J Immunol 182(10): 6562-6568.
Maneechotesuwan, K., S. Supawita, K. Kasetsinsombat, A. Wongkajornsilp and P. J. Barnes (2008). "Sputum indoleamine-2, 3-dioxygenase activity is increased in asthmatic airways by using inhaled corticosteroids." J Allergy Clin Immunol 121(1): 43-50.
Marshall, J. C. and K. Reinhart (2009). "Biomarkers of sepsis." Crit Care Med 37(7): 2290-2298.
Martens-Lobenhoffer, J. and S. M. Bode-Boger (2006). "Measurement of asymmetric dimethylarginine (ADMA) in human plasma: from liquid chromatography estimation to liquid chromatography-mass spectrometry quantification." Eur J Clin Pharmacol 62(Supplement 13): 61-68.
Martens-Lobenhoffer, J. and S. M. Bode-Boger (2007). "Chromatographic-mass spectrometric methods for the quantification of L-arginine and its methylated metabolites in biological fluids." J Chromatogr B Analyt Technol Biomed Life Sci 851(1-2): 30-41.
Marti-Carvajal, A., G. Salanti and A. F. Cardona (2007). "Human recombinant activated protein C for severe sepsis." Cochrane Database Syst Rev(3): CD004388.
Martin, G. S., D. M. Mannino, S. Eaton and M. Moss (2003). "The epidemiology of sepsis in the United States from 1979 through 2000." N Engl J Med 348(16): 1546-1554.
Matsuda, M., A. K. Ulfgren, R. Lenkei, M. Petersson, A. C. Ochoa, S. Lindblad, P. Andersson, L. Klareskog and R. Kiessling (1998). "Decreased expression of signal-transducing CD3 zeta chains in T cells from the joints and peripheral blood of rheumatoid arthritis patients." Scand J Immunol 47(3): 254-262.
209
Mazzoni, A., V. Bronte, A. Visintin, J. H. Spitzer, E. Apolloni, P. Serafini, P. Zanovello and D. M. Segal (2002). "Myeloid suppressor lines inhibit T cell responses by an NO-dependent mechanism." J Immunol 168(2): 689-695.
McDonald, K. K., S. Zharikov, E. R. Block and M. S. Kilberg (1997). "A caveolar complex between the cationic amino acid transporter 1 and endothelial nitric-oxide synthase may explain the "arginine paradox"." J Biol Chem 272(50): 31213-31216.
McGown, C. C. and Z. L. Brookes (2007). "Beneficial effects of statins on the microcirculation during sepsis: the role of nitric oxide." Br J Anaesth 98(2): 163-175.
Meakins, J. L., J. B. Pietsch, O. Bubenick, R. Kelly, H. Rode, J. Gordon and L. D. MacLean (1977). "Delayed hypersensitivity: indicator of acquired failure of host defenses in sepsis and trauma." Ann Surg 186(3): 241-250.
Medzhitov, R. (2010). "Inflammation 2010: new adventures of an old flame." Cell 140(6): 771-776.
Mellion, B. T., L. J. Ignarro, E. H. Ohlstein, E. G. Pontecorvo, A. L. Hyman and P. J. Kadowitz (1981). "Evidence for the inhibitory role of guanosine 3', 5'-monophosphate in ADP-induced human platelet aggregation in the presence of nitric oxide and related vasodilators." Blood 57(5): 946-955.
Merino, E., R. A. Jensen and C. Yanofsky (2008). "Evolution of bacterial trp operons and their regulation." Curr Opin Microbiol 11(2): 78-86.
Mitchell, J. A., F. Ali, L. Bailey, L. Moreno and L. S. Harrington (2008). "Role of nitric oxide and prostacyclin as vasoactive hormones released by the endothelium." Exp Physiol 93(1): 141-147.
Mittermayer, F., K. Namiranian, J. Pleiner, G. Schaller and M. Wolzt (2004). "Acute Escherichia coli endotoxaemia decreases the plasma l-arginine/asymmetrical dimethylarginine ratio in humans." Clin Sci (Lond) 106(6): 577-581.
Moali, C., J. L. Boucher, M. A. Sari, D. J. Stuehr and D. Mansuy (1998). "Substrate specificity of NO synthases: detailed comparison of L-arginine, homo-L-arginine, their N omega-hydroxy derivatives, and N omega-hydroxynor-L-arginine." Biochemistry 37(29): 10453-10460.
Mohib, K., Q. Guan, H. Diao, C. Du and A. M. Jevnikar (2007). "Proapoptotic activity of indoleamine 2,3-dioxygenase expressed in renal tubular epithelial cells." Am J Physiol Renal Physiol 293(3): F801-812.
Moncada, S. and E. A. Higgs (2006). "Nitric oxide and the vascular endothelium." Handb Exp Pharmacol(176 Pt 1): 213-254.
Mookerjee, R. P., M. Malaki, N. A. Davies, S. J. Hodges, R. N. Dalton, C. Turner, S. Sen, R. Williams, J. Leiper, P. Vallance and R. Jalan (2007). "Increasing dimethylarginine levels are associated with adverse clinical outcome in severe alcoholic hepatitis." Hepatology 45(1): 62-71.
Mori, M. and T. Gotoh (2004). "Arginine metabolic enzymes, nitric oxide and infection." J Nutr 134(10 Suppl): 2820S-2825S; discussion 2853S.
Morris, C. R., G. J. Kato, M. Poljakovic, X. Wang, W. C. Blackwelder, V. Sachdev, S. L. Hazen, E. P. Vichinsky, S. M. Morris, Jr. and M. T. Gladwin (2005). "Dysregulated arginine metabolism, hemolysis-associated pulmonary hypertension, and mortality in sickle cell disease." Jama 294(1): 81-90.
210
Morris, S. M., Jr. (2007). "Arginine metabolism: boundaries of our knowledge." J Nutr 137(6 Suppl 2): 1602S-1609S.
Morris, S. M., Jr. (2009). "Recent advances in arginine metabolism: roles and regulation of the arginases." Br J Pharmacol 157(6): 922-930.
Morrison, R. D. and J. W. Dolan (2005). "Peak Fronting, Column Life, and Column Conditioning." LCGC North America Jun.
Movahedi, K., M. Guilliams, J. Van den Bossche, R. Van den Bergh, C. Gysemans, A. Beschin, P. De Baetselier and J. A. Van Ginderachter (2008). "Identification of discrete tumor-induced myeloid-derived suppressor cell subpopulations with distinct T cell-suppressive activity." Blood 111(8): 4233-4244.
Moyer, E. D., R. H. McMenamy, F. B. Cerra, R. A. Reed, L. Yu, R. Chenier, J. Caruana and J. R. Border (1981). "Multiple systems organ failure: III Contrasts in plasma amino acid profiles in septic trauma patients who subsequently survive and do not survive-effects of intravenous amino acids." J Trauma 21(4): 263-274.
Muhling, J., M. E. Campos, A. Sablotzki, M. Krull, M. G. Dehne, J. Gonther, S. Weiss, M. Fuchs and G. Hempelmann (2002). "Effects of propofol and taurine on intracellular free amino acid profiles and immune function markers in neutrophils in vitro." Clin Chem Lab Med 40(2): 111-121.
Muller, A. J., M. D. Sharma, P. R. Chandler, J. B. Duhadaway, M. E. Everhart, B. A. Johnson, 3rd, D. J. Kahler, J. Pihkala, A. P. Soler, D. H. Munn, G. C. Prendergast and A. L. Mellor (2008). "Chronic inflammation that facilitates tumor progression creates local immune suppression by inducing indoleamine 2,3 dioxygenase." Proc Natl Acad Sci U S A 105(44): 17073-17078.
Munder, M., K. Eichmann, J. M. Moran, F. Centeno, G. Soler and M. Modolell (1999). "Th1/Th2-regulated expression of arginase isoforms in murine macrophages and dendritic cells." J Immunol 163(7): 3771-3777.
Munder, M., F. Mollinedo, J. Calafat, J. Canchado, C. Gil-Lamaignere, J. M. Fuentes, C. Luckner, G. Doschko, G. Soler, K. Eichmann, F.-M. Muller, A. D. Ho, M. Goerner and M. Modolell (2005). "Arginase I is constitutively expressed in human granulocytes and participates in fungicidal activity." Blood 105(6): 2549-2556.
Munder, M., H. Schneider, C. Luckner, T. Giese, C. D. Langhans, J. M. Fuentes, P. Kropf, I. Mueller, A. Kolb, M. Modolell and A. D. Ho (2006). "Suppression of T-cell functions by human granulocyte arginase." Blood 108(5): 1627-1634.
Munn, D. H., E. Shafizadeh, J. T. Attwood, I. Bondarev, A. Pashine and A. L. Mellor (1999). "Inhibition of T cell proliferation by macrophage tryptophan catabolism." J Exp Med 189(9): 1363-1372.
Munn, D. H., M. D. Sharma, B. Baban, H. P. Harding, Y. Zhang, D. Ron and A. L. Mellor (2005). "GCN2 kinase in T cells mediates proliferative arrest and anergy induction in response to indoleamine 2,3-dioxygenase." Immunity 22(5): 633-642.
Munn, D. H., M. D. Sharma, J. R. Lee, K. G. Jhaver, T. S. Johnson, D. B. Keskin, B. Marshall, P. Chandler, S. J. Antonia, R. Burgess, C. L. Slingluff, Jr. and A. L. Mellor (2002). "Potential regulatory function of human dendritic cells expressing indoleamine 2,3-dioxygenase." Science 297(5588): 1867-1870.
211
Munn, D. H., M. Zhou, J. T. Attwood, I. Bondarev, S. J. Conway, B. Marshall, C. Brown and A. L. Mellor (1998). "Prevention of allogeneic fetal rejection by tryptophan catabolism." Science 281(5380): 1191-1193.
Murphy, K., P. Travers and M. Walport (2008). Janeway's Immunobiology. New York, Garland Science.
Nagaraj, S., M. Collazo, C. A. Corzo, J. I. Youn, M. Ortiz, D. Quiceno and D. I. Gabrilovich (2009). "Regulatory myeloid suppressor cells in health and disease." Cancer Res 69(19): 7503-7506.
Nakamura, T., E. Sato, N. Fujiwara, Y. Kawagoe, T. Suzuki, Y. Ueda, S. Yamada, H. Shoji, M. Takeuchi, S. Ueda, T. Matsui, H. Adachi, S. Okuda and S. I. Yamagishi (2009). "Circulating levels of advanced glycation end products (AGE) and interleukin-6 (IL-6) are independent determinants of serum asymmetric dimethylarginine (ADMA) levels in patients with septic shock." Pharmacol Res.
Neviere, R., D. Mathieu, J. L. Chagnon, N. Lebleu, J. P. Millien and F. Wattel (1996). "Skeletal muscle microvascular blood flow and oxygen transport in patients with severe sepsis." Am J Respir Crit Care Med 153(1): 191-195.
Nijveldt, R. J., M. P. Siroen, T. Teerlink and P. A. van Leeuwen (2004). "Elimination of asymmetric dimethylarginine by the kidney and the liver: a link to the development of multiple organ failure?" J Nutr 134(10 Suppl): 2848S-2852S; discussion 2853S.
Nijveldt, R. J., T. Teerlink, M. P. Siroen, A. A. van Lambalgen, J. A. Rauwerda and P. A. van Leeuwen (2003). "The liver is an important organ in the metabolism of asymmetrical dimethylarginine (ADMA)." Clin Nutr 22(1): 17-22.
Nijveldt, R. J., T. Teerlink, B. Van Der Hoven, M. P. Siroen, D. J. Kuik, J. A. Rauwerda and P. A. van Leeuwen (2003). "Asymmetrical dimethylarginine (ADMA) in critically ill patients: high plasma ADMA concentration is an independent risk factor of ICU mortality." Clin Nutr 22(1): 23-30.
Nijveldt, R. J., T. Teerlink, C. van Guldener, H. A. Prins, A. A. van Lambalgen, C. D. Stehouwer, J. A. Rauwerda and P. A. van Leeuwen (2003). "Handling of asymmetrical dimethylarginine and symmetrical dimethylarginine by the rat kidney under basal conditions and during endotoxaemia." Nephrol Dial Transplant 18(12): 2542-2550.
Nijveldt, R. J., T. Teerlink and P. A. van Leeuwen (2003). "The asymmetrical dimethylarginine (ADMA)-multiple organ failure hypothesis." Clin Nutr 22(1): 99-104.
Nohria, A., M. Gerhard-Herman, M. A. Creager, S. Hurley, D. Mitra and P. Ganz (2006). "Role of nitric oxide in the regulation of digital pulse volume amplitude in humans." J Appl Physiol 101(2): 545-548.
Nonaka, S., M. Tsunoda, K. Imai and T. Funatsu (2005). "High-performance liquid chromatographic assay of N(G)-monomethyl-L-arginine, N(G),N(G)-dimethyl-L-arginine, and N(G),N(G)'-dimethyl-L-arginine using 4-fluoro-7-nitro-2, 1,3-benzoxadiazole as a fluorescent reagent." J Chromatogr A 1066(1-2): 41-45.
Nuttall, K. L., M. Chen and G. Komaromy-Hiller (1998). "Delayed separation and the plasma amino acids arginine and ornithine." Ann Clin Lab Sci 28(6): 354-359.
Nuttall, M. E., A. J. Patton, D. L. Olivera, D. P. Nadeau and M. Gowen (1998). "Human trabecular bone cells are able to express both osteoblastic and adipocytic
212
phenotype: implications for osteopenic disorders." J Bone Miner Res 13(3): 371-382.
O'Dwyer, M. J., F. Dempsey, V. Crowley, D. P. Kelleher, R. McManus and T. Ryan (2006). "Septic shock is correlated with asymmetrical dimethyl arginine levels, which may be influenced by a polymorphism in the dimethylarginine dimethylaminohydrolase II gene: a prospective observational study." Crit Care 10(5): R139.
Oberholzer, A., S. M. Souza, S. K. Tschoeke, C. Oberholzer, A. Abouhamze, J. P. Pribble and L. L. Moldawer (2005). "Plasma cytokine measurements augment prognostic scores as indicators of outcome in patients with severe sepsis." Shock 23(6): 488-493.
Ochoa, A. C., A. H. Zea, C. Hernandez and P. C. Rodriguez (2007). "Arginase, prostaglandins, and myeloid-derived suppressor cells in renal cell carcinoma." Clin Cancer Res 13(2 Pt 2): 721s-726s.
Ochoa, J. B., A. O. Udekwu, T. R. Billiar, R. D. Curran, F. B. Cerra, R. L. Simmons and A. B. Peitzman (1991). "Nitrogen oxide levels in patients after trauma and during sepsis." Ann Surg 214(5): 621-626.
Odemuyiwa, S. O., A. Ghahary, Y. Li, L. Puttagunta, J. E. Lee, S. Musat-Marcu and R. Moqbel (2004). "Cutting edge: human eosinophils regulate T cell subset selection through indoleamine 2,3-dioxygenase." J Immunol 173(10): 5909-5913.
Ogawa, T., M. Kimoto and K. Sasaoka (1989). "Purification and properties of a new enzyme, NG,NG-dimethylarginine dimethylaminohydrolase, from rat kidney." J Biol Chem 264(17): 10205-10209.
Oh, G. S., H. O. Pae, B. M. Choi, S. C. Chae, H. S. Lee, D. G. Ryu and H. T. Chung (2004). "3-Hydroxyanthranilic acid, one of metabolites of tryptophan via indoleamine 2,3-dioxygenase pathway, suppresses inducible nitric oxide synthase expression by enhancing heme oxygenase-1 expression." Biochem Biophys Res Commun 320(4): 1156-1162.
Orabona, C., M. T. Pallotta, C. Volpi, F. Fallarino, C. Vacca, R. Bianchi, M. L. Belladonna, M. C. Fioretti, U. Grohmann and P. Puccetti (2008). "SOCS3 drives proteasomal degradation of indoleamine 2,3-dioxygenase (IDO) and antagonizes IDO-dependent tolerogenesis." Proc Natl Acad Sci U S A 105(52): 20828-20833.
Oreiro-Garcia, M. T., M. D. Vazquez-Illanes, G. Sierra-Paredes and G. Sierra-Marcuno (2005). "Analysis of neuroactive amino acids from microdialysate samples by fluorescence detection using a modification of the 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate method." Biomed Chromatogr 19(10): 720-724.
Ostrand-Rosenberg, S. and P. Sinha (2009). "Myeloid-derived suppressor cells: linking inflammation and cancer." J Immunol 182(8): 4499-4506.
Palmer, R. M., A. G. Ferrige and S. Moncada (1987). "Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor." Nature 327(6122): 524-526.
Pasini, A. F., U. Garbin, C. Stranieri, V. Boccioletti, C. Mozzini, S. Manfro, A. Pasini, M. Cominacini and L. Cominacini (2008). "Nebivolol treatment reduces serum levels of asymmetric dimethylarginine and improves endothelial dysfunction in essential hypertensive patients." Am J Hypertens 21(11): 1251-1257.
213
Pawlak, D., A. Tankiewicz, T. Matys and W. Buczko (2003). "Peripheral distribution of kynurenine metabolites and activity of kynurenine pathway enzymes in renal failure." J Physiol Pharmacol 54(2): 175-189.
Pellegrin, K., G. Neurauter, B. Wirleitner, A. W. Fleming, V. M. Peterson and D. Fuchs (2005). "Enhanced enzymatic degradation of tryptophan by indoleamine 2,3-dioxygenase contributes to the tryptophan-deficient state seen after major trauma." Shock 23(3): 209-215.
Petersen, P. H., C. G. Fraser, L. Jorgensen, I. Brandslund, M. Stahl, E. Gowans, J. C. Libeer and C. Ricos (2002). "Combination of analytical quality specifications based on biological within- and between-subject variation." Ann Clin Biochem 39(Pt 6): 543-550.
Popovic, P. J., H. J. Zeh, 3rd and J. B. Ochoa (2007). "Arginine and immunity." J Nutr 137(6 Suppl 2): 1681S-1686S.
Prins, H. A., A. P. Houdijk, R. J. Nijveldt, T. Teerlink, P. Huygens, L. G. Thijs and P. A. van Leeuwen (2001). "Arginase release from red blood cells: possible link in transfusion induced immune suppression?" Shock 16(2): 113-115.
Puchau, B., H. H. Hermsdorff, M. A. Zulet and J. A. Martinez (2009). "DDAH2 mRNA expression is inversely associated with some cardiovascular risk-related features in healthy young adults." Dis Markers 27(1): 37-44.
Remick, D. G. (2007). "Pathophysiology of sepsis." Am J Pathol 170(5): 1435-1444. Reverter, M., T. Lundh and J. E. Lindberg (1997). "Determination of free amino acids in
pig plasma by precolumn derivatization with 6-N-aminoquinolyl-N-hydroxysuccinimidyl carbamate and high-performance liquid chromatography." Journal of Chromatography B: Biomedical Sciences and Applications 696(1): 1-8.
Richir, M. C., A. A. van Lambalgen, T. Teerlink, W. Wisselink, E. Bloemena, H. A. Prins, T. P. de Vries and P. A. van Leeuwen (2009). "Low arginine/asymmetric dimethylarginine ratio deteriorates systemic hemodynamics and organ blood flow in a rat model." Crit Care Med 37(6): 2010-2017.
Rodriguez, P. C., M. S. Ernstoff, C. Hernandez, M. Atkins, J. Zabaleta, R. Sierra and A. C. Ochoa (2009). "Arginase I-producing myeloid-derived suppressor cells in renal cell carcinoma are a subpopulation of activated granulocytes." Cancer Res 69(4): 1553-1560.
Rodriguez, P. C., C. P. Hernandez, D. Quiceno, S. M. Dubinett, J. Zabaleta, J. B. Ochoa, J. Gilbert and A. C. Ochoa (2005). "Arginase I in myeloid suppressor cells is induced by COX-2 in lung carcinoma." J Exp Med 202(7): 931-939.
Rodriguez, P. C., D. G. Quiceno and A. C. Ochoa (2007). "L-arginine availability regulates T-lymphocyte cell-cycle progression." Blood 109(4): 1568-1573.
Rodriguez, P. C., A. H. Zea, K. S. Culotta, J. Zabaleta, J. B. Ochoa and A. C. Ochoa (2002). "Regulation of T cell receptor CD3zeta chain expression by L-arginine." J Biol Chem 277(24): 21123-21129.
Rodriguez, P. C., A. H. Zea, J. DeSalvo, K. S. Culotta, J. Zabaleta, D. G. Quiceno, J. B. Ochoa and A. C. Ochoa (2003). "L-arginine consumption by macrophages modulates the expression of CD3 zeta chain in T lymphocytes." J Immunol 171(3): 1232-1239.
Romagnani, S. (2006). "Regulation of the T cell response." Clin Exp Allergy 36(11): 1357-1366.
214
Rose, W. C. (1937). "The Nutritive Significance of the Amino Acids and Certain Related Compounds." Science 86(2231): 298-300.
Rosemblatt, M. and M. R. Bono (2004). "Functional consequences of immune cell adhesion to endothelial cells." Curr Pharm Des 10(2): 109-120.
Rustum, A. M. and V. Estrada (1998). "Separation and quantitation of the S-(+)-enantiomer in the bulk drug tiagabine x HCl by chiral high-performance-liquid chromatography using a Chiralcel-OD column." J Chromatogr B Biomed Sci Appl 705(1): 111-117.
Sahai, S. and S. Uhlhaas (1985). "Stability of amino acids in human plasma." Clin Chim Acta 148(3): 255-259.
Samelson-Jones, B. J. and S. R. Yeh (2006). "Interactions between nitric oxide and indoleamine 2,3-dioxygenase." Biochemistry 45(28): 8527-8538.
Sander, L. E., S. D. Sackett, U. Dierssen, N. Beraza, R. P. Linke, M. Muller, J. M. Blander, F. Tacke and C. Trautwein (2010). "Hepatic acute-phase proteins control innate immune responses during infection by promoting myeloid-derived suppressor cell function." J Exp Med.
Satriano, J., D. Schwartz, S. Ishizuka, M. J. Lortie, S. C. Thomson, F. Gabbai, C. J. Kelly and R. C. Blantz (2001). "Suppression of inducible nitric oxide generation by agmatine aldehyde: beneficial effects in sepsis." J Cell Physiol 188(3): 313-320.
Schaefer, A., F. Piquard and P. Haberey (1987). "Plasma amino-acids analysis: effects of delayed samples preparation and of storage." Clin Chim Acta 164(2): 163-169.
Schefold, J. C., J. P. Zeden, C. Fotopoulou, S. von Haehling, R. Pschowski, D. Hasper, H. D. Volk, C. Schuett and P. Reinke (2009). "Increased indoleamine 2,3-dioxygenase (IDO) activity and elevated serum levels of tryptophan catabolites in patients with chronic kidney disease: a possible link between chronic inflammation and uraemic symptoms." Nephrol Dial Transplant 24(6): 1901-1908.
Schefold, J. C., J. P. Zeden, R. Pschowski, B. Hammoud, C. Fotopoulou, D. Hasper, G. Fusch, S. Von Haehling, H. D. Volk, C. Meisel, C. Schutt and P. Reinke (2010). "Treatment with granulocyte-macrophage colony-stimulating factor is associated with reduced indoleamine 2,3-dioxygenase activity and kynurenine pathway catabolites in patients with severe sepsis and septic shock." Scand J Infect Dis 42(3): 164-171.
Schmielau, J. and O. J. Finn (2001). "Activated granulocytes and granulocyte-derived hydrogen peroxide are the underlying mechanism of suppression of t-cell function in advanced cancer patients." Cancer Res 61(12): 4756-4760.
Schnabel, R., S. Blankenberg, E. Lubos, K. J. Lackner, H. J. Rupprecht, C. Espinola-Klein, N. Jachmann, F. Post, D. Peetz, C. Bickel, F. Cambien, L. Tiret and T. Munzel (2005). "Asymmetric dimethylarginine and the risk of cardiovascular events and death in patients with coronary artery disease: results from the AtheroGene Study." Circ Res 97(5): e53-59.
Schroecksnadel, K., G. Weiss, O. Stanger, T. Teerlink and D. Fuchs (2007). Increased Asymmetric Dimethylarginine Concentrations in Stimulated Peripheral Blood Mononuclear Cells. Scandinavian Journal of Immunology. 0: ???-???
Schulze, F., H. Lenzen, C. Hanefeld, A. Bartling, K. J. Osterziel, L. Goudeva, C. Schmidt-Lucke, M. Kusus, R. Maas, E. Schwedhelm, D. Strodter, B. C. Simon,
215
A. Mugge, W. G. Daniel, H. Tillmanns, B. Maisch, T. Streichert and R. H. Boger (2006). "Asymmetric dimethylarginine is an independent risk factor for coronary heart disease: results from the multicenter Coronary Artery Risk Determination investigating the Influence of ADMA Concentration (CARDIAC) study." Am Heart J 152(3): 493 e491-498.
Schwulst, S. J., J. T. Muenzer, O. M. Peck-Palmer, K. C. Chang, C. G. Davis, J. S. McDonough, D. F. Osborne, A. H. Walton, J. Unsinger, J. E. McDunn and R. S. Hotchkiss (2008). "Bim siRNA decreases lymphocyte apoptosis and improves survival in sepsis." Shock 30(2): 127-134.
Searles, C. D. (2006). "Transcriptional and posttranscriptional regulation of endothelial nitric oxide synthase expression." Am J Physiol Cell Physiol 291(5): C803-816.
Seki, T., M. Naruse, K. Naruse, T. Yoshimoto, A. Tanabe, T. Imaki, H. Hagiwara, S. Hirose and H. Demura (1997). "Interrelation between nitric oxide synthase and heme oxygenase in rat endothelial cells." Eur J Pharmacol 331(1): 87-91.
Sekkai, D., O. Guittet, G. Lemaire, J. P. Tenu and M. Lepoivre (1997). "Inhibition of nitric oxide synthase expression and activity in macrophages by 3-hydroxyanthranilic acid, a tryptophan metabolite." Arch Biochem Biophys 340(1): 117-123.
Singer, M. (2008). "Cellular dysfunction in sepsis." Clin Chest Med 29(4): 655-660, viii-ix.
Sinha, P., V. K. Clements, A. M. Fulton and S. Ostrand-Rosenberg (2007). "Prostaglandin E2 promotes tumor progression by inducing myeloid-derived suppressor cells." Cancer Res 67(9): 4507-4513.
Sinha, P., C. Okoro, D. Foell, H. H. Freeze, S. Ostrand-Rosenberg and G. Srikrishna (2008). "Proinflammatory S100 proteins regulate the accumulation of myeloid-derived suppressor cells." J Immunol 181(7): 4666-4675.
Skotty, D. R., W.-Y. Lee and T. A. Nieman (1996). "Determination of Dansyl Amino Acids and Oxalate by HPLC with Electrogenerated Chemiluminescence Detection Using Tris(2,2‘-bipyridyl)ruthenium(II) in the Mobile Phase." Analytical Chemistry 68(9): 1530-1535.
Slade, E., P. S. Tamber and J. L. Vincent (2003). "The Surviving Sepsis Campaign: raising awareness to reduce mortality." Crit Care 7(1): 1-2.
Sonoki, T., A. Nagasaki, T. Gotoh, M. Takiguchi, M. Takeya, H. Matsuzaki and M. Mori (1997). "Coinduction of nitric-oxide synthase and arginase I in cultured rat peritoneal macrophages and rat tissues in vivo by lipopolysaccharide." J Biol Chem 272(6): 3689-3693.
Sprung, C. L., F. B. Cerra, H. R. Freund, R. M. Schein, F. N. Konstantinides, E. H. Marcial and M. Pena (1991). "Amino acid alterations and encephalopathy in the sepsis syndrome." Crit Care Med 19(6): 753-757.
Srivastava, M. K., P. Sinha, V. K. Clements, P. Rodriguez and S. Ostrand-Rosenberg (2009). "Myeloid-Derived Suppressor Cells Inhibit T-Cell Activation by Depleting Cystine and Cysteine." Cancer Res.
Stapleton, P. P., T. M. Mahon, P. Nowlan and F. J. Bloomfield (1994). "Effects of in-vivo administration of taurine and HEPES on the inflammatory response in rats." J Pharm Pharmacol 46(9): 745-750.
Stephens, D. P., J. H. Thomas, A. Higgins, M. Bailey, N. M. Anstey, B. J. Currie and A. C. Cheng (2008). "Randomized, double-blind, placebo-controlled trial of
216
granulocyte colony-stimulating factor in patients with septic shock." Crit Care Med 36(2): 448-454.
Stromberg, P. E., C. A. Woolsey, A. T. Clark, J. A. Clark, I. R. Turnbull, K. W. McConnell, K. C. Chang, C. S. Chung, A. Ayala, T. G. Buchman, R. S. Hotchkiss and C. M. Coopersmith (2009). "CD4+ lymphocytes control gut epithelial apoptosis and mediate survival in sepsis." Faseb J 23(1).
Strydom, D. J. and S. A. Cohen (1994). "Comparison of Amino Acid Analyses by Phenylisothiocyanate and 6-Aminoquinolyl-N-hydroxysuccinimidyl Carbamate Precolumn Derivatization." Analytical Biochemistry 222(1): 19-28.
Sud, N., S. M. Wells, S. Sharma, D. A. Wiseman, J. Wilham and S. M. Black (2008). "Asymmetric dimethylarginine inhibits HSP90 activity in pulmonary arterial endothelial cells: role of mitochondrial dysfunction." Am J Physiol Cell Physiol 294(6): C1407-1418.
Surdacki, A., M. Nowicki, J. Sandmann, D. Tsikas, R. H. Boeger, S. M. Bode-Boeger, O. Kruszelnicka-Kwiatkowska, F. Kokot, J. S. Dubiel and J. C. Froelich (1999). "Reduced urinary excretion of nitric oxide metabolites and increased plasma levels of asymmetric dimethylarginine in men with essential hypertension." J Cardiovasc Pharmacol 33(4): 652-658.
Suzuki, Y., T. Suda, K. Furuhashi, M. Suzuki, M. Fujie, D. Hahimoto, Y. Nakamura, N. Inui, H. Nakamura and K. Chida (2010). "Increased serum kynurenine/tryptophan ratio correlates with disease progression in lung cancer." Lung Cancer 67(3): 361-365.
Taheri, F., J. B. Ochoa, Z. Faghiri, K. Culotta, H. J. Park, M. S. Lan, A. H. Zea and A. C. Ochoa (2001). "L-Arginine regulates the expression of the T-cell receptor zeta chain (CD3zeta) in Jurkat cells." Clin Cancer Res 7(3 Suppl): 958s-965s.
Takala, A., I. Jousela, S. E. Jansson, K. T. Olkkola, O. Takkunen, A. Orpana, S. L. Karonen and H. Repo (1999). "Markers of systemic inflammation predicting organ failure in community-acquired septic shock." Clin Sci (Lond) 97(5): 529-538.
Takenaga, N., Y. Ishii, S. Monden, Y. Sasaki and S. Hata (1995). "Simultaneous determination of a new anticancer agent (NB-506) and its active metabolite in human plasma and urine by high-performance liquid chromatography with ultraviolet detection." J Chromatogr B Biomed Appl 674(1): 111-117.
Takikawa, O., R. Yoshida, R. Kido and O. Hayaishi (1986). "Tryptophan degradation in mice initiated by indoleamine 2,3-dioxygenase." J Biol Chem 261(8): 3648-3653.
Tanriover, M. D., G. S. Guven, D. Sen, S. Unal and O. Uzun (2006). "Epidemiology and outcome of sepsis in a tertiary-care hospital in a developing country." Epidemiol Infect 134(2): 315-322.
Tattevin, P., D. Monnier, O. Tribut, J. Dulong, N. Bescher, F. Mourcin, F. Uhel, Y. Le Tulzo and K. Tarte (2010). "Enhanced indoleamine 2,3-dioxygenase activity in patients with severe sepsis and septic shock." J Infect Dis 201(6): 956-966.
Teale, D. M. and A. M. Atkinson (1992). "Inhibition of nitric oxide synthesis improves survival in a murine peritonitis model of sepsis that is not cured by antibiotics alone." J Antimicrob Chemother 30(6): 839-842.
217
Teerlink, T. (2005). "Measurement of asymmetric dimethylarginine in plasma: methodological considerations and clinical relevance." Clin Chem Lab Med 43(10): 1130-1138.
Teerlink, T. (2007). "HPLC analysis of ADMA and other methylated l-arginine analogs in biological fluids." J Chromatogr B Analyt Technol Biomed Life Sci.
Teerlink, T., R. J. Nijveldt, S. de Jong and P. A. van Leeuwen (2002). "Determination of arginine, asymmetric dimethylarginine, and symmetric dimethylarginine in human plasma and other biological samples by high-performance liquid chromatography." Anal Biochem 303(2): 131-137.
Teerlink, T., P. A. van Leeuwen and A. Houdijk (1994). "Plasma amino acids determined by liquid chromatography within 17 minutes." Clin Chem 40(2): 245-249.
Terblanche, M., Y. Almog, R. S. Rosenson, T. S. Smith and D. G. Hackam (2006). "Statins: panacea for sepsis?" The Lancet Infectious Diseases 6(4): 242-248.
Thomas, S. R., D. Mohr and R. Stocker (1994). "Nitric oxide inhibits indoleamine 2,3-dioxygenase activity in interferon-gamma primed mononuclear phagocytes." J Biol Chem 269(20): 14457-14464.
Tinder, T. L., D. B. Subramani, G. D. Basu, J. M. Bradley, J. Schettini, A. Million, T. Skaar and P. Mukherjee (2008). "MUC1 enhances tumor progression and contributes toward immunosuppression in a mouse model of spontaneous pancreatic adenocarcinoma." J Immunol 181(5): 3116-3125.
Torres, M. I., M. A. Lopez-Casado, P. Lorite and A. Rios (2007). "Tryptophan metabolism and indoleamine 2,3-dioxygenase expression in coeliac disease." Clin Exp Immunol 148(3): 419-424.
Trzeciak, S., I. Cinel, R. Phillip Dellinger, N. I. Shapiro, R. C. Arnold, J. E. Parrillo and S. M. Hollenberg (2008). "Resuscitating the microcirculation in sepsis: the central role of nitric oxide, emerging concepts for novel therapies, and challenges for clinical trials." Acad Emerg Med 15(5): 399-413.
Tschaikowsky, K., M. Hedwig-Geissing, A. Schiele, F. Bremer, M. Schywalsky and J. Schuttler (2002). "Coincidence of pro- and anti-inflammatory responses in the early phase of severe sepsis: Longitudinal study of mononuclear histocompatibility leukocyte antigen-DR expression, procalcitonin, C-reactive protein, and changes in T-cell subsets in septic and postoperative patients." Crit Care Med 30(5): 1015-1023.
Tsunoda, M., S. Nonaka and T. Funatsu (2005). "Determination of methylated arginines by column-switching high-performance liquid chromatography-fluorescence detection." Analyst 130(10): 1410-1413.
Ulloa, L., M. Ochani, H. Yang, M. Tanovic, D. Halperin, R. Yang, C. J. Czura, M. P. Fink and K. J. Tracey (2002). "Ethyl pyruvate prevents lethality in mice with established lethal sepsis and systemic inflammation." Proc Natl Acad Sci U S A 99(19): 12351-12356.
Valkonen, V. P., J. Laakso, H. Paiva, T. Lehtimaki, T. A. Lakka, M. Isomustajarvi, I. Ruokonen, J. T. Salonen and R. Laaksonen (2003). "Asymmetrical dimethylarginine (ADMA) and risk of acute coronary events. Does statin treatment influence plasma ADMA levels?" Atheroscler Suppl 4(4): 19-22.
218
Valkonen, V. P., H. Paiva, J. T. Salonen, T. A. Lakka, T. Lehtimaki, J. Laakso and R. Laaksonen (2001). "Risk of acute coronary events and serum concentration of asymmetrical dimethylarginine." Lancet 358(9299): 2127-2128.
Vallance, P., J. Collier and S. Moncada (1989). "Nitric oxide synthesised from L-arginine mediates endothelium dependent dilatation in human veins in vivo." Cardiovasc Res 23(12): 1053-1057.
Vallance, P., A. Leone, A. Calver, J. Collier and S. Moncada (1992). "Accumulation of an endogenous inhibitor of nitric oxide synthesis in chronic renal failure." Lancet 339(8793): 572-575.
Vallet, B. (2003). "Bench-to-bedside review: endothelial cell dysfunction in severe sepsis: a role in organ dysfunction?" Crit Care 7(2): 130-138.
van den Akker, E. L., C. C. Baan, B. van den Berg, H. Russcher, K. Joosten, A. C. Hokken-Koelega, S. W. Lamberts and J. W. Koper (2008). "Ficoll-separated mononuclear cells from sepsis patients are contaminated with granulocytes." Intensive Care Med 34(5): 912-916.
van der Sluijs, K. F., M. Nijhuis, J. H. Levels, S. Florquin, A. L. Mellor, H. M. Jansen, T. van der Poll and R. Lutter (2006). "Influenza-induced expression of indoleamine 2,3-dioxygenase enhances interleukin-10 production and bacterial outgrowth during secondary pneumococcal pneumonia." J Infect Dis 193(2): 214-222.
van Wandelen, C. and S. A. Cohen (1997). "Using quaternary high-performance liquid chromatography eluent systems for separating 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate-derivatized amino acid mixtures." Journal of Chromatography A 763(1-2): 11-22.
van Wandelen, C. and S. A. Cohen (1997). "Using quaternary high-performance liquid chromatography eluent systems for separating 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate-derivatized amino acid mixtures. ." J Chromatogr A 763: 11-22.
van Wandelen, C. and S. A. Cohen (1997). "Using quaternary high-performance liquid chromatography eluent systems for separating 6-aminoquionolyl-N-hydroxysuccinyl carbamate-derivatized amino acid mixtures. ." J Chromatogr A 763: 11-22.
van Wissen, M., M. Snoek, B. Smids, H. M. Jansen and R. Lutter (2002). "IFN-gamma amplifies IL-6 and IL-8 responses by airway epithelial-like cells via indoleamine 2,3-dioxygenase." J Immunol 169(12): 7039-7044.
Vaudo, G., S. Marchesi, D. Siepi, M. Brozzetti, R. Lombardini, M. Pirro, A. Alaeddin, A. R. Roscini, G. Lupattelli and E. Mannarino (2007). "Human endothelial impairment in sepsis." Atherosclerosis 197(2): 747-752.
Vaudo, G., S. Marchesi, D. Siepi, M. Brozzetti, R. Lombardini, M. Pirro, A. Alaeddin, A. R. Roscini, G. Lupattelli and E. Mannarino (2008). "Human endothelial impairment in sepsis." Atherosclerosis 197(2): 747-752.
Vignali, D. A., L. W. Collison and C. J. Workman (2008). "How regulatory T cells work." Nat Rev Immunol 8(7): 523-532.
Villalpando, S., J. Gopal, A. Balasubramanyam, V. P. Bandi, K. Guntupalli and F. Jahoor (2006). "In vivo arginine production and intravascular nitric oxide synthesis in hypotensive sepsis." Am J Clin Nutr 84(1): 197-203.
219
Vincent, J. L., A. de Mendonca, F. Cantraine, R. Moreno, J. Takala, P. M. Suter, C. L. Sprung, F. Colardyn and S. Blecher (1998). "Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on "sepsis-related problems" of the European Society of Intensive Care Medicine." Crit Care Med 26(11): 1793-1800.
Vladimirova-Kitova, L. G. and T. Deneva-Koycheva (2010). "Asymmetric dimethylarginine--a determinant of the effect of the high dose Simvastatin." Clin Biochem 43(10-11): 843-850.
von Muller, L., A. Klemm, M. Weiss, M. Schneider, H. Suger-Wiedeck, N. Durmus, W. Hampl and T. Mertens (2006). "Active cytomegalovirus infection in patients with septic shock." Emerg Infect Dis 12(10): 1517-1522.
Waage, A., A. Halstensen and T. Espevik (1987). "Association between tumour necrosis factor in serum and fatal outcome in patients with meningococcal disease." Lancet 1(8529): 355-357.
Wang, S. D., K. J. Huang, Y. S. Lin and H. Y. Lei (1994). "Sepsis-induced apoptosis of the thymocytes in mice." J Immunol 152(10): 5014-5021.
Wang, Y., H. Liu, G. McKenzie, P. K. Witting, J. P. Stasch, M. Hahn, D. Changsirivathanathamrong, B. J. Wu, H. J. Ball, S. R. Thomas, V. Kapoor, D. S. Celermajer, A. L. Mellor, J. F. Keaney, Jr., N. H. Hunt and R. Stocker (2010). "Kynurenine is an endothelium-derived relaxing factor produced during inflammation." Nat Med 16: 279-285.
Warren, H. S. (1997). "Strategies for the treatment of sepsis." N Engl J Med 336(13): 952-953.
Wheeler, A. P. (2007). "Recent developments in the diagnosis and management of severe sepsis." Chest 132(6): 1967-1976.
Widner, B., N. Sepp, E. Kowald, U. Ortner, B. Wirleitner, P. Fritsch, G. Baier-Bitterlich and D. Fuchs (2000). "Enhanced tryptophan degradation in systemic lupus erythematosus." Immunobiology 201(5): 621-630.
Wu, G. and S. M. Morris, Jr. (1998). "Arginine metabolism: nitric oxide and beyond." Biochem J 336 ( Pt 1): 1-17.
Xia, Y., L. J. Roman, B. S. Masters and J. L. Zweier (1998). "Inducible nitric-oxide synthase generates superoxide from the reductase domain." J Biol Chem 273(35): 22635-22639.
Yamashita, T., S. Kawashima, M. Ozaki, M. Namiki, S. Satomi-Kobayashi, T. Seno, Y. Matsuda, N. Inoue, K. Hirata, H. Akita, K. Umetani, E. Tanaka, H. Mori and M. Yokoyama (2001). "Role of endogenous nitric oxide generation in the regulation of vascular tone and reactivity in small vessels as investigated in transgenic mice using synchrotron radiation microangiography." Nitric Oxide 5(5): 494-503.
Yanagawa, Y., K. Iwabuchi and K. Onoe (2009). "Co-operative action of interleukin-10 and interferon-gamma to regulate dendritic cell functions." Immunology 127(3): 345-353.
Yanofsky, C., V. Horn and P. Gollnick (1991). "Physiological studies of tryptophan transport and tryptophanase operon induction in Escherichia coli." J Bacteriol 173(19): 6009-6017.
Yeo, T. W., D. A. Lampah, R. Gitawati, E. Tjitra, E. Kenangalem, Y. R. McNeil, C. J. Darcy, D. L. Granger, J. B. Weinberg, B. K. Lopansri, R. N. Price, S. B. Duffull,
220
D. S. Celermajer and N. M. Anstey (2007). "Impaired nitric oxide bioavailability and L-arginine reversible endothelial dysfunction in adults with falciparum malaria." J Exp Med 204(11): 2693-2704.
Yeo, T. W., D. A. Lampah, R. Gitawati, E. Tjitra, E. Kenangalem, K. Piera, R. N. Price, S. B. Duffull, D. S. Celermajer and N. M. Anstey (2008). "Angiopoietin-2 is associated with decreased endothelial nitric oxide and poor clinical outcome in severe falciparum malaria." Proc Natl Acad Sci U S A 105(44): 17097-17102.
Yeo, T. W., D. A. Lampah, E. Tjitra, R. Gitawati, C. J. Darcy, C. Jones, E. Kenangalem, Y. R. McNeil, D. L. Granger, B. K. Lopansri, J. B. Weinberg, R. N. Price, S. B. Duffull, D. S. Celermajer and N. M. Anstey (2010). "Increased asymmetric dimethylarginine in severe falciparum malaria: association with impaired nitric oxide bioavailability and fatal outcome." PLoS Pathog 6(4): e1000868.
Yoshida, R., J. Imanishi, T. Oku, T. Kishida and O. Hayaishi (1981). "Induction of pulmonary indoleamine 2,3-dioxygenase by interferon." Proc Natl Acad Sci U S A 78(1): 129-132.
Young, J. D. and E. M. Cameron (1995). "Dynamics of skin blood flow in human sepsis." Intensive Care Med 21(8): 669-674.
Zangerle, R., B. Widner, G. Quirchmair, G. Neurauter, M. Sarcletti and D. Fuchs (2002). "Effective antiretroviral therapy reduces degradation of tryptophan in patients with HIV-1 infection." Clin Immunol 104(3): 242-247.
Zani, B. G. and H. G. Bohlen (2005). "Transport of extracellular l-arginine via cationic amino acid transporter is required during in vivo endothelial nitric oxide production." Am J Physiol Heart Circ Physiol 289(4): H1381-1390.
Zanotti-Cavazzoni, S. L. and R. D. Goldfarb (2009). "Animal models of sepsis." Crit Care Clin 25(4): 703-719, vii-viii.
Zea, A. H., B. D. Curti, D. L. Longo, W. G. Alvord, S. L. Strobl, H. Mizoguchi, S. P. Creekmore, J. J. O'Shea, G. C. Powers, W. J. Urba and et al. (1995). "Alterations in T cell receptor and signal transduction molecules in melanoma patients." Clin Cancer Res 1(11): 1327-1335.
Zea, A. H., P. C. Rodriguez, M. B. Atkins, C. Hernandez, S. Signoretti, J. Zabaleta, D. McDermott, D. Quiceno, A. Youmans, A. O'Neill, J. Mier and A. C. Ochoa (2005). "Arginase-producing myeloid suppressor cells in renal cell carcinoma patients: a mechanism of tumor evasion." Cancer Res 65(8): 3044-3048.
Zea, A. H., P. C. Rodriguez, K. S. Culotta, C. P. Hernandez, J. DeSalvo, J. B. Ochoa, H. J. Park, J. Zabaleta and A. C. Ochoa (2004). "L-Arginine modulates CD3zeta expression and T cell function in activated human T lymphocytes." Cell Immunol 232(1-2): 21-31.
Ziegler, E. J., C. J. Fisher, Jr., C. L. Sprung, R. C. Straube, J. C. Sadoff, G. E. Foulke, C. H. Wortel, M. P. Fink, R. P. Dellinger, N. N. Teng and et al. (1991). "Treatment of gram-negative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin. A randomized, double-blind, placebo-controlled trial. The HA-1A Sepsis Study Group." N Engl J Med 324(7): 429-436.
Zoccali, C., S. Bode-Boger, F. Mallamaci, F. Benedetto, G. Tripepi, L. Malatino, A. Cataliotti, I. Bellanuova, I. Fermo, J. Frolich and R. Boger (2001). "Plasma concentration of asymmetrical dimethylarginine and mortality in patients with end-stage renal disease: a prospective study." Lancet 358(9299): 2113-2117.
221
Zoccali, C., R. Maas, S. Cutrupi, P. Pizzini, P. Finocchiaro, F. Cambareri, V. Panuccio, C. Martorano, F. Schulze, G. Enia, G. Tripepi and R. Boger (2007). "Asymmetric dimethyl-arginine (ADMA) response to inflammation in acute infections." Nephrol Dial Transplant 22(3): 801-806.
Zoccali, C., F. Mallamaci and G. Tripepi (2006). "Asymmetric dimethylarginine (ADMA) as a cardiovascular risk factor in end-stage renal disease (ESRD)." Eur J Clin Pharmacol 62 Suppl 1: 131-135.
222
Appendix: Published papers from this thesis
BioMed CentralBMC Clinical Pathology
ss
Open AcceResearch articleEx-vivo changes in amino acid concentrations from blood stored at room temperature or on ice: implications for arginine and taurine measurementsJoshua S Davis*1,2, Christabelle J Darcy1, Kim Piera1, Yvette R McNeil1, Tonia Woodberry1 and Nicholas M Anstey1,2Address: 1International Health Division, Menzies School of Health Research and Charles Darwin University, Darwin, NT 0810, Australia and 2Division of Medicine, Royal Darwin Hospital, Darwin, NT, 0810, Australia
Email: Joshua S Davis* - [email protected]; Christabelle J Darcy - [email protected]; Kim Piera - [email protected]; Yvette R McNeil - [email protected]; Tonia Woodberry - [email protected]; Nicholas M Anstey - [email protected]
* Corresponding author
AbstractBackground: Determination of the plasma concentrations of arginine and other amino acids isimportant for understanding pathophysiology, immunopathology and nutritional supplementationin human disease. Delays in processing of blood samples cause a change in amino acidconcentrations, but this has not been precisely quantified. We aimed to describe the concentrationtime profile of twenty-two amino acids in blood from healthy volunteers, stored at roomtemperature or on ice.
Methods: Venous blood was taken from six healthy volunteers and stored at room temperatureor in an ice slurry. Plasma was separated at six time points over 24 hours and amino acid levelswere determined by high-performance liquid chromatography.
Results: Median plasma arginine concentrations decreased rapidly at room temperature, with a6% decrease at 30 minutes, 25% decrease at 2 hours and 43% decrease at 24 hours. Plasmaornithine increased exponentially over the same period. Plasma arginine was stable in blood storedon ice, with a < 10% change over 24 hours. Plasma taurine increased by 100% over 24 hours, andthis change was not prevented by ice. Most other amino acids increased over time at roomtemperature but not on ice.
Conclusion: Plasma arginine concentrations in stored blood fall rapidly at room temperature, butremain stable on ice for at least 24 hours. Blood samples taken for the determination of plasmaamino acid concentrations either should be placed immediately on ice or processed within 30minutes of collection.
BackgroundQuantification of plasma amino acids is not routinelyoffered by clinical laboratories and thus plasma often
needs to be transported to research or reference laborato-ries for testing. In order to accurately assess the concentra-tion of plasma amino acids, it is important to know their
Published: 27 November 2009
BMC Clinical Pathology 2009, 9:10 doi:10.1186/1472-6890-9-10
Received: 26 June 2009Accepted: 27 November 2009
This article is available from: http://www.biomedcentral.com/1472-6890/9/10
© 2009 Davis et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Page 1 of 7(page number not for citation purposes)
BMC Clinical Pathology 2009, 9:10 http://www.biomedcentral.com/1472-6890/9/10
stability in human blood which has been stored or trans-ported prior to testing. Previous studies addressing thisquestion have been small and the rate of degradation hasnot been precisely quantified.
Arginine, the precursor of nitric oxide (NO) [1], is impor-tant for endothelial [2] and immunological [3] functionand is acutely decreased in sepsis [4,5], malaria [6] andtrauma [7], and was thus the focus of this study. Themajor routes for arginine metabolism in humans aremetabolism by arginase to urea and ornithine; use for cre-atine synthesis; and metabolism by nitric oxide synthaseto NO and citrulline [8]. Both red blood cells (RBCs) [9]and macrophages [10] are rich in arginase. In storedpacked RBCs, arginase is released and the resulting degra-dation of plasma arginine is thought to be a mechanismof transfusion-associated immunosuppression [9,11].Other amino acids which are commonly added to supple-mentary nutrition for critically ill patients may also playan important role in immune function including tryp-tophan [12] glutamine [13] and taurine [14,15].
Hainque and colleagues studied eight healthy volunteersand found a "significant degradation" of plasma argininefollowing 4 hours at room temperature but this was notquantified and no other time points were reported [16].Schaefer et al. studied one volunteer and found a 50%decrease in plasma arginine after 6 hours at room temper-ature compared with a 10% decrease after 6 hours at 4degrees centigrade, with earlier time points not reported[17]. Nutall and colleagues reported time profile datafrom one volunteer, which showed an approximate 33%decrease in plasma arginine by 2 hours at room tempera-ture [18].
To determine the impact of delayed processing we under-took a study to estimate the rate of arginine degradationin human plasma at room temperature and on ice. Wehypothesised that this degradation would be primarilydue to plasma arginase activity and that there would beless than 10% degradation at 2 hours in samples placedimmediately on ice. We also sought to determine theeffect of delayed separation and freezing of plasma on theconcentration of other amino acids.
MethodsThe study was considered by the Chair of the HumanResearch Ethics Committee of the Menzies School ofHealth Research and Northern Territory Department ofHealth and Families, and was approved as a laboratoryquality assurance activity which did not require full ethi-cal review. Following written informed consent, sixhealthy normotensive fasting volunteers had venousblood collected into 12 × 2 mL lithium heparin tubes(Vacutainer, Becton Dickinson, Franklin Lakes, New Jer-
sey) using a 21 gauge needle and vacutainer system. Foreach subject, the first six tubes were immediately placedinto an ice slurry and the second six were left at room tem-perature (25° Celsius (C)) in an air conditioned labora-tory. After intervals of 0 minutes, 30 minutes, 2 hours, 4hours, 8 hours and 24 hours from the time of venepunc-ture, the tubes were centrifuged at 3000 rpm for 10 min-utes (either at 4°C or at room temperature as appropriate)and the plasma immediately separated and stored at -80°C.
Subsequently, following thawing, plasma amino acidswere extracted with ethanol, then derivatized with AccQ-Fluor (Waters, Milford, MA). Amino acid concentrationswere then determined by reverse-phase high performanceliquid chromatography (HPLC; Shimadzu corporation,Kyoto, Japan) with UV (250 nm) and fluorescence (exci-tation 250 nm, emission 395 nm) detection, using amethod modified from van Wandelen and Cohen [19].
The data were analysed using Stata 10 (Statacorp, CollegeStation, Texas) and GraphPad Prism 5 (Graphpad soft-ware, San Diego, California). Due to the small number ofsubjects, data were summarized using median and inter-quartile range. Median amino acid concentrations overtime were compared using a paired Wilcoxon test, with ap-value of < 0.05 considered significant. The arginine deg-radation curve was fitted using a one-phase exponentialdecay model. The sample size was determined using datafrom an earlier experiment (unpublished data), whichfound that there was 31.8% (std dev = 14%) degradationof arginine at room temperature by 2 hours. Using apower of 80% and a significance level of 5%, five subjectsin each group would be needed to detect a difference of22% degradation at 2 hours, meaning less than 10% deg-radation in the ice group. To allow for sample wastage anderrors, we recruited six subjects.
ResultsOf the six study subjects, half were male, and the medianage was 37.5 years, with a range of 19-47 years (table 1).All were healthy, of normal weight and normotensive,and none had cardiovascular disease or diabetes mellitus.The median baseline plasma arginine concentration was74.9 μmol/L, similar to previously reported mean plasmaarginine concentrations from healthy volunteers, themajority of which are between 60 and 80 μmol/L [20].
Arginine and ornithine time profiles at room temperaturePlasma arginine concentration decreased rapidly at roomtemperature (Figures 1a, 2, Table 2) with 6% degradationwithin 30 minutes, 25% degradation within 2 hours and43% degradation within 24 hours. A non-linear model ofthe plasma arginine profile over time was defined by theequation Y = ((Y0-P)*e-kt)+P, where t = time in hours, P =
Page 2 of 7(page number not for citation purposes)
BMC Clinical Pathology 2009, 9:10 http://www.biomedcentral.com/1472-6890/9/10
the plateau value, Y0 = initial value. The parameters of themodel were Y0 = 81.3, P = 37.8, and k = 0.6273. Thismodel fitted the data well, with an R2 of 0.73. Plasmaornithine concentration increased exponentially at roomtemperature (Table 1, Figure 2), with a 4% increase at 30minutes, a 62% increase at 2 hours, and a 183% increaseat 24 hours.
Arginine time profile on ice compared with room temperaturePlasma arginine was very stable on ice, with a less than10% change over a 24 hour period. At 2 hours, the medianplasma arginine concentration had decreased by 6% inthe ice specimens compared with 25% in the room tem-perature specimens (p < 0.001) (Figure 1). At 24 hours,the change in arginine was negligible for the ice specimenscompared with a 43% decrease at room temperature (p <0.001). Ornithine was also more stable on ice, with a 24%increase over the 24 hour period, compared with a 183%increase at room temperature.
Time profile of other amino acidsFor the majority of other amino acids, concentrationsincreased by > 10% over 24 hours at room temperature(Table 3). The majority of these changes were largely orcompletely prevented in the blood that was placed on ice.The most notable room temperature concentrationincreases at 24 hours were seen with taurine (which dou-bled) and glutamate (which increased more than five-fold). The change in taurine was unusual in that it wasmore marked in the blood placed on ice (a 126%increase) than the room temperature specimens (a 100%increase), suggesting that the increase in taurine may bedue to release from lysed cells rather than to an enzymatic
Table 1: Characteristics of study subjects
Subject Age (years) Gender Ethnicity
1 36 F Caucasian
2 39 M Caucasian
3 47 F Caucasian
4 27 F Caucasian
5 19 M Caucasian
6 44 M Caucasian
Plasma arginine time profile at room temperature and on iceFigure 1Plasma arginine time profile at room temperature and on ice. Each curve represents an individual subject. Fig-ure 1a depicts results from whole blood stored at room tem-perature (25°C). Figure 1b depicts results from aliquots of the same blood samples which were stored in an ice slurry.
Time profile of median plasma arginine and ornithine concen-trations in blood stored at room temperatureFigure 2Time profile of median plasma arginine and ornithine concentrations in blood stored at room temperature. Each point represents the median value for that time, and the error bars represent the interquartile range. Median plasma arginine is indicated by triangles, and ornithine by solid cir-cles.
Page 3 of 7(page number not for citation purposes)
BMC Clinical Pathology 2009, 9:10 http://www.biomedcentral.com/1472-6890/9/10
process (Figure 3). Tryptophan was very stable both atroom temperature and on ice.
DiscussionPlasma arginine concentration decreases rapidly in wholeblood held at room temperature, and this decrease isgreatly attenuated by placing the blood on ice. Ornithine,the metabolic product of arginine metabolism by argin-ase, rises exponentially at room temperature, and this risedoes not occur on ice, suggesting that it is due to an enzy-matic process. Thus, it is likely that arginase is the primarymechanism of arginine degradation in ex-vivo blood sam-ples. This arginase could come from either lysed RBCs or
lysed leucocytes, but we did not evaluate the source ofarginase, and thus cannot determine which of these wasmore important. In-vitro hemolysis is difficult to meas-ure, as the released cell-free haemoglobin is immediatelybound by haptoglobin. While we have not proven thishypothesis, our observations strongly suggest it.
Most other amino acids increase at room temperature butnot on ice, which also implies an enzymatic reaction.Tryptophan is very stable both at room temperature andon ice. Taurine and glutamine are unusual, in that theyincrease markedly both at room temperature and on ice;this may be due to cellular release rather than enzymaticcatabolism.
The rate of decrease of plasma arginine which we found inblood held at room temperature is similar to that foundby Nuttall and colleagues in the only published paper tohave reported plasma arginine concentrations at roomtemperature at more than two time points [18]. The lackof early time points in other papers makes it difficult toestimate the rate of decline and whether it is linear orexponential. Nuttall et al. reported data in graphical form,from a single subject up to 2.5 hours post venepuncture.They found a fall from 89 μmol/L to approximately 60μmol/L at 2 hours (a 33% drop), similar to our reporteddecrease of 25% at 2 hours.
The large increases seen in taurine and glutamate in ourstudy have not previously been reported. Sahai et al.measured amino acid levels in whole blood from twenty-two volunteers, stored on ice for 1 hour or 2 hours, andfound a less than 10% decrease in plasma taurine andglutamate at 1 and 2 hours [21]. Shaeffer et al. reported a
Table 2: Median (IQR) arginine and ornithine plasma concentrations over time from blood stored at room temperature compared with stored on ice
Baseline 30 minutes 2 hours 4 hours 8 hours 24 hours
Arginine RTa 74.9 70.3 49.6 40.4 37.3 42.6
73.2-87.8 63.4-75.5 46.0-53.6 35.8-45.8 32.1-42.6 25.5-42.8
Arginine Ice 79.6 77.1 74.8 78.6 80.4 81.0
76.8-93.0 74.6-90.8 73.4-86.9 74.6-86.1 79.4-86.7 79.9-83.0
Ornithine RT 44.7 45.6 72.6 87.4 101.6 114.1
32.9-60.8 39.5-69.8 58.7-94.2 69.1-112.3 79.4-125.9 100.9-153.6
Ornithine Ice 38.6 31.6 39.2 40.5 36.3 43.1
29.4-57.3 38.3-59.2 30.2-60.0 30.5-61.9 32.8-61.5 36.2-68.1
a. RT = Room Temperature
Time profile of median plasma taurine concentrations in blood stored at room temperature and on iceFigure 3Time profile of median plasma taurine concentra-tions in blood stored at room temperature and on ice. Each point represents the median value for that time, and the error bars represent the interquartile range. Median plasma taurine at room temperature is represented by solid circles, and median plasma taurine on ice is represented by triangles.
Page 4 of 7(page number not for citation purposes)
BMC Clinical Pathology 2009, 9:10 http://www.biomedcentral.com/1472-6890/9/10
Page 5 of 7(page number not for citation purposes)
Table 3: Change in amino acid concentrations in whole blood after 24 hours at room temperature and on ice.
% Change at 24 h at RTa, b % Change at 24 h on iceb
Group 1 - ≤ 10% change at RTa over 24 h
Citrulline -4 (-6, 4) -7 (-9, -4)
Glutamine -10 (-13, -10) -5 (-5, -4)
Hydroxyproline 8 (7,9) -3 (-4,-3)
Methionine 0 (-2, 1) 1 (1, 6)
Tryptophan 7 (5, 8) 4 (4, 6)
Tyrosine 8 (5, 12) -2 (-3, -1)
Valine 8 (4, 11) 1 (0, 1)
Group 2 - > 10% increase at RTa over 24 h
Alanine 18 (16,20) 0 (-1, 0)
Asparagine 17 (12, 21) 0 (-1, +3)
Glutamate 593 (563, 612) 38 (92, 186)
Glycine 26 (24, 34) 3 (2, 4)
Histidine 23 (17, 27) 1 (0, 1)
Isoleucine 16 (10, 21) 0 (-1, 2)
Leucine 23 (17, 34) 2 (1, 5)
Lysine 19 (18, 19) 2 (1, 5)
Ornithine 183 (180, 224) 24 (23,25)
Phenylalanine 15 (14,22) 1 (1, 3)
Proline 11 (6,13) 1 (-1,2)
Serine 18 (17,28) 6 (2,6)
Taurine 100 (94, 102) 126 (120, 147)
Threonine 11 (10, 14) -2 (-5, 0)
Group 3 - > 10% decrease at RTa over 24 h
Arginine -43 (-65, -43) -1 (-5, 4)
a. RT - Room temperatureb. Expressed as median % change (Interquartile range)Note - % change values are an increase (positive change) unless otherwise specified.
BMC Clinical Pathology 2009, 9:10 http://www.biomedcentral.com/1472-6890/9/10
< 10% decrease in plasma taurine and glutamate at 6hours in blood held at room temperature from onehealthy volunteer [17]. The reason for this discrepancy isunclear. Both papers used different methods for aminoacid quantification than we did. Sahai et al did not meas-ure time points beyond 2 hours, and most of the increasein both taurine and glutamine in our study occurredbeyond 2 hours. However, until this finding is reproducedby other investigators, it should be regarded with caution.
The primary limitations of this study are the relativelysmall number of subjects and the lack of subjects sufferingfrom sepsis, trauma or other conditions of interest. Alarger number of subjects would allow a more accurateestimate of the time profile of arginine degradation overtime. Considering arginase activity is increased in severesepsis [22] and trauma [23], it is unclear if blood frompatients with these conditions would yield the sameresults as we observed. We did not directly measure argin-ase activity in blood or plasma, and thus our inferencethat plasma arginase is primarily responsible for theobserved ex-vivo arginine degradation is based on indirectevidence. However, the only other significant mechanismfor arginine degradation likely to occur ex-vivo is thebreakdown of arginine to NO and citrulline by nitricoxide synthase, which accounts for less than 5% ofarginine metabolism in healthy humans [24].
One potential implication of these data is that wholeblood stored for the purpose of transfusion is likely tocontain non-physiological concentrations of amino acids,which may have unintended immunosuppressive effects.These data also reinforce the importance of accurate meth-odological descriptions in papers reporting plasma aminoacid levels. In a hospital setting, it is not always possibleto process samples within 30 minutes of collection. It istherefore essential to note the time between collectionand freezing when reporting concentrations of plasmaamino acids. This is particularly important if the samplecannot be kept on ice - for example, if the blood is to beused for both peripheral blood mononuclear cell (PBMC)collection and amino acid analysis. As PBMCs are dam-aged by freezing, these samples must be kept at room tem-perature and processed as soon as possible to allowaccurate analysis of both PBMC function and amino acidconcentrations. Furthermore, where plasma amino acidsare being measured for clinical applications, our dataemphasise the importance of timely separation and freez-ing of plasma to avoid potential diagnostic errors.
ConclusionIn conclusion, arginine undergoes rapid ex-vivo degrada-tion at room temperature but this does not occur on ice;plasma tryptophan is stable for at least 24 hours both atroom temperature and on ice; plasma taurine concentra-
tions show large increases both at room temperature andon ice. Blood collected for the purposes of plasma aminoacid determination should be placed immediately on ice;if this is not possible, plasma should be frozen with 30minutes of collection.
AbbreviationsHPLC: High Performance Liquid Chromatography; NO:Nitric Oxide; PBMC: Peripheral Blood Mononuclear Cell;IQR: Interquartile range.
Competing interestsThe authors declare that they have no competing interests.
Authors' contributionsAll authors took part in study design and contributed tothe final draft of the paper. In addition, JSD participatedin interpretation of HPLC results, performed the dataanalysis and wrote the first draft of the paper. CJD, KP,and TW performed sample preparation. YM performedand analysed the HPLC. NA secured the funding. Allauthors read and approved the final manuscript.
Funding sourcesThe study was funded by the National Health and MedicalResearch Council of Australia (NHMRC Program Grants290208, 496600; Practitioner Fellowship to NMA, Schol-arship to JSD). The funders had no role in study design,data collection and analysis, decision to publish, or prep-aration of the manuscript.
AcknowledgementsWe wish to thank Alex Humphrey, Ric Price, Mark McMillan, Suresh Sharma, Barbara Molanus, and Jacqui Hughes for assistance; and Barbara MacHunter and Catherine Jones for help with HPLC assays
References1. Boger RH: The pharmacodynamics of L-arginine. J Nutr 2007,
137:1650S-1655S.2. Ganz P, Vita JA: Testing endothelial vasomotor function: nitric
oxide, a multipotent molecule. Circulation 2003, 108:2049-2053.3. Bogdan C: Nitric oxide and the immune response. Nat Immunol
2001, 2:907-916.4. Luiking YC, Poeze M, Dejong CH, Ramsay G, Deutz NE: Sepsis: an
arginine deficiency state? Crit Care Med 2004, 32:2135-2145.5. Davis JS, Yeo TW, Thomas JH, McMillan M, Darcy CJ, McNeil YR,
Cheng AC, Celermajer DS, Stephens DP, Anstey NM: Sepsis-asso-ciated microvascular dysfunction measured by peripheralarterial tonometry: an observational study. Crit Care 2009,13:R155.
6. Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, McNeil YR,Darcy CJ, Granger DL, Weinberg JB, Lopansri BK, et al.: Impairednitric oxide bioavailability and L-arginine reversibleendothelial dysfunction in adults with falciparum malaria. JExp Med 2007, 204:2693-2704.
7. Ochoa JB, Udekwu AO, Billiar TR, Curran RD, Cerra FB, SimmonsRL, Peitzman AB: Nitrogen oxide levels in patients aftertrauma and during sepsis. Ann Surg 1991, 214:621-626.
8. Boger RH, Bode-Boger SM: The clinical pharmacology of L-arginine. Annu Rev Pharmacol Toxicol 2001, 41:79-99.
9. Bernard A, Meier C, Lopez N, May J, Chang P, Boulanger B, KearneyP: Packed red blood cell-associated arginine depletion ismediated by arginase. J Trauma 2007, 63:1108-1112.
Page 6 of 7(page number not for citation purposes)
BMC Clinical Pathology 2009, 9:10 http://www.biomedcentral.com/1472-6890/9/10
Publish with BioMed Central and every scientist can read your work free of charge
"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral
10. Mori M, Gotoh T: Arginine metabolic enzymes, nitric oxideand infection. J Nutr 2004, 134:2820S-2825S.
11. Prins HA, Houdijk AP, Nijveldt RJ, Teerlink T, Huygens P, Thijs LG,van Leeuwen PA: Arginase release from red blood cells: possi-ble link in transfusion induced immune suppression? Shock2001, 16:113-115.
12. Munn DH, Zhou M, Attwood JT, Bondarev I, Conway SJ, Marshall B,Brown C, Mellor AL: Prevention of allogeneic fetal rejection bytryptophan catabolism. Science 1998, 281:1191-1193.
13. Ardawi MS, Newsholme EA: Glutamine metabolism in lym-phocytes of the rat. Biochem J 1983, 212:835-842.
14. Stapleton PP, Mahon TM, Nowlan P, Bloomfield FJ: Effects of in-vivoadministration of taurine and HEPES on the inflammatoryresponse in rats. J Pharm Pharmacol 1994, 46:745-750.
15. Muhling J, Fuchs M, Fleck C, Sablotzki A, Krull M, Dehne MG, GonterJ, Weiss S, Engel J, Hempelmann G: Effects of arginine, L-alanyl-L-glutamine or taurine on neutrophil (PMN) free amino acidprofiles and immune functions in vitro. Amino Acids 2002,22:39-53.
16. Hainque B, Gerbet D, Roisin JP, Le Moel G, Troupel S, Galli A:[Course of the concentration of serum free amino acids as afunction of time and the method of preservation]. Ann Biol Clin(Paris) 1985, 43:221-226.
17. Schaefer A, Piquard F, Haberey P: Plasma amino-acids analysis:effects of delayed samples preparation and of storage. ClinChim Acta 1987, 164:163-169.
18. Nuttall KL, Chen M, Komaromy-Hiller G: Delayed separation andthe plasma amino acids arginine and ornithine. Ann Clin Lab Sci1998, 28:354-359.
19. van Wandelen C, Cohen SA: Using quaternary high-perform-ance liquid chromatography eluent systems for separating 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate-deriva-tized amino acid mixtures. J Chromatogr A 1997, 763:11-22.
20. Martens-Lobenhoffer J, Bode-Boger SM: Measurement of asym-metric dimethylarginine (ADMA) in human plasma: fromliquid chromatography estimation to liquid chrmoatgora-phy-mass spectrometry quantification. Eur J Clin Pharmacol2006, 62:61-68.
21. Sahai S, Uhlhaas S: Stability of amino acids in human plasma.Clin Chim Acta 1985, 148:255-259.
22. Argaman Z, Young VR, Noviski N, Castillo-Rosas L, Lu XM, Zura-kowski D, Cooper M, Davison C, Tharakan JF, Ajami A, Castillo L:Arginine and nitric oxide metabolism in critically ill septicpediatric patients. Crit Care Med 2003, 31:591-597.
23. Bernard AC, Mistry SK, Morris SM Jr, O'Brien WE, Tsuei BJ, MaleyME, Shirley LA, Kearney PA, Boulanger BR, Ochoa JB: Alterationsin arginine metabolic enzymes in trauma. Shock 2001,15:215-219.
24. Castillo L, Beaumier L, Ajami AM, Young VR: Whole body nitricoxide synthesis in healthy men determined from [15N]arginine-to-[15N]citrulline labeling. Proc Natl Acad Sci USA1996, 93:11460-11465.
Pre-publication historyThe pre-publication history for this paper can be accessedhere:
http://www.biomedcentral.com/1472-6890/9/10/prepub
Page 7 of 7(page number not for citation purposes)
Available online http://ccforum.com/content/13/5/R155
Open AccessVol 13 No 5ResearchSepsis-associated microvascular dysfunction measured by peripheral arterial tonometry: an observational studyJoshua S Davis1,2, Tsin W Yeo1, Jane H Thomas3, Mark McMillan1, Christabelle J Darcy1, Yvette R McNeil1, Allen C Cheng1,2, David S Celermajer4, Dianne P Stephens3 and Nicholas M Anstey1,2
1International Health Division, Menzies School of Health Research and Charles Darwin University, Rocklands Drive, Darwin, NT 0810, Australia2Division of Medicine, Royal Darwin Hospital, Rocklands Drive, Darwin, NT 0810, Australia3Intensive Care Unit, Royal Darwin Hospital, Rocklands Drive, Darwin, NT 0810, Australia4Department of Medicine, University of Sydney and Department of Cardiology, Royal Prince Alfred Hospital, Missenden Road, Sydney, NSW 2006, Australia
Corresponding author: Nicholas M Anstey, [email protected]
Received: 20 Apr 2009 Revisions requested: 30 Jun 2009 Revisions received: 6 Aug 2009 Accepted: 25 Sep 2009 Published: 25 Sep 2009
Critical Care 2009, 13:R155 (doi:10.1186/cc8055)This article is online at: http://ccforum.com/content/13/5/R155© 2009 Davis et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Introduction Sepsis has a high mortality despite advances inmanagement. Microcirculatory and endothelial dysfunctioncontribute to organ failure, and better tools are needed toassess microcirculatory responses to adjunctive therapies. Wehypothesised that peripheral arterial tonometry (PAT), a noveluser-independent measure of endothelium-dependentmicrovascular reactivity, would be impaired in proportion tosepsis severity and related to endothelial activation and plasmaarginine concentrations.
Methods Observational cohort study in a 350-bed teachinghospital in tropical Australia. Bedside microvascular reactivitywas measured in 85 adults with sepsis and 45 controls atbaseline and 2-4 days later by peripheral arterial tonometry.Microvascular reactivity was related to measures of diseaseseverity, plasma concentrations of L-arginine (the substrate fornitric oxide synthase), and biomarkers of endothelial activation.
Results Baseline reactive hyperaemia index (RH-PAT index),measuring endothelium-dependent microvascular reactivity;(mean [95% CI]) was lowest in severe sepsis (1.57 [1.43-1.70]), intermediate in sepsis without organ failure (1.85 [1.67-2.03]) and highest in controls (2.05 [1.91-2.19]); P < 0.00001.Independent predictors of baseline RH-PAT index in sepsiswere APACHE II score and mean arterial pressure, but notplasma L-arginine or markers of endothelial activation. Lowbaseline RH-PAT index was significantly correlated with anincrease in SOFA score over the first 2-4 days (r = -0.37, P =0.02).Conclusions Endothelium-dependent microvascular reactivityis impaired in proportion to sepsis severity and suggestsdecreased endothelial nitric oxide bioavailability in sepsis.Peripheral arterial tonometry may have a role as a user-independent method of monitoring responses to noveladjunctive therapies targeting endothelial dysfunction in sepsis.
IntroductionMortality from severe sepsis remains high, despite advances inits management [1]. Organ failure commonly occurs despitethe achievement of normal haemodynamics in response tofluid resuscitation, vasopressors and the treatment of infec-tion. This may be due to impaired vasomotor regulation of themicrocirculation [2]. In sepsis, the endothelium has key rolesin regulating vascular tone and permeability and its activation
is pivotal in initiating both the inflammatory and coagulationcascades [3].
Endothelial function is assessed clinically by the ability ofblood vessels to vasodilate in response to pharmacologicalstimuli or to shear stress, and is primarily dependent onendothelial nitric oxide (NO) production [4]. As a result, manyclinical studies investigating the endothelium in sepsis have
Page 1 of 9(page number not for citation purposes)
APACHE: Acute Physiology and Chronic Health Evaluation; CI: confidence interval; ELISA: enzyme-linked immunosorbent assay; ICAM-1: intra-cel-lular adhesion molecule-1; ICU: intensive care unit; IL: interleukin; MAP: mean arterial pressure; NIRS: near infrared spectroscopy; NO: nitric oxide; NOS: nitric oxide synthase; NS: not significant; OR: odds ratio; RH-PAT: reactive hyperaemia peripheral arterial tonometry; SOFA: Sequential Organ Failure Assessment.
Critical Care Vol 13 No 5 Davis et al.
measured circulating endothelial activation markers, as a sur-rogate for endothelial function. Current techniques for meas-urement of endothelial function, such as laser Doppler,plethysmography and flow-mediated dilatation of the brachialartery, require skilled operators and are technically difficult toperform at the bedside. Some studies have assessedendothelial function by measuring reactive hyperaemia inhuman sepsis using these operator-dependant techniques [5-10]. These studies have generally shown normal baselineblood flow and impaired reactive hyperaemic responses insepsis, but have been small (n = 8 to 45) and have not corre-lated reactive hyperaemia with L-arginine or circulating mark-ers of endothelial activation. More recently, investigators usingdynamic near-infrared spectroscopy (NIRS) have foundimpaired microvascular responses in sepsis; however, thenature of the relation between NIRS and endothelial NO activ-ity is unclear [11].
Reactive hyperaemia peripheral arterial tonometry (RH-PAT) isa novel, simple and user-independent bedside technique usedto measure microvascular endothelial function [12] (Figure 1).It is increasingly being used to measure endothelial function asa cardiovascular risk assessment tool in ambulatory patients[12-16], including in the third-generation Framingham HeartStudy cohort [17]. RH-PAT has been shown to be at least50% dependent on endothelial NO activity [18]. RH-PAT uses
finger probes to measure digital pulse wave amplitudedetected by a pressure transducer, and has been validatedagainst the operator-dependent flow-mediated dilatationmethod [19,20] and with endothelial function in other vascularbeds, including the coronary arteries [13]. Using RH-PAT, wehave demonstrated endothelial dysfunction in subjects withsevere malaria [21] but it has not previously been evaluated insubjects with sepsis.
Vasodilatory shock in sepsis has been hypothesized to reflecta state of NO excess. However, several recent isotope studieshave shown no net increase in NO synthesis in humans withsepsis [22-24]. To explain this, it has been proposed that sep-sis may be a state of imbalance between the NOS isoformsinducible NOS and endothelial NOS in the microvasculature[25]. This could lead to a relative deficiency of endothelial NO,which is required to maintain the microvascular endothelium ina healthy, quiescent state.
Another possible reason for endothelial NO deficiency isdecreased availability of L-arginine, the substrate for NOS andthe precursor for NO [26]. Sepsis has been hypothesised tobe an arginine-deficient state [27], although plasma L-argininelevels in humans with sepsis have been variably reported to behigh [28], normal [29,30] or low [22,31,32]. Decreased
Figure 1
Representative normal and abnormal peripheral arterial tonometry tracesRepresentative normal and abnormal peripheral arterial tonometry traces. The tracings represent the pulse wave amplitude from a fingertip over a 15-minute period. The y axis is pulse wave amplitude in arbitrary units (derived from millivolts). The top trace was taken from a control subject whose reactive hyperaemia peripheral arterial tonometry; (RH-PAT) index was 1.98, and the bottom from a severe sepsis subject whose RH-PAT index was 1.16. The horizontal axis is time. The first shaded section is averaged as a baseline signal. The middle section is arterial occlusion, with consequent loss of the pulse wave signal. The final section is the pulse wave amplitude following release of the cuff. The random vertical spikes are movement artefacts. In the top trace there is reactive hyperaemia, with an increase in average pulse wave amplitude. The shaded post-occlusion section is com-pared with the shaded baseline section to give a ratio -- the RH-PAT index.
Page 2 of 9(page number not for citation purposes)
Available online http://ccforum.com/content/13/5/R155
plasma L-arginine has been linked to decreased NO produc-tion in animal and in vitro models [33].
We hypothesised that RH-PAT would be a feasible techniqueto measure microvascular reactivity in sepsis and that microv-ascular reactivity would be impaired in subjects with sepsis inproportion to disease severity. Our secondary hypotheseswere that microvascular reactivity would correlate with plasmaL-arginine and measures of endothelial activation, and thatplasma L-arginine concentrations would be decreased insepsis.
Materials and methodsStudy design and settingWe performed a prospective observational cohort study in a350-bed teaching hospital in tropical northern Australia, withan 18-bed mixed intensive care unit (ICU). Approval wasobtained from Human Research Ethics Committee of the Men-zies School of Health Research and the Department of Healthand Community Services, Darwin. Written informed consentwas obtained from all participants or next of kin.
ParticipantsBetween March 2006 and November 2007, all adult subjects(≥ 18 years) admitted to the hospital were screened regardingeligibility for the study. Inclusion criteria for sepsis subjectswere: suspected or proven infection; presence of two or morecriteria for the systemic inflammatory response syndromewithin the past four hours [34]; and admission to ICU withinthe preceding 24 hours or to the wards within the preceding36 hours. Exclusion criteria were coagulopathy (platelets ≤ 20× 109/L, activated partial thromboplastin time ≥ 70 seconds,international normalized ratio ≥ 2.0); smoking of tobaccowithin the preceding four hours; and current administration ofintravenous nitrates. Control subjects were recruited from hos-pital patients with no clinical or laboratory evidence of inflam-mation or infection, and who had not met systemicinflammatory response syndrome criteria within the preceding30 days. Severe sepsis was defined as sepsis with organ dys-function or shock at the time of enrolment according to Amer-ican College of Chest Physicians/Society of Critical CareMedicine consensus criteria [34,35].
Measurement of microvascular reactivitySepsis subjects underwent standardised demographic andclinical data collection, bedside RH-PAT measurement(Endopat 2000, Itamar Medical, Caesarea, Israel), and bloodcollection at days 0 and 2 to 4. All studies were performedafter resuscitation and at least one hour of hemodynamic sta-bility (defined as no change in vasopressor dose or need forfluid boluses) in a quiet room at 25°C, with the patient recum-bent. Control subjects had the same assessment at a singletime point.
In this study, probes were placed on the index fingers of bothhands of all patients, or on other fingers if the index fingerswere not suitable. Digital pulse wave amplitude was recordedfrom both hands for a resting baseline period of five minutesand then a blood pressure cuff was rapidly inflated on thestudy arm up to 200 mmHg, or 50 mmHg above systolic bloodpressure, whichever was greater. After five minutes ± 10 sec-onds, the cuff was deflated. Pulse wave amplitude was thenrecorded for a further five minutes. An automated computer-ised algorithm provided by the manufacturer (Endo-PAT 2000software version 3.1.2, Itamar Medical, Caesarea, Israel) wasused to calculate a post occlusion-pre occlusion ratio (RH-PAT index), thus making the measurements user independent.The software also normalises the RH-PAT index to the controlarm to correct for changes in systemic vascular tone (Figure1).
There was no systematic difference between RH-PAT indicesgenerated by different observers. We have previously exam-ined the reproducibility of RH-PAT measurements by repeat-ing them after 0.5 to 0.75 hours in 37 healthy adults [21].Reproducibility was acceptable according to the method ofBland and Altman [36], and was comparable with previousreproducibility results for RH-PAT [37] and with thoseobtained with the flow-mediated dilatation method [38].
Laboratory assaysBlood was collected in lithium heparin tubes at each time pointand the plasma was frozen. Plasma arginine concentrationswere determined using high-performance liquid chromatogra-phy, with a method modified from van Wandelen and Cohen[39]. To assess circulating measures of endothelial activation,intra-cellular adhesion molecule-1 (ICAM1) and E-selectinwere measured by ELISA (R&D Systems, Minneapolis, Min-nestoa, USA). Plasma IL-6 was measured by flow cytometryusing a cytokine bead array (BD Biosciences, San Jose, Cali-fornia, USA). Ex vivo plasma arginase activity causes signifi-cant degradation of L-arginine at room temperature [40], thusonly L-arginine levels derived from blood frozen within 30 min-utes of collection were included in the analysis.
Statistical methodsPredefined groups for analysis were sepsis without organ fail-ure, severe sepsis and controls. Continuous variables werecompared using Student's t-test and analysis of variance orMann Whitney U test for parametric and non-parametric varia-bles, respectively. Categorical variables were compared usingFisher's exact test. Correlates with baseline RH-PAT indexwere determined using Pearson's (parametric) or Spearman's(non-parametric) coefficient for univariate analysis. For multi-variate analysis, linear regression with backward selection wasused. To examine longitudinal correlations, linear mixed-effectsmodels were used. A two-sided P value of < 0.05 was consid-ered significant. All analyses were performed using Stata ver-sion 10 (Stata Corp, College Station, Texas, USA).
Page 3 of 9(page number not for citation purposes)
Critical Care Vol 13 No 5 Davis et al.
ResultsParticipantsOver the 19-month study period, 85 subjects with sepsis and45 control subjects were enrolled. Of the sepsis subjects, 54had organ failure due to sepsis at baseline (severe sepsisgroup) and 31 did not (sepsis without organ failure). The threegroups were well matched in terms of risk factors for endothe-lial dysfunction and other baseline characteristics (Table 1). Ofthe 85 sepsis subjects, 92% had community-acquired sepsis,with no preceding trauma or surgery, and pneumonia was themost common focus of infection.
Baseline microvascular reactivityBaseline microvascular reactivity was impaired in sepsis sub-jects compared with controls (P < 0.0001; Table 2). MeanRH-PAT index was lowest in the severe sepsis group (1.57,95% confidence interval (CI): 1.43 to 1.70), intermediate in
the sepsis without organ failure group (1.85, 95% CI: 1.67 to2.03), and highest in the control group (2.05, 95% CI: 1.91 to2.19; P < 0.00001; Figure 2). Subjects with severe sepsiswere more likely to have endothelial dysfunction than controlsubjects (odds ratio (OR) 9.4, 95% CI: 3.5 to 25.0). This rela-tion persisted after controlling for known associations with andrisk factors for endothelial dysfunction (diabetes, smoking,ischaemic heart disease, chronic renal disease, hypercholes-terolaemia, hypertension, statin use and age; adjusted OR17.0, 95% CI: 5.0 to 58.0). Within the severe sepsis group,mean RH-PAT index was not significantly different in the 27subjects requiring vasopressors (1.48, 95% CI: 1.30 to 1.66)than in those not requiring vasopressors (1.64, 95% CI: 1.39to 1.89; P = not significant (NS)). In those receiving noradren-aline (n = 25), there was no correlation between RH-PAT indexand noadrenaline dose (r = 0.19, P = NS). There was also norelation between body temperature and RH-PAT index. Males
Table 1
Baseline characteristics of participants
Severe sepsis Sepsis without organ failure Control P valuea
N 54 31 45
Ageb 52.4 (48.3-56.5) 50.8 (46.5-55.2) 47.2 (43.1-51.4) NS
Male n (%) 33 (61) 21 (68) 30 (67) NS
Diabetic n (%) 18 (33) 7 (23) 14 (31) NS
Smoker n (%) 28 (57) 12 (39) 18 (41) NS
IHD n (%) 9 (17) 6 (19) 6 (13) NS
On statin n (%) 13 (24) 9 (29) 13 (29) NS
APACHE IIc 19.0 (15-23) 7.5 (5-11) < 0.0001
SOFA scorec 6 (3-9) 1 (0-2) < 0.0001
Focus of infection -- n (%)
Pleuropulmonary n (%) 26 (48) 16 (52)
Skin/soft tissue n (%) 9 (17) 9 (29)
Intra-abdominal n (%) 6 (11) 1 (3)
Urinary n (%) 4 (7) 3 (10)
Other n (%) 9 (17) 2(6)
Causative organism
None cultured n (%) 25 (46) 20 (65)
Gram positive bacterium n (%) 15 (28) 5 (16)
Gram negative bacterium n (%) 14 (26) 6 (19)
Origin of sepsis
Community-acquired n (%) 47 (87) 30 (97)
Nosocomial n (%) 7 (13) 1 (3)
a. For difference between all three groups by one way analysis of varianceb. Mean (95% confidence interval)c. Median (interquartile range)APACHE II = Acute Physiology and Chronic Health Evaluation II; IHD = ischaemic heart disease; NS = not significant; SOFA = Sequential Organ Failure Assessment score
Page 4 of 9(page number not for citation purposes)
Available online http://ccforum.com/content/13/5/R155
(1.76, 95% CI: 1.62 to 1.89) had higher baseline microvascu-lar reactivity than females (1.50, 95% CI: 1.32 to 1.68; P =0.02).
RH-PAT was well tolerated by all subjects. In 18 of 227 meas-urements (8%), a result was not obtainable. This occurred in15 of 182 measurements (8%) in sepsis subjects and 3 of 45(7%) in controls and was due either to inability to obtain abaseline pulse wave reading, or failure to completely occludeforearm blood flow due to oedema.
Plasma markers of endothelial activation (ICAM-1 and E-selec-tin) were both significantly raised in sepsis subjects comparedwith controls (Table 2); however, they did not correlate withRH-PAT index. Blood lactate levels were routinely measuredonly in subjects with severe sepsis, in whom the baselinemedian lactate was 1.6 mmol/L (range 0.5 to 12.7; interquar-tile range (IQR) 1.0 to 2.3). Among severe sepsis subjects,lactate correlated inversely with RH-PAT index, but this wasnot statistically significant (r = -0.28, P = 0.06).
Among all sepsis subjects, baseline RH-PAT index correlatedwith mean arterial pressure (MAP; r = 0.55, P < 0.0001) andserum albumin (r = 0.27, P = 0.03), and was inversely related
Figure 2
Baseline microvascular reactivity is impaired in sepsis, in proportion to disease severityBaseline microvascular reactivity is impaired in sepsis, in proportion to disease severity. Solid circles represent mean values, with error bars representing 95% confidence intervals (CI). P values indicate pairwise comparisons between groups. RH-PAT = reactive hyperaemia periph-eral arterial tonometry.
Table 2
RH-PAT index and related variables at time of initial measurement
Severe sepsis Sepsis without organ failure
Control P value pooled sepsis v control
P value severe sepsis vs SWOF
N 54 31 45
RH-PAT indexa 1.57 (1.43-1.70) 1.85 (1.67-2.03) 2.05 (1.91-2.19) < 0.00001 0.01
Plasma L-arginine (μmol/L)
35.8 (30.2-41.4) 40.9 (33.5-48.3) 80.4 (72.3-88.6) < 0.00001 NS
MAP (mmHg)a 77 (74-81) 89 (83-95) 83 (79-87) NS 0.0006
Receiving vasopressors n (%)
27 (50) 0
Noradrenaline dose (μg/kg/min)b, c 0.08 (0.03-0.42)
Receiving assisted ventilation n (%)
20 (37) 0
CVP (cmH20)a 12.2 (10.3-14.1)
Plasma ICAM-1 (ng/ml)b 811 (500-1502) 507 (368-673) 323 (252-397) < 0.00001 0.003
Plasma E-selectin (ng/ml)b 329 (138-502) 90 (51-164) 38 (26-63) < 0.00001 0.0003
Plasma IL 6 (pg/ml)b 385 (124-996) 148 (46-315) 5 (2-8) < 0.00001 0.009
White blood cell counta16.7 (14.2-19.2) 15.5 (13.3-17.7) 8.4 (6.9-9.8) < 0.00001 NS
C-reactive proteinb 190 (131-255) 102 (84-234) 7 (3-24) < 0.00001 NS
a. mean (95% confidence interval)b. Median (interquartile range)c. Of 27 patients receiving vasopressors, 25 were receiving noradrenalined. Severe sepsis n = 30, sepsis without organ failure n = 26, control n = 2.CVP = central venous pressure; ICAM = intra-cellular adhesion molecule-1; NS = not significant; MAP = mean arterial pressure; RH-PAT = reactive hyperaemia peripheral arterial tonometry; SWOF = sepsis without organ failure.
Page 5 of 9(page number not for citation purposes)
Critical Care Vol 13 No 5 Davis et al.
to Acute Physiology and Chronic Health Evaluation(APACHE) II score (r = -0.36, P = 0.002), C-reactive protein(r = -0.30, P = 0.02) and the cardiovascular component of theSequential Organ Failure Assessment (SOFA) score (r = -0.29, P = 0.01), but not with total SOFA score. Independentpredictors of baseline RH-PAT index on multivariate analysiswere APACHE II score (β = -0.014, P = 0.03) and MAP (β =0.012, P < 0.0001).
Baseline plasma L-arginineIn the subjects whose blood samples were processed within30 minutes of collection, baseline mean plasma L-arginineconcentration was significantly lower in sepsis subjects (38.6μmol/L, 95% CI: 34.2 to 43.1; n = 56) than in controls (80.3μmol/L, 95% CI: 72.5 to 88.1; n = 27; P < 0.0001). There wasno significant difference in L-arginine levels between severesepsis and sepsis without organ failure (Figure 3). When allsubjects including controls were considered, baseline plasmaL-arginine correlated with baseline RH-PAT index (r = 0.32, P= 0.007); however, this association was no longer significantwhen stratified by disease severity.
Longitudinal changes in RH-PAT and L-arginineLongitudinal RH-PAT readings were only available in 70% ofsubjects. There was no difference in disease severity, asmeasured by APACHE II score, in those with (median 14, IQR8 to 23) and without (median 15.5, IQR 8.5 to 20.5; P = NS)longitudinal data. In sepsis subjects, there was no statisticallysignificant change in mean RH-PAT index from baseline to day2 to 4 (95% CI: 1.67 to 1.85, P = NS; Figure 3). The samewas true in the severe sepsis subgroup (95% CI: 1.57 to 1.76,P = NS). In contrast, mean plasma L-arginine concentrationssignificantly increased from baseline to day 2 to 4 (95% CI:38.2 to 49.9 μmol/L, P = 0.01). In a mixed-effects linearregression model, change in microvascular reactivity over thefirst 2 to 4 days of treatment correlated significantly withincreasing MAP and decreasing C-reactive protein, but notwith change in plasma L-arginine.
Subject outcomesLow baseline RH-PAT index was significantly correlated withan increase in SOFA score over the first 2 to 4 days (r = -0.37,P = 0.02). In subjects whose SOFA score worsened over thefirst 2 to 4 days, the median RH-PAT index was 1.54, com-pared with 1.74 in those whose SOFA score improved or didnot change (P = 0.01). At both hospital discharge and 28-dayfollow-up, 8 of 85 (9%) subjects with sepsis had died. Amongthose with septic shock at baseline, 6 of 29 (21%) had died at28-day follow-up. The mean baseline RH-PAT index was 1.67among survivors and 1.60 among non-survivors (P = NS). Thestrongest baseline predictors of death on univariate analysiswere APACHE II score (P = 0.008), SOFA score (P = 0.002)and IL-6 level (P = 0.004).
DiscussionTo the authors' knowledge, this is the largest published studyto date assessing reactive hyperaemia in human sepsis andthe first to use peripheral arterial tonometry. We have foundthat endothelium-dependent microvascular reactivity isimpaired in sepsis, in proportion to disease severity, even aftercontrolling for known associations with endothelial dysfunc-tion, suggesting that sepsis itself is the explanation for theobserved impairment in microvascular reactivity, rather thantraditional cardiovascular risk factors. Furthermore, the degreeof impairment of baseline microvascular reactivity predictedsubsequent deterioration in organ function.
RH-PAT proved to be a practical and feasible method of meas-uring microvascular reactivity at the bedside in critically ill sep-tic subjects, with a low proportion of technical failures, whichwere no more common in sepsis subjects than in controls, andwhich showed no relation with noradrenaline dose. The find-ings of this study are generally consistent with those of theprevious small studies of reactive hyperaemia in adult subjectswith sepsis using other methods, which were generally user-dependant and of limited availability.
Figure 3
Longitudinal change in microvascular reactivity and plasma arginine in sepsis subjectsLongitudinal change in microvascular reactivity and plasma arginine in sepsis subjects. Solid circles represent mean values, with error bars representing 95% confidence intervals (CI). RH-PAT = reactive hyper-aemia peripheral arterial tonometry.
Page 6 of 9(page number not for citation purposes)
Available online http://ccforum.com/content/13/5/R155
Plethysmographic measures of forearm blood flow in sepsishave found a post occlusion-pre occlusion ratio of 1.6 [9] andforearm skin laser Doppler studies have found a ratio of 1.4[5]. These results are very similar to our observed ratio of 1.57,suggesting that the finding of impaired reactive hyperaemia inadults with sepsis is a true phenomenon, which is independentof the method used to measure it.
Compared with laser Doppler flowmetry, venous plethysmog-raphy and flow-mediated dilatation of the brachial artery, PATrequires less staff training and simpler equipment, has lesspotential for inter-observer variability, and is easier to performon uncooperative patients. PAT has also been validated withregards to accuracy [13,19,20] and reproducibility [37,41].Disadvantages of PAT include the expense of disposable fin-ger probes.
Because RH-PAT is at least 50% NO-dependent [18],impaired RH-PAT responses in sepsis suggest reducedendothelial NO bioavailability. Our results are in accord withincreasing data from radiolabelled arginine flux studies sug-gesting that NO synthesis is decreased in sepsis [22-24].Impaired RH-PAT has been demonstrated to be reversiblewith L-arginine infusion in malaria caused by Plasmodium fal-ciparum, providing direct evidence for NO dependence inacute inflammatory states [21]. However, we cannot excludecontributions by other mechanisms, including impaired pro-duction of prostacyclin and endothelium-derived hyperpolariz-ing factor [42,43].
There was a significant correlation between plasma L-arginineand microvascular reactivity when all subjects were consid-ered together, but this was not significant within groups. Fur-thermore, the improvement of plasma L-arginine over the first2 to 4 days was not significantly correlated with change inmicrovascular reactivity. These findings suggest that NO pro-duction and endothelial function in sepsis are influenced byother factors in addition to circulating L-arginine. Such factorsmay include an increase in competitive inhibitors of NOS, suchas asymmetric dimethylarginine [44]; deficiency of NOScofactors such as tetrahydrobiopterin; NO quenching bymicrovascular reactive oxygen intermediates [45]; and theenhanced local expression and activity of endothelial cell argi-nase [46]. The observation of higher microvascular reactivity inmales compared with females is an unexpected finding; previ-ous studies have found better microvascular function infemales than males, both in non-inflammatory states [47] andin response to infusion of lipopolysaccharide [48]. However,gender-specific microvascular function has not previouslybeen reported in sepsis.
The marked hypoargininaemia, which we found in subjectswith sepsis, supports the hypothesis that L-arginine isdecreased in sepsis, independent of trauma [27]. This findingis strengthened by the fact that we only included subjects
within 24 to 36 hours of admission, with standardised sepsiscriteria and with more than 90% having community-acquiredsepsis.
Targeting tissue oxygen delivery [49] or the splanchnic micro-circulation [50] as resuscitation goals in sepsis have not beenshown to improve outcomes. What, then, is the significance ofmonitoring the microvascular endothelium in sepsis? Endothe-lial cells have multiple roles in sepsis pathophysiology, includ-ing the regulation of microcirculatory vasomotor tone and theregulation of coagulation, immune and inflammatoryresponses and microvascular barrier function. Preliminarystudies aimed at increasing endothelial NO bioavailability insepsis have shown promising results [51] and the interven-tions which have been demonstrated to improve outcomes insepsis (activated protein C [52], early goal directed therapy[53] and intensive insulin therapy [54]) could all potentially bemediated, at least in part, via attenuation of endothelial celldysfunction [55]. Thus, monitoring of microvascular andendothelial function are likely to be important components offuture trials of adjunctive treatments in sepsis.
Our study has several potential limitations. Baseline blood flowmeasurements were not available, and it is possible that theapparent decrease in reactive hyperaemia in sepsis is an arte-fact of marked baseline vasodilatation. This could potentiallylimit the subjects' ability to respond to ischaemia by increasedblood flow, because they already have near-maximal vasodila-tation. This is unlikely to be the case because baseline forearmblood flow in septic subjects has been found to be normal ordecreased by multiple investigators [6,7,10,56]. Furthermore,skeletal muscle has the capacity to increase blood flow by upto 10-fold [57], which greatly exceeds the increase seen inboth healthy and septic subjects in this and other studies.
Although we controlled for the major factors influencingendothelial function, we cannot exclude minor influences ofaltered thyroid or adrenal function. Due to variations in sampleprocessing time, we were unable to determine accurateplasma arginine values for all subjects. Thus the reportedarginine values may not be fully representative of the groups asa whole. Of the subjects who had an initial measurement ofRH-PAT index, 70% had a repeat measurement 2 to 4 dayslater. Although those who were not followed up had a similarbaseline APACHE II score to those who were followed up, thismay not have been a representative population, because sub-jects who rapidly improved and were discharged home did nothave repeat measurements. Thus the observed degree ofrecovery in microvascular reactivity is likely to be anunderestimate.
The mortality rate in this cohort was low (hospital and 28-daymortality 9% overall and 21% among those with septic shock).Although this is consistent with the relatively low mortality ratein severe sepsis previously documented in our ICU [35], it
Page 7 of 9(page number not for citation purposes)
Critical Care Vol 13 No 5 Davis et al.
does mean that the study may have been underpowered todetect associations of measured variables with mortality.
ConclusionsIn summary, we have found that peripheral arterial tonometry isa feasible tool for measuring microvascular reactivity in sepsis,and that it is impaired in sepsis in proportion to disease sever-ity, suggesting reduced endothelial function and decreasedendothelial NO bioavailability. Baseline RH-PAT was useful inpredicting subsequent deterioration in organ dysfunction,although this should be reproduced by other investigatorsbefore its clinical utility can be confirmed. Given the growinginterest in HMG CoA reductase inhibitors [58] and otherpotential adjunctive therapies targeting the endothelium insepsis [55], better tools for monitoring the response of theendothelium in clinical trials are needed. RH-PAT is an attrac-tive option for such studies, as other current methods are user-dependent and have limited availability.
Competing interestsDC has received research support (as equipment) from ItamarMedical, the manufacturer of the RH-PAT device, and hasreceived speaker's fees (less than US$1000 per year) forspeaking at Itamar-sponsored educational events. The otherauthors have no competing interests.
Authors' contributionsStudy design was performed by JSD, NMA, TWY, DPS andDSC. Patient recruitment was carried out by JHT, MM, JSDand DPS. The data was processed by JSD and MM, and wasanalysed by JSD with help from ACC, TWY and NMA. Labo-ratory sample processing and HPLC assays were performedby CJD and YRM. The manuscript was drafted by JSD andNMA. All authors had access to all data and contributed to thefinal draft of the paper. All authors read and approved the finalmanuscript.
AcknowledgementsWe would like to thank Kim Piera, Tonia Woodberry, Barbara Mac-Hunter and Catherine Jones for laboratory assistance; Karl Blenk, Antony Van Asche, Steven Tong and Paulene Kittler for RH-PAT meas-urements; Craig Boutlis for help with initial study design; Ric Price and
Joseph McDonnell for statistical advice; and the medical and nursing staff of the Royal Darwin Hospital Intensive Care and Hospital in the Home units.
Funding sources: The study was funded by the National Health and Medical Research Council of Australia (NHMRC Program Grants 290208, 496600; Practitioner Fellowship to NMA, Scholarship to JSD). The funding source played no role in the design or conduct of the study, nor in the drafting of the manuscript or the decision to submit it for publication.
References1. Angus DC, Pereira CA, Silva E: Epidemiology of severe sepsis
around the world. Endocr Metab Immune Disord Drug Targets2006, 6:207-212.
2. Ince C, Sinaasappel M: Microcirculatory oxygenation andshunting in sepsis and shock. Crit Care Med 1999,27:1369-1377.
3. Aird WC: The role of the endothelium in severe sepsis andmultiple organ dysfunction syndrome. Blood 2003,101:3765-3777.
4. Deanfield JE, Halcox JP, Rabelink TJ: Endothelial function anddysfunction: testing and clinical relevance. Circulation 2007,115:1285-1295.
5. Young JD, Cameron EM: Dynamics of skin blood flow in humansepsis. Intensive Care Med 1995, 21:669-674.
6. Neviere R, Mathieu D, Chagnon JL, Lebleu N, Millien JP, Wattel F:Skeletal muscle microvascular blood flow and oxygen trans-port in patients with severe sepsis. Am J Respir Crit Care Med1996, 153:191-195.
7. Kubli S, Boegli Y, Ave AD, Liaudet L, Revelly JP, Golay S, BroccardA, Waeber B, Schaller MD, Feihl F: Endothelium-dependentvasodilation in the skin microcirculation of patients with septicshock. Shock (Augusta, Ga) 2003, 19:274-280.
8. Hartl WH, Gunther B, Inthorn D, Heberer G: Reactive hyperemiain patients with septic conditions. Surgery 1988, 103:440-444.
9. Astiz ME, DeGent GE, Lin RY, Rackow EC: Microvascular func-tion and rheologic changes in hyperdynamic sepsis. Crit CareMed 1995, 23:265-271.
10. Vaudo G, Marchesi S, Siepi D, Brozzetti M, Lombardini R, Pirro M,Alaeddin A, Roscini AR, Lupattelli G, Mannarino E: Humanendothelial impairment in sepsis. Atherosclerosis 2007,197:747-752.
11. Creteur J, Carollo T, Soldati G, Buchele G, De Backer D, VincentJL: The prognostic value of muscle StO(2) in septic patients.Intensive Care Med 2007, 33:1549-1556.
12. Celermajer DS: Reliable endothelial function testing: at ourfingertips? Circulation 2008, 117:2428-2430.
13. Bonetti PO, Pumper GM, Higano ST, Holmes DR Jr, Kuvin JT, Ler-man A: Noninvasive identification of patients with early coro-nary atherosclerosis by assessment of digital reactivehyperemia. J Am Coll Cardiol 2004, 44:2137-2141.
14. Chenzbraun A, Levin G, Scheffy J, Keren A, Stern S, Goor D: Theperipheral vascular response to exercise is impaired inpatients with risk factors for coronary artery disease. Cardiol-ogy 2001, 95:126-130.
15. Haller MJ, Stein J, Shuster J, Theriaque D, Silverstein J, Schatz DA,Earing MG, Lerman A, Mahmud FH: Peripheral artery tonometrydemonstrates altered endothelial function in children withtype 1 diabetes. Pediatr Diabetes 2007, 8:193-198.
16. Kuvin JT, Mammen A, Mooney P, Alsheikh-Ali AA, Karas RH:Assessment of peripheral vascular endothelial function in theambulatory setting. Vasc Med 2007, 12:13-16.
17. Hamburg NM, Keyes MJ, Larson MG, Vasan RS, Schnabel R,Pryde MM, Mitchell GF, Sheffy J, Vita JA, Benjamin EJ: Cross-sec-tional relations of digital vascular function to cardiovascularrisk factors in the Framingham Heart Study. Circulation 2008,117:2467-2474.
18. Nohria A, Gerhard-Herman M, Creager MA, Hurley S, Mitra D,Ganz P: Role of nitric oxide in the regulation of digital pulsevolume amplitude in humans. J Appl Physiol 2006,101:545-548.
19. Kuvin JT, Patel AR, Sliney KA, Pandian NG, Sheffy J, Schnall RP,Karas RH, Udelson JE: Assessment of peripheral vascular
Key messages
• Current tools for assessing endothelial function in patients with sepsis are generally user dependant and are not widely available.
• Peripheral arterial tonometry, a simple, user-independ-ent technique for measuring endothelium-dependent microvascular reactivity is feasible in patients with sepsis.
• Endothelium-dependent microvascular reactivity is impaired in sepsis, in proportion to disease severity, and may predict subsequent deterioration in organ function.
Page 8 of 9(page number not for citation purposes)
Available online http://ccforum.com/content/13/5/R155
endothelial function with finger arterial pulse wave amplitude.Am Heart J 2003, 146:168-174.
20. Dhindsa M, Sommerlad SM, DeVan AE, Barnes JN, Sugawara J,Ley O, Tanaka H: Interrelationships among noninvasive meas-ures of postischemic macro- and microvascular reactivity. JAppl Physiol 2008, 105:427-432.
21. Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, McNeilYR, Darcy CJ, Granger DL, Weinberg JB, Lopansri BK, Price RN,Duffull SB, Celermajer DS, Anstey NM: Impaired nitric oxide bio-availability and L-arginine reversible endothelial dysfunction inadults with falciparum malaria. J Exp Med 2007,204:2693-2704.
22. Luiking YC, Poeze M, Ramsay G, Deutz NE: Reduced citrullineproduction in sepsis is related to diminished de novo arginineand nitric oxide production. Am J Clin Nutr 2009, 89:142-152.
23. Kao CC, Bandi V, Guntupalli KK, Wu M, Castillo L, Jahoor F:Arginine, citrulline, and nitric oxide metabolism in sepsis. ClinSci (Lond) 2009, 117:23-30.
24. Villalpando S, Gopal J, Balasubramanyam A, Bandi VP, GuntupalliK, Jahoor F: In vivo arginine production and intravascular nitricoxide synthesis in hypotensive sepsis. Am J Clin Nutr 2006,84:197-203.
25. McGown CC, Brookes ZL: Beneficial effects of statins on themicrocirculation during sepsis: the role of nitric oxide. Br JAnaesth 2007, 98:163-175.
26. Hecker M, Sessa WC, Harris HJ, Anggard EE, Vane JR: Themetabolism of L-arginine and its significance for the biosyn-thesis of endothelium-derived relaxing factor: culturedendothelial cells recycle L-citrulline to L-arginine. Proc NatlAcad Sci USA 1990, 87:8612-8616.
27. Luiking YC, Poeze M, Dejong CH, Ramsay G, Deutz NE: Sepsis:an arginine deficiency state? Crit Care Med 2004,32:2135-2145.
28. Chiarla C, Giovannini I, Siegel JH, Boldrini G, Castagneto M: Therelationship between plasma taurine and other amino acid lev-els in human sepsis. J Nutr 2000, 130:2222-2227.
29. Ochoa JB, Udekwu AO, Billiar TR, Curran RD, Cerra FB, SimmonsRL, Peitzman AB: Nitrogen oxide levels in patients after traumaand during sepsis. Ann Surg 1991, 214:621-626.
30. Askanazi J, Carpentier YA, Michelsen CB, Elwyn DH, Furst P,Kantrowitz LR, Gump FE, Kinney JM: Muscle and plasma aminoacids following injury. Influence of intercurrent infection. AnnSurg 1980, 192:78-85.
31. Sprung CL, Cerra FB, Freund HR, Schein RM, Konstantinides FN,Marcial EH, Pena M: Amino acid alterations and encephalopa-thy in the sepsis syndrome. Crit Care Med 1991, 19:753-757.
32. Druml W, Heinzel G, Kleinberger G: Amino acid kinetics inpatients with sepsis. Am J Clin Nutr 2001, 73:908-913.
33. Hallemeesch MM, Lamers WH, Deutz NE: Reduced arginineavailability and nitric oxide production. Clin Nutr 2002,21:273-279.
34. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA,Schein RM, Sibbald WJ: Definitions for sepsis and organ failureand guidelines for the use of innovative therapies in sepsis.The ACCP/SCCM Consensus Conference Committee. Ameri-can College of Chest Physicians/Society of Critical CareMedicine. Chest 1992, 101:1644-1655.
35. Stephens DP, Thomas JH, Higgins A, Bailey M, Anstey NM, CurrieBJ, Cheng AC: Randomized, double-blind, placebo-controlledtrial of granulocyte colony-stimulating factor in patients withseptic shock. Crit Care Med 2008, 36:448-454.
36. Bland JM, Altman DG: Statistical methods for assessing agree-ment between two methods of clinical measurement. Lancet1986, 1:307-310.
37. Bonetti PO, Barsness GW, Keelan PC, Schnell TI, Pumper GM,Kuvin JT, Schnall RP, Holmes DR, Higano ST, Lerman A:Enhanced external counterpulsation improves endothelialfunction in patients with symptomatic coronary artery disease.J Am Coll Cardiol 2003, 41:1761-1768.
38. Jarvisalo MJ, Jartti L, Marniemi J, Ronnemaa T, Viikari JS, LehtimakiT, Raitakari OT: Determinants of short-term variation in arterialflow-mediated dilatation in healthy young men. Clin Sci (Lond)2006, 110:475-482.
39. van Wandelen C, Cohen SA: Using quaternary high-perform-ance liquid chromatography eluent systems for separating 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate-derivatizedamino acid mixtures. J Chromatogr A 1997, 763:11-22.
40. Nuttall KL, Chen M, Komaromy-Hiller G: Delayed separation andthe plasma amino acids arginine and ornithine. Ann Clin LabSci 1998, 28:354-359.
41. Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, Piera K,Price RN, Duffull SB, Celermajer DS, Anstey NM: Angiopoietin-2is associated with decreased endothelial nitric oxide and poorclinical outcome in severe falciparum malaria. Proc Natl AcadSci USA 2008, 105:17097-17102.
42. Bellien J, Thuillez C, Joannides R: Contribution of endothelium-derived hyperpolarizing factors to the regulation of vasculartone in humans. Fundam Clin Pharmacol 2008, 22:363-377.
43. Mitchell JA, Ali F, Bailey L, Moreno L, Harrington LS: Role of nitricoxide and prostacyclin as vasoactive hormones released bythe endothelium. Exp Physiol 2008, 93:141-147.
44. O'Dwyer MJ, Dempsey F, Crowley V, Kelleher DP, McManus R,Ryan T: Septic shock is correlated with asymmetrical dimethylarginine levels, which may be influenced by a polymorphism inthe dimethylarginine dimethylaminohydrolase II gene: a pro-spective observational study. Crit Care 2006, 10:R139.
45. Xia Y, Roman LJ, Masters BS, Zweier JL: Inducible nitric-oxidesynthase generates superoxide from the reductase domain. JBiol Chem 1998, 273:22635-22639.
46. Argaman Z, Young VR, Noviski N, Castillo-Rosas L, Lu XM, Zura-kowski D, Cooper M, Davison C, Tharakan JF, Ajami A, Castillo J:Arginine and nitric oxide metabolism in critically ill septic pedi-atric patients. Crit Care Med 2003, 31:591-597.
47. Kneale BJ, Chowienczyk PJ, Brett SE, Coltart DJ, Ritter JM: Gen-der differences in sensitivity to adrenergic agonists of forearmresistance vasculature. J Am Coll Cardiol 2000, 36:1233-1238.
48. van Eijk LT, Dorresteijn MJ, Smits P, Hoeven JG van der, NeteaMG, Pickkers P: Gender differences in the innate immuneresponse and vascular reactivity following the administrationof endotoxin to human volunteers. Crit Care Med 2007,35:1464-1469.
49. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D:Elevation of systemic oxygen delivery in the treatment of criti-cally ill patients. New Engl J Med 1994, 330:1717-1722.
50. Palizas F, Dubin A, Regueira T, Bruhn A, Knobel E, Lazzeri S, Bare-des N, Hernandez G: Gastric tonometry versus cardiac index asresuscitation goals in septic shock: a multicenter, randomized,controlled trial. Crit Care 2009, 13:R44.
51. Spronk PE, Ince C, Gardien MJ, Mathura KR, OudemansvanStraaten HM, Zandstra DF: Nitroglycerin in septic shock afterintravascular volume resuscitation. Lancet 2002,360:1395-1396.
52. Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut JF,Lopez-Rodriguez A, Steingrub JS, Garber GE, Helterbrand JD, ElyEW, Fisher CJ: Efficacy and safety of recombinant human acti-vated protein C for severe sepsis. New Engl J Med 2001,344:699-709.
53. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B,Peterson E, Tomlanovich M: Early goal-directed therapy in thetreatment of severe sepsis and septic shock. New Engl J Med2001, 345:1368-1377.
54. Berghe G van den, Wouters P, Weekers F, Verwaest C, Bruyn-inckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouil-lon R: Intensive insulin therapy in the critically ill patients. NewEngl J Med 2001, 345:1359-1367.
55. Aird WC: Endothelium as a therapeutic target in sepsis. CurrDrug Targets 2007, 8:501-507.
56. Astiz ME, Tilly E, Rackow ED, Weil MH: Peripheral vascular tonein sepsis. Chest 1991, 99:1072-1075.
57. Hudlicka O: Regulation of muscle blood flow. Clin Physiol1985, 5:201-229.
58. Terblanche M, Almog Y, Rosenson RS, Smith TS, Hackam DG:Statins: panacea for sepsis? Lancet Infect Dis 2006,6:242-248.
Page 9 of 9(page number not for citation purposes)
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Journal of Chromatography B, 878 (2010) 8–12
Contents lists available at ScienceDirect
Journal of Chromatography B
journa l homepage: www.e lsev ier .com/ locate /chromb
HPLC analysis of asymmetric dimethylarginine, symmetric dimethylarginine,homoarginine and arginine in small plasma volumes using a Gemini-NX columnat high pH
Catherine E. Jones1, Christabelle J. Darcy ∗,1, Tonia Woodberry, Nicholas M. Anstey, Yvette R. McNeilMenzies School of Health Research, Rocklands Drive, Tiwi, Darwin, NT, Australia
a r t i c l e i n f o
Article history:Received 5 August 2009Accepted 30 October 2009Available online 6 November 2009
Keywords:Asymmetric dimethylarginineSymmetric dimethylarginineHomoarginineArginineHigh performance liquid chromatography
a b s t r a c t
There is increasing recognition of the clinical importance of endogenous nitric oxide synthase inhibitorsin critical illness. This has highlighted the need for an accurate high performance liquid chromatography(HPLC) method for detection of asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine(SDMA) in small volumes of blood. Here, the validation of an accurate, precise HPLC method for thedetermination of ADMA, SDMA, homoarginine and arginine concentrations in plasma is described. Solidphase extraction is followed by derivatisation with AccQ-FluorTM and reversed phase separation on aGemini-NX column at pH 9. Simultaneous detection by both UV–vis and fluorescence detectors affordsextra validation. This solid phase extraction method gives absolute recoveries of more than 85% for ADMAand SDMA and relative recoveries of 102% for ADMA and 101% for SDMA. The intra-assay relative stan-dard deviations are 2.1% and 2.3% for ADMA and SDMA, respectively, with inter-assay relative standarddeviations of 2.7% and 3.1%, respectively. Advantages of this method include improved recovery of allanalytes using isopropanol in the solid phase extraction; sharp, well-resolved chromatographic peaksusing a high pH mobile phase; a non-endogenous internal standard, n-propyl l-arginine; and accurateand precise determination of methylated arginine concentrations from only 100 L of plasma.
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
The clinical importance of endogenous nitric oxide synthase(NOS) inhibitors has long been recognised in chronic disease [1].Nitric oxide (NO) is important in the maintenance of normalendothelial function [2] and the prevention of platelet aggrega-tion [3]. NO synthesis from l-arginine is reduced in the presenceof asymmetric dimethylarginine (ADMA) and symmetric dimethy-larginine (SDMA), which are products of methylated proteindegradation.
ADMA and homoarginine compete with arginine for specificbinding sites on NOS. Homoarginine is an alternative but less effi-cient substrate for NOS [4] whereas ADMA directly inhibits nitricoxide synthases. ADMA, SDMA and homoarginine each competewith arginine for transport into the cell [5] and may therefore, alsolimit the amount of arginine available to NOS [6,7]. High concen-trations of methylated arginines have been associated with a broadrange of chronic diseases, including hypertension [8], renal failure
∗ Corresponding author at: International Health Division, Menzies School ofHealth Research, PO Box 41096, Casuarina, Darwin, NT 0811, Australia.Tel.: +61 889228839; fax: +61 889275187.
E-mail address: [email protected] (C.J. Darcy).1 These authors contributed equally.
[1], hypercholesterolemia [9] and diabetes [10]. Indeed, elevatedADMA is an independent risk factor for both cardiovascular disease[11] and all-cause mortality [12].
In addition to the importance of ADMA in chronic disease, thereis increasing recognition of its important role in acute critical ill-ness [13,14] and acute inflammatory conditions such septic shock[15]. As limited blood is available from critically ill patients, thereis a need for an accurate high performance liquid chromatography(HPLC) method for detection of ADMA and SDMA in small volumesof blood.
This paper describes a reversed phase HPLC method for the mea-surement of arginine, ADMA, SDMA and homoarginine from 100 Lof plasma. The chromatography utilised a Gemini-NX column witha novel, high pH borate buffer-acetonitrile gradient, and the non-endogenous internal standard n-propyl l-arginine (NPLA). Samplepreparation utilised solid phase extraction (SPE) and fluorescentderivatisation. The extraction procedure and HPLC method giveaccurate and precise results from a small volume of plasma.
2. Experimental
2.1. Materials
l-Arginine-HCl, l-homo-arginine-HCl, NG,NG di-methyl-l-arginine and NG,NG′
di-methyl-l-arginine were purchased from
1570-0232/$ – see front matter © 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.jchromb.2009.10.035
Author's personal copy
C.E. Jones et al. / J. Chromatogr. B 878 (2010) 8–12 9
Calbiochem (La Jolla, CA, USA). n-Propyl l-arginine was a productof Cayman Chemicals (Ann Arbor, MI, USA). Sodium tetra boratedecahydrate and boric acid were obtained from Sigma–Aldrich (St.Louis, MO, USA). Oasis Mixed Mode Cation Exchange (MCX) car-tridges (1 mL, 30 cm3) were purchased from Waters (Milford, MAUSA). Isopropanol and ammonia solution 28–30% were purchasedfrom Merck (Darmstadt, Germany). HPLC-grade acetonitrile wasobtained from Burdick and Jackson (Muskego, MI, USA). Highpurity water was used to prepare all aqueous solutions (Milli-Qwater system, Milli-Pore, Billerica, MA, USA). The AccQ-FluorTM kitfrom Waters (Milford, MA, USA) contained the fluorescent reagent6-aminoquinolyl-N-hydroxysuccinimidyl, a vial of acetonitrilediluent, and a vial of aqueous borate buffer (0.2 M, pH 8.8) for thederivatisation reaction.
2.2. Plasma samples
Venous blood from healthy volunteers or patients was collectedinto lithium heparin tubes, centrifuged (492 × g for 8 min) within120 min of collection and the plasma were frozen at −80 C untilanalysis. A pool of plasma from Australian Red Cross blood donorswas used as quality control plasma.
Plasma from 30 apparently healthy volunteers was used todetermine healthy concentrations of ADMA and SDMA. 8 of thesevolunteers were laboratory staff (blood collected as above) and 22were blood bank donors (blood collected according to standardAustralian Red Cross blood bank procedures). Blood from bloodbank donors was usually separated the day after collection. Theage range of the healthy volunteers was 16–61; 18 were female and12 were male. The use of this plasma was approved by the EthicsCommittees of the Australian Red Cross and the Menzies School ofHealth Research.
2.3. Extraction
Oasis MCX cartridges were affixed to a vacuum manifold andpre-equilibrated with 1 mL of isopropanol, followed by 1 mL of50 mM borate buffer (pH 9). 100 L of plasma or calibrator wasmixed with 100 L 15 M NPLA and diluted with 800 L 50 mMborate buffer (pH 9) and then loaded onto the cartridge. Cartridgeswere then washed with 1 mL of water and then 1 mL of isopropanol.Extracts were eluted from the cartridges into glass collection tubeswith 1 mL of eluting solvent (isopropanol:water:28–30% ammoniasolution (5:4:1)). Flow rates were controlled by vacuum adjust-ment. The vacuum manifold pressure was less than 254 mm Hg forthe pre-equilibration and wash steps, and less than 127 mm Hg forthe loading and eluting steps.
Extracts were dried under nitrogen at 75 C (for approximately1 h). Dried eluates were reconstituted in 0.2 mL water and trans-ferred to glass storage vials.
2.4. Derivatisation
Extracts were derivatised with Waters AccQ-FluorTM kit priorto chromatography. In a 250 L HPLC vial insert; 20 L of extract,diluted with 70 L of Waters’ borate buffer, was reacted with 10 LAccQ-FluorTM reagent by immediate vortexing for 10 s.
2.5. Chromatography
The Shimadzu VP series HPLC system consisted of a gradientpump, degasser, column oven (42 C) and UV-vis and fluores-cence detectors. The detectors were connected in series forsimultaneous detection of UV (absorption wavelength = 250 nm)and fluorescence (excitation wavelength = 250 nm, emission wave-length = 395 nm). Extracts were separated on a C18 Gemini-NX
Table 1Mobile phase delivery program.
Time (min) Eluenta Value (%) Event
0.00–18.00 A:B 93:7 Isocratic18.01–21.00 A:B 93:7 92:8 Gradient 7–8% over 3 min21.01–29.00 A:B 92:8 Isocratic29.01–40.00 A:B 87:13 Isocratic40.01–52.00 B:C 65:35 Wash
a Eluents: 20 mM borate buffer pH 9 (A), acetonitrile (B) and water (C).
analytical column (150 mm × 4.6 mm, 3 m) protected by a C18Gemini-NX security guard cartridge (4.0 mm × 3.0 mm), both fromPhenomenex (Lane Cove, NSW, Australia). Mobile phase flow ratewas 1 mL min−1.
A 100 mM stock solution of sodium tetra borate/boric acid wasprepared and filtered (0.2 m) into a sterile container. The stockwas kept at room temperature. Eluent A was a 1:5 dilution of theborate buffer stock solution.
The mobile phase delivery program of 20 mM borate buffer pH9 (A), acetonitrile (B) and water (C) is shown in Table 1. All eluentswere filtered through 0.45 m filters before use.
2.6. Calibration and validation
Stock solutions of arginine (2.5 mM), homoarginine (500 M),ADMA (100 M), SDMA (100 M) and NPLA (2.5 mM) were pre-pared, aliquoted and stored at −80 C. Seven calibration standardswere made to encompass physiological and disease-associatedconcentration ranges. Arginine covered the range of 7.5–200 M,homoarginine 0.5–12 M, ADMA 0.25–6 M and SDMA 0.25–6 M.The calibration standards were extracted and derivatised in thesame manner as plasma samples. Identification of analytes withinplasma samples was based on the retention time of the correspond-ing standard. A seven level calibration curve for each analyte, usingpeak area/amount ratios of the analytes to internal standard wasconstructed from integrated chromatograms.
Analyte recovery during the extraction process was determinedby calculating the relative recovery and absolute concentra-tions recovered after calibration standards were subjected to SPEcompared with un-extracted calibrator concentrations. Seven stan-dards were run without undergoing SPE in parallel with aliquots ofthe same standards subjected to SPE. Absolute recovery was calcu-lated by comparing the area of the extracted peaks to the area of theun-extracted peaks. This ensured no particular analyte was prefer-entially lost through extraction. Relative recovery was calculated byplotting the extracted calibrators onto the curve of the un-extractedcalibrators. The percent recovery was calculated by dividing themeasured concentration by the theoretical concentration from theun-extracted curve.
The HPLC method was validated by calculating the intra-assayand inter-assay precision of pooled quality control plasma and bydetermining the spike recovery of analyte added to control plasma.The intra-assay precision of the HPLC method was determinedby running a single extract of control plasma 10 times consecu-tively and calculating the concentration of the analytes of interest.Inter-assay precision was calculated by extracting and running 30separate control plasmas over 2 months. In order to determine theaccuracy of the HPLC method, the pooled quality control plasmawas spiked with known concentrations of arginine, homoarginine,ADMA and SDMA. The percent spike recovery was expressed asthe recovery of added analyte from spiked plasma samples. Thisprocess was repeated three times in 6 months.
Limit of detection (LOD) was determined by a signal to noiseratio of 2:1 and the limit of quantification (LOQ) was determinedby a signal to noise ratio of 10:1.
Author's personal copy
10 C.E. Jones et al. / J. Chromatogr. B 878 (2010) 8–12
Fig. 1. Fluorescence detection of a calibration standard (A) with 30 M arginine,2 M homoarginine, 1 M ADMA and 1 M SDMA and (B) the pooled quality con-trol plasma (black) with 23.68 M arginine, 1.82 M homoarginine, 0.48 M ADMAand 0.39 M SDMA, overlaid with a chromatogram from a patient with falciparummalaria (red) without internal standard added. Peak identity: (1) arginine; (2)homoarginine, (3) ADMA, (4) SDMA, (5) NPLA. Inset B: region 27–31 min magnified40×.
3. Results and discussion
3.1. Chromatography
Homoarginine, ADMA and SDMA were detected simultaneouslyusing UV and fluorescence detection. Arginine was out of range offluorescence detection once above 30 M and was therefore pri-marily detected by UV. There was less than 5% deviation betweenADMA and SDMA values measured by either fluorescence or UV.Validation data presented in this paper is from the fluorescentdetection of ADMA, SDMA and homoarginine and the UV detectionof arginine.
This method provided excellent separation of arginine,homoarginine, ADMA, SDMA and NPLA. Fig. 1 shows the separa-tion of analytes in a standard, the pooled quality control plasmaand plasma from a malaria patient. Blank samples of water alsounderwent the extraction and derivatisation processes and werechromatographed to ensure there were no co-eluting peaks origi-nating from the SPE method or the derivatising agent. The pooledquality control plasma and plasma from 2 patients with bacterialsepsis and 2 patients with falciparum malaria were subjected toSPE without the addition of internal standard, to ensure there wasa flat baseline under NPLA (see Fig. 1B).
The coefficient of determination (r2) for each analyte was>0.999. Limit of detection was 0.04 M for arginine, 0.06 M forhomoarginine, 0.04 M for ADMA and 0.03 M for SDMA. The limit
Table 2Average absolute and relative recovery of analytes from 7 level calibration standardsafter solid phase extraction (n = 4).
Analyte (conc. range) Absolute recoverymean ± SD %
Relative recoverymean ± SD %
Arginine (7.5–200 M) 80.9 ± 5.6 98.9 ± 2.5Homoarginine (0.5–12 M) 78.1 ± 5.6 94.9 ± 3.2ADMA (0.25–6 M) 85.1 ± 6.5 101.6 ± 1.3SDMA (0.25–6 M) 86.3 ± 5.2 101.4 ± 2.4NPLA (15 M) 83.4 ± 5.5 100.0
of quantification was 0.20 M for arginine, 0.30 M for homoargi-nine, 0.20 M for ADMA and 0.15 M for SDMA.
Borate was chosen as the mobile phase buffer in this method asit is also the matrix of the derivatised samples and greatest reten-tion time reproducibility is obtained when samples are dissolvedin a similar solution to the mobile phase. The borate buffer wasprepared to pH 9 as the pKa of borate buffer is 9.2 and buffers aremost effective within 0.5 pH units of their pKa. The combinationof high pH and acetonitrile resulted in sharp, well-resolved chro-matographic peaks. The Gemini-NX column was selected for thismethod as it has a large pH stability range of 1–12.
3.2. Extraction and derivatisation
A number of different extraction solvents and procedures weretrialled, including the procedures recommended in the Oasis MCXcartridge literature. Most published methods use methanol inthe final eluting solution and/or during the pre-equilibration andwash stages. However, optimal recovery of all analytes, especiallyNPLA, was obtained by substituting methanol with the slightly lesspolar alcohol, isopropanol. The cleanest extracts were producedwhen the cartridges were pre-equilibrated with the sample matrix(50 mM borate pH 9). Water was added to the eluting mixture toincrease arginine recovery [16]. The absolute and relative recover-ies of the SPE method are shown in Table 2.
As the fluorescent adducts of AccQ-FluorTM are stable for at least7 days [17], large batches of samples can be efficiently extracted andderivatised.
3.3. Method validation
Method precision was evaluated using the pooled quality con-trol plasma. The inter-assay percent relative standard deviations(RSDs) (n = 10) were less than 2.3% for all analytes. The inter-assayRSDs for ADMA (2.7%) and SDMA (3.1%) compare very well to otherHPLC assays using fluorescence detection [16–20] and to HPLC orgas chromatography mass spectrometry methods [21,22]. As ADMAand SDMA have a very narrow concentration range in the gen-eral population, high analytical precision is required to produceclinically useful results [23]. Blackwell et al. [24] recently deter-mined the intra-individual variability for ADMA and SDMA to be7.4% and 5.8%, respectively in healthy European volunteers. Theminimum required precision of an assay is defined as 0.75 timesthe intra-individual variability [24,25]. This definition requires thatinter-assay RSDs be ≤5.6% for ADMA and ≤4.4% for SDMA. Desirableimprecision goals are defined as 0.5 times the intra-individual vari-ability [25] which is ≤3.7% for ADMA and ≤2.9% for SDMA [24]. Theinter-assay RSDs for ADMA with this method are within the desir-able imprecision goals. The inter-assay RSDs for SDMA come closeto the desirable imprecision goals and are well within the minimumrequirements. As Blackwell et al. note, few published methods formeasuring ADMA and SDMA meet these desirable precision goals.Data on the precision of this method are presented in Table 3.
An aliquot of pooled quality control plasma was analysed byHPLC at an independent research laboratory with an established,
Author's personal copy
C.E. Jones et al. / J. Chromatogr. B 878 (2010) 8–12 11
Table 3Intra-assay (n = 10) and inter-assay (n = 30) precision calculated from pooled quality control plasma.
Analyte Intra-assaymean (M) ± SD
Intra-assayRSD (%)
Inter-assaymean (M) ± SD
Inter-assayRSD (%)
Arginine 21.06 ± 0.2 0.93 23.68 ± 1.86 7.88Homoarginine 1.87 ± 0.02 1.22 1.88 ± 0.09 4.57ADMA 0.49 ± 0.01 2.06 0.48 ± 0.01 2.69SDMA 0.39 ± 0.01 2.26 0.38 ± 0.01 3.07
Table 4Assay accuracy calculated from spiked plasma (n = 3)a.
Analyte Concentration (M) RSD (%) Mean spikerecovered (M)
Accuracy/spikerecovery (%)
Mean unspiked plasma Spike added Mean spiked plasma SD
Arginine 11.70 3.78 15.58 0.41 2.63 3.88 102.87.55 19.86 0.91 4.61 8.16 108.1
12.60 25.47 1.01 3.95 13.78 109.415.10 26.91 0.82 3.05 15.22 100.825.20 37.50 0.48 1.27 25.80 102.450.50 64.14 1.81 2.82 52.44 103.8
Homoarginine 0.94 0.50 1.35 0.26 18.94 0.41 81.30.75 1.63 0.27 16.34 0.69 92.01.00 1.86 0.29 15.83 0.92 91.71.50 2.42 0.31 12.97 1.48 98.43.00 4.03 0.43 10.65 3.09 103.0
ADMA 0.25 0.13 0.36 0.02 4.81 0.12 92.00.25 0.51 0.03 5.70 0.26 104.70.38 0.62 0.02 2.45 0.38 100.90.50 0.75 0.02 2.05 0.50 100.30.75 1.00 0.06 6.09 0.76 101.11.50 1.78 0.10 5.55 1.53 102.1
SDMA 0.20 0.13 0.32 0.03 7.78 0.13 102.70.25 0.46 0.05 9.96 0.27 106.00.38 0.58 0.03 5.51 0.39 103.60.50 0.70 0.04 5.71 0.51 101.00.75 0.96 0.03 3.13 0.77 102.01.50 1.72 0.07 4.08 1.53 101.9
a Calculated as a percentage of spike recovered from spiked plasma after subtraction of the unspiked plasma concentration.
validated method [17]. This laboratory reported mean values of0.48 M ADMA and 0.35 M SDMA, which concurred with theresults obtained using this method.
Data on accuracy, expressed as recovery of added analyte fromspiked quality control plasma (n = 3), are presented in Table 4.
This assay has since been used successfully to measure plasmadimethylarginines in over 194 patients with critical illness. It isimportant to note that of these patients, only 15 had ADMA morethan 1 M (unpublished data). Hence this assay was optimised tobe accurate and precise at low concentrations of ADMA and SDMA.
3.4. Healthy plasma levels
Thirty apparently healthy volunteers provided plasma samples.The mean and standard deviation of each analyte of interest areshown in Table 5. These values were within the healthy rangereported by others [24,26], with the exception of l-arginine con-centration, which was lower than expected due to the delay inprocessing blood from blood bank donors [27].
Table 5Healthy plasma arginine, homoarginine and methylated arginine values (n = 30).
Arginine (M) Homoarginine (M) ADMA (M) SDMA (M)
Min 23.40 0.86 0.30 0.20Max 152.92 3.95 0.58 0.54Mean 66.91 2.15 0.45 0.40SD 33.46 0.75 0.07 0.09
3.5. Limitations and strengths of the assay
A limitation of this assay is the need to condition new HPLCcolumns before retention times stabilise, a requirement notedin other methods [28–31]. After conditioning the new columnwith repeated injections of either standards or the quality controlplasma, retention times stabilised and excellent retention timeswere then obtained for the duration of the column life. This methodhas been used with three Gemini-NX columns, each lasting approx-imately 900 injections.
This method is not as short as a number of other publishedmethods because it uses AccQ-FluorTM derivatisation and a non-endogenous internal standard. AccQ-FluorTM derivatisation leadsto longer chromatography [32,33], however the stable adducts pro-duced by AccQ-FluorTM give accurate results without requiringon-line derivatisation. Furthermore, the shorter published methodstend to use either monomethylarginine (MMA) or homoarginine asinternal standards, concentrations of which may be altered in dis-ease states [20,34]. Using a non-endogenous internal standard givesmore accurate results and also allows all analytes to be quantitatedin plasma.
This method has several strengths. Firstly, the substitution ofmethanol with isopropanol in the SPE method gives improvedrecovery of all analytes. Secondly, a combination of the ace-tonitrile gradient and borate buffer at pH 9 on the Gemini-NXcolumn produced clearly defined chromatographic peaks. Thirdly,the average accuracy of ADMA was 100.2 ± 4.3% while for SDMAit was 102.9 ± 1.8%. Finally, the inter-assay RSDs for ADMA arewithin the desirable precision goals set out by Blackwell et
Author's personal copy
12 C.E. Jones et al. / J. Chromatogr. B 878 (2010) 8–12
al. [24] while SDMA measurements easily meet the minimumstandards and come close to achieving the desirable precisiongoals.
Importantly, as this method achieves accurate and preciseresults from small volumes of plasma it is particularly useful forresearch into critical illness.
Conflict of interest statement
The authors do not have a commercial or other association thatmight pose a conflict of interest.
Funding sources
The study was funded by the National Health and MedicalResearch Council of Australia (NHMRC Program Grants 290208,496600 and Practitioner Fellowship 490307) and National Insti-tutes of Health (AI041764).
Acknowledgments
We thank Tamila Heresztyn from the Cardiology Unit at the Uni-versity of Adelaide for kindly analysing an aliquot of the pooledquality control plasma. We also thank Dr Tsin Yeo who provided theplasma for the chromatogram from a malaria patient. We gratefullyacknowledge the support of the Australian Red Cross Blood Service.
References
[1] P. Vallance, A. Leone, A. Calver, J. Collier, S. Moncada, Lancet 339 (1992)572.
[2] P. Vallance, J. Collier, S. Moncada, Cardiovasc. Res. 23 (1989) 1053.[3] B.T. Mellion, L.J. Ignarro, E.H. Ohlstein, E.G. Pontecorvo, A.L. Hyman, P.J. Kad-
owitz, Blood 57 (1981) 946.[4] C. Moali, J.L. Boucher, M.A. Sari, D.J. Stuehr, D. Mansuy, Biochemistry 37 (1998)
10453.[5] K.K. McDonald, S. Zharikov, E.R. Block, M.S. Kilberg, J. Biol. Chem. 272 (1997)
31213.[6] E.I. Closs, F.Z. Basha, A. Habermeier, U. Forstermann, Nitric Oxide 1 (1997)
65.
[7] M. Kakoki, H.S. Kim, C.J. Edgell, N. Maeda, O. Smithies, D.L. Mattson, Am. J.Physiol. Renal Physiol. 291 (2006) F297.
[8] C.D. Goonasekera, D.D. Rees, P. Woolard, A. Frend, V. Shah, M.J. Dillon, J. Hyper-tens. 15 (1997) 901.
[9] R.H. Boger, D. Tsikas, S.M. Bode-Boger, L. Phivthong-Ngam, E. Schwedhelm, J.C.Frolich, Nitric Oxide 11 (2004) 1.
[10] A. Fard, C.H. Tuck, J.A. Donis, R. Sciacca, M.R. Di Tullio, H.D. Wu, T.A. Bryant, N.T.Chen, M. Torres-Tamayo, R. Ramasamy, L. Berglund, H.N. Ginsberg, S. Homma,P.J. Cannon, Arterioscler. Thromb. Vasc. Biol. 20 (2000) 2039.
[11] F. Schulze, H. Lenzen, C. Hanefeld, A. Bartling, K.J. Osterziel, L. Goudeva, C.Schmidt-Lucke, M. Kusus, R. Maas, E. Schwedhelm, D. Strodter, B.C. Simon, A.Mugge, W.G. Daniel, H. Tillmanns, B. Maisch, T. Streichert, R.H. Boger, Am. HeartJ. 152 (2006) 493, e1.
[12] R.H. Boger, L.M. Sullivan, E. Schwedhelm, T.J. Wang, R. Maas, E.J. Benjamin, F.Schulze, V. Xanthakis, R.A. Benndorf, R.S. Vasan, Circulation 119 (2009) 1592.
[13] R.J. Nijveldt, T. Teerlink, P.A. van Leeuwen, Clin. Nutr. 22 (2003) 99.[14] R.J. Nijveldt, T. Teerlink, B. Van Der Hoven, M.P. Siroen, D.J. Kuik, J.A. Rauwerda,
P.A. van Leeuwen, Clin. Nutr. 22 (2003) 23.[15] M.J. O’Dwyer, F. Dempsey, V. Crowley, D.P. Kelleher, R. McManus, T. Ryan, Crit.
Care 10 (2006) R139.[16] T. Teerlink, R.J. Nijveldt, S. de Jong, P.A. van Leeuwen, Anal. Biochem. 303 (2002)
131.[17] T. Heresztyn, M.I. Worthley, J.D. Horowitz, J. Chromatogr. B Analyt. Technol.
Biomed. Life Sci. 805 (2004) 325.[18] M. Tsunoda, S. Nonaka, T. Funatsu, Analyst 130 (2005) 1410.[19] S. Nonaka, M. Tsunoda, K. Imai, T. Funatsu, J. Chromatogr. A 1066 (2005) 41.[20] S. Blackwell, D.S. O’Reilly, D.K. Talwar, Clin. Chim. Acta 401 (2009) 14.[21] L.F. Huang, F.Q. Guo, Y.Z. Liang, B.Y. Li, B.M. Cheng, Anal. Bioanal. Chem. 380
(2004) 643.[22] J. Albsmeier, E. Schwedhelm, F. Schulze, M. Kastner, R.H. Boger, J. Chromatogr.
B Analyt. Technol. Biomed. Life Sci. 809 (2004) 59.[23] T. Teerlink, Clin. Chem. Lab. Med. 43 (2005) 1130.[24] S. Blackwell, S. O’Reilly, D.D. Talwar, Eur. J. Clin. Invest. 37 (2007) 364.[25] P.H. Petersen, C.G. Fraser, L. Jorgensen, I. Brandslund, M. Stahl, E. Gowans, J.C.
Libeer, C. Ricos, Ann. Clin. Biochem. 39 (2002) 543.[26] J.D. Horowitz, T. Heresztyn, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci.
851 (2007) 42.[27] K.L. Nuttall, M. Chen, G. Komaromy-Hiller, Ann. Clin. Lab. Sci. 28 (1998) 354.[28] R.D. Morrison, J.W. Dolan, LCGC North America June (2005).[29] D.R. Skotty, W.-Y. Lee, T.A. Nieman, Anal. Chem. 68 (1996) 1530.[30] N. Takenaga, Y. Ishii, S. Monden, Y. Sasaki, S. Hata, J. Chromatogr. B Biomed.
Appl. 674 (1995) 111.[31] A.M. Rustum, V. Estrada, J. Chromatogr. B Biomed. Sci. Appl. 705 (1998) 111.[32] M.T. Oreiro-Garcia, M.D. Vazquez-Illanes, G. Sierra-Paredes, G. Sierra-Marcuno,
Biomed. Chromatogr. 19 (2005) 720.[33] N. Ahmed, O.K. Argirov, H.S. Minhas, C.A. Cordeiro, P.J. Thornalley, Biochem. J.
364 (2002) 1.[34] J. Martens-Lobenhoffer, S.M. Bode-Boger, J. Chromatogr. B Analyt. Technol.
Biomed. Life Sci. 851 (2007) 30.
Asymmetric Dimethylarginine, Endothelial Nitric OxideBioavailability and Mortality in SepsisJoshua S. Davis1,2*., Christabelle J. Darcy1., Tsin W. Yeo1,2, Catherine Jones1, Yvette R. McNeil1,
Dianne P. Stephens4, David S. Celermajer3, Nicholas M. Anstey1,2
1 International Health Division, Menzies School of Health Research and Charles Darwin University, Darwin, Northern Territory, Australia, 2 Division of Medicine, Royal
Darwin Hospital, Darwin, Northern Territory, Australia, 3 Department of Medicine, University of Sydney and Department of Cardiology, Royal Prince Alfred Hospital,
Sydney, New South Wales, Australia, 4 Intensive Care Unit, Royal Darwin Hospital, Darwin, Northern Territory, Australia
Abstract
Background: Plasma concentrations of asymmetric dimethylarginine (ADMA), an endogenous inhibitor of nitric oxidesynthase, are raised in patients with chronic vascular disease, causing increased cardiovascular risk and endothelialdysfunction, but the role of ADMA in acute inflammatory states is less well defined.
Methods and Results: In a prospective longitudinal study in 67 patients with acute sepsis and 31 controls, digitalmicrovascular reactivity was measured by peripheral arterial tonometry and blood was collected at baseline and 2–4 dayslater. Plasma ADMA and L-arginine concentrations were determined by high performance liquid chromatography. Baselineplasma L-arginine: ADMA ratio was significantly lower in sepsis patients (median [IQR] 63 [45–103]) than in hospital controls(143 [123–166], p,0.0001) and correlated with microvascular reactivity (r = 0.34, R2 = 0.12, p = 0.02). Baseline plasma ADMAwas independently associated with 28-day mortality (Odds ratio [95% CI] for death in those in the highest quartile($0.66 mmol/L) = 20.8 [2.2–195.0], p = 0.008), and was independently correlated with severity of organ failure. Increase inADMA over time correlated with increase in organ failure and decrease in microvascular reactivity.
Conclusions: Impaired endothelial and microvascular function due to decreased endothelial NO bioavailability is a potentialmechanism linking increased plasma ADMA with organ failure and death in sepsis.
Citation: Davis JS, Darcy CJ, Yeo TW, Jones C, McNeil YR, et al. (2011) Asymmetric Dimethylarginine, Endothelial Nitric Oxide Bioavailability and Mortality inSepsis. PLoS ONE 6(2): e17260. doi:10.1371/journal.pone.0017260
Editor: Pieter Reitsma, Leiden University Medical Center, Netherlands
Received August 31, 2010; Accepted January 27, 2011; Published February 18, 2011
Copyright: 2011 Davis et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was funded by the National Health and Medical Research Council of Australia (NHMRC Program Grants 290208, 496600; Fellowships to NMAand TWY, scholarship to JSD). The funding source played no role in the design or conduct of the study, nor in the drafting of the manuscript or the decision tosubmit it for publication.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: [email protected]
. These authors contributed equally to this work.
Introduction
Asymmetric dimethylarginine (ADMA), an endogenous non-
specific nitric oxide synthase (NOS) inhibitor, is associated with
chronic endothelial dysfunction [1] and increased cardiovascular
risk [2], but its role in the setting of acute infections has been less
well characterised.
Severe sepsis (acute infection resulting in organ dysfunction) is
the leading cause of death in intensive care units in the USA [3],
and is increasing in incidence globally [4]. Microvascular and
endothelial dysfunction are key contributors to organ failure and
death in sepsis but the mechanisms linking sepsis with vascular
dysfunction remain incompletely understood [5]. A relative
deficiency of constitutively expressed endothelial nitric oxide
(NO), essential to maintain a quiescent and functional endothe-
lium, may underlie sepsis-associated endothelial and microvascular
dysfunction [6,7]. NO is produced by NOS from its primary
substrate, L-arginine. ADMA competitively inhibits the produc-
tion of NO by NOS and additionally, along with symmetrical
dimethylarginine (SDMA) and L-lysine, competes with L-arginine
for transport across the cell membrane [8]. Hence the L-arginine:
ADMA ratio is considered a better indicator of the availability of
L-arginine to NOS than is plasma L-arginine concentration alone
[9].
Infusion of ADMA in both rats [10] and humans [11] acutely
decreases NO production, resulting in endothelial dysfunction.
Plasma ADMA concentrations are increased in patients with
chronic renal disease [12], hypertension [13], diabetes mellitus
[14] and peripheral vascular disease [15]. Furthermore, ADMA
has been shown to be an independent predictor of cardiovascular
events in patients with existing coronary artery disease [16] and
end-stage renal disease [17].
In contrast, few studies have examined the role of ADMA in
humans with sepsis, and none have reported L-arginine: ADMA
ratios or examined microvascular reactivity in this context. The
few clinical studies that have reported plasma ADMA concentra-
tions during acute infection have had conflicting results
[18,19,20,21]. Using peripheral arterial tonometry, we have
previously shown that digital microvascular reactivity, a measure
of endothelial NO bioavailability [22], is decreased in patients with
PLoS ONE | www.plosone.org 1 February 2011 | Volume 6 | Issue 2 | e17260
sepsis [7]. However, we did not find a correlation between
concentrations of plasma L-arginine and microvascular reactivity.
We also found that despite an increase in plasma L-arginine
concentrations over time, there was no corresponding improve-
ment in microvascular reactivity. A potential explanation for these
findings in sepsis is competitive inhibition of NOS by ADMA.
We hypothesised that plasma L-arginine: ADMA ratio would be
decreased in sepsis, in proportion to disease severity, and would
correlate with reactive hyperaemia peripheral arterial tonometry
(RH-PAT) index, an in vivo measure of endothelial NO
bioavailability. Furthermore, we hypothesised that increased
plasma ADMA would be associated with mortality.
Methods
Study design and settingWe performed a prospective observational study at a 350-bed
Australian teaching hospital, with an 18-bed mixed intensive care
unit (ICU). Approval was obtained from the Human Research
Ethics Committee of the Menzies School of Health Research and
the Department of Health and Community Services. Written
informed consent was obtained from all participants or next of kin
where necessary.
ParticipantsThe study subjects were adults ($18 years) hospitalised with
sepsis, who were enrolled in a previously-reported study of
microvascular reactivity; more detail of subject recruitment and
study procedures are provided in this paper [7]. Sepsis was defined
as a proven or suspected infection plus at least 2 criteria for the
systemic inflammatory response syndrome (SIRS) present within
the last 4 hours [23]; these include tachycardia (heart rate.90
beats per minute), tachypnoea (respiratory rate .20 breaths/
minute), abnormal temperature (body temperature .38uC or
,36uC), and abnormal white blood cell count (,4,000 cells/ml or
.12,000 cells/ml or .10% band forms). Septic patients were
eligible for enrolment within 24 hours of their admission to the
ICU, or within 36 hours of admission to the ward. Control
subjects were adults recruited from hospitalised patients with no
clinical or laboratory evidence of inflammation or infection, and
who had not met SIRS criteria within the last 30 days. Septic
patients were classified as septic shock, or sepsis without shock.
Septic shock was defined at the time of enrolment as systolic blood
pressure ,90 mmHg or a reduction of $40 mmHg from baseline
despite adequate fluid resuscitation, or the need for vasopressors to
maintain these targets [23]. Disease severity was assessed by the
Acute Physiology and Chronic Health Evaluation (APACHE) II
score and organ failure was determined using the Sequential
Organ Failure Assessment (SOFA) score [24].
Laboratory assaysBlood from arterial lines if present, or venepuncture if not, was
collected in lithium heparin tubes at baseline and 2–4 days later,
and plasma was separated and stored at 270uC within 2 hours of
blood collection. Control patients had blood collected at baseline
only.
ADMA and SDMA were measured by reverse phase HPLC
with simultaneous fluorescence and UV-visible detection, as
previously described [25]. The method precision, represented by
percent relative standard deviation was 2.0% for ADMA and
2.3% for SDMA. Method accuracy measured by percent spike
recovery was 98% for ADMA and 99% for SDMA. Arginine was
measured using a method modified from van Wandelen and
Cohen [26]. Angiopoietin-2 (Ang-2) and intracellular adhesion
molecule-1 (ICAM-1) were measured by ELISA (R&D systems).
IL-6 and TNFa were measured by flow cytometry using a cytokine
bead array (BD Biosciences, CA, USA).
Measurement of microvascular reactivityMicrovascular reactivity was measured at the bedside by RH-
PAT (Itamar Medical, Caesarea, Israel), a non-invasive method of
assessing endothelial function [27–28] which is at least 50%
dependent on endothelial NO production [22]. Peripheral arterial
tonometry (PAT) was measured in a fingertip before and after a 5-
minute ischemic stress at the forearm, generating an RH-PAT
index, normalized to the control arm, as previously reported [7].
Statistical methodsContinuous variables were compared using Mann Whitney U test,
and categorical variables using Fisher’s exact test. Correlates with
Table 1. Baseline characteristics.
Septic Shock Sepsis without shock Controls p valuea
n 20 47 31
Ageb 51.5(12.0) 52.5 (14.4) 45.4 (12.7) NS
Malec 11 (55) 30 (63) 24 (75) NS
Diabeticc 6 (30) 13 (27) 10 (31) NS
Smokerc 8 (40) 22 (46) 14 (44) NS
IHDc 4 (20) 8 (17) 4 (13) NS
Hypertensionc 5 (25) 17 (35) 9 (28) NS
Hyperlipidemiac 4 (20) 11 (22) 11 (34) NS
Chronic renal diseasec 4 (20) 4 (8) 3 (10) NS
APACHE II scored 20.0 (16–23) 10.0 (6–16) ,0.0001
SOFA scored 6 (3–9) 2.0 (0.5–4.0) ,0.0001
a – by Chi2 test for difference between all 3 groups.b – Mean (sd).c – n (%).d – Median (Interquartile range).doi:10.1371/journal.pone.0017260.t001
ADMA in Sepsis
PLoS ONE | www.plosone.org 2 February 2011 | Volume 6 | Issue 2 | e17260
baseline ADMA and arginine:ADMA ratio were determined using
Spearman’s coefficient for univariate analysis. Day 2 values were
compared with baseline values using paired Wilcoxon signed-rank test.
In an a priori analytical plan, the relationship between baseline ADMA
and mortality among sepsis patients was examined using logistic
regression, with ADMA divided into quartiles as previously described
[29]. To examine longitudinal correlations, linear mixed-effects models
were used. A 2-sided p-value of ,0.05 was considered significant. All
analyses were performed using Intercooled Stata 10 (Statacorp, Texas).
Results
There were 20 subjects with septic shock, 47 with sepsis without
shock and 31 controls. The three groups were well-matched in
terms of age, sex and known associations with chronically raised
ADMA (Table 1).
Arginine:ADMA ratio and disease severityBaseline plasma L-arginine: ADMA ratio was significantly lower
in sepsis patients (median [IQR] 63 [45–103]) than in hospital
controls (143 [123–166], p,0.0001) (Table 2). Furthermore, septic
shock patients had significantly lower L-arginine: ADMA ratio
(median [IQR] 43 [34–73]) than sepsis patients without shock (91
[56–108], p,0.0001) (Figure 1a). The plasma L-arginine: ADMA
ratio inversely correlated with severity of illness as measured by
APACHE II score (r = 20.4, R2 = 0.16, p = 0.003) and organ failure
as measured by SOFA score (r = 20.5, R2 = 0.25, p = 0.0001).
ADMA, disease severity and mortalityThe median [IQR] plasma concentration of ADMA was
significantly higher in septic shock patients (0.64 [0.54–0.85] mM)
than sepsis patients without shock (0.47 [0.38–0.57] mM) (p = 0.008)
(Table 2) and correlated with SOFA score (r = 0.45, R2 = 0.20,
p,0.001). Six of 67 sepsis patients (9%) had died by day 28 of follow-
up, 5 of whom were in the septic shock subgroup. Median [IQR]
baseline ADMA was approximately twice as high in those who died
(1.07 [0.75–1.31]) as in survivors (0.51 [0.39–0.61]), p = 0.001. Sepsis
patients with a baseline plasma ADMA concentration in the highest
quartile ($0.66 mmol/L) had an odds ratio for death of 20.8 (95% CI
2.2–195.0, p = 0.008). In a multivariate model incorporating SOFA
score, age, gender, creatinine and IL-6 concentration, baseline
ADMA was the only significant predictor of death (p = 0.04).
SDMA, renal function and disease severitySDMA was highest in septic shock, intermediate in sepsis
without shock and lowest in controls (Table 2). Predominantly
renally excreted [30], SDMA correlated strongly with serum
creatinine (r = 0.70, R2 = 0.49, p,0.001), whereas ADMA did not
(r = 0.16, R2 = 0.03, p = NS). On univariate analysis, sepsis
patients with a plasma SDMA concentration in the highest
quartile ($1.30 mmol/L) had an odds ratio for death of 8.12 (95%
CI 1.33–50.0), however this became insignificant on controlling
for renal function.
Arginine, ADMA and microvascular reactivityThere was a modest but significant correlation between baseline
L-arginine: ADMA ratio and NO-dependent microvascular
reactivity as measured by RH-PAT (figures 1a and 1b) both
on univariate analysis (r = 0.34, R2 = 0.12, p = 0.02), and in a
multivariate linear regression model adjusting for serum creatinine
(Wald p-value for L-arginine: ADMA ratio = 0.03). The L-
arginine: ADMA ratio was significantly lower in sepsis patients
who required vasopressors (median [IQR] = 42 [32–55])
compared to those who did not (74 [54–108], p = 0.002). Baseline
plasma ADMA concentration correlated with markers of endo-
thelial activation including Ang-2 (r = 0.45, R2 = 0.20, p = 0.0002)
and ICAM-1 (r = 0.47, R2 = 0.22, p = 0.0001). This relationship
persisted after controlling for disease severity (using APACHE II
score) in a multivariate analysis.
Over the first 2–4 days of follow up, plasma ADMA increased in
the sepsis patients (0.53 to 0.64, p = 0.002) (Table 3), and also in
the septic shock subgroup (0.64 to 0.85, p = 0.03). Plasma L-
arginine concentrations also increased, but due to the increase in
ADMA, there was no significant change in the L-arginine: ADMA
ratio. In a mixed effects linear regression model examining change
from baseline to day 2–4, increase in ADMA over time
Table 2. Plasma asymmetric dimethylarginine and related variables at time of initial measurement.
All sepsis Septic shock Sepsis without shock Control
p valuepooledsepsis vcontrol
p valueseptic shockvs control
n 67 20 47 31
Plasma ADMA(mmol/L)a 0.52 (0.39–0.65) 0.64 (0.54–0.85) 0.47 (0.38–0.57) 0.57 (0.50–0.62) 0.10 0.09
Plasma L-arginine (mmol/L)a,b 35.5 (27.3–51.2) 31.0 (23.7–40.4) 38.1 (29.4–51.7) 81.8 (68.9–91.3) ,0.001 ,0.001
Plasma L-arginine/ADMA ratioa,b 63.2 (45.3–103.4) 43.4(33.6–73.3) 91.4 (55.5–108.3) 142.9 (123.0–165.7) ,0.001 ,0.001
Plasma SDMA(mmol/L)a 0.66 (0.50–1.29) 1.05 (0.77–1.45) 0.56 (0.45–0.80) 0.47 (0.43–0.65) 0.002 ,0.001
Plasma lysine(mmol/L)a 128 (100–171) 129 (90–190) 128 (104–162) 184 (157–216) ,0.001 0.006
Receiving mechanical ventilationc 14 (21) 9 (47) 5 (26) - - -
RH-PAT indexd 1.70 (0.47) 1.47 (0.40) 1.78 (0.47) 2.05 (0.46) 0.001 ,0.001
Plasma Interleukin 6 (pg/ml)a 223 (76.6–563) 885 (298–2412) 148 (46.0–322) 4.7 (2.2–9.5) ,0.001 ,0.001
White blood cell counta 15.2 (10.1–20.2) 17.5 (11.0–27.8) 15.2 (9.1–17.8) 7.7 (5.7–9.0) ,0.001 ,0.001
C-reactive proteina 180 (87.3–259) 202 (126–297) 143(84–259) 7 (4–22) ,0.001 ,0.001
a. median (Interquartile range);b. n = septic shock 19, sepsis without shock 37, controls 27;c. n (%).d. mean (sd).doi:10.1371/journal.pone.0017260.t002
ADMA in Sepsis
PLoS ONE | www.plosone.org 3 February 2011 | Volume 6 | Issue 2 | e17260
significantly correlated with increase in SOFA score (p,0.001)
and decrease in RH-PAT index (p = 0.03), but not with change in
IL-6 or CRP. It also correlated with increase in the liver (p,0.001)
but not the renal (p = 0.09) components of the SOFA score.
Discussion
The plasma L-arginine: ADMA ratio is significantly reduced in
sepsis, in proportion to disease severity. Plasma ADMA concen-
tration correlates with the degree of organ failure and predicts
mortality in patients with sepsis. Increase in ADMA over time is
associated with worsening microvascular reactivity and organ
dysfunction. Our results suggest a possible mechanism underlying
these associations: impairment of microvascular function due to
inhibition of endothelial NO production by ADMA.
Decreased L-arginine: ADMA ratio may contribute to organ
failure in sepsis by reducing microvascular reactivity. Impaired
microvascular and endothelial function have been shown to be
important contributors to organ dysfunction and death in animals
and humans with sepsis [31]. ADMA causes both acute [11] and
chronic [2] endothelial dysfunction by inhibiting NOS and
decreasing endothelial NO bioavailability. The L-arginine:
ADMA ratio is a marker of the availability of L-arginine to
NOS [9]. In severe malaria, plasma ADMA is increased and is
associated with endothelial dysfunction and reduced exhaled nitric
oxide [32]. In this study we found that baseline L-arginine: ADMA
ratio, but not arginine or ADMA alone, correlated with
endothelial nitric oxide dependent microvascular reactivity.
Furthermore, plasma ADMA concentrations correlated with
increased plasma concentrations of Ang-2 and ICAM-1, both of
which are associated with reduced endothelial nitric oxide
bioavailablity [7,33,34]. Together, these findings suggest that a
decreased L-arginine: ADMA ratio reduces endothelial nitric
oxide bioavailability and thus impairs microvascular reactivity in
sepsis. This may provide a mechanistic explanation for the
observed association of plasma ADMA concentrations with organ
failure and death in this and other studies [20,29]. However,
although significant, this association was not strong and further
work is needed to confirm this preliminary observation.
A recent small study in human volunteers injected with
lipopolysaccharide also found an acute increase in plasma ADMA
and decrease in NO-dependant vasodilatation, but did not find a
correlation between NO-dependant vasodilatation and L-argini-
ne:ADMA ratio [35]. These differing findings may be due to the
fact that in Engelberger’s study, the sample size (n = 7) was too
small to detect such a correlation. In addition, sepsis is a highly
complex pathophysiological state, and the findings from sepsis
models may be difficult to apply to clinical sepsis.
The increase in plasma ADMA concentrations over time
observed in this study agrees with the findings of another recent
observational study in septic humans [21]. This increase may in
part explain the lack of significant improvement in microvascular
reactivity as patients recover [7], despite an increase in plasma L-
arginine. This may be because the L-arginine: ADMA ratio (and
Figure 1. Ratio of L-arginine to asymmetric dimethylarginine inbaseline plasma samples, according to disease category,compared with baseline microvascular reactivity according todisease category. Panel A shows plasma arginine: ADMA ratio andpanel B shows reactive hyperaemia peripheral arterial tonometry index.P values represent comparisons between groups. Solid circles representindividual sepsis subjects and solid triangles represent individualcontrol subjects. Horizontal lines represent median group values, anderror bars represent interquartile range. In panel B, solid circlesrepresent mean group values for sepsis subjects, and the solid trianglefor control subjects. Error bars represent standard error of the mean.doi:10.1371/journal.pone.0017260.g001
Table 3. Longitudinal results in subjects with sepsis.
Day 0 Day 2 P Day 0 to 2
n 67 47
ADMA 0.53 (0.39–0.66) 0.64 (0.51–0.78) 0.002
L-arginine 35.5 (27.3–51.2) 47.2 (30.8–58.1) 0.03
L-arginine: ADMAratio
63.2 (45.3–103.4) 63.0 (41.7–108.0) NS
RH-PAT index 1.70 (1.57–1.82) 1.81 (1.65–1.96) NS
SDMA 0.66 (0.50–1.30) 0.71 (0.47–1.36) NS
IL-6 223 (78.2–530) 54.5 (16.1–201) ,0.001
SOFA score 3 (1–7) 2 (1–7) 0.04
Note: ADMA = Asymmetric dimethylarginine. RH-PAT index = Reactivehyperaemia peripheral arterial tonometry index. SDMA = Symmetricdimethylarginine. IL-6 = Interleukin 6. SOFA score = Sequential Organ FailureAssessment Score.doi:10.1371/journal.pone.0017260.t003
ADMA in Sepsis
PLoS ONE | www.plosone.org 4 February 2011 | Volume 6 | Issue 2 | e17260
thus the availability of L-arginine to NOS within endothelial cells)
does not change over time. The mechanism behind the change in
ADMA over time cannot be determined from these data, however
there are several possibilities. Protein catabolism in patients with
sepsis could lead to progressive release of methylated L-arginine
residues into the plasma. However, this is unlikely to be the case
because endogenous leucine flux (a measure of protein catabolism)
does not correlate with plasma ADMA concentrations in septic
humans [36]. NO causes direct inhibition of dimethylarginine
dimethylaminohydrolase (DDAH) activity by S-nitrosylation of an
active cysteine residue [37]. Thus it is possible that as patients
recover from sepsis and endothelial NO bioavailability increases,
DDAH activity is inhibited, resulting in an increase in plasma
ADMA concentrations. Finally, the longitudinal inverse associa-
tion between liver function and plasma ADMA suggests that
worsening liver function due to sepsis progression, and thus
decreased metabolism of ADMA, may also explain these findings.
The disparity between ADMA concentrations in shock and
without shock may be due to different mechanisms within these
two states. Early sepsis is a hyperdynamic state, with increased
cardiac output and liver and kidney blood flow [38,39]. This may
lead to increased degradation of ADMA in the liver by DDAH
and, to a lesser extent, increased renal excretion. This hypothesis is
supported by a study which found that the liver fractional
extraction rate for ADMA is significantly higher and circulating
ADMA is significantly lower in endotoxemic rats compared to
controls [40]. Patients with septic shock have generally developed
multiple organ failure and down-regulation of cellular functions
[41] and thus hepatic metabolism and renal excretion of ADMA
may drop back to baseline concentrations. This hypothesis is
supported by our finding that ADMA concentrations inversely
correlate with liver function, both at baseline and longitudinally.
Similar findings have recently been reported in patients with
malaria, in whom plasma ADMA concentrations were raised in
those with severe malaria but low normal in those with moderately
severe malaria [32].
Our study helps to clarify the inconsistencies reported in
previous clinical studies measuring ADMA in acute infections. It
demonstrates that while patients with septic shock have increased
ADMA, patients without shock have decreased ADMA resulting
in no significant difference between the ADMA concentrations in
pooled sepsis and hospital controls – a potentially misleading
finding unless patients are stratified by sepsis severity. The
previous studies that found that ADMA was increased in sepsis
[18,20,21] primarily enrolled patients with septic shock. The only
other published study to enrol sepsis patients without shock also
found no overall difference in plasma ADMA concentrations
between sepsis and control patients [19]; however, they did not
consider patients with and without shock separately.
This study has several limitations. Although it is at least 50%
dependant on endothelial NO [22], peripheral arterial tonometry
is not a direct measure of NO activity. Other factors are likely to
contribute to endothelial NO bioavailability in addition to the L-
arginine:ADMA ratio, including CAT transport inhibitors (such as
SDMA) and oxidative stress resulting in NO-quenching. The 67
sepsis patients were not all followed up on day 2-4, largely because
of hospital discharge; thus the longitudinal results may underes-
timate the degree of improvement in microvascular and organ
function.
Raised plasma ADMA concentrations are a strong predictor of
death in patients with sepsis and thus may be useful as a prognostic
marker. Impaired endothelial and microvascular function due to
decreased endothelial NO production may be a mechanism
linking ADMA with organ dysfunction and mortality. The
DDAH-ADMA axis is a potential therapeutic target and may be
important in individual tailoring of therapy. Agents which
compete with ADMA for NOS (such as L-arginine) or which
potentiate DDAH activity should be further investigated in sepsis.
Acknowledgments
We would like to thank Kim Piera and Barbara MacHunter for technical
assistance; Jane Thomas, Mark McMillan, Karl Blenk, Antony Van Asche,
Steven Tong and Paulene Kittler for RH-PAT measurements; Ric Price
and Joseph McDonnell for statistical advice; and the medical and nursing
staff of the Royal Darwin Hospital Intensive Care and Hospital in the
Home units.
Author Contributions
Conceived and designed the experiments: JSD CJD TWY DSC NMA.
Performed the experiments: JSD CJD CJ YRM. Analyzed the data: JSD
CJD. Contributed reagents/materials/analysis tools: DPS NMA. Wrote
the manuscript: JSD CJD TWY DPS DSC NMA.
References
1. Juonala M, Viikari JS, Alfthan G, Marniemi J, Kahonen M, et al. (2007)
Brachial artery flow-mediated dilation and asymmetrical dimethylarginine in the
cardiovascular risk in young Finns study. Circulation 116: 1367–1373.
2. De Gennaro Colonna V, Bianchi M, Pascale V, Ferrario P, Morelli F, et al.
(2009) Asymmetric dimethylarginine (ADMA): an endogenous inhibitor of nitric
oxide synthase and a novel cardiovascular risk molecule. Med Sci Monit 15:
RA91–101.
3. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, et al. (2001)
Epidemiology of severe sepsis in the United States: analysis of incidence,
outcome, and associated costs of care. Crit Care Med 29: 1303–1310.
4. Martin GS, Mannino DM, Eaton S, Moss M (2003) The epidemiology of sepsis
in the United States from 1979 through 2000. N Engl J Med 348: 1546–1554.
5. Aird WC (2003) The role of the endothelium in severe sepsis and multiple organ
dysfunction syndrome. Blood 101: 3765–3777.
6. Trzeciak S, Cinel I, Phillip Dellinger R, Shapiro NI, Arnold RC, et al. (2008)
Resuscitating the microcirculation in sepsis: the central role of nitric oxide,
emerging concepts for novel therapies, and challenges for clinical trials. Acad
Emerg Med 15: 399–413.
7. Davis JS, Yeo TW, Thomas JH, McMillan M, Darcy CJ, et al. (2009) Sepsis-
associated microvascular dysfunction measured by peripheral arterial tonometry:
an observational study. Crit Care 13: R155.
8. Closs EI, Basha FZ, Habermeier A, Forstermann U (1997) Interference of L-
arginine analogues with L-arginine transport mediated by the y+ carrier hCAT-
2B. Nitric Oxide 1: 65–73.
9. Bode-Boger SM, Scalera F, Ignarro LJ (2007) The L-arginine paradox:
Importance of the L-arginine/asymmetrical dimethylarginine ratio. Pharmacol
Ther 114: 295–306.
10. De Gennaro Colonna V, Bonomo S, Ferrario P, Bianchi M, Berti M, et al.
(2007) Asymmetric dimethylarginine (ADMA) induces vascular endothelium
impairment and aggravates post-ischemic ventricular dysfunction in rats.
Eur J Pharmacol 557: 178–185.
11. Vallance P, Leone A, Calver A, Collier J, Moncada S (1992) Accumulation of an
endogenous inhibitor of nitric oxide synthesis in chronic renal failure. Lancet
339: 572–575.
12. Kielstein JT, Boger RH, Bode-Boger SM, Frolich JC, Haller H, et al.
(2002) Marked increase of asymmetric dimethylarginine in patients
with incipient primary chronic renal disease. J Am Soc Nephrol 13:
170–176.
13. Surdacki A, Nowicki M, Sandmann J, Tsikas D, Boeger RH, et al. (1999)
Reduced urinary excretion of nitric oxide metabolites and increased plasma
levels of asymmetric dimethylarginine in men with essential hypertension.
J Cardiovasc Pharmacol 33: 652–658.
14. Abbasi F, Asagmi T, Cooke JP, Lamendola C, McLaughlin T, et al. (2001)
Plasma concentrations of asymmetric dimethylarginine are increased in patients
with type 2 diabetes mellitus. Am J Cardiol 88: 1201–1203.
15. Boger RH, Bode-Boger SM, Thiele W, Junker W, Alexander K, et al. (1997)
Biochemical evidence for impaired nitric oxide synthesis in patients with
peripheral arterial occlusive disease. Circulation 95: 2068–2074.
ADMA in Sepsis
PLoS ONE | www.plosone.org 5 February 2011 | Volume 6 | Issue 2 | e17260
16. Valkonen VP, Paiva H, Salonen JT, Lakka TA, Lehtimaki T, et al. (2001) Risk
of acute coronary events and serum concentration of asymmetrical dimethy-larginine. Lancet 358: 2127–2128.
17. Zoccali C, Bode-Boger S, Mallamaci F, Benedetto F, Tripepi G, et al. (2001)
Plasma concentration of asymmetrical dimethylarginine and mortality inpatients with end-stage renal disease: a prospective study. Lancet 358:
2113–2117.18. O’Dwyer MJ, Dempsey F, Crowley V, Kelleher DP, McManus R, et al. (2006)
Septic shock is correlated with asymmetrical dimethyl arginine levels, which may
be influenced by a polymorphism in the dimethylarginine dimethylaminohy-drolase II gene: a prospective observational study. Crit Care 10: R139.
19. Zoccali C, Maas R, Cutrupi S, Pizzini P, Finocchiaro P, et al. (2007) Asymmetricdimethyl-arginine (ADMA) response to inflammation in acute infections.
Nephrol Dial Transplant 22: 801–806.20. Nakamura T, Sato E, Fujiwara N, Kawagoe Y, Suzuki T, et al. (2009)
Circulating levels of advanced glycation end products (AGE) and interleukin-6
(IL-6) are independent determinants of serum asymmetric dimethylarginine(ADMA) levels in patients with septic shock. Pharmacol Res.
21. Iapichino G, Umbrello M, Albicini M, Spanu P, Bellani G, et al. (2010) Timecourse of endogenous nitric oxide inhibitors in severe sepsis in humans. Minerva
Anestesiol 76: 325–333.
22. Nohria A, Gerhard-Herman M, Creager MA, Hurley S, Mitra D, et al. (2006)Role of nitric oxide in the regulation of digital pulse volume amplitude in
humans. J Appl Physiol 101: 545–548.23. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, et al. (1992) Definitions
for sepsis and organ failure and guidelines for the use of innovative therapies insepsis. The ACCP/SCCM Consensus Conference Committee. American
College of Chest Physicians/Society of Critical Care Medicine. Chest 101:
1644–1655.24. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, et al. (1996) The
SOFA (Sepsis-related Organ Failure Assessment) score to describe organdysfunction/failure. On behalf of the Working Group on Sepsis-Related
Problems of the European Society of Intensive Care Medicine. Intensive Care
Med 22: 707–710.25. Jones CE, Darcy CJ, Woodberry T, Anstey NM, McNeil YR (2010) HPLC
analysis of asymmetric dimethylarginine, symmetric dimethylarginine, homo-arginine and arginine in small plasma volumes using a Gemini-NX column at
high pH. J Chromatogr B Analyt Technol Biomed Life Sci 878: 8–12.26. van Wandelen C, Cohen SA (1997) Using quaternary high-performance liquid
chromatography eluent systems for separating 6-aminoquinolyl-N-hydroxysuc-
cinimidyl carbamate-derivatized amino acid mixtures. J Chromatogr A 763:11–22.
27. Kuvin JT, Mammen A, Mooney P, Alsheikh-Ali AA, Karas RH (2007)Assessment of peripheral vascular endothelial function in the ambulatory setting.
Vasc Med 12: 13–16.
28. Hamburg NM, Benjamin EJ (2009) Assessment of endothelial function using
digital pulse amplitude tonometry. Trends Cardiovasc Med 19: 6–11.
29. Nijveldt RJ, Teerlink T, Van Der Hoven B, Siroen MP, Kuik DJ, et al. (2003)
Asymmetrical dimethylarginine (ADMA) in critically ill patients: high plasma
ADMA concentration is an independent risk factor of ICU mortality. Clin Nutr
22: 23–30.
30. Kielstein JT, Salpeter SR, Bode-Boeger SM, Cooke JP, Fliser D (2006)
Symmetric dimethylarginine (SDMA) as endogenous marker of renal function–a
meta-analysis. Nephrol Dial Transplant 21: 2446–2451.
31. Vallet B (2003) Bench-to-bedside review: endothelial cell dysfunction in severe
sepsis: a role in organ dysfunction? Crit Care 7: 130–138.
32. Yeo TW, Lampah DA, Tjitra E, Gitawati R, Darcy CJ, et al. (2010) Increased
asymmetric dimethylarginine in severe falciparum malaria: association with
impaired nitric oxide bioavailability and fatal outcome. PLoS Pathog 6:
e1000868.
33. Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, et al. (2008)
Angiopoietin-2 is associated with decreased endothelial nitric oxide and poor
clinical outcome in severe falciparum malaria. Proc Natl Acad Sci U S A 105:
17097–17102.
34. Davis JS, Yeo TW, Piera KA, Woodberry T, Celermajer DS, et al. (2010)
Angiopoietin-2 is increased in sepsis and inversely associated with nitric oxide-
dependent microvascular reactivity. Crit Care 14: R89.
35. Engelberger RP, Pittet YK, Henry H, Delodder F, Hayoz D, et al. (2010) Acute
Endotoxemia Inhibits Microvascular Nitric Oxide-Dependent Vasodilation in
Humans. Shock.
36. Kao CC, Bandi V, Guntupalli KK, Wu M, Castillo L, et al. (2008) Arginine,
citrulline, and nitric oxide metabolism in sepsis. Clin Sci (Lond).
37. Leiper J, Murray-Rust J, McDonald N, Vallance P (2002) S-nitrosylation of
dimethylarginine dimethylaminohydrolase regulates enzyme activity: further
interactions between nitric oxide synthase and dimethylarginine dimethylami-
nohydrolase. Proc Natl Acad Sci U S A 99: 13527–13532.
38. Di Giantomasso D, May CN, Bellomo R (2003) Vital organ blood flow during
hyperdynamic sepsis. Chest 124: 1053–1059.
39. Lang CH, Bagby GJ, Ferguson JL, Spitzer JJ (1984) Cardiac output and
redistribution of organ blood flow in hypermetabolic sepsis. Am J Physiol 246:
R331–337.
40. Nijveldt RJ, Teerlink T, van Guldener C, Prins HA, van Lambalgen AA, et al.
(2003) Handling of asymmetrical dimethylarginine and symmetrical dimethy-
larginine by the rat kidney under basal conditions and during endotoxaemia.
Nephrol Dial Transplant 18: 2542–2550.
41. Singer M (2008) Cellular dysfunction in sepsis. Clin Chest Med 29: 655–660,
viii–ix.
ADMA in Sepsis
PLoS ONE | www.plosone.org 6 February 2011 | Volume 6 | Issue 2 | e17260
An Observational Cohort Study of the Kynurenine toTryptophan Ratio in Sepsis: Association with ImpairedImmune and Microvascular FunctionChristabelle J. Darcy1., Joshua S. Davis1,2., Tonia Woodberry1, Yvette R. McNeil1, Dianne P. Stephens3,
Tsin W. Yeo1,2, Nicholas M. Anstey1,2*
1 Global Health Division, Menzies School of Health Research and Charles Darwin University, Darwin, Northern Territory, Australia, 2 Division of Medicine, Royal Darwin
Hospital, Darwin, Northern Territory, Australia, 3 Intensive Care Unit, Royal Darwin Hospital, Darwin, Northern Territory, Australia
Abstract
Both endothelial and immune dysfunction contribute to the high mortality rate in human sepsis, but the underlyingmechanisms are unclear. In response to infection, interferon-c activates indoleamine 2,3-dioxygenase (IDO) whichmetabolizes the essential amino acid tryptophan to the toxic metabolite kynurenine. IDO can be expressed in endothelialcells, hepatocytes and mononuclear leukocytes, all of which contribute to sepsis pathophysiology. Increased IDO activity(measured by the kynurenine to tryptophan [KT] ratio in plasma) causes T-cell apoptosis, vasodilation and nitric oxidesynthase inhibition. We hypothesized that IDO activity in sepsis would be related to plasma interferon-c, interleukin-10, Tcell lymphopenia and impairment of microvascular reactivity, a measure of endothelial nitric oxide bioavailability. In anobservational cohort study of 80 sepsis patients (50 severe and 30 non-severe) and 40 hospital controls, we determined therelationship between IDO activity (plasma KT ratio) and selected plasma cytokines, sepsis severity, nitric oxide-dependentmicrovascular reactivity and lymphocyte subsets in sepsis. Plasma amino acids were measured by high performance liquidchromatography and microvascular reactivity by peripheral arterial tonometry. The plasma KT ratio was increased in sepsis(median 141 [IQR 64–235]) compared to controls (36 [28–52]); p,0.0001), and correlated with plasma interferon-c andinterleukin-10, and inversely with total lymphocyte count, CD8+ and CD4+ T-lymphocytes, systolic blood pressure andmicrovascular reactivity. In response to treatment of severe sepsis, the median KT ratio decreased from 162 [IQR 100–286]on day 0 to 89 [65–139] by day 7; p = 0.0006) and this decrease in KT ratio correlated with a decrease in the SequentialOrgan Failure Assessment score (p,0.0001). IDO-mediated tryptophan catabolism is associated with dysregulated immuneresponses and impaired microvascular reactivity in sepsis and may link these two fundamental processes in sepsispathophysiology.
Citation: Darcy CJ, Davis JS, Woodberry T, McNeil YR, Stephens DP, et al. (2011) An Observational Cohort Study of the Kynurenine to Tryptophan Ratio in Sepsis:Association with Impaired Immune and Microvascular Function. PLoS ONE 6(6): e21185. doi:10.1371/journal.pone.0021185
Editor: Jane Deng, University of California Los Angeles, United States of America
Received November 30, 2010; Accepted May 23, 2011; Published June 22, 2011
Copyright: 2011 Darcy et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was funded by the National Health and Medical Research Council of Australia (Program Grants 290208, 496600; Practitioner Fellowship toNMA, Scholarship to JSD) and an Australian Postgraduate Award to CJD. The funders had no role in study design, data collection and analysis, decision to publish,or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: [email protected]
. These authors contributed equally to this work.
Introduction
Sepsis is a systemic inflammatory response to infection [1].
Despite advances in its management, severe sepsis still has a
mortality rate of 30–50% [2,3,4]. Both immune and endothelial
dysfunction are thought to contribute to the high mortality rate in
sepsis [5,6], however the underlying mechanisms are not
completely understood.
Tryptophan is an essential amino acid that is central to cellular
respiration [7] and neurotransmission [8], and is a key immune
mediator. During inflammation, tryptophan is metabolised by
indoleamine 2,3-dioxygenase (IDO) to the toxic metabolite
kynurenine. IDO activity is measured by the ratio of kynurenine
to tryptophan (the KT ratio). Endothelial cells, monocytes, renal
tubular epithelial cells and hepatocytes express IDO in response to
interferon-c [9,10,11,12,13] and IL10 stabilises IDO expression
[14].
IDO activity regulates a number of immune responses.
Increased IDO activity inhibits T cell function [15] and
proliferation [14,16,17] and contributes to T cell apoptosis [18].
Furthermore, elevated IDO activity inhibits nitric oxide synthase
and vice versa [19,20,21]. Recent isotope studies have shown that
systemic NO production is either reduced or unchanged in human
sepsis compared with healthy controls [22,23,24].
In addition to regulating the immune response, IDO activity may
also regulate endothelial function. Kynurenine, a metabolite of
IDO, has recently been described as an endogenous vasorelaxing
factor [9]. Increased IDO activity would therefore be expected to
directly decrease systemic vascular resistance. Additionally, as IDO
inhibits NOS, IDO may indirectly affect endothelial function by
impairing NO-dependent microvascular reactivity. NO is essential
for normal endothelial function and NO-dependent microvascular
reactivity has been previously shown to be impaired in patients with
sepsis, in proportion to disease severity [25,26]. Finally, plasma
PLoS ONE | www.plosone.org 1 June 2011 | Volume 6 | Issue 6 | e21185
kynurenine concentrations have been associated with markers of
endothelial dysfunction in patients with end-stage renal disease [27].
IDO activity correlates with disease severity in patients with
chronic inflammatory diseases such as human immunodeficiency
virus [28], systemic lupus erythematosus [29] and malignancy
[30], but little is known about IDO activity in acute inflammatory
states. A raised KT ratio has recently been reported in patients
with bacteremia [31].
We investigated the relationship between the KT ratio and
disease severity in sepsis. We hypothesised that the KT ratio would
be related to IFN-c and IL10 concentrations, and inversely related
to both T cell lymphopenia and microvascular reactivity, a
measure of endothelial NO bioavailability.
Methods
ParticipantsWe evaluated patients with sepsis and hospital controls who were
part of a previously reported study of endothelial function in sepsis
[25]. Sepsis patients had suspected or proven infection and the
presence of two or more criteria for the systemic inflammatory
response syndrome (SIRS) within the last 4 hours [1]. Severe sepsis
patients had organ dysfunction or shock at the time of enrolment
according to the American College of Chest Physicians/Society of
Critical Care Medicine criteria [1,32]. Sepsis severity was estimated
using the Acute Physiology and Chronic Health Evaluation
(APACHE) II score from the first 24 hours of admission and daily
modified Sequential Organ Failure Assessment (SOFA) score [33].
Patients were enrolled within 24 hours of ICU admission or within
36 hours of ward admission. Control subjects were recruited from
hospital patients who had not met SIRS criteria within the last 30
days and who had no clinical or laboratory evidence of inflammation
or infection. Written informed consent was obtained from all
participants or next of kin. All sepsis patients had undergone
resuscitation and were haemodynamically stable at the time of study
enrolment. The study was approved by the Human Research Ethics
Committee of Menzies School of Health Research and the
Department of Health and Community Services.
Blood collection and lymphocyte countsVenous blood was collected in lithium heparin tubes at
enrolment, day 2–4, and day 7 until discharge from the hospital
or death. Whole blood differential white cell counts were
measured by Coulter Counter. Lymphopenia was defined as an
absolute lymphocyte count less than 1.26103/mL [34]. Plasma was
separated and stored at 280uC.
Lymphocytes were analysed in more detail in a subset of
patients from whom samples were processed within 30 minutes of
collection, matched for age and gender. Peripheral blood
mononuclear cells were separated using Ficoll-PaqueTM Plus
(GE Healthcare Biosciences, Uppsala, Sweden) and cryopreserved
in fetal calf serum and dimethyl sulfoxide. Cells were thawed and
stained with appropriate antibodies and analysed on a FACSCa-
libur flow cytometer (Becton Dickinson Immunocytometry
Systems, MA, USA). Antibodies were sourced from Biolegend,
California, USA (CD3, CD16 and CD56) or BD Biosciences
Pharmingen, California, USA (CD4 and CD8). Results were
analysed using Flow Jo software (Tree Star, Oregon, USA). T cells
were defined as CD3+ lymphocytes and natural killer cells were
defined as CD32CD16+CD56+ lymphocytes.
Tryptophan and kynurenine measurementsPlasma tryptophan and kynurenine concentrations were
measured by High Pressure Liquid Chromatography (HPLC;
Shimadzu, Kyoto, Japan) with UV (250 nm) and fluorescence
(excitation 250 nm, emission 395 nm) detection, using a method
modified from van Wandelen and Cohen [35]. The kynurenine to
tryptophan (KT) ratio was calculated by dividing the kynurenine
concentration (mmol/L) by the tryptophan concentration (mmol/L)
and multiplying the quotient by 1000 [28,36,37].
Plasma cytokine measurementsConcentrations of plasma IFN-c, IL6 and IL10 were deter-
mined using a cytometric bead array (Human Th1/Th2 Cytokine
Kit II, BD Biosciences Pharmingen, CA, USA) and a FACSCa-
libur flow cytometer (Becton Dickinson Immunocytometry
Systems, MA, USA). Results were analysed using FCAP array
version 1.0.1 (Soft Flow Hungary for Becton Dickinson Biosci-
ences). The lower limits of detection (LLD) of the assay were
2.5 pg/mL for IFN-c and 10 pg/mL for IL6 and IL10. Values
below the LLD were assigned a value halfway between zero and
the LLD for statistical analysis. Cytokines were only measured if
plasma had been frozen within 2 hours of collection.
Measurement of endothelial functionSepsis patients underwent serial bedside reactive hyperemia
peripheral arterial tonometry (RH-PAT) measurements at enrol-
ment, day 2–4, and day 7 [25]. Control patients had the same
assessment at a single time point. RH-PAT (Itamar Medical,
Caesarea, Israel) is a non-invasive operator-independent method
of assessing endothelial function. Endothelial function is defined by
the ability of blood vessels to vasodilate in response to an ischemic
stress, which invasive studies have demonstrated to be dependent
on endothelial cell NO production [38]. RH-PAT is at least 50%
NO-dependent [39]. RH-PAT uses finger probes to measure
digital pulse wave amplitude detected by a pressure transducer
[40], and has been validated against the more operator-dependent
flow-mediated dilatation method [41] and with endothelial
function in other vascular beds [42].
Statistical methodsPredefined groups for analysis were severe sepsis, non-severe
sepsis (meaning sepsis without evidence of organ dysfunction or
shock at enrolment), and hospital controls. Continuous parametric
variables were compared using Student’s t-test or ANOVA while
continuous non-parametric variables were compared using Mann-
Whitney, Kruskal-Wallis or Wilcoxon tests as appropriate.
Correlations were examined using Pearson’s or Spearman’s tests
for parametric and non-parametric data respectively. As SOFA
score was highly right-skewed and no transformation gave a
normal distribution, Kendall’s tau coefficient for partial correla-
tion was used for multivariate analysis involving SOFA [43].
Linear mixed-effects models were used to examine longitudinal
correlations. A 2-sided p-value of ,0.05 was considered
significant. Analyses were performed using Stata version 10.0
(Stata Corp TX, USA) and Prism version 5.01 (GraphPad
Software, CA, USA).
Results
PatientsThe study included 50 patients with severe sepsis, 30 with non-
severe sepsis and 40 hospital controls. The three groups did not
differ significantly in age or gender (Table 1). Ninety percent of
severe sepsis patients and all non-severe sepsis patients were either
orally or enterally fed at the time of enrolment; none were
receiving parenteral nutrition.
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 2 June 2011 | Volume 6 | Issue 6 | e21185
IDO activity and sepsis severityPlasma tryptophan concentrations were significantly reduced in
patients with sepsis (p,0.0001, Figure 1 and Table 2). In all
sepsis patients, plasma tryptophan was inversely related to SOFA
score (r = 20.45, p,0.0001). There was no difference in the
baseline plasma tryptophan concentrations among severe sepsis
patients who were orally fed (n = 29), enterally fed (n = 16) or who
were nil by mouth (n = 5).
Conversely, plasma kynurenine concentrations were elevated in
sepsis patients compared to hospital controls (p,0.0001, Figure 1and Table 2). In all sepsis patients, plasma kynurenine correlated
with SOFA score (r = 0.34, p = 0.005). As kynurenine is renally
excreted and accumulates in renal failure [44,45], kynurenine
concentrations were tested for relationships with renal impair-
ment. Kynurenine concentrations were significantly higher in
patients requiring continuous renal replacement therapy (CRRT)
(median 4.5 mmol [IQR 4–5.3]) than in patients not receiving
CRRT (2.8 mmol [2.1–4.4]; p = 0.03). In all sepsis patients,
kynurenine concentration correlated with plasma creatinine
(r = 0.41, p = 0.0002). Nevertheless, the association between
plasma kynurenine concentration and SOFA score remained
significant even after controlling for creatinine (ktau = 0.24,
p,0.01).
IDO activity was significantly increased in sepsis patients
(median KT ratio 141 [IQR 64–235]) compared to controls (36
[28–52]) (p,0.0001) and in severe sepsis compared to non-severe
sepsis (p = 0.0006, Table 2). The baseline KT ratio correlated
with APACHE II (rs = 0.51, p,0.0001) and total SOFA scores
(rs = 0.54, p,0.0001) in sepsis patients. The KT ratio positively
correlated with the hepatic (rs = 0.28, p = 0.01), renal (rs = 0.53,
p,0.0001), cardiovascular (rs = 0.42, p,0.0001) and respiratory
(rs = 0.36, p = 0.0009) components of the SOFA score but not the
coagulation component (rs = 0.13, p = ns).
Of the 80 sepsis patients, 6 died by day 28 of the study. The
baseline KT ratio in patients who died (median 270 [IQR 102–
431] was not statistically significantly different to those who
survived (138 [63–232]; p = 0.2).
In longitudinal analysis of severe sepsis, the KT ratio
significantly decreased between day 0 (median 162 [IQR 100–
286]) and day 7 (89 [65–139]), p = 0.0006); Figure 1D. Among all
sepsis patients, decrease in KT ratio correlated with decrease in
SOFA score over time (p,0.0001).
IDO activity and plasma cytokinesPlasma IFN-c, IL6 and IL10 were all significantly increased in
patients with sepsis (Table 2). Plasma concentrations of IL1, IL2,
IL4 and tumour necrosis factor-a were not significantly increased
in this cohort and were not analysed further. Both IL6 and IL10
positively correlated with SOFA score (rs = 0.55, p,0.0001 and
rs = 0.55, p,0.0001 respectively) but there was no association
between IFN-c and SOFA score.
In sepsis patients, the KT ratio correlated with plasma IFN-c(rs = 0.44, p = 0.0002), IL6 (rs = 0.49, p,0.0001) and IL10
(rs = 0.62, p,0.0001). The associations between KT ratio and IL6
and IL10 remained significant after controlling for SOFA score
(ktau = 0.30, p,0.003 and ktau = 0.45, p,0.0001 respectively).
Table 1. Baseline clinical characteristics of participants.
Severe sepsis Non-severe sepsis Controls p value*
Subjects (n) 50 30 40
Age 52 (48–57) 50 (46–55) 48 (44–52) NS
Male – n (%) 29 (58%) 20 (67%) 27 (68%) NS
Diabetic – n (%) 16 (32%) 7 (23%) 13 (33%) NS
Mean Arterial Pressure 74 (70–82)n = 50
88 (77–104)n = 30
80 (73–93)n = 37
0.001
Systolic Blood Pressure 113 (105–132)n = 49
123 (110–140)n = 24
115 (110–128)n = 37
NS
Diastolic Blood Pressure 60 (54–68)n = 49
70 (60–90)n = 24
60 (60–75)n = 37
0.002
APACHE II 19 (15–23) 7 (5–12) ,0.0001
SOFA score (day 0) 6 (3–9) 1 (0–2) ,0.0001
RH-PAT index 1.59 (1.45–1.73)n = 45
1.86 (1.67–2.05)n = 26
2.04 (1.91–2.18)n = 36
,0.0001
Causative Organism – n (%) NS
None Cultured 23 (46%) 20 (67%)
Gram Positive Bacterium 14 (28%) 4 (13%)
Gram Negative Bacterium 13 (26%) 6 (20%)
Nutrition – n (%)
Oral feeding 29 (58%) 29 (97%)
Enteral feeding 16 (32%) 1 (3%)
Nil By Mouth 5 (10%)
*For difference between all 3 groups by one way analysis of variance.Mean (95% confidence interval).Median (interquartile range).doi:10.1371/journal.pone.0021185.t001
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 3 June 2011 | Volume 6 | Issue 6 | e21185
In a univariate mixed effects model, the decrease in KT ratio
over time correlated with the decrease in IL6 (p,0.0001) and IL10
(p,0.0001) between day 0 and day 7. In a multivariate model,
these relationships remained significant after controlling for
change in SOFA score (IL6 p = 0.009; IL10 p = 0.02).
IDO activity and lymphocyte countsSepsis patients had increased white blood cell counts
(p,0.0001) primarily due to increased circulating neutrophils
(p,0.05; Table 2), which proliferate in response to bacterial
infections [46]. Conversely, sepsis patients had significantly lower
total lymphocyte counts compared with hospital controls
(p,0.0001, Table 2). In all sepsis patients the baseline KT ratio
was weakly associated with absolute lymphocyte count (rp = 0.26,
p = 0.02). In a linear mixed effects model, absolute lymphocyte
count increased as the KT ratio decreased over time (p = 0.001).
This relationship persisted after controlling for SOFA score
(p = 0.008). When all subjects were grouped according to
lymphopenia, lymphopenic patients (n = 63) had a median KT
ratio of 128 [IQR 63–236], compared with 59 [33–86] in non-
lymphopenic patients (n = 57) (p,0.0001).
As IDO activity contributes to T cell apoptosis [18], we
examined the relationship between KT ratio and lymphocyte
subsets. Peripheral blood mononuclear cells were analysed from 23
of the 80 sepsis patients whose blood had been processed within
30 minutes of collection. This subset of patients was representative
of the cohort in terms of age, gender distribution, total lymphocyte
count and KT ratio. In this subset of patients, the KT ratio
negatively correlated with absolute numbers of lymphocytes
(rp = 20.54, p = 0.007), T cells (rp = 20.53, p = 0.01), CD4+ T
cells (rp = 20.50, p = 0.01), CD8+ T cells (rp = 20.49, p = 0.02)
and natural killer cells (rp = 20.46, p = 0.03) (Table 2).
Figure 1. Plasma assessment of tryptophan catabolism. The concentration of plasma tryptophan (Fig. 1A), kynurenine (Fig. 1B) and the KTratio (Fig. 1C) in 50 severe sepsis patients, 30 non-severe sepsis patients and 40 hospital controls. Fig. 1D shows the KT ratio in severe sepsis patientson admission (n = 50), day 2 (n = 34) and day 7 (n = 16). The KT ratio is determined by dividing the plasma kynurenine concentration (mmol/L) by theplasma tryptophan concentration (mmol/L) and multiplying the quotient by 1000. Horizontal lines represent median values for the group. P valueanalysis in Figs. 1A–C used a Mann Whitney test, and in Fig. 1D, a paired Wilcoxon test.doi:10.1371/journal.pone.0021185.g001
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 4 June 2011 | Volume 6 | Issue 6 | e21185
IDO activity and endothelial functionIn sepsis, the KT ratio at baseline correlated inversely with NO-
dependent microvascular reactivity (rs = 20.45, p = 0.001) even
after controlling for disease severity (using SOFA score; p = 0.001).
In a multivariate mixed effects model controlling for SOFA score,
improvement in KT ratio between day 0 and day 7 correlated with
improvement in microvascular reactivity (p = 0.001). In all sepsis
patients, there was an inverse association between the baseline KT
ratio and mean arterial pressure (rs = 20.29, p = 0.009) and
diastolic blood pressure (rs = 20.29, p = 0.01) but no association
with systolic blood pressure.
Discussion
IDO activity is increased in sepsis, in proportion to disease
severity. IDO-mediated tryptophan catabolism is associated with
dysregulated immune responses and impaired microvascular
reactivity in sepsis. IFN-c and IL10 are associated with, and
may contribute to, increased IDO activity in sepsis. The
independent inverse longitudinal association with total lymphocyte
counts suggests a potential role in sepsis-associated lymphopenia.
Similarly, the independent inverse association between the KT
ratio and microvascular reactivity suggests that IDO activity may
also contribute to impaired endothelial function in sepsis. Based on
these associations we propose a model of interpretation outlined in
Figure 2.
Increased expression of IFN-c [47], IL6 [48,49] and IL10 [14]
have each been associated with increased tryptophan catabolism
by IDO in other disease states. In sepsis patients in our study, IFN-
c concentration correlated with the KT ratio only at baseline,
whereas IL6 and IL10 correlated with the KT ratio both at
baseline and longitudinally. Our findings agree with the in vitro
literature, where IFN-c induces IDO [10,47]. Although under
certain conditions, IL-10 has been reported to suppress IDO
activity [50], our findings support the majority of in vitro studies
which have shown that IL-10 induces or stabilises IDO
[14,51,52,53]. The high IFN-c associated with early sepsis [54]
may lead to increased IDO activity while high IL10 may sustain or
potentially enhance IDO activity [53] throughout the course of the
disease. The role of IL6 in IDO expression is unclear. Orabona et
al. suggest that IL6 inhibits IDO activity by increasing murine
dendritic cell SOCS3 expression, which drives IDO breakdown
[55]. On the other hand, a low tryptophan environment created
by IDO activity stabilises IL6 mRNA and increases IL6 responses
[56]. Given the conflicting evidence in these and other studies
regarding IL6 and IDO, we investigated the relationship between
the KT ratio and IL6 in sepsis patients. The strong positive
correlation between plasma KT ratio and IL6 concentration
is consistent with findings in murine models of sepsis where
IDO2/2 mice or mice treated with IDO inhibitors have lower
plasma IL6 concentrations [57,58].
We report that the high KT ratio in sepsis is associated with a
decreased lymphocyte count, independent of disease severity, a
Table 2. Immunological characteristics of participants (median and interquartile range).
Severe sepsis Non-severe sepsis Combined sepsis ControlsSepsis vsControl*
n 50 30 80 40
Plasma tryptophan mmol/L 21 (13–29) 31 (23–37) 24 (14–35) 49 (40–55) ,0.0001
Plasma kynurenine mmol/L 3.5 (2.4–5.2) 2.3 (1.9–3.9) 3.1 (2.1–4.7) 1.9 (1.5–2.3) ,0.0001
KT ratio 162 (100–286) 82 (55–159) 141 (64–235) 36 (28–52) ,0.0001
Plasma IFN-c pg/mL 8 (1.3–20.1) n = 38 27 (3–84) n = 29 9 (3–48) n = 67 1.3 (1.3–7) n = 37 ,0.0001
Plasma IL6 pg/mL 380 (121–979) n = 38 136 (44–320) n = 29 222 (75–596) n = 67 5 (5-5) n = 37 ,0.0001
Plasma IL10 pg/mL 23 (13–64) n = 38 5 (5–25) n = 29 16 (5–41) n = 67 5 (5 - 5) n = 37 ,0.0001
Neutrophils 6103/mL 13.5 (8.7–20.4) n = 49 14.1 (9.2–16.3) 14 (8.8–16.6) n = 79 5.1 (3.2–6.5) n = 20 0.049
Lymphocytes x 6103/mL 0.9 (0.5–1.2) n = 49 1.0 (0.7–1.3) 0.9 (0.5–1.2) n = 79 2.1 (1.2–2.2) n = 20 ,0.0001
Lymphocyte subsets n = 11 n = 12 n = 23 n = 4
T cells 6103/mL 0.65 (0.34–1.8) 0.67 (0.34–1.0) 0.65 (0.34–1.1) 1.49 (1.0–1.7) NS
CD4+ T cells 6103/mL 0.35 (0.17–0.85) 0.35 (0.17–0.59) 0.35 (0.18–0.67) 0.89 (0.52–1.2) NS
CD8+ T cells 6103/mL 0.18 (0.07–0.72) 0.16 (0.10–0.33) 0.18 (0.1–0.34) 0.46 (0.31–0.61) NS
NK cells 6103/mL 0.07 (0.03–0.12) 0.06 (0.03–0.17) 0.06 (0.03–0.11) 0.08 (0.04–0.20) NS
*p values, all sepsis vs controls, Mann Whitney test.Performed in a subset of patients representative of the entire cohort, as described in methods and results. Severe sepsis n = 11, non-severe sepsis n = 12, control n = 4.doi:10.1371/journal.pone.0021185.t002
Figure 2. Proposed model of tryptophan catabolism in sepsis.IDO = Indoleamine 2,3-dioxygenase, IL6 = interleukin-6, IL10 = interleu-kin-10, IFN-c= interferon gamma and NO = nitric oxide.doi:10.1371/journal.pone.0021185.g002
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 5 June 2011 | Volume 6 | Issue 6 | e21185
finding similar to that found in patients with trauma [37], human
immunodeficiency virus [28] and cancer [59]. Previous studies in
sepsis have associated lymphopenia with disease severity [60],
duration of ICU stay [60] and mortality [61] and prevention of
lymphocyte apoptosis improves survival in animal models of sepsis
[62,63,64,65]. T cells co-cultured with IDO-producing cells have
reduced proliferation and increased death [66,67]. Both high
kynurenine concentrations and low tryptophan concentrations
appear to contribute to T cell death. In vivo, kynurenine treatment
in mice depletes overall thymocyte counts and, in vitro, thymocytes
die of apoptosis when cultured in media with kynurenines [18].
Furthermore, T cells cultured in low tryptophan media have
reduced proliferation and increased apoptosis via activated GCN2
kinase [68,69]. These in vitro studies suggest a potential mechanism
through which increased IDO activity may contribute to
lymphopenia and its deleterious consequences in sepsis.
IDO activity regulates vascular tone in sepsis. In this study IDO
activity in sepsis patients correlated with diastolic blood pressure
but not systolic blood pressure. This is consistent with the recent
finding that kynurenine is a vascular relaxation factor [9]. Another
important regulator of endothelial function in sepsis is NO. There
is significant cross-talk between IDO and NOS, with IDO activity
inhibiting both expression and activity of NOS [19,20,21] and vice
versa. We found the KT ratio in sepsis is inversely associated with
microvascular reactivity as measured by RH-PAT, which is at least
50% dependent on endothelial NO production [70]. Increased
IDO activity in sepsis may regulate vascular tone directly, via the
vasorelaxing effects of kynurenine, and indirectly, by impairing
NO-dependent microvascular reactivity. Increased plasma kyn-
urenine concentrations may further impede endothelial function in
sepsis by mediating adhesion of monocytes and neutrophils to the
vascular endothelium [71].
A limitation of this study is that we did not directly measure
IDO expression. However, the KT ratio is an established measure
of systemic IDO activity [28,72] with tissue IDO expression and
activity directly correlated with plasma KT ratio in multiple
human disease states, including celiac disease [73], hepatitis C [11]
and pre-eclampsia [74]. There are several possible sources of IDO
activity in sepsis patients including the endothelium, kidney, liver,
lungs and leukocytes [9,10,11,12,13,53], although a recent study
was unable to detect spontaneous IDO expression in PBMC from
sepsis patients [75]. Importantly, the effects of the high KT ratio in
sepsis on immune function and endothelial function would be the
same whether the high KT ratio was the result of increased IDO
activity alone or in combination with decreased feeding and
impaired renal excretion of kynurenine. Furthermore, it is unlikely
that nutritional deficiency and renal impairment accounted for the
differences we found, because controlling for these factors made no
difference to our results.
In our study the KT ratio was not significantly associated with
mortality. Consistent with the previously reported low mortality
from severe sepsis in our ICU [32,76], there were few deaths in
our study. This suggests that our study was under-powered to
examine the relationship between IDO activity and mortality.
However, in a study with higher numbers of deaths, Hattunen and
colleagues found a clear association between plasma KT ratio and
risk of death in sepsis [31].
The generation of a low tryptophan environment may be a
maladaptive host response to infection. While growth of some
bacterial species is inhibited by low tryptophan [77], most can
synthesize tryptophan [78] and others have specialized tryptophan
transport systems [79]. In murine models of sepsis, IDO2/2 mice
have significantly increased survival compared to wild type mice
[58] and treatment of wild-type mice with IDO inhibitors such as
1-methyl-tryptophan [58] or ethyl pyruvate also significantly
increase survival [57]. The KT ratio is significantly higher in
bacteremic patients with a fatal outcome [31] and we and others
have demonstrated that the KT ratio is associated with disease
severity in sepsis [31,75,80]. Together, this evidence supports the
hypothesis that increased IDO activity is a deleterious host
response in human sepsis. IDO inhibitors are being considered as
potential adjunctive cancer treatments [81] and these treatments
may also have therapeutic potential in sepsis.
ConclusionIDO activity is elevated in sepsis and associated with disease
severity, T cell lymphopenia and microvascular dysfunction.
Because excessive IDO activity is associated with both immune
and endothelial dysfunction, increased tryptophan catabolism may
link these two key aspects of sepsis pathophysiology. Modulation of
IDO activity warrants investigation as a therapeutic strategy in
sepsis.
Acknowledgments
We thank Kim Piera, Catherine Jones and Barbara MacHunter for
laboratory assistance; Jane Thomas, Mark McMillan, Karl Blenk, Antony
Van Asche, Steven Tong and Paulene Kittler for RH-PAT measurements
and sample collection; Alex Humphrey for database assistance; Joseph
McDonnell for statistical advice; David Celermajer for advice on vascular
function assessments and contribution to the design of the original study,
and the medical and nursing staff of the Royal Darwin Hospital Intensive
Care Unit, Division of Medicine and Hospital in the Home.
Author Contributions
Conceived and designed the experiments: CJD JSD TW YRM DPS TWY
NMA. Performed the experiments: JSD DPS CJD TW YRM. Analyzed
the data: CJD JSD TW NMA. Wrote the paper: CJD JSD TW YRM DPS
TWY NMA.
References
1. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, et al. (1992) Definitions for
sepsis and organ failure and guidelines for the use of innovative therapies in sepsis.
The ACCP/SCCM Consensus Conference Committee. American College of
Chest Physicians/Society of Critical Care Medicine. Chest 101: 1644–1655.
2. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, et al. (2001)
Epidemiology of severe sepsis in the United States: analysis of incidence,
outcome, and associated costs of care. Crit Care Med 29: 1303–1310.
3. Finfer S, Bellomo R, Lipman J, French C, Dobb G, et al. (2004) Adult-
population incidence of severe sepsis in Australian and New Zealand intensive
care units. Intensive Care Med 30: 589–596.
4. Blanco J, Muriel-Bombin A, Sagredo V, Taboada F, Gandia F, et al. (2008)
Incidence, organ dysfunction and mortality in severe sepsis: a Spanish
multicentre study. Crit Care 12: R158.
5. Hotchkiss RS, Karl IE (2003) The pathophysiology and treatment of sepsis.
N Engl J Med 348: 138–150.
6. Aird WC (2003) The role of the endothelium in severe sepsis and multiple organ
dysfunction syndrome. Blood 101: 3765–3777.
7. Ellinger P, Abdel Kader MM (1947) Tryptophane as precursor of nicotinamide
in mammals. Nature 160: 675.
8. Fernstrom JD, Wurtman RJ (1971) Brain serotonin content: physiological
dependence on plasma tryptophan levels. Science 173: 149–152.
9. Wang Y, Liu H, McKenzie G, Witting PK, Stasch JP, et al. (2010) Kynurenine is
an endothelium-derived relaxing factor produced during inflammation. Nat
Med 16: 279–285.
10. Carlin JM, Borden EC, Sondel PM, Byrne GI (1989) Interferon-induced
indoleamine 2,3-dioxygenase activity in human mononuclear phagocytes.
J Leukoc Biol 45: 29–34.
11. Larrea E, Riezu-Boj JI, Gil-Guerrero L, Casares N, Aldabe R, et al. (2007)
Upregulation of indoleamine 2,3-dioxygenase in hepatitis C virus infection.
J Virol 81: 3662–3666.
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 6 June 2011 | Volume 6 | Issue 6 | e21185
12. Iwamoto N, Ito H, Ando K, Ishikawa T, Hara A, et al. (2009) Upregulation ofindoleamine 2,3-dioxygenase in hepatocyte during acute hepatitis caused by
hepatitis B virus-specific cytotoxic T lymphocytes in vivo. Liver Int 29: 277–283.
13. Mohib K, Guan Q, Diao H, Du C, Jevnikar AM (2007) Proapoptotic activity ofindoleamine 2,3-dioxygenase expressed in renal tubular epithelial cells.
Am J Physiol Renal Physiol 293: F801–812.
14. Munn DH, Sharma MD, Lee JR, Jhaver KG, Johnson TS, et al. (2002) Potentialregulatory function of human dendritic cells expressing indoleamine 2,3-
dioxygenase. Science 297: 1867–1870.
15. Fallarino F, Grohmann U, You S, McGrath BC, Cavener DR, et al. (2006) Thecombined effects of tryptophan starvation and tryptophan catabolites down-
regulate T cell receptor zeta-chain and induce a regulatory phenotype in naive Tcells. J Immunol 176: 6752–6761.
16. Munn DH, Shafizadeh E, Attwood JT, Bondarev I, Pashine A, et al. (1999)
Inhibition of T cell proliferation by macrophage tryptophan catabolism. J ExpMed 189: 1363–1372.
17. Boasso A, Herbeuval JP, Hardy AW, Anderson SA, Dolan MJ, et al. (2007) HIV
inhibits CD4+ T-cell proliferation by inducing indoleamine 2,3-dioxygenase inplasmacytoid dendritic cells. Blood 109: 3351–3359.
18. Fallarino F, Grohmann U, Vacca C, Bianchi R, Orabona C, et al. (2002) T cell
apoptosis by tryptophan catabolism. Cell Death Differ 9: 1069–1077.
19. Sekkai D, Guittet O, Lemaire G, Tenu JP, Lepoivre M (1997) Inhibition of nitric
oxide synthase expression and activity in macrophages by 3-hydroxyanthranilic
acid, a tryptophan metabolite. Arch Biochem Biophys 340: 117–123.
20. Chiarugi A, Rovida E, Dello Sbarba P, Moroni F (2003) Tryptophan availability
selectively limits NO-synthase induction in macrophages. J Leukoc Biol 73:
172–177.
21. Samelson-Jones BJ, Yeh SR (2006) Interactions between nitric oxide and
indoleamine 2,3-dioxygenase. Biochemistry 45: 8527–8538.
22. Luiking YC, Poeze M, Ramsay G, Deutz NE (2009) Reduced citrullineproduction in sepsis is related to diminished de novo arginine and nitric oxide
production. Am J Clin Nutr 89: 142–152.
23. Kao CC, Bandi V, Guntupalli KK, Wu M, Castillo L, et al. (2009) Arginine,citrulline and nitric oxide metabolism in sepsis. Clin Sci (Lond) 117: 23–30.
24. Villalpando S, Gopal J, Balasubramanyam A, Bandi VP, Guntupalli K, et al.
(2006) In vivo arginine production and intravascular nitric oxide synthesis inhypotensive sepsis. Am J Clin Nutr 84: 197–203.
25. Davis JS, Yeo TW, Thomas JH, McMillan M, Darcy CJ, et al. (2009) Sepsis-
associated microvascular dysfunction measured by peripheral arterial tonometry:an observational study. Crit Care 13: R155.
26. Vaudo G, Marchesi S, Siepi D, Brozzetti M, Lombardini R, et al. (2008) Humanendothelial impairment in sepsis. Atherosclerosis 197: 747–752.
27. Pawlak K, Domaniewski T, Mysliwiec M, Pawlak D (2009) Kynurenines and
oxidative status are independently associated with thrombomodulin and vonWillebrand factor levels in patients with end-stage renal disease. Thromb Res
124: 452–457.
28. Huengsberg M, Winer JB, Gompels M, Round R, Ross J, et al. (1998) Serumkynurenine-to-tryptophan ratio increases with progressive disease in HIV-
infected patients. Clin Chem 44: 858–862.
29. Widner B, Sepp N, Kowald E, Ortner U, Wirleitner B, et al. (2000) Enhancedtryptophan degradation in systemic lupus erythematosus. Immunobiology 201:
621–630.
30. Huang A, Fuchs D, Widner B, Glover C, Henderson DC, et al. (2002) Serumtryptophan decrease correlates with immune activation and impaired quality of
life in colorectal cancer. Br J Cancer 86: 1691–1696.
31. Huttunen R, Syrjanen J, Aittoniemi J, Oja SS, Raitala A, et al. (2009) Highactivity of indoleamine 2,3 dioxygenase enzyme predicts disease severity and
case fatality in bacteremic patients. Shock.
32. Stephens DP, Thomas JH, Higgins A, Bailey M, Anstey NM, et al. (2008)Randomized, double-blind, placebo-controlled trial of granulocyte colony-
stimulating factor in patients with septic shock. Crit Care Med 36: 448–454.
33. Vincent JL, de Mendonca A, Cantraine F, Moreno R, Takala J, et al. (1998) Useof the SOFA score to assess the incidence of organ dysfunction/failure in
intensive care units: results of a multicenter, prospective study. Working groupon ‘‘sepsis-related problems’’ of the European Society of Intensive Care
Medicine. Crit Care Med 26: 1793–1800.
34. Hotchkiss RS, Swanson PE, Freeman BD, Tinsley KW, Cobb JP, et al. (1999)Apoptotic cell death in patients with sepsis, shock, and multiple organ
dysfunction. Crit Care Med 27: 1230–1251.
35. van Wandelen C, Cohen SA (1997) Using quaternary high-performance liquidchromatography eluent systems for separating 6-aminoquionolyl-N-hydroxysuc-
cinyl carbamate-derivatized amino acid mixtures. J Chromatogr A 763: 11–22.
36. Zangerle R, Widner B, Quirchmair G, Neurauter G, Sarcletti M, et al. (2002)Effective antiretroviral therapy reduces degradation of tryptophan in patients
with HIV-1 infection. Clin Immunol 104: 242–247.
37. Pellegrin K, Neurauter G, Wirleitner B, Fleming AW, Peterson VM, et al. (2005)
Enhanced enzymatic degradation of tryptophan by indoleamine 2,3-dioxygenase
contributes to the tryptophan-deficient state seen after major trauma. Shock 23:209–215.
38. Deanfield JE, Halcox JP, Rabelink TJ (2007) Endothelial function and
dysfunction: testing and clinical relevance. Circulation 115: 1285–1295.
39. Kuvin JT, Mammen A, Mooney P, Alsheikh-Ali AA, Karas RH (2007)
Assessment of peripheral vascular endothelial function in the ambulatory setting.
Vasc Med 12: 13–16.
40. Celermajer DS (2008) Reliable endothelial function testing: at our fingertips?
Circulation 117: 2428–2430.
41. Kuvin JT, Patel AR, Sliney KA, Pandian NG, Sheffy J, et al. (2003) Assessmentof peripheral vascular endothelial function with finger arterial pulse wave
amplitude. Am Heart J 146: 168–174.
42. Bonetti PO, Pumper GM, Higano ST, Holmes DR, Jr., Kuvin JT, et al. (2004)Noninvasive identification of patients with early coronary atherosclerosis by
assessment of digital reactive hyperemia. J Am Coll Cardiol 44: 2137–2141.
43. Gibbons J, Chakraborti S (2003) Nonparametric Statistical Inference Marcel
Dekker.
44. Pawlak D, Tankiewicz A, Matys T, Buczko W (2003) Peripheral distribution ofkynurenine metabolites and activity of kynurenine pathway enzymes in renal
failure. J Physiol Pharmacol 54: 175–189.
45. Schefold JC, Zeden JP, Fotopoulou C, von Haehling S, Pschowski R, et al.(2009) Increased indoleamine 2,3-dioxygenase (IDO) activity and elevated serum
levels of tryptophan catabolites in patients with chronic kidney disease: a possible
link between chronic inflammation and uraemic symptoms. Nephrol DialTransplant 24: 1901–1908.
46. Nelson S (1994) Role of granulocyte colony-stimulating factor in the immune
response to acute bacterial infection in the nonneutropenic host: an overview.Clin Infect Dis 18 Suppl 2: S197–204.
47. Yoshida R, Imanishi J, Oku T, Kishida T, Hayaishi O (1981) Induction of
pulmonary indoleamine 2,3-dioxygenase by interferon. Proc Natl Acad Sci U S A
78: 129–132.
48. Maes M, Meltzer HY, Scharpe S, Bosmans E, Suy E, et al. (1993) Relationshipsbetween lower plasma L-tryptophan levels and immune-inflammatory variables
in depression. Psychiatry Res 49: 151–165.
49. Bonaccorso S, Lin A, Song C, Verkerk R, Kenis G, et al. (1998) Serotonin-immune interactions in elderly volunteers and in patients with Alzheimer’s
disease (DAT): lower plasma tryptophan availability to the brain in the elderlyand increased serum interleukin-6 in DAT. Aging (Milano) 10: 316–323.
50. MacKenzie CR, Gonzalez RG, Kniep E, Roch S, Daubener W (1999) Cytokinemediated regulation of interferon-gamma-induced IDO activation. Adv Exp
Med Biol 467: 533–539.
51. van der Sluijs KF, Nijhuis M, Levels JH, Florquin S, Mellor AL, et al. (2006)Influenza-induced expression of indoleamine 2,3-dioxygenase enhances inter-
leukin-10 production and bacterial outgrowth during secondary pneumococcalpneumonia. J Infect Dis 193: 214–222.
52. Maneechotesuwan K, Supawita S, Kasetsinsombat K, Wongkajornsilp A,
Barnes PJ (2008) Sputum indoleamine-2, 3-dioxygenase activity is increased in
asthmatic airways by using inhaled corticosteroids. J Allergy Clin Immunol 121:43–50.
53. Yanagawa Y, Iwabuchi K, Onoe K (2009) Co-operative action of interleukin-10
and interferon-gamma to regulate dendritic cell functions. Immunology 127:345–353.
54. Hunsicker A, Kullich W, Weissenhofer W, Lorenz D, Petermann J, et al. (1997)
Correlations between endotoxin, interferon-gamma, biopterin and serum
phospholipase A2-activities during lethal gram negative sepsis in rats.Eur J Surg 163: 379–385.
55. Orabona C, Pallotta MT, Volpi C, Fallarino F, Vacca C, et al. (2008) SOCS3
drives proteasomal degradation of indoleamine 2,3-dioxygenase (IDO) andantagonizes IDO-dependent tolerogenesis. Proc Natl Acad Sci U S A 105:
20828–20833.
56. van Wissen M, Snoek M, Smids B, Jansen HM, Lutter R (2002) IFN-gamma
amplifies IL-6 and IL-8 responses by airway epithelial-like cells via indoleamine2,3-dioxygenase. J Immunol 169: 7039–7044.
57. Ulloa L, Ochani M, Yang H, Tanovic M, Halperin D, et al. (2002) Ethyl
pyruvate prevents lethality in mice with established lethal sepsis and systemicinflammation. Proc Natl Acad Sci U S A 99: 12351–12356.
58. Jung ID, Lee MG, Chang JH, Lee JS, Jeong YI, et al. (2009) Blockade of
indoleamine 2,3-dioxygenase protects mice against lipopolysaccharide-induced
endotoxin shock. J Immunol 182: 3146–3154.
59. Ino K, Yamamoto E, Shibata K, Kajiyama H, Yoshida N, et al. (2008) Inversecorrelation between tumoral indoleamine 2,3-dioxygenase expression and
tumor-infiltrating lymphocytes in endometrial cancer: its association with diseaseprogression and survival. Clin Cancer Res 14: 2310–2317.
60. Le Tulzo Y, Pangault C, Gacouin A, Guilloux V, Tribut O, et al. (2002) Early
circulating lymphocyte apoptosis in human septic shock is associated with poor
outcome. Shock 18: 487–494.
61. Felmet KA, Hall MW, Clark RS, Jaffe R, Carcillo JA (2005) Prolongedlymphopenia, lymphoid depletion, and hypoprolactinemia in children with
nosocomial sepsis and multiple organ failure. J Immunol 174: 3765–3772.
62. Hotchkiss RS, Tinsley KW, Swanson PE, Chang KC, Cobb JP, et al. (1999)Prevention of lymphocyte cell death in sepsis improves survival in mice. Proc
Natl Acad Sci U S A 96: 14541–14546.
63. Hotchkiss RS, Chang KC, Swanson PE, Tinsley KW, Hui JJ, et al. (2000)
Caspase inhibitors improve survival in sepsis: a critical role of the lymphocyte.Nat Immunol 1: 496–501.
64. Bommhardt U, Chang KC, Swanson PE, Wagner TH, Tinsley KW, et al. (2004)
Akt decreases lymphocyte apoptosis and improves survival in sepsis. J Immunol172: 7583–7591.
65. Schwulst SJ, Muenzer JT, Peck-Palmer OM, Chang KC, Davis CG, et al. (2008)
Bim siRNA decreases lymphocyte apoptosis and improves survival in sepsis.
Shock 30: 127–134.
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 7 June 2011 | Volume 6 | Issue 6 | e21185
66. Fallarino F, Vacca C, Orabona C, Belladonna ML, Bianchi R, et al. (2002)
Functional expression of indoleamine 2,3-dioxygenase by murine CD8 alpha(+)
dendritic cells. Int Immunol 14: 65–68.
67. Odemuyiwa SO, Ghahary A, Li Y, Puttagunta L, Lee JE, et al. (2004) Cutting
edge: human eosinophils regulate T cell subset selection through indoleamine
2,3-dioxygenase. J Immunol 173: 5909–5913.
68. Lee GK, Park HJ, Macleod M, Chandler P, Munn DH, et al. (2002) Tryptophan
deprivation sensitizes activated T cells to apoptosis prior to cell division.
Immunology 107: 452–460.
69. Forouzandeh F, Jalili RB, Germain M, Duronio V, Ghahary A (2008) Skin cells,
but not T cells, are resistant to indoleamine 2, 3-dioxygenase (IDO) expressed by
allogeneic fibroblasts. Wound Repair Regen 16: 379–387.
70. Nohria A, Gerhard-Herman M, Creager MA, Hurley S, Mitra D, et al. (2006)
Role of nitric oxide in the regulation of digital pulse volume amplitude in
humans. J Appl Physiol 101: 545–548.
71. Barth MC, Ahluwalia N, Anderson TJ, Hardy GJ, Sinha S, et al. (2009)
Kynurenic acid triggers firm arrest of leukocytes to vascular endothelium under
flow conditions. J Biol Chem.
72. Suzuki Y, Suda T, Furuhashi K, Suzuki M, Fujie M, et al. (2010) Increased
serum kynurenine/tryptophan ratio correlates with disease progression in lung
cancer. Lung Cancer 67: 361–365.
73. Torres MI, Lopez-Casado MA, Lorite P, Rios A (2007) Tryptophan metabolism
and indoleamine 2,3-dioxygenase expression in coeliac disease. Clin Exp
Immunol 148: 419–424.
74. Kudo Y, Boyd CA, Sargent IL, Redman CW (2003) Decreased tryptophan
catabolism by placental indoleamine 2,3-dioxygenase in preeclampsia.Am J Obstet Gynecol 188: 719–726.
75. Tattevin P, Monnier D, Tribut O, Dulong J, Bescher N, et al. (2010) Enhanced
indoleamine 2,3-dioxygenase activity in patients with severe sepsis and septicshock. J Infect Dis 201: 956–966.
76. Davis JS, Cheng AC, McMillan M, Anstey NM (2011) Sepsis in the tropicalNorthern Territory of Australia: high disease burden with disproportionate
impact on Indigenous Australians. Med J Aust In press.
77. MacKenzie CR, Hadding U, Daubener W (1998) Interferon-gamma-inducedactivation of indoleamine 2,3-dioxygenase in cord blood monocyte-derived
macrophages inhibits the growth of group B streptococci. J Infect Dis 178:875–878.
78. Merino E, Jensen RA, Yanofsky C (2008) Evolution of bacterial trp operons andtheir regulation. Curr Opin Microbiol 11: 78–86.
79. Yanofsky C, Horn V, Gollnick P (1991) Physiological studies of tryptophan
transport and tryptophanase operon induction in Escherichia coli. J Bacteriol173: 6009–6017.
80. Schefold JC, Zeden JP, Pschowski R, Hammoud B, Fotopoulou C, et al. (2010)Treatment with granulocyte-macrophage colony-stimulating factor is associated
with reduced indoleamine 2,3-dioxygenase activity and kynurenine pathway
catabolites in patients with severe sepsis and septic shock. Scand J Infect Dis 42:164–171.
81. Lob S, Konigsrainer A, Rammensee HG, Opelz G, Terness P (2009) Inhibitorsof indoleamine-2,3-dioxygenase for cancer therapy: can we see the wood for the
trees? Nat Rev Cancer 9: 445–452.
KT Ratio in Sepsis
PLoS ONE | www.plosone.org 8 June 2011 | Volume 6 | Issue 6 | e21185
