ORIGINAL ARTICLE

Features of Immune Senescence in LiverTransplant Recipients with Established GraftsWilliam Gelson,1 Matthew Hoare,1 Sarah Vowler,2 Arun Shankar,1 Paul Gibbs,3 Arne N. Akbar,4

and Graeme J. M. Alexander11Department of Medicine, 2Centre for Applied Medical Statistics, and 3Department of Surgery,University of Cambridge, Cambridge, UK; and 4Department of Immunology,University College London, London, UK

Immune senescence is the normal process whereby the human immune system ages, but becomes less effective. Weinvestigated whether liver transplant recipients have features of immune senescence. Lymphocytes from 97 liver transplantrecipients with established grafts and 41 age-matched and sex-matched controls were subjected to an 8-color flow cytome-try assay that measured expression of killer cell lectin-like receptor subfamily G member 1, cluster of differentiation 127(CD127), CD45RO, CD27, CD28, CD4, CD8, and CD57. Lymphocyte telomere length was assessed by flow-fluorescencein situ hybridization. Cases were compared with controls for each marker of immune senescence using a Mann-Whitney Utest. For liver transplant recipients, linear regression analyses identified associations between markers of immune senes-cence and clinical or demographic characteristics. Lymphocytes from liver transplant recipients expressed more phenotypicmarkers of maturity than did lymphocytes from controls. Lymphocyte telomeres were shorter in liver transplant recipientsthan in controls. Age, hepatocellular carcinoma at transplantation, and skin malignancy developing after transplantationwere associated independently with shortened lymphocyte telomeres. Increasing age and previous cytomegalovirus infec-tion were associated independently with phenotypic markers of lymphocyte maturity. Thus, lymphocytes from liver transplantrecipients are older ‘‘biologically’’ than lymphocytes from age-matched and sex-matched controls. Hepatocellular carcinomaat transplantation, subsequent skin malignancy, and previous cytomegalovirus infection are associated with lymphocyte se-nescence in liver transplant recipients. Liver Transpl 16:577-587, 2010. VC 2010 AASLD.

Received November 14, 2009; accepted January 16, 2010.

See Editorial on Page 548

The peripheral T lymphocyte pool is maintained by acombination of antigenic stimulation and other, lesswell-defined homeostatic mechanisms.1 Peripheral Tlymphocytes turn over constantly. This may be part ofa slow homeostatic process, with bursts of ‘‘activity’’when a target antigen is experienced, or throughchronic antigenic stimulation, for example in the pres-ence of chronic viral infection.2-4 The result of con-

stant turnover is immune senescence, characterizedby a population of ‘‘exhausted’’ lymphocytes with amature cell surface phenotype that demonstrate repli-cative senescence.5,6

All healthy cells, including T lymphocytes, have a fi-nite lifespan. Human cluster of differentiation 4-posi-tive (CD4þ) lymphocytes sustain around 33 popula-tion doublings in culture; CD8þ cells sustain fewer,around 23.7,8 As T cells reach replicative senescence,they stop dividing due to the development of cell-cyclearrest,9 at which point the cell becomes resistant to

Abbreviations: ALD, alcohol-related liver disease; AIH, autoimmune hepatitis; APC, allophycocyanin; bp, base pair; CD, cluster ofdifferentiation; CI, confidence interval; CMV, cytomegalovirus; Cy5, cyanine 5; EBV, Epstein-Barr virus; FISH, fluorescent in situhybridization; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HIV, human immunodeficiencyvirus; KLRG1, killer cell lectin-like receptor subfamily G member 1; mfi, mean fluorescence intensity; PBC, primary biliarycirrhosis; PBMC, peripheral blood mononuclear cell; PBS, phosphate-buffered saline; PE, phycoerythrin; PSC, primary sclerosingcholangitis.The British Transplantation Society supported Dr. William Gelson; The Wellcome Trust supported Dr. Matthew Hoare. The FrankLitchfield Charitable Trust and Cambridge Hepatology Endowment Fund provided financial support for the running costs of thestudy.Address reprint requests to Dr. Graeme J. M. Alexander, Department of Medicine (Box 157), Cambridge University Hospitals NHS FoundationTrust, Cambridge, CB0 2QQ, United Kingdom. Telephone: 01223 586614; FAX: 01223 216111; E-mail: gja1000@doctors.org.uk

DOI 10.1002/lt.22033Published online in Wiley InterScience (www.interscience.wiley.com).

LIVER TRANSPLANTATION 16:577-587, 2010

VC 2010 American Association for the Study of Liver Diseases.

apoptosis and there are significant changes inimmune function.10 In animals, aged naı̈ve T cellsproduce less interleukin-2 and more interferon-c thantheir young counterparts and express fewer mem-brane activation markers (CD25, CD62L, and CD154)when stimulated in vitro.11-13 The ability of aged naı̈veT cells to provide B cell help also wanes.14 Further,aged memory T cells derived from aged naı̈ve T cellsfail to proliferate or provide help in vitro.3

Telomeres are formed by a repeated hexamericsequence of nucleotides (TTAGGG),15 which are foundat the ends of chromosomes and shorten by 50-100base pairs (bp) in most somatic cells (including lym-phocytes) at each cellular division.16 When telomerelength becomes critically short, the cell becomessenescent.17 Telomere length thus provides a surro-gate in vivo marker for the assessment of immunesenescence in different cell populations, includinglymphocytes.

Many previous studies in nontransplant popula-tions have demonstrated that cardiac disease, malig-nancy, cerebrovascular disease, and infections areassociated closely with shortened telomeres in periph-eral blood mononuclear cells (PBMCs); furthermore,prospective studies in healthy elderly populationsreveal that shortened telomeres identify a cohort witha subsequent increase in morbidity and mortal-ity.16,18-21 These observations suggest first thatimmune senescence carries an increased risk andsecond that cardiac disease, malignancy, cerebrovas-cular disease, and infections are disorders associatedwith the process of immune senescence.

There is a marked increase in the prevalence of car-diac disease, malignancy, cerebrovascular disease,and infections in patients with established liver grafts,eventually affecting a majority of cases and which inthe past have been attributed to agents used to sup-press immune responses.22-31 However, an alternative(and not exclusive) hypothesis is that liver transplantrecipients develop premature immune senescence,which is associated with an increased risk of cardiacdisease, malignancy, cerebrovascular disease, andinfection, perhaps consequent to chronic alloanti-genic stimulation.16 To our knowledge, immune se-nescence has not been studied in the context oftransplantation.

Based on these observations, we investigated keyfeatures of immune senescence (lymphocyte telomerelength and lymphocyte phenotypic markers) in livertransplant recipients with established liver allografts.The relation between immune senescence and bothclinical features of alloimmunity and complications oftransplantation were also investigated.

PATIENTS AND METHODS

Consent

Informed consent was obtained from patients andcontrols.

Patient Selection

A total of 97 patients with established liver allograftswere selected for study from consecutive routine livertransplant clinics. An established graft was defined asbeing at least 3 years from engraftment. All thepatients were well at the time of recruitment. Patientcharacteristics are shown in Table 1. The study wascarried out with the approval of the Local Researchand Ethics Committee.

Controls

Controls comprised healthy individuals who attendedclinic with patients (n ¼ 41) in an attempt to obtain agroup with similar demographic characteristics; theirdetails are also shown in Table 1.

TABLE 1. Patient and Control Details

Characteristic Cases Controls

Mean age (standarddeviation)*

55.0 (15.8) 55.9 (13.2)

% male* 51.4% 39.0%n 97 41Etiology:Alcohol-related liver disease 12 (12.4%) –Cryptogenic 8 (8.2%) –Hepatitis C virus 6 (6.2%) –Hepatitis B virus 5 (5.2%) –Primary biliary cirrhosis 16 (16.5%) –Autoimmune hepatitis 2 (2.1%) –Primary sclerosing

cholangitis11 (11.3%) –

Acute 11 (11.3%) –Metabolic 7 (7.2%) –Other 19 (19.6%) –Hepatocellular carcinoma

at engraftment9 (9.2%) –

% cytomegalovirusantibody-positive

80.4% –

% with normal liverbiochemistry

76.3% –

% on calcineurin inhibitor 67.0% –% on sirolimus 24.7% –Mean no. of immune

suppressants1.3 –

Mean no. of infectiousepisodes

2.5 –

Mean no. of treatedepisodes of acuterejection

0.7 –

Number of skinmalignancies

12 in10 patients

–

Number of othermalignancies

9 in7 patients

–

*There was no significant difference in age (P ¼ 0.93,Mann-Whitney U Test) or sex (P ¼ 0.18, chi-squaredtest) between cases and controls.

578 GELSON ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

Preparation of PBMCs

PBMCs were obtained by centrifugation of citrated pe-ripheral blood over Lymphoprep (Axis-Shield, Cam-bridgeshire, UK). Cell number and viability wereassessed, the latter by trypan blue exclusion. Theywere cryopreserved at �80�C in 80% fetal bovine se-rum, 10% Roswell Park Memorial Institute 1640(RPMI-1640) medium and 10% dimethyl sulfoxide inaliquots of 5 million cells. Prior to analysis, PBMCswere snap-thawed to room temperature in a waterbath at 37�C before slow resuspension in 10 mLRPMI-1640/10% fetal bovine serum that had beenwarmed to 37�C.

Flow Cytometry

Samples of 0.5 � 106 to 1 � 106 cells were washed andresuspended in 100 lL phosphate-buffered saline(PBS). Cells were then incubated at 4�C for 30 minuteswith saturating concentrations of biotinylated, fluoro-chrome-conjugated antibody or appropriate isotypecontrols and washed with 1 mL PBS. For biotinylatedantibodies, cells were resuspended in 100 lL PBS,incubated at 4�C for 30 minutes with a saturating con-centration of streptavidin-Qdot605 (Invitrogen) andwashed with 1 mL PBS. Samples were then resus-pended in 200 lL BD Cytofix solution (Becton Dickin-son, UK) for fixing prior to analysis. Flow cytometrywas performed on a FACSCanto (BD) and data were an-alyzed using FACSDiva software (BD). An 8-color assaywas used to assess lymphocyte phenotype. Antibodiesused were CD4 Pacific Blue and CD8 conjugated tophycoerythrin and cyanine 5 (PE-Cy5) to identify helperand cytotoxic lymphocytes; CD127-PE, CD27 conju-gated to allophycocyanin and AlexaFluor450 (APC-Alexa450) and CD28-APC to identify immature lym-phocytes; killer cell lectin-like receptor G1 (KLRG1)conjugated to fluorescein isothiocyanate (FITC),32

CD45RO-PE-Cy7, and biotinylated CD57 with streptavi-din Qdot 655 conjugate to identify mature lymphocytes.

Assessment of Telomere Length Using

Flow Cytometry (Flow-Fluorescent

In Situ Hybridization)

Telomere length of CD4þ and CD8þ T cells was meas-ured in a 5-color flow-fluorescent in situ hybridization(FISH) assay, a technique that has been describedpreviously.33,34 PBMCs were resuspended in PBS andstained with CD4-Qdot655 (Invitrogen), CD8-Qdot605(Invitrogen), CD45RO-FITC (Dako) and biotinylatedCD57 (BD) followed by streptavidin-Cy3 (CedarlaneLaboratories). These markers were chosen to allowassessment of telomere length in mature T cells(CD4þ CD45ROþ and CD8þ CD57þ), immature Tcells (CD4þ CD45RO� and CD8þ CD57�) and all Tcells (CD4þ and CD8þ). Samples were then fixed (Fixand Perm cell permeabilization kit; CALTAG Laborato-ries, Burlingame, CA). After washing in hybridizationbuffer (containing 70% formamide [VWR], 1% bovine

serum albumin, 150 mM NaCl, 20 mM Tris-HCl) cellswere divided into 3 equal aliquots and incubated with200 lL 0.75 lg/mL Cy5-conjugated telomere probe((C3TA2)3, Panagene, UK) in a water bath at 82�C for10 minutes. After rapid cooling on ice and hybridiza-tion in the dark at room temperature for 60 minutes,samples were washed in posthybridization buffer(70% formamide, 10 mM Tris-HCl, 150 mM NaCl,0.1% bovine serum albumin, 0.1% Tween 20) and im-mediately analyzed in triplicate by flow cytometry.PBMCs from the same healthy individual were ana-lyzed in each experiment as an internal control.Results are expressed as mean fluorescence intensity(mfi). For each individual, this is a weighted mean ofthe 3 experimental samples, normalized to the CD4þPBMCs from the healthy individual.

Statistics

Statistical analysis was performed using SPSS version13.0 for Macintosh. Spearman correlations confirmedsignificant relationships of telomere length and lympho-cyte cell surface markers with age. Mann-Whitney Utests compared telomere length with lymphocyte cellsurface marker expression between cases and controls.Simple linear regression with study variables as out-comes were screened for associations with clinical anddemographic factors (P < 0.1). These factors were car-ried through to multiple linear regressions where P <0.05 was considered significant. Clinical and demo-graphic factors entered included age, sex, pediatric oradult recipient at engraftment, number of transplants,underlying disease etiology (immune [primary biliarycirrhosis, primary sclerosing cholangitis, or autoim-mune hepatitis], viral [chronic hepatitis B virus or hep-atitis C virus infection], metabolic [alcohol-related liverdisease, cryptogenic, or other metabolic causes such ashemochromatosis and a1-antitrypsin deficiency], acute,or other), the presence of hepatocellular carcinoma(HCC) at transplantation, anti-cytomegalovirus (CMV)antibody status at sampling, having normal liver bio-chemistry at sampling, calcineurin inhibitor or siroli-mus-based immune suppression, number of prescribedimmune-suppressive agents at sampling, number of in-fective episodes, number of treated episodes of acuterejection, malignancy (skin and other), and precancer-ous lesions. There were only 2 cerebrovascular and 2cardiovascular events in the series, so these could notbe included in the statistical analysis.

RESULTS

An example of flow cytometry output for PBMC cell-surface phenotype is given in Fig. 1.

The Relation Between Age with Telomere

Length and Lymphocyte Surface Phenotype

in Cases and Controls

Age was associated inversely and significantly withtelomere length in both cases and controls in all

IMMUNE SENESCENCE IN LIVER TRANSPLANT RECIPIENTS 579

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

lymphocyte subsets studied (Fig. 2 and Table 2).There was correlation between age and cell-surfacemarkers of cellular senescence, which was significantfor all lymphocyte subsets studied in the cases. Fur-ther, there was close correlation between lymphocytetelomere length and cell-surface markers of cellularsenescence in cases but not controls. These observa-tions confirm that age was associated with increasedlymphocyte maturation, which was most marked incases; thus lymphocyte telomeres shortened withincreasing age, the proportion of lymphocytes with amature lymphocyte phenotype increased (KLRG1,CD57, and CD45RO), and the proportion of lympho-cytes with an immature phenotype decreased (CD27,CD28, and CD127).

Comparison of Liver Transplant Recipients

with Healthy Controls

There was no evidence of an age (P ¼ 0.93) or sex (P ¼0.18) difference between cases and controls. Table 3shows a comparison of telomere length in differentlymphocyte subsets in liver transplant recipients (n ¼97) with healthy controls (n ¼ 41). Telomere lengthwas shorter in liver transplant recipients than con-trols in all T cell subsets. These differences were sig-nificant in lymphocytes overall (P ¼ 0.004), CD4þ Tcells (P ¼ 0.01), CD8þ T cells (P ¼ 0.01), less matureT cells (CD4þ CD45RO� [P ¼ 0.004] and CD8þCD57� [P ¼ 0.003]). Using a regression line for alldata in Fig. 2, the differences were interpreted interms of absolute age, indicating a difference betweenliver transplant recipients and controls of between 4and 5 years of additional immune aging. The observa-tion that the difference was more marked in immaturethan mature T cell subsets (and was significant inimmature but not mature T cell subsets) was surpris-ing but noteworthy.

To assess whether accelerated immune senescenceis an ongoing process in liver transplant recipients,telomere length for CD4þ cells were plotted againstage for cases and controls (Fig. 2). The slopes andintercepts of the regression lines were compared withan analysis of covariance test. The slopes were notsignificantly different (P ¼ 0.95); the intercepts weresignificantly different (P < 0.001).

Table 4 shows a comparison of lymphocyte pheno-typic markers of maturity in liver transplant recipi-ents (n ¼ 97) with healthy controls (n ¼ 41). For allmarkers, there was a higher proportion of lympho-cytes with a mature phenotype (CD4þ KLRG1þ,CD4þ CD45ROþ, CD8þ KLRG1þ and CD8þ CD57þ)

Figure 1. Representative flow cytometry plots for lymphocytecell-surface markers related to senescence. Lymphocytes werestained with anti-CD8 PE-Cy5, anti-CD4 pacific blue, anti-CD27 APC AlexaFluor750, anti-CD28 APC, anti-CD127 PE,anti-KLRG1 FITC, anti-CD45RO PE-Cy7, and anti-CD57Qdot655. (A) A lymphocyte gate was first set on all PBMCs(gray shape at lower left). (B) CD81 and CD41 cells werethen studied. (C) Histograms are shown of markers onCD41 cells (left) and CD81 cells (right). Each marker wasexpressed as a proportion of CD41 or CD81 cells. Thus,the final output was: proportion of CD8 cells expressingKLRG1, CD28, CD27, CD127, or CD57, and theproportion of CD4 cells expressing KLRG1, CD28, CD27,CD127, or CD45RO.

Figure 2. A scatter plot demonstrates the correlationbetween age and CD41 lymphocyte telomere length in cases(n 5 97) and controls (n 5 41). Regression lines are shownfor cases (gray) and controls (black). The slopes of theregression lines were not significantly different (P 5 0.95);the intercepts differed significantly (P < 0.001) (analysis ofcovariance test). This suggests that the cases and controlswere aging at a similar rate, but that the cases had a lower‘‘baseline age’’. When both cases and controls are plotted(not shown), Spearman q 5 20.6 (P < 0.0001).

580 GELSON ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

TABLE

2.TheRelationBetw

eenAgeandTelomere

Length

orLymphocyte

Phenotypein

Cases(n

597)andControls

(n5

41)

AgeVersu

sTelomereLen

gth

Ageve

rsusLymphocy

tePhen

otype

TelomereLen

gth

Versu

sLymphocy

tePhen

otype

Telomere

length

in

celltype

Case

sControls

Lymphocy

te

phen

otype

Case

sControls

Lymphocy

te

phen

otype

Case

sControls

Spea

rmanq

(95%

CI)

PValue

Spea

rmanq

(95%

CI)

PValue

Spea

rmanq

(95%

CI)

PValue

Spea

rmanq

(95%

CI)

PValue

Spea

rmanq

(95%

CI)

PValue

Spea

rman

q

(95%

CI)

PValue

Lymphocy

te�0

.60

(�0.72,�0

.46)

<0.0001

�0.61

(�0.78,�0

.38)

<0.0001

CD8þ

KLRG1þ

0.44

(0.26,0.59)

<0.0001

0.02

(�0.29,0.33)

0.90

CD8þ

KLRG1þ

�0.31

(�0.48,�0

.12)

0.002

0.02

(�0.29,0.33)

0.90

CD4þ

�0.60

(�0.71,�0

.45)

<0.0001

�0.60

(�0.77,�0

.36)

<0.0001

CD8þ

CD127þ

�0.44

(�0.59,�0

.27)

<0.0001

�0.38

(�0.62,�0

.08)

0.02

CD8þ

CD127þ

0.53

(0.38,0.66)

<0.0001

0.33

(0.02,0.58)

0.05

CD4þ

CD45ROþ

�0.61

(�0.72,�0

.47)

<0.0001

�0.63

(�0.78,�0

.39)

<0.0001

CD8þ

CD28þ

�0.30

(�0.48,�0

.11)

0.003

�0.27

(�0.54,0.04)

0.08

CD8þ

CD28þ

0.42

(0.24,0.57)

<0.0001

0.22

(�0.09,0.5)

0.16

CD4þ

CD45RO�

�0.60

(�0.71,�0

.46)

<0.0001

�0.58

(�0.76,�0

.34)

<0.0001

CD8þ

CD27þ

�0.42

(�0.57,�0

.24)

<0.0001

�0.37

(�0.61,�0

.07)

0.02

CD8þ

CD27þ

0.55

(0.4,0.68)

<0.0001

0.30

(0,0.56)

0.05

CD8þ

�0.61

(�0.72,�0

.47)

<0.0001

�0.69

(�0.82,�0

.48)

<0.0001

CD8þ

CD57þ

0.24

(0.04,0.42)

0.02

0.49

(0.21,0.69)

0.001

CD8þ

CD57þ

�0.29

(�0.47,�0

.1)

0.004

�0.18

(�0.46,0.14)

0.26

CD8þ

CD57þ

�0.41

(�0.56,�0

.23)

<0.0001

�0.45

(�0.67,�0

.17)

0.003

CD4þ

CD45ROþ

0.25

(0.05,0.42)

0.016

0.17

(�0.15,0.45)

0.29

CD4þ

CD45ROþ

�0.54

(�0.67,�0

.38)

<0.0001

�0.22

(�0.49,0.09)

0.17

CD8þ

CD57�

�0.63

(�0.73,�0

.49)

<0.0001

�0.66

(�0.8,�0

.44)

<0.0001

CD4þ

CD28þ

�0.37

(�0.53,�0

.19)

0.0002

�0.02

(�0.32,0.29)

0.91

CD4þ

CD28þ

0.49

(0.33,0.63)

<0.0001

0.19

(�0.12,0.47)

0.23

CD4þ

CD27þ

�0.41

(�0.57,�0

.23)

<0.0001

�0.07

(�0.37,0.24)

0.67

CD4þ

CD27þ

0.59

(0.44,0.71)

<0.0001

0.15

(�0.17,0.43)

0.37

CD4þ

CD127þ

�0.34

(�0.51,�0

.15)

0.001

�0.24

(�0.51,0.08)

0.17

CD4þ

CD127þ

0.52

(0.35,0.65)

<0.0001

0.38

(0.08,0.61)

0.02

CD4þ

KLRG1þ

0.34

(0.15,0.51)

0.001

�0.14

(�0.43,0.17)

0.38

CD4þ

KLRG1þ

�0.31

(�0.48,�0

.12)

0.002

�0.15

(�0.44,0.17)

0.36

Leftpanel:Telomerelength

issh

own

inCD4þ

and

CD8þ

Tce

lls.

CD45RO

and

CD57,mark

ersofmatu

rity,wereuse

dto

allow

ass

essm

entoftelomerelength

inmatu

reandim

matu

reCD4þ

andCD8þ

Tce

lls,

resp

ective

ly.A

significa

ntneg

ative

ass

ociation

betwee

ntelomerelength

andagewasfoundforalllymphocy

tesu

bse

tsstudied.Middle

panel:PeripheralT

cellswerech

ara

cterized

usingmark

ersass

ociated

with

immunese

nes

cence

.CD27,CD28,and

CD127

are

found

on

immatu

reT

cells,

KLRG1

on

matu

reT

cells,

CD57

on

matu

reCD8þ

cellsand

CD45RO

on

matu

reCD4þ

cells.

Significa

ntneg

ative

ass

ociationsbetwee

nageand

mark

ersof

cellularmatu

rity

(KLRG1,CD57,and

CD45RO)and

positive

ass

ociationsbetwee

nageand

mark

ersofim

matu

rity

(CD28,CD28,and

CD127)werefound

inca

ses.

Sim

ilarass

ociationswerefoundin

controls.Thes

eweresignifica

ntforCD8þ

CD127þ,

CD8þ

CD27þ,

andCD8þ

CD57þ.

Rightpanel:significa

ntpositive

ass

ociations

betwee

nlymphocy

tetelomerelength

andmark

ersofce

llularmatu

rity

(KLRG1,CD57andCD45RO)andneg

ative

ass

ociationsbetwee

ntelomerelength

andmark

ersof

immatu

rity

(CD28,CD28andCD127)werefoundin

case

s.Sim

ilarass

ociationswerenotfoundin

controls.

IMMUNE SENESCENCE IN LIVER TRANSPLANT RECIPIENTS 581

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

in liver transplant recipients than controls and alower proportion of lymphocytes with an immaturephenotype (CD4þ CD27þ, CD4þ CD28þ, CD4þCD127þ, CD8þ CD27þ, CD8þ CD28þ, and CD8þ

CD127þ) in liver transplant recipients than controls.These differences were significant for CD4þCD45ROþ (P ¼ 0.004), CD4þ CD27þ (P ¼ 0.003),CD4þ CD28þ (P ¼ 0.02), CD4þ CD127þ (P ¼ 0.04),

TABLE 3. A Comparison of Telomere Length Between Liver Transplant Recipients with Established

Grafts (n 5 97) and Healthy Controls (n 5 41)

Lymphocyte Population Group

Median Telomere

Length (mfi)

Difference

(Case-Control)

Effective Additional

Aging (Years)

Significance

(Mann-Whitney U Test)

All lymphocytes Case 115.12 �7.83 4.07 0.004Control 122.95

CD4þ cells Case 118.83 �6.12 3.18 0.011Control 124.96

CD4þ 45ROþ Case 109.53 �1.50 0.78 0.13Control 111.03

CD4þ 45RO� Case 119.17 �6.01 3.12 0.004Control 125.19

CD8þ cells Case 111.93 �8.42 4.37 0.011Control 120.35

CD8þ 57þ Case 98.13 �4.00 2.08 0.17Control 102.13

CD8þ 57� Case 118.04 �9.93 5.16 0.003Control 127.98

A Mann-Whitney test was used to compare cases with controls. For each study variable, the following data are given:median values, difference between median for case and control, stratification of the difference in terms of age (ascertainedfrom Fig. 2) and significance of difference (P < 0.05 taken as significant).

TABLE 4. A Comparison of Cell-Surface Senescence Markers Between Liver Transplant Recipients

with Established Grafts (n 5 97) and Healthy Controls (n 5 41)

Lymphocyte Population Group

Lymphocyte Phenotype (% Positive)

within CD8þ or CD4 Subset

Difference

(Case-Control)

Significance

(Mann-Whitney)

CD8þ CD27þ Case 47.3% �15.1% 0.034Control 62.4%

CD8þ CD28þ Case 44.0% �12.5% 0.045Control 56.5%

CD8þ CD127þ Case 45.1% �12.7% 0.006Control 57.8%

CD8þ CD57þ Case 25.8% 2.0% 0.35Control 23.7%

CD8þ KLRG1þ Case 46.9% 2.1% 0.81Control 44.8%

CD4þ CD27þ Case 83.7% �5.2% 0.003Control 88.9%

CD4þ CD28þ Case 89.3% �4.6% 0.02Control 93.8%

CD4þ CD127þ Case 76.3% �4.9% 0.04Control 81.3%

CD4þ CD45ROþ Case 69.2% 10.3% 0.004Control 58.9%

CD4þ KLRG1þ Case 9.6% 1.3% 0.092Control 8.3%

A Mann-Whitney test was used to compare cases with controls. For each study variable the following data are given:median values, difference between median for case and control and significance of difference (P < 0.05 taken assignificant). The difference between median values for cases and controls was negative for markers of immaturity andpositive for markers of maturity. Thus, the lymphocyte populations from cases are more mature than those from controls.

582 GELSON ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

CD8þ CD27þ (P ¼ 0.03), CD8þ CD28þ (P ¼ 0.045),and CD8þ CD127þ (P ¼ 0.006) and were not signifi-cant for CD4þ KLRG1þ (P ¼ 0.09), CD8þ KLRG1þ (P¼ 0.81), and CD8þ CD57þ (P ¼ 0.35) lymphocytes.

Taken together, these observations suggest thatliver transplant recipients have more lymphocyteswith a cell-surface phenotype of increased aging thancontrols and that immature lymphocytes from livertransplant recipients are biologically older than thosefrom controls. The fact that regression line slopes forcases and controls were similar, but that the inter-cepts were not, suggests that cases and controls wereaging at a similar rate, but that cases had a lower‘‘baseline age’’.

Relation Between Clinical Characteristics

and Features of Immune Senescence in Liver

Transplant Recipients

With each of the experimental variables as outcomes,simple linear regression was used to screen for associ-ated clinical and demographic characteristics (P <0.10). These characteristics were taken through intomultiple linear regressions, with experimental varia-bles as outcomes. Table 5 summarizes the clinicalcharacteristics that were associated independentlywith experimental variables.

Associations with Telomere Length in Liver

Transplant Recipients

Age was associated with telomere length in all lym-phocyte subsets. The average b coefficient for age andtelomere length over all lymphocyte subsets was�0.50 (standard deviation [SD] ¼ 0.10); thus, as aliver transplant recipient with an established graftgets 1 year older, their telomere length shortens by0.5 mfi.

HCC at transplantation was associated negativelywith telomere length in all lymphocyte subsets stud-ied, except for CD4þ CD45RO� T cells, where therewas no association. The average b coefficient for HCCat transplantation and telomere length was �12.40(SD ¼ 2.91); telomere length in patients with HCCwho underwent transplantation would be expected tobe 12.40 mfi shorter than those transplanted withoutan HCC (which equates to about 6.2 years of addi-tional aging).

Skin malignancy subsequent to transplantation wasassociated negatively with telomere length in lympho-cytes, CD4þ CD45ROþ T cells, CD8þ CD57þ T cells,and CD8þ CD57� T cells. The average b coefficientfor skin malignancy after transplantation and telo-mere length was �7.35 (SD ¼ 0.43); telomere lengthin patients with skin malignancy after transplantationwould be expected to be 7.35 mfi shorter than thosewithout skin malignancy after transplantation (whichequates to around 3.7 years of additional aging).

There were no consistent independent associationsbetween telomere length and: sex; pediatric or adultrecipient at engraftment; number of transplants;

underlying disease etiology; CMV status; having nor-mal liver biochemistry at sampling; calcineurin inhibi-tor or sirolimus-based immune suppression; numberof prescribed immune-suppressive agents at sam-pling; number of infective episodes; number of treatedepisodes of acute rejection; and history of malignancy(other than skin) and precancerous lesions. A relationto cardiovascular or cerebrovascular events could notbe addressed because the events were rare in thisseries.

Associations with Lymphocyte Cell-Surface

Phenotype in Liver Transplant Recipients

Age was associated consistently and independentlywith markers of lymphocyte surface phenotype. Theassociations were negative for markers of immaturity(CD28, CD27, and CD127) and positive for markers ofmaturity (CD45RO, CD57, and KLRG1). The b coeffi-cient varied in magnitude from 0.26 and 0.59, equat-ing to changes in cell surface expression of 0.26%and 0.59% per year.

Evidence of previous exposure to CMV was associ-ated independently and negatively with expression ofCD8þ CD127þ, CD8þ CD28þ, CD8þ CD27þ, CD4þCD28þ, and CD4þ CD27þ and associated positivelywith CD8þ CD57þ expression. The b coefficient wasgreater in CD8þ cells than CD4þ cells, accounting forup to 29.95% reduction in CD8þ CD27þ cells.

Normal liver biochemistry was associated independ-ently and positively with expression of CD8þ CD127þand CD8þ CD28þ and negatively with CD8þ CD57þexpression. The greatest b coefficient was for CD8þCD57þ cells at �16.81.

There were no consistent independent associationsbetween lymphocyte cell surface phenotype and: sex;pediatric or adult recipient at engraftment; number oftransplants; underlying disease etiology; the presenceof HCC at transplantation; having normal liver bio-chemistry at sampling; calcineurin inhibitor or siroli-mus-based immune suppression; number of pre-scribed immune suppressive agents at sampling;number of infective episodes; number of treated epi-sodes of acute rejection; and history of malignancyand precancerous lesions.

DISCUSSION

Lymphocyte telomere length was associated negativelywith age in all lymphocyte subsets studied, consistentwith the fact that telomeres shorten with age. Lym-phocyte markers of maturity (CD45RO, CD57, andKLRG1) were associated positively with age, whilelymphocyte markers of immaturity (CD28, CD27 andCD127) were associated negatively with age. Lympho-cyte telomere length was associated positively withmarkers of lymphocyte maturity and associated nega-tively with markers of immaturity in cases, consistentwith the concept of lymphocyte telomere loss withthe proliferation associated with antigen experience.2

IMMUNE SENESCENCE IN LIVER TRANSPLANT RECIPIENTS 583

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

Telomere length was also significantly shorter inmature than immature CD4þ and CD8þ T cells.

We believe this is the first study of immune senes-cence in the field of solid organ transplantation. Liver

transplant recipients with established allografts hadshorter lymphocyte telomeres (in both CD4þ andCD8þ T cell subsets) when compared with the healthycontrol group with similar demographics. Lymphocyte

TABLE 5. A Summary of Significant Clinical and Demographic Characteristics in Liver Transplant Recipients

Identified by Multiple Linear Regression Analysis

Study Variable

Significant Factors in

Multiple Linear Regressions Coefficient b

95% Confidence

Interval P Value

Telomere LengthLymphocyte telomere length Age �0.49 �0.71, �0.27 <0.0001

HCC at transplantation �12.55 �21.29, �3.8 0.005Malignancy (skin) �6.92 �13.51, �0.32 0.040

CD4þ telomere length Age �0.55 �0.79, �0.32 <0.0001HCC at transplantation �9.41 �18.68, �0.14 0.047

CD4þ CD45ROþ telomere length Age �0.48 �0.68, �0.27 <0.0001Malignancy (skin) �7.03 �13.32, �0.74 0.029

CD4þ CD45RO� telomere length Age �0.54 �0.77, �0.3 <0.0001HCC at transplantation �9.99 �19.21, �0.77 0.034

CD8þ telomere length Age �0.59 �0.84, �0.33 <0.0001HCC at transplantation �13.98 �24.83, �3.13 0.012

CD8þ CD57þ telomere length Age �0.28 �0.49, �0.08 0.007Number of transplants �9.64 �16.18, �3.1 0.004HCC at transplantation �17.26 �27.2, �7.33 0.001

Malignancy (skin) �7.69 �15.18, �0.21 0.044CD8þ CD57� telomere length Age �0.56 �0.82, �0.31 <0.0001

HCC at transplantation �11.22 �21.69, �0.76 0.036Malignancy (skin) �7.75 �15.23, �0.26 0.043

Lymphocyte Cell Surface PhenotypeCD8þ KLRG1þ Age 0.46 0.19, 0.74 0.001CD8þ CD127þ Age �0.54 �0.77, �0.32 <0.0001

CMV-positive �20.31 �28.14, �12.47 <0.0001Normal liver biochemistry 7.84 0.63, 15.06 0.034

CD8þ CD28þ Age �0.50 �0.75, �0.26 <0.0001Metabolic etiology �8.01 �15.3, �0.72 0.032

CMV-positive �25.30 �33.4, �17.2 <0.0001Normal liver biochemistry 8.84 1.3, 16.37 0.022

CD8þ CD27þ Age �0.59 �0.84, �0.34 <0.0001CMV-positive �29.95 �37.69, �22.21 <0.0001

CD8þ CD57þ HCC at transplantation 12.14 1.2, 23.07 0.030CMV-positive 22.86 15, 30.72 <0.0001

Normal liver biochemistry �16.81 �24.2, �9.42 <0.0001CD4þ CD45ROþ Age 0.26 0, 0.52 0.049

Viral 14.57 3.22, 25.92 0.012CD4þ CD28þ Age �0.26 �0.4, �0.11 0.001

CMV-positive �6.65 �12.22, �1.08 0.020CD4þ CD27þ Age �0.34 �0.52, �0.15 0.001

CMV-positive �8.79 �14.85, �2.73 0.005CD4þ CD127þ Age �0.19 �0.35, �0.04 0.014

Calcineurin inhibitor-based IS 5.51 0.24, 10.77 0.041CD4þ KLRG1þ Age 0.28 0.13, 0.44 0.001

Precancerous lesion 6.98 1.27, 12.69 0.017

Clinical and demographic characteristics that were entered into the multiple linear regressions for each experimentalvariable were age, sex, pediatric or adult recipient at engraftment, number of transplants, underlying disease etiology, thepresence of HCC at transplantation, anti-CMV antibody status at sampling, having normal liver biochemistry at sampling,calcineurin inhibitor or sirolimus-based immune suppression, number of prescribed immune suppressive agents atsampling, number of infective episodes, number of treated episodes of acute rejection, malignancy, and precancerouslesions. Each experimental variable with significant clinical or demographic associations is included. For each variable,significant (P < 0.05) and independently associated clinical and demographic characteristics are given with thecorresponding b coefficient and P value.

584 GELSON ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

telomere length was similar in liver transplant recipi-ents and healthy controls when analysis was restrictedto aged lymphocytes (CD4þ CD45ROþ and CD8þCD57þ T cells). In addition to having lymphocytes withshorter telomeres, liver transplant recipients also hadmore mature CD4þ T cells (CD45ROþ) and fewerimmature CD4þ and CD8þ T cells (CD27þ, CD28þ,and CD127þ). Taken together, these observationssuggest first that liver transplant recipients withestablished grafts have fewer immature lymphocytesand second that their immature lymphocytes aremore aged than healthy controls. This is likely todecrease the potency of immune responses to novelantigen.2

Several studies demonstrate expansion and telo-mere shortening of antigen-specific cells to chronic vi-ral antigens.3,4,35 However, antigen-naı̈ve lymphocyteswere also more aged in liver transplant recipientsthan healthy controls. It seems improbable that thisis due to chronic antigenic exposure, and alternativeexplanations must be sought; either non-antigen de-pendent lymphocyte turnover is responsible,1 or thereis a preferential antigen-dependent selection/survivalof naı̈ve cells with long telomeres. We have not pur-sued these hypotheses in this study.

The presence of HCC at engraftment and skinmalignancy after transplantation were associated in-dependently with shortened lymphocyte telomeres.HCC and skin malignancy have not been studied inrelation to lymphocyte or PBMC telomere length. How-ever, patients were found to have shorter telomeresthan age-matched controls in all groups in a study of92 patients with head and neck cancer, 135 withbladder cancer, 54 with lung cancer, and 32 withrenal cell carcinoma.36 The differences were 0.9 kb,0.2 kb, 0.4 kb, and 0.2 kb, respectively. Given thatPBMC telomere loss is between 15 and 55 bp/year inadults,37 this equates to between 6 and 26 years ofadditional aging (taking a mean value of 35 bp/year).In our patient group, the difference was smaller, atabout 6 years of additional aging for HCC and 4 yearsfor cancer that developed following liver transplanta-tion. However, the patients in the present series werequite different. None of the patients with HCC attransplantation was known to have recurrent HCC atsampling and all were at least 3 years from ‘‘definitivetreatment’’ by liver transplantation. Skin malignancyin our study included squamous cell carcinoma, basalcell carcinoma, and one melanoma. Only one patienthad an ‘‘active’’ skin malignancy at the time of lym-phocyte sampling (head and neck squamous cellcarcinoma).

Previous CMV infection (as determined by positivefor anti-CMV antibody at sampling) was found to beassociated with decreased expression of CD8þCD127þ, CD8þ CD28þ, CD8þ CD27þ, CD4þ CD28þ,and CD4þ CD27þ and increased expression of CD8þCD57þ, but not with telomere length. There is a recog-nized association of low CD28þ, CD27þ, and CD127þexpression and high CD57þ expression with previousCMV infection.4,38,39 KLRG1 may be up-regulated by

CMV infection,4 but this study did not confirm thatobservation.

Normal liver biochemistry was associated with animmature CD8þ phenotype (CD8þ CD127þ, CD8þCD28þ, and CD8þ CD57�). Normal liver biochemis-try may identify a group that includes those withoperational tolerance; expression of these markersmay, in some, therefore relate to clinical tolerance. Astudy of immune suppression withdrawal would berequired to investigate this further. Another possibilityis that patients with more naı̈ve lymphocytes mayrespond to the complications of transplantation moreeffectively than those who have an exhausted immunesystem, or conversely that those patients with healthygrafts have had fewer complications and thereforehave not ‘‘worn out’’ their immune system. A longitu-dinal study would be required to assess this phenom-enon further and is underway. This will allow anassessment of the utility of these biomarkers in riskstratification.

Outside the transplant setting, cardiac disease,hematological malignancy, cerebrovascular disease,and infections are known to be associated with short-ened telomere length in PBMCs.16,18-21 Such associa-tions were not found in this study. However, this se-ries was small and based largely on healthy patientsattending a liver transplant clinic for a routineappointment. A larger study is required to investigatethese associations further in the field of livertransplantation.

The hallmarks of immune senescence are replicativesenescence (characterized by short telomeres, alteredimmune function, and poor proliferative ability inlong-term cell culture), the expansion of lymphocytepopulations with a mature, antigen-experienced cell-surface phenotype and a diminished T cell receptorrepertoire.10,16 In vitro studies of proliferative capacityand immune function and assessment of T cell recep-tor repertoire would help to further develop our under-standing of immune senescence in liver transplantation.

The field of immune senescence consists mainly ofcross-sectional studies. Longitudinal studies that exam-ine telomere dynamics over time have not yet beenreported, although no doubt many are underway.40 Fac-tors thought to cause accelerated immune senescenceinclude chronic viral infections (CMV, EBV, and humanimmunodeficiency virus), smoking, obesity, stress, diabe-tes mellitus, and sarcoidosis.3,4,41-44 It is not possible toascertain whether these are causal associations.

This study has shown, for the first time, an associa-tion of solid organ transplantation with features ofimmune senescence (short lymphocyte telomeresand mature cell surface phenotype). Because immunesenescence may predispose to diseases that are com-mon in liver transplant recipients, namely cardiovas-cular disease, malignancy, and infections, this may bean important finding.

If there was a causal relationship between livertransplantation and immune senescence, one mightexpect a correlation between the number of trans-plants performed and features of immune senescence.

IMMUNE SENESCENCE IN LIVER TRANSPLANT RECIPIENTS 585

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

This correlation was found only for telomere length inCD8þ CD57þ cells; thus, such a relationship mightnot exist. However, only 13 of 97 recipients receivedmore than 1 graft which may be too few for analyticalpurposes; the hypothesis assumes that lymphocytetelomeres shorten but never lengthen, which may notbe correct.

Many factors may contribute to immune senescencein liver transplantation, including chronic liver dis-ease before and after engraftment, alloimmuneresponses, infection, and immune suppression. Wehave embarked upon a longitudinal study compar-ing features of immune senescence before and afterengraftment to investigate a causal relationship be-tween these factors and immune senescence.

REFERENCES

1. Goronzy JJ, Weyand CM. T cell development and recep-tor diversity during aging. Curr Opin Immunol 2005;17:468-475.

2. Weng NP. Aging of the immune system: how much canthe adaptive immune system adapt? Immunity 2006;24:495-499.

3. Fletcher JM, Vukmanovic-Stejic M, Dunne PJ, Birch KE,Cook JE, Jackson SE, et al. Cytomegalovirus-specific CD4þT cells in healthy carriers are continuously driven to replica-tive exhaustion. J Immunol 2005;175:8218- 8225.

4. Ibegbu CC, Xu YX, Harris W, Maggio D, Miller JD, Kour-tis AP. Expression of killer cell lectin-like receptor G1 onantigen-specific human CD8þ T lymphocytes duringactive, latent, and resolved infection and its relation withCD57. J Immunol 2005;174:6088-6094.

5. Pawelec G, Akbar A, Caruso C, Effros R, Grubeck-Loe-benstein B, Wikby A. Is immunosenescence infectious?Trends Immunol 2004;25:406-410.

6. Akbar AN, Soares MV, Plunkett FJ, Salmon M. Differen-tial regulation of CD8þ T cell senescence in mice andmen. Mech Ageing Dev 2000;121:69-76.

7. Pawelec G, Sansom D, Rehbein A, Adibzadeh M, Beck-man I. Decreased proliferative capacity and increasedsusceptibility to activation-induced cell death in late-passage human CD4þ TCR2þ cultured T cell clones.Exp Gerontol 1996;31:655-668.

8. Perillo NL, Naeim F, Walford RL, Effros RB. In vitro cellu-lar aging in T-lymphocyte cultures: analysis of DNA con-tent and cell size. Exp Cell Res 1993;207:131-135.

9. Ben-Porath I, Weinberg RA. The signals and pathwaysactivating cellular senescence. Int J Biochem Cell Biol2005;37:961-976.

10. Aw D, Silva AB, Palmer DB. Immunosenescence: emerg-ing challenges for an ageing population. Immunology2007;120:435-436.

11. Haynes L, Swain SL. Why aging T cells fail: implicationsfor vaccination. Immunity 2006;24:663-666.

12. Swain S, Clise-Dwyer K, Haynes L. Homeostasis and theage-associated defect of CD4 T cells. Semin Immunol2005;17:370-377.

13. Haynes L, Linton PJ, Eaton SM, Tonkonogy SL, SwainSL. Interleukin 2, but not other common gamma chain-binding cytokines, can reverse the defect in generation ofCD4 effector T cells from naive T cells of aged mice. JExp Med 1999;190:1013-1024.

14. Eaton SM, Burns EM, Kusser K, Randall TD, Haynes L.Age-related defects in CD4 T cell cognate helper functionlead to reductions in humoral responses. J Exp Med2004;200:1613-1622.

15. Blackburn EH. Switching and signaling at the telomere.Cell 2001;106:661-673.

16. Akbar AN, Beverley PC, Salmon M. Will telomere erosionlead to a loss of T-cell memory? Nat Rev Immunol 2004;4:737-743.

17. Hodes RJ, Hathcock KS, Weng NP. Telomeres in T and Bcells. Nat Rev Immunol 2002;2:699-706.

18. Cawthon RM, Smith KR, O’Brien E, Sivatchenko A, Ker-ber RA. Association between telomere length in bloodand mortality in people aged 60 years or older. Lancet2003;361:393-395.

19. Pommier JP, Gauthier L, Livartowski J, Galanaud P,Boue F, Dulioust A, et al. Immunosenescence in HIVpathogenesis. Virology 1997;231:148-154.

20. Ohyashiki JH, Sashida G, Tauchi T, Ohyashiki K. Telo-meres and telomerase in hematologic neoplasia. Onco-gene 2002;21:680-687.

21. Martin-Ruiz C, Dickinson HO, Keys B, Rowan E, KennyRA, Von Zglinicki T. Telomere length predicts poststrokemortality, dementia, and cognitive decline. Ann Neurol2006;60:174-180.

22. Johnston SD, Morris JK, Cramb R, Gunson BK, Neu-berger J. Cardiovascular morbidity and mortality afterorthotopic liver transplantation. Transplantation 2002;73:901-906.

23. Rabkin JM, Corless CL, Rosen HR, Olyaei AJ. Immuno-suppression impact on long-term cardiovascular compli-cations after liver transplantation. Am J Surg 2002;183:595-599.

24. Kizilisik AT, Grewal HP, Shokouh-Amiri MH, Vera SR,Hathaway DK, Gaber AO. Impact of long-term immuno-suppressive therapy on psychosocial and physical wellbeing in liver transplant recipients. Prog Transplant2003;13:278-283.

25. Sheiner PA, Magliocca JF, Bodian CA, Kim-Schluger L,Altaca G, Guarrera JV, et al. Long-term medical compli-cations in patients surviving > or ¼ 5 years after livertransplant. Transplantation 2000;69:781-789.

26. Kniepeiss D, Iberer F, Schaffellner S, Jakoby E, DullerD, Tscheliessnigg K. Dyslipidemia during sirolimus ther-apy in patients after liver transplantation. ClinTrans-plant 2004;18:642-646.

27. Neal DA, Tom BD, Luan J, Wareham NJ, Gimson AE,Delriviere LD, et al. Is there disparity between risk andincidence of cardiovascular disease after liver trans-plant? Transplantation 2004;77:93-99.

28. Canzanello VJ, Schwartz L, Taler SJ, Textor SC, WiesnerRH, Porayko MK, et al. Evolution of cardiovascular riskafter liver transplantation: a comparison of cyclosporineA and tacrolimus (FK506). Liver Transpl Surg 1997;3:1-9.

29. Aberg F, Pukkala E, Hockerstedt K, Sankila R, IsoniemiH. Risk of malignant neoplasms after liver transplanta-tion: a population-based study. Liver Transpl 2008;14:1428-1436.

30. Carey WD, Dumot JA, Pimentel RR, Barnes DS, HobbsRE, Henderson JM, et al. The prevalence of coronary ar-tery disease in liver transplant candidates over age 50.Transplantation 1995;59:859-864.

31. Singh N, Wagener MM, Gayowski T. Seasonal pattern ofearly mortality and infectious complications in livertransplant recipients. Liver Transpl 2001;7:884-889.

32. Voehringer D, Koschella M, Pircher H. Lack of prolifera-tive capacity of human effector and memory T cellsexpressing killer cell lectinlike receptor G1 (KLRG1).Blood 2002;100:3698-3702.

33. Hultdin M, Gronlund E, Norrback K, Eriksson-LindstromE, Just T, Roos G. Telomere analysis by fluorescence insitu hybridization and flow cytometry. Nucleic Acids Res1998;26:3651-3656.

586 GELSON ET AL.

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases

34. Baerlocher GM, Vulto I, de Jong G, Lansdorp PM.Flow cytometry and FISH to measure the averagelength of telomeres (flow FISH). Nat Protoc 2006;1:2365-2376.

35. Akbar AN, Fletcher JM. Memory T cell homeostasis andsenescence during aging. Curr Opin Immunol 2005;17:480-485.

36. Wu X, Amos CI, Zhu Y, Zhao H, Grossman BH, Shay JW,et al. Telomere dysfunction: a potential cancer predispo-sition factor. J Natl Cancer Inst 2003;95:1211-1218.

37. Unryn BM, Cook LS, Riabowol KT. Paternal age is posi-tively linked to telomere length of children. Aging Cell2005;4:97-101.

38. Koch S, Larbi A, Ozcelik D, Solana R, Gouttefangeas C,Attig S, et al. Cytomegalovirus infection: a driving forcein human T cell immunosenescence. Ann N Y Acad Sci2007;1114:23-35.

39. Khan N, Shariff N, Cobbold M, Bruton R, Ainsworth JA,Sinclair AJ, et al. Cytomegalovirus seropositivity drivesthe CD8 T cell repertoire toward greater clonality in

healthy elderly individuals. J Immunol 2002;169:1984-1992.

40. Meyer TD, Rietzschel ER, Buyzere ML, Criekinge WV,Bekaert S. Studying telomeres in a longitudinal popula-tion based study. Front Biosci 2008;13:2960-2970.

41. Valdes AM, Andrew T, Gardner JP, Kimura M, Oelsner E,Cherkas LF, et al. Obesity, cigarette smoking, and telo-mere length in women. Lancet 2005;366:662-664.

42. Damjanovic AK, Yang Y, Glaser R, Kiecolt-Glaser JK,Nguyen H, Laskowski B, et al. Accelerated telomere ero-sion is associated with a declining immune function ofcaregivers of Alzheimer’s disease patients. J Immunol2007;179:4249-4254.

43. Adaikalakoteswari A, Balasubramanyam M, RavikumarR, Deepa R, Mohan V. Association of telomere shorteningwith impaired glucose tolerance and diabetic macroangi-opathy. Atherosclerosis 2007;195:83-89.

44. Guan JZ, Maeda T, Sugano M, Oyama JI, Higuchi Y,Suzuki T, et al. An analysis of telomere length in sarcoid-osis. J Gerontol A Biol Sci Med Sci 2007;62:1199- 1203.

IMMUNE SENESCENCE IN LIVER TRANSPLANT RECIPIENTS 587

LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases