1
In-silico immune cell deconvolution of the airway proteomes of
infants with pneumonia reveals a link between reduced airway
eosinophils and an increased risk of mortality
Charles J Sande1†, Jacqueline M Waeni, James M Njunge1, Martin N Mutunga1, Elijah Gicheru1, Nelson
K Kibinge1, Agnes Gwela1
1. KEMRI-Wellcome Trust Research Programme, Kilifi, Kenya
† Corresponding author email: [email protected]
Keywords: Respiratory syncytial virus, proteome, pneumonia, deconvolution
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
2
Abstract
Rationale: Pneumonia is a leading cause of mortality in infants and young children. The
mechanisms that lead to mortality in these children are poorly understood. Studies of the
cellular immunology of the infant airway have traditionally been hindered by the limited
sample volumes available from the young, frail children who are admitted to hospital
with pneumonia. This is further compounded by the relatively low frequencies of certain
immune cell phenotypes that are thought to be critical to the clinical outcome of
pneumonia. To address this, we developed a novel in-silico deconvolution method for
inferring the frequencies of immune cell phenotypes in the airway of children with
different survival outcomes using proteomic data.
Methods: Using high-resolution mass spectrometry, we identified > 1,000 proteins
expressed in the airways of children who were admitted to hospital with clinical
pneumonia. 61 of these children were discharged from hospital and survived for more
than 365 days after discharge, while 19 died during admission. We used machine learning
by random forest to derive protein features that could be used to deconvolve immune
cell phenotypes in paediatric airway samples. We applied these phenotype-specific
signatures to identify airway-resident immune cell phenotypes that were differentially
enriched by survival status and validated the findings using a large retrospective
pneumonia cohort.
Main Results: We identified immune-cell phenotype classification features for 33
immune cell types. Eosinophil-associated features were significantly elevated in airway
samples obtained from pneumonia survivors and were downregulated in children who
subsequently died. To confirm these results, we analyzed clinical parameters from
>10,000 children who had been admitted with pneumonia in the previous 10 years. The
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
3
results of this retrospective analysis mirrored airway deconvolution data and showed that
survivors had significantly elevated eosinophils at admission compared to fatal
pneumonia.
Conclusions: Using a proteomics bioinformatics approach, we identify airway
eosinophils as a critical factor for pneumonia survival in infants and young children.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
4
Introduction
Pneumonia is a leading cause of paediatric mortality word-wide. A recent study on the
global burden of paediatric pneumonia conducted in seven countries found that viruses
account for about 61% of all paediatric pneumonia infections, while about 27% of
infection were attributed to bacterial pathogens, with RSV accounting for the largest
etiological fraction of paediatric pneumonia1. More than 90% of the deaths that occur
due to pneumonia in children under 5, occur in low resource settings, mainly due to the
lack of paediatric intensive care facilities2. Very young infants especially those with
comorbidities such as HIV and malnutrition have a poor survival prognosis following
pneumonia infection. HIV-infected infants who develop a pneumonia infection are up to
10 times more likely to die from the infection than non-HIV infected children3, while those
with malnutrition are more than 15 times more likely to die after admission4. The damage
to the lungs caused by severe pneumonia appears to persist even after discharge from
hospital, with recent estimates showing that post-discharge mortality in African children
previously admitted to hospital with pneumonia being eight times greater than those
discharged with other diagnoses5.
The difference in the immunological response to pneumonia between children who
succumb to infection and those who survive it is not clear. An improved understanding of
the immunobiology of this elevated post infection mortality risk will be crucial in
identifying prognostic biomarkers of poor outcome, which will be critical in guiding care
decisions in the first critical hours after admission. Most studies on the mechanisms of
severe pneumonia in infants have been done using blood samples6, and whist these
studies have provided significant insights into disease pathology, they might not fully
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
5
recapitulate the immune response to infection in the airway. In the case of RSV, there has
only been one study in the last 60 years that examined the lung samples of children who
died from pneumonia. Archived post-mortem lung tissue from three children who died in
the pre-intensivist era in the USA in the 1930s and 1940s was evaluated and airway
obstruction with necrotic cells, fibrin, mucus, and fluid identified as a prominent feature
in RSV lung infection7. Staining for immune cell populations was not technically feasible
in these long term archival samples.
Due to the paucity of mechanistic data on the immunological response to pneumonia in
the airway there is broad interest in understanding the dynamics of airway-resident
immune cells following pneumonia infection in infants. This effort has been hindered by
the unsuitability of routine airway sampling techniques for conventional cytometric
analyses. Samples from nasopharyngeal washings and naso- and oropharyngeal swabbing
are the most common methods of sampling the airways of sick children, and whilst ideal
for molecular diagnostics, they are less suitable for phenotyping of airway resident
immune cells using conventional cytometry techniques. Sampling of the airway by these
methods typically results in limited cell yields and the cells that are recovered are
generally highly enriched for granulocytes, complicating detailed characterisation of less
abundant phenotypes using traditional flow cytometry-based tools8. Recent analysis of
the cellular composition of upper airway by flow cytometry showed that the typical
abundances of critical effector cells like T Cells, B Cells, Mast cells, Dendritic cells and NK
cells to be less than 0.5% of all airway cells, while granulocytes were present at a median
frequency of >90%9.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
6
A potential way to address these shortcomings is to monitor changes in marker proteins
that are uniquely or predominantly expressed by specific immune cell phenotypes and to
use this information to infer the dynamics of the underlying cell types. We have recently
described a high-resolution mass-spectrometry-based proteomics approach for
characterizing the total proteome of the infant airway to a depth of more than 1,000
proteins10,11. The airway proteome characterized by this technique represents an unbiased
snapshot of the underlying cell populations and can be leveraged to infer changes in the
frequencies of the contributing cell phenotypes. Here, we describe an in-silico immune
cell phenotype deconvolution approach where we use protein markers that are uniquely
or predominantly expressed by specific immune cell subsets to infer the dynamics of
those phenotypes in airways of children who survived or died from pneumonia.
Phenotype-classification features were derived from a previously published data set
containing the individual proteomes of purified immune cell populations12 (deconvolution
data set) and these were then applied to airway proteome data from children with
different pneumonia outcomes (Figure 1 contains a graphical description of the study
design). Using this information, we identified an eosinophil-related protein signature that
was elevated in the airways of children who survived pneumonia but that was
downregulated in fatal pneumonia and in well controls. We subsequently validated this
association using a large retrospective pneumonia cohort of >10,000 children.
Results
To determine whether the deconvolution proteome data set contained sufficient
resolution to distinguish individual immune cell phenotypes, we used protein expression
levels from the data set to visualise phenotype segregation by nonmetric
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
7
multidimensional scaling (NMDS). The analysis showed that major immune cell
phenotypes including B-cells, T cells, natural killer (NK) cells, dendritic cells(DC),
monocytes (MO), basophils, eosinophils and neutrophils could be distinctly segregated
on the basis of differential protein expression (figure 2a). We then set out to identify
individual proteins that could be used to accurately distinguish major and sub immune
phenotypes (sub phenotypes defined as lower-level hierarchies of the major phenotypes
– e.g. plasmacytoid and myeloid DCs within the major DC phenotype) within the
deconvolution data set. Using the recursive feature elimination-paired random forest (RF)
algorithm, boruta13, we identified classification features for 33 major and sub immune
cell phenotypes (supplementary table 1). Since only a subset of these feature proteins are
likely to be expressed within the specific context of the inflamed airway, we set out to
determine which of the features identified by the model were also expressed in the
airways of children admitted with pneumonia. For some cell types such as
monocytes/macrophages and neutrophils, a substantial proportion (>30%) of the
phenotype classification features from the feature selection model were also expressed
in the airway, while for others (e.g. NK cells), a lower proportion of the classifiers were
identified in the airway (figures 2B, D, F & H). Supplementary table 1 contains a complete
list of all the identified features and their respective expression levels in the infant airway.
Next, we undertook a more detailed analysis of the phenotype classification features that
were identified by machine learning to determine whether they could be used to
recapitulate the phenotypes in an unsupervised prediction analysis. We reasoned that the
classification features of a particular phenotype would be broadly related to its functional
properties and that when they are subjected to an independent unsupervised enrichment
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
8
analysis, they would successfully predict the phenotypes from which they were initially
derived. Using the unsupervised enrichment platform, enrichR14,15, we found that 20 of
the 33 immune cell phenotypes were successfully predicted by their respective feature
proteins (examples in figures 2 C, E, G & I). Subsequent analysis was restricted to these
phenotypes. We visualised the performance of individual classification features by
comparing their expression levels between different immune cell phenotypes. An
example of this analysis is shown in figure 3A, where the monocyte classification feature,
SERPINB2, is shown as being expressed at substantially higher levels in activated
monocytes relative to all other cell phenotypes; the median expression level of SERPINB2
in classical activated monocytes was >104 fold higher in monocytes, relative to all other
phenotypes (figure 3B). Of the 64 classification features identified by machine learning
for activated monocytes, 17 were detected in the airway samples from children with
pneumonia (figure 3C).
To enhance the power of the identified features to resolve the constituent cell
phenotypes of a complex sample, we generated a phenotype classification profile, in
which all the classification features of a particular phenotype were aggregated and their
combined expression was plotted relative to all other phenotypes. An example of this
analysis is shown in figure 3D, where the combined expression of all monocyte-specific
features (plotted in red) was compared to the expression level of the same proteins in all
other phenotypes (plotted in gray). The results of this analysis showed that these proteins
were expressed at a significantly higher level in monocytes compared to all other
phenotypes (p<0.0001). Similar analysis for all other phenotypes is presented in
supplementary figure 1. We then used t-SNE dimensional reduction analysis to visualise
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
9
the segregation of major and sub phenotypes on the basis of the phenotype classification
features (figure 4A). The results of this analysis showed that the feature classifier proteins
could clearly distinguish most immune cell types; phenotypes like plasma cells
(B.plasma), pDC, mDC, eosinophils, basophils, neutrophils and different monocyte sub
phenotypes were clearly distinguishable from the rest of the phenotypes. Some sub-
phenotypes such as central memory CD4 T cells (T4.CM) and effector memory CD4 T cells
(T4.EM), could not be clearly disaggregated on the basis of feature expression alone.
We then applied the phenotype classification features identified by machine learning to
airway proteome data obtained from naso/oropharyngeal samples from children who
were admitted to hospital with clinical pneumonia and who either survived and were
discharged from hospital(N=61) or ultimately died (N=19) in the course of admission. We
also obtained similar data from age-matched well controls, who were sampled from the
community (N=10). Using a false discovery rate (FDR) of 5%, we identified >1,000
proteins in the airways of these children. We then compared the expression levels of
phenotype-specific feature proteins between these groups. We found that the expression
levels of eosinophil-specific classification features were significantly overexpressed in
children who survived infection, relative to non-survivors (P<0.0001) and controls (figure
4B). This was in contrast to neutrophils, where the expression level of phenotype-specific
proteins did not vary by survival status (figure 4B). The expression level of proteins
predominantly expressed by naïve CD8 T cells (T8.naive) were also significantly elevated
in pneumonia survivors compared to non-survivors (P<0.05). We observed no significant
differences for the remaining phenotypes (figure 4C). To validate these findings, we
reviewed the clinical records of >10,000 children who had been admitted to Kilifi County
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
10
Hospital Kenya over an approximate 10-year period with clinical pneumonia and for
whom haematological data (including eosinophil and neutrophil counts) had been
collected at admission. These data were stratified by survival status and the difference in
eosinophils and neutrophil counts between these groups was tested. We found that
eosinophil counts in children who survived pneumonia in the retrospective validation
cohort (N=10,859) were significantly greater than those of non-survivors (N=1,604,
p=0.0004 ) – Figure 5A, first panel. We then evaluated the risk of mortality in the first 10
days after admission in a subset of children who are at an elevated risk of pneumonia
mortality (undernourished children), in order to determine the association between
eosinophil counts and pneumonia survival in this group. Eosinophils counts in this group
were stratified into two groups on the basis of median counts. Undernourished children
who were admitted with pneumonia and with low (below median) eosinophil counts,
were significantly more likely to die in the first 10 days of admission, relative to those with
high eosinophil counts (Figure 5A, second panel). In contrast, no association with
mortality was observed when neutrophils were analysed using a similar strategy (Figure
5B).
Discussion
We report on a new method of deconvolving immune cell populations that are resident
in the airways of infants and children with different survival outcomes of severe
pneumonia. The study mucosal cellular immunity during very severe pneumonia has been
hindered by a number of important hurdles including low sample volumes as well as the
relatively low abundance of immune cell phenotypes that may be critical in directing the
clinical course of pneumonia. In this study, we addressed these problems by using a
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
11
machine learning approach to identify protein markers that could be used to deconvolve
mixed immune cells. We then applied these markers to airway proteome data obtained
from children with different survival outcomes of clinical pneumonia. Our results show
that protein markers associated with eosinophils are elevated in the airways of survivors
and are diminished in children who later died from infection. The airway levels of these
eosinophil markers were no different between children who died and well controls,
suggesting that the failure to mount an appropriate eosinophil response is a potential
mechanism of pneumonia-related mortality, especially in high risk populations such as
undernourished children. To validate the findings of the airway proteome analysis, we
reviewed the hospitalisation records of >10,000 children who had been admitted to
hospital in the previous 10 years with clinical pneumonia. The results of this retrospective
analysis confirmed the observations made from analysis of the airway proteome and
showed that children who died from pneumonia, had significantly lower blood eosinophil
counts relative to those who survived. In contrast, neutrophil levels were not different
between survivors and non survivors in both the airway proteome and in the retrospective
validation cohort.
Previous studies have shown that eosinophils are activated in the airway shortly after
pneumonia infection appear to contribute significantly to airway recovery. An increase in
the expression of eosinophil-related markers in children with severe pneumonia has been
associated with a reduced requirement for supplemental oxygen16, suggesting that these
cells are a critical component in the host’s response to infecting pathogens in the airway.
Airway-resident eosinophils have not been extensively characterised in previous
pneumonia studies, although proteins expressed predominantly by eosinophils including
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
12
leukotriene C4, eosinophil-derived neurotoxin (RNASE2) and eosinophil cationic protein
(ECP) are expressed at high levels during viral pneumonia17-19. Taken together with the
results of the current study, there appears to be a clear role for eosinophils in the response
to pneumonia infection and that the failure to sufficiently recruit them to the airway may
be a significant factor in mortality. A potential limitation of our study is that it was carried
out using samples collected from the upper airway and not the lung. While lung samples
would have undoubtedly been more informative of responses at the site of disease, the
collection of these samples is a highly invasive process and exposes children to
substantial additional risk without providing additional diagnostic value above
nasopharyngeal or peripheral blood sampling. As a result, these samples are generally
only available in settings with paediatric intensive care facilities, and even here, they are
generally collected in children with atypically severe disease.
In addition to eosinophils, previous studies have shown infections that cause severe
pneumonia such as RSV, trigger a strong cellular response, characterised by the influx of
innate immune cells20,21. The initial response to infection is characterised by the airway
recruitment of neutrophils, which express markers such as CD11b (ITGAM) and neutrophil
granule proteins such as neutrophil elastase (ELANE)22,23. Other innate immune cells such
as NK cells which expressed granzyme B are recruited in the lower airway and can be
detected in in the lungs of mechanically-ventilated children with very severe
pneumonia24,25. In addition to these cells, both myeloid and plasmacytoid dendritic cells
are typically recruited into the airways of children with pneumonia in the early stages of
infection24,26 and exhibit an activated proinflammatory phenotype24. Adaptive immune
cells including CD4+ and CD8+ T cells are present in airway samples for children with viral
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
13
pneumonia20,27. During infection, airway frequencies of granzyme B-secreting, activated
CD8 T cells is greater in the airways of children with severe viral pneumonia25. In the
present study, we found that protein markers of different phenotypes of CD8 T cells were
significantly elevated in the airways of children who survived pneumonia relative to those
who died or well controls. Although data was not available for validation of these
phenotypes in the retrospective cohort, the data suggests a role for these cells in
pneumonia survival. Previous studies in mechanically ventilated children showed that the
frequency of lung-associated T cells increased as children recovered from infection,
suggesting that these cells are an important component of effective local immunity
against pneumonia and that deficits in the airways of children with pneumonia is a
significant risk factor for mortality. In summary, the present data identifies a new
approach for characterising the cellular immune response to pneumonia in the airway and
identified critical immune populations that appear to be critical for survival. Future
studies should aim to replicate these findings in samples collected from the lower
respiratory tract.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
14
Methods
Study site and population
This study recruited 80 infants and children who were admitted to Kilifi County Hospital
with clinical pneumonia, defined using World Health Organisation syndromic criteria.
Nasal samples were collected from each child for proteomic analysis. The microbial
etiology of pneumonia was determined using both blood cultures and using a 15-target
multiplex PCR panel for the detection of respiratory syncytial virus (RSV - A & B),
rhinovirus, parainfluenza virus (1, 2, 3 & 4) adenovirus, influenza (A, B & C), coronavirus
(OC43 & e229), human metapneumovirus and Mycoplasma pneumoniae. Children with
clinical pneumonia signs and a positive diagnostic result from any of these tests were
included in the analysis. Children were stratified into the survival group if they were alive
for at least 365 days after discharge(n=61) while the mortality group comprised of
children who died within 72 hours of admission (n=19). In addition to these groups, we
recruited 10 age-matched well controls as a comparator group. Written informed consent
was sought from the parents and legal guardians of all children who were sampled in this
study. Ethical approval for the conduct of this study was granted by the Kenya Medical
Research Institute’s Scientific and ethical research unit (SERU). All study procedures were
conducted in accordance with Good Clinical Laboratory Practise (GCLP) standards.
Analysis of airway proteomes using mass spectrometry
Naso- and oropharyngeal swab samples were centrifuged at 17,000xg for 10 mins at 4°C
to obtain cell pellets which were washed once using PBS and lysed by bead-vortexing for
10 minutes in cell lysis buffer (RLT, Qiagen, Germany). Proteins (as well as DNA and RNA)
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
15
were then extracted from the lysate using the AllPrep DNA/RNA/Protein Mini Kit
(Qiagen, Germany) following manufacturers instructions. The concentration of total
protein obtained was determined using the Bradford assay (Bio-Rad, USA). Thirty
micrograms (30µg) of total protein from each sample was then reduced with 10mM
tris(2-carboxyethyl)phosphine (TCEP, Sigma-Aldrich, USA) at 55°C for 1h and
subsequently alkylated with 18mM IAA (Sigma-Aldrich, USA) for 30 minutes at room
temperature, while keeping the reaction protected from light. Proteins were precipitated
overnight at -20°C with six volumes of pre-chilled (-20°C) acetone (Sigma-Aldrich, USA).
The samples were centrifuged at 8,000xg for 10 minutes at 4°C to obtain the protein
pellets and supernatants were discarded. The protein pellet was resuspended in 100µl of
50mM Triethylammonium bicarbonate (TEAB, Sigma-Aldrich, USA). Trypsin (Sigma-
Aldrich, USA) was added to the protein samples at a trypsin-protein sample ratio of 1:10
and protein digestion was allowed to proceed overnight at 37°C with shaking. The peptide
samples were randomly assigned to 10 individual batches: each containing nine patient
samples and one pooled control sample. The pooled control sample consisted of a pool
of peptides from all patient samples. The peptide samples derived from individual
patients were then individually labelled using the TMT10plex mass tag kit (Thermo
scientific, USA) according to manufacturer’s instructions, with one isobaric tag being
exclusively used to label the pooled control sample. The labelled peptides for each 10plex
were subsequently combined to generate 10 individual pools. The labelled peptide pools
were desalted using P10 C18 pipette ZipTips (Millipore, USA) according to the
manufacturer’s instructions. Eluted peptides were dried in a Speedvac concentrator
(Thermo Scientific, USA). Peptides (8 μl) were loaded using a Dionex Ultimate 3000 nano-
flow ultra-high-pressure liquid chromatography system (Thermo Scientific, USA) on to a
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
16
75µm x 2 cm C18 trap column (Thermo Scientific, USA) and separated on a 75µm x 50 cm
C18 reverse-phase analytical column (Thermo Scientific) at heated at 40°C. For LFQ
protein quantification; elution was carried out with mobile phase B (80% acetonitrile
with 0.1% formic acid) gradient (4 to 30%) over 310 min at a flow rate of 0.25 μl/min.
Each LC run was finished by washout with 98% B for 10 min and re-equilibration in 2% B
for 30 min. Five blanks of 40 min each were run on the column between each injection
comprising of two wash cycles with 90% B and an equilibration phase of 15 min to avoid
sample carryover. Peptides were measured using a Q Exactive Orbitrap mass
spectrometer (Thermo Scientific, USA) coupled to the chromatography system via a
nano-electrospray ion source (Thermo Scientific). On the Q Exactive , the ms^1 settings
for peptides were: Resolution, 70000; AGC target, 3e6; maximum IT, 120 ms; scan range,
400-1800 m/z; while the ms^2 settings for fragmentation spectra of peptides were:
Resolution, 17000 (35000 for labelled peptides); AGC target, 5e4; maximum IT, 120 ms ;
isolation window, 1.6 m/z. MS data were acquired by data dependent acquisition where
the top 12 (15 for labelled peptides) most intense precursor ions in positive mode were
selected for ms^2 Higher-energy C-trap dissociation fragmentation which were
subsequently excluded for the next 45 s following fragmentation event. Charge exclusion
was set to ignore peptide spectrum matches that were unassigned, singly charged, and
those with ≥+8 charges. Raw mass spectrometer files were analysed by MaxQuant
software version 1.6.0.1. by searching against the human Uniprot FASTA database
(downloaded February 2014) using the Andromeda search engine..
Airway samples were centrifuged at 17,000xg for 10 mins at 4°C to obtain cell pellets
which were washed once using PBS and lysed by bead-vortexing for 10 minutes in cell lysis
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
17
buffer (RLT, Qiagen, Germany). Proteins (as well as DNA and RNA) were then extracted
from the lysate using the AllPrep DNA/RNA/Protein Mini Kit (Qiagen, Germany)
following manufacturers instructions. The concentration of total protein obtained was
determined using the Bradford assay (Bio-Rad, USA). Thirty micrograms (30µg) of total
protein from each sample was then reduced with 10mM tris(2-carboxyethyl)phosphine
(TCEP, Sigma-Aldrich, USA) at 55°C for 1h and subsequently alkylated with 18mM IAA
(Sigma-Aldrich, USA) for 30 minutes at room temperature, while keeping the reaction
protected from light. Proteins were precipitated overnight at -20°C with six volumes of
pre-chilled (-20°C) acetone (Sigma-Aldrich, USA). The samples were centrifuged at
8,000xg for 10 minutes at 4°C to obtain the protein pellets and supernatants were
discarded. The protein pellet was resuspended in 100µl of 50mM Triethylammonium
bicarbonate (TEAB, Sigma-Aldrich, USA). Trypsin (Sigma-Aldrich, USA) was added to the
protein samples at a trypsin-protein sample ratio of 1:10 and protein digestion was
allowed to proceed overnight at 37°C with shaking. The peptide samples were randomly
assigned to 10 individual batches: each containing nine patient samples and one pooled
control sample. The pooled control sample consisted of a pool of peptides from all
patient samples. The peptide samples derived from individual patients were then
individually labelled using the TMT10plex mass tag kit (Thermo scientific, USA) according
to manufacturer’s instructions, with one isobaric tag being exclusively used to label the
pooled control sample. The labelled peptides for each 10plex were subsequently
combined to generate 10 individual pools. The labelled peptide pools were desalted using
P10 C18 pipette ZipTips (Millipore, USA) according to the manufacturer’s instructions.
Eluted peptides were dried in a Speedvac concentrator (Thermo Scientific, USA). Peptides
(8 μl) were loaded using a Dionex Ultimate 3000 nano-flow ultra-high-pressure liquid
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
18
chromatography system (Thermo Scientific, USA) on to a 75µm x 2 cm C18 trap column
(Thermo Scientific, USA) and separated on a 75µm x 50 cm C18 reverse-phase analytical
column (Thermo Scientific) at heated at 40°C. For LFQ protein quantification; elution was
carried out with mobile phase B (80% acetonitrile with 0.1% formic acid) gradient (4 to
30%) over 310 min at a flow rate of 0.25 μl/min. Each LC run was finished by washout
with 98% B for 10 min and re-equilibration in 2% B for 30 min. Five blanks of 40 min each
were run on the column between each injection comprising of two wash cycles with 90%
B and an equilibration phase of 15 min to avoid sample carryover. Peptides were measured
using a Q Exactive Orbitrap mass spectrometer (Thermo Scientific, USA) coupled to the
chromatography system via a nano-electrospray ion source (Thermo Scientific). On the Q
Exactive , the ms^1 settings for peptides were: Resolution, 70000; AGC target, 3e6;
maximum IT, 120 ms; scan range, 400-1800 m/z; while the ms^2 settings for
fragmentation spectra of peptides were: Resolution, 17000 (35000 for labelled peptides);
AGC target, 5e4; maximum IT, 120 ms ; isolation window, 1.6 m/z. MS data were acquired
by data dependent acquisition where the top 12 (15 for labelled peptides) most intense
precursor ions in positive mode were selected for ms^2 Higher-energy C-trap dissociation
fragmentation which were subsequently excluded for the next 45 s following
fragmentation event. Charge exclusion was set to ignore peptide spectrum matches that
were unassigned, singly charged, and those with ≥+8 charges. Raw mass spectrometer
files were analysed by MaxQuant software version 1.6.0.1. by searching against the human
Uniprot FASTA database (downloaded February 2014) using the Andromeda search
engine.
Analysis of airway-resident immune cells using flow cytometry
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
19
1ml of nasopharyngeal and oropharyngeal swab samples obtained from children was
centrifuged at 17,000xg for 7 minutes, after which 800�l of the supernatant was removed
and discarded. The remaining 200µl were split into two aliquots of 100µl each. The first
aliquot was used for neutrophil phenotyping assays and the other was used for neutrophil
phagocytosis assays. 20µl of a pre-constituted cocktail of the following antibodies (from
ThermoFisher) was used to label both aliquots – CD45, CD16, CD14, CD3, CD19, HLA-DR,
CD66b, CD11b and a Live-dead marker. With the exception of the live/dead marker, all
other antibodies were diluted 1:100 in FACS buffer. The live/dead marker was prepared at
a 1:1000 dilution in FACS buffer. For the phagocytosis assay tube, 20�l of opsonised
Escherichia coli (E.coli) was added to the tube (pHrodo Red E. coli BioParticles;
ThermoFisher). The bacteria was initially prepared by mixing the E.coli strain with new-
born calf sera followed by a 30-minute incubation at 37oC. After this step, both tubes
were incubated at 37oC for 35 minutes. After the incubation, 20�l of a live-dead marker
was added to each tube and incubated for 10 minutes at 37oC. The reaction was stopped
by adding 500�l of 1X RBC lysis buffer followed by a 5-minute incubation. Cells in each
tube were then spun down at 2,700xg for 1 minute and the supernatant discarded. Cells
were then washed twice with FACS buffer, after which 350�l of FACS flow was added.
Cells were then analysed immediately on a BD LSR Fortessa instrument. The following
gating strategy was used to detect airway-resident neutrophils: Debris were excluded on
the basis of their forward (FSC-A) and side (SSC-A) scatter characteristics, doublets were
excluded using FSC-A versus FSC-H and dead cells were excluded using the live-dead
marker. Data was analysed using FlowJo software.
Statistical data analysis
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
20
Data were analyzed in R. deconvolution analysis by random forest classification was done
using the Boruta package, with a maximum of 300 iterations. Protein classifiers that
whose mean expression in the test phenotype was significantly greater than alternative
phenotypes were taken forward for further analysis. T-SNE analysis was carried out using
the Rtsne, with the iterations parameter set to a maximum of 300 and a perplexity value
of 30. Analysis was carried out in two dimensional space. Cell type enrichment analysis
was done one the enrichR platform. The input search term used for enrichment analysis
was the RF protein classifier lists derived from random forest classification. All pairwise
comparisons between the expression level of RF classifiers between phenotypes and
frequency of immune cell types between survival states was done using t-tests on log10
normalised data. The deconvolution data set that was used to identify phenotype-specific
protein classifiers was obtained from a previously published paper by Rieckmann et al.12
Data availability
The proteomics data reported in this paper are available at the ProteomeXchange
Consortium database (Accession numbers: PXD009403) .
Acknowledgements
This study was supported by fellowship funding to CJS from the Wellcome Trust
(WT105882MA). The funder played no role in the conceptualization, design, data
collection, analysis, decision to publish, or preparation of the manuscript
Conflicts of interest
The authors declare that they have no conflicting interests
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
21
Author contributions
CJS designed the study
CJS, JMN, MNM, ETG, AG conducted the experiments
CJS,NKK analysed the data
CJS wrote the manuscript
All authors reviewed and approved the manuscript
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
22
Figures
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
23
Figure 1: Analysis of the profile of airway-resident immune cells using flow cytometry and in-silico deconvolution. (A) To characterize the
phenotype diversity of airway-resident immune cells, fresh naso- and oropharyngeal flocked swabs were obtained from children admitted
to hospital with pneumonia and processed immediately by flow cytometry. Here the gating strategy is shown, where live, singlet cells were
initially gated on CD45 expression, followed by fluorescent antibody staining for different cell surface markers. CD3+ and CD19- cells were
gated as T-cells, CD66-, CD14+ and HLA-DR- cells were gated as macrophages/monocytes while neutrophils were gated on the basis of
CD66 and CD16 double expression. The functional activity of neutrophils was characterized by co-culturing the cells with E.coli particles
that were labelled with pHRhodo. Neutrophils that contained phagocytosed particles are shown in green in the overlay dot plot. The
bottom represented FSC and SSC plots of samples obtained from 6 different infants and demonstrates a high level of diversity in the
immune cell populations present in the airways of children with pneumonia. (B) In silico deconvolution analysis was based on high
resolution mass spectrometry data obtained from the airway. Cells were obtained from the naso and oropharyngeal sites using separate
swabs, which were both eluted in a common transport media. As shown in A above, these samples contained a broad diversity of immune
cells including T-cells, monocytes/macrophages and neutrophils. These cells were isolated from the sample by centrifugation and
processed for mass spectrometry analysis using a standard protocol (see methods). (C) In silico phenotype deconvolution was conducted
initially by identifying phenotype specific protein features in a previously-published deconvolution data set using random forest feature
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
24
selection. These features were then applied on airway proteome data set obtained from children admitted to hospital with severe
pneumonia, and used to construct a detailed map of airway immunophenotypes and their associations with clinical outcomes.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
25
NEUTROPHIL
PERIPHERAL BLOOD
GRANULOCYTE
CORD BLOOD
SPLEEN (BULK TISSUE)
0 10 20
B
Basophil
DC
Eosinophil
MO
Neutrophil
NKT
−0.2
−0.1
0.0
0.1
−0.6 −0.4 −0.2 0.0 0.2NMDS1
NMDS
2
12.5%
37.5%67.5%
87.5%
0%
25%
50%
75%
NK cells
CAPN
5
CD48
RRAS
2GNLY
ORAI2
B4GA
LT4
PNOC
KLRD1
GOLIM4
FCGR3A
GOLGA8R
GZMB
MATK
TYROBP
NCR1
SH2D1B
NCR3
TPST2
CST7
EOMES
CD8A
ABCB1GZMA
NCAM1PRF1
IL2RBGZMHCD247ENPP1AKAP5
KLRC1
AOAH
TXKKIR3DL1
KIR3DL2
KIR2DS2
AFM
RGS3
ENTP
D1
ITGA1
CTSW
NCALD
GNAI1
SATB1
NFATC2IL18R1RUNX3DH
CR24
AGKMC
TP2KDE
LC1ERM
P1GPR1
14SAMD3
HAVCR2
DPF3
NEIL1
ABHD17A
CD244
LAT2
OSBPL5
IRF2BPL
APOBEC3G
GOPC
TOR4A
TBX21
MAFFSIGLEC7
HDGFRP3TNFRSF18FCGR3B
AKR1C3ANXA1
A1BG
SYK
12.5%
37.5%67.5%
87.5%
0%
25%
50%
75%
Neutrophils
POM121C
ZFP36L2
THOC1
RMND
1ATXN
3AS
AP1
ANO9
ZMYM
3RPA3OS
CLDN
D1MAR
K2TP53I11
GPSM
3CTAG
E5CALM1
PCBD2
LUC7L3
MSL1
SNAP23
TOR1A
ADAM10
NDC80
SCAMP1
SIGLEC5
SLC16A3
AQP9
RANBP6
B3GAT3
RTN3
LRATSVILGGPS1
ABL1
GLAMME
MMP7
SLC2A3
TIMP2
CBLRPS6KB
1
CXCR2
CFPPKLRWASKIR2D
L3RGS19HDGFGNG2SPCS3NTHL1CDK17SPTBN1PLAURCD47SIGLEC14APOBRSTX4MAP3K1ROCK1FAM53BNBR1MESDC2PPP2R5AVPS72TMEM132A
MTHFSDLRRFIP1C5orf22FGD3SMEK2TMEM63BHGSNATADNP2PPP1R18RICTORTAOK1GOLGA7ARHGAP30
YTHDF3RBM23
NT5DC3
HOOK3
SFR1
GPR97
RAB43
TMC8
SLC44A2
LMTK2
ORMDL3
SLC15A4
NFATC2IP
NEK7SH3TC1
BCL7CMITD1BRI3BPRAD54LDDX12PMCUR1
ARL8ACHM
P6JAKM
IP1TEFMPHF12CSF3RCIB1CRACR2A
DCTN
5RB
CK1
EDEM
3TM
6SF1
CYSTM1
DNAJC5
DEF6
CCDC
134
ZFYV
E28
POLM
HPS4
DROS
HA
AGPAT4
ARL8B
NDE1
SACS
TMOD
2
ZNF282
ADD3
LAMTOR3
NAGPA
AMFRINTS6
RAB26
PACSIN2
SMC3PL
AAATP8A1
ZNF281AK6
TSC22D4
UBIAD
1LRR
FIP2CD
177FCGR3BCEA
CAM8LSP1GNB
2L1UQCRBANXA6CASTMGAMCR1CAPN1RAP1B
OLR1PDIA3PFN1CAMP
CRISP3TNC
U2AF2SDCBPARPC2CPNE3
PGLYRP1RP2DYSFFLOT1VNN1RAB3D
HPORM1GNAI2ARG1
S100A9EPHX1P4HBMGPGNAI3
HMGB1RALB
MMP9
ITGB5
AZU1
FCAR
PRTN3
STOMGRN
ICAM3MYH9
TALDO1GYG1PGDIST1
RAB2ATMSB4XLCN2GBE1
CYP4F3LMAN2
STIM1
SPTAN1MAPRE1
MAPK14SYTL1MICU2MCU
NCSTNELM
O1
ERO1L
VAT1CHP1SLK
TMX1
APMAP
12.5%
37.5%67.5%
87.5%
0%
25%
50%
75%
DC
C12orf45
DOCK
1CY
P51A1
SRGAP
2DC
KCLAS
P2SN
CAMAP
1APLD4
CDK2
VEZF1
IRF7
HIP1
TRAFD1
IFIT3
CLIC2
TBC1D4
LY75
PFKFB2
RPS6KA4
FCHSD2
WDR47
PTBP3
ADAFBP
1
MRC1
PTPRE
NFKBIA
ITGB8
CD1CEPHA
2
PRKAR2B
GPR183
S100A3CDKN1AETV3ETV6ECE1GCLCDUSP3AKR1D1MAP2K6GABARAPL2DAB2PLCB3IRF8ZFP36L1SCRN1PAK1TARBP1IRF4CCDC6CSRP2CALCRL
DUSP5CCDC88A
ZNF672
CMPK2
SPECC1
ZNF618
TNS3
TIFAB
WDFY4
SMIM5
RUFY3
SGK223
ZDHHC17
CCDC50
PM20D2
LILRB2
KRTCAP2
OTUD6BZNF579ZNRF2LILRB4PIK3R5NR4A3
NSMAF
ISG20
UBE2W
DHX58
PPP1R14B
ACY3
DHTKD1
NGLY1
EPSTI1
UBE2E2
LGMN
REPIN1
PACSIN1
IFIH1
MTMR
12
PARP12
FAM49A
BCL11AALG2
TMEM206
CXorf21
TMEM9BPD
LIM7HE
BP1
TCP11L
1TBC
1D13SPH
K1TLR7ZNFX1EIF2A
K4DNAJB4
FEZ2HERC5PALD1USP18SH2B3SCML2CD2APST14
KIAA1598PON2
PLXNB2CTNND1ATP1B1ISG15PYGLPLEK
GSTM1IFIT2IFIT1CKB
LAMP2
PLS3
NFKB1
MX2
COMT
LAP3
CTNNA1TAGLN2RPL3
ALDH7A1IDH3A
RPL27MYO1E
TCIRG1ALCAM
DPYSL2TRIM
22CYP2S1
CORO
1BFAM
213ATUBB6
H2AFY2NDRG
2SAM
HD1HYO
U1
12.5%
37.5%67.5%
87.5%
0%
25%
50%
75%
Monocytes
SULT1A4
TSPO
CD36
PICA
LMBACH
1DS
CR3
SPTLC1
SLC31A1
TLR2
STX11
MFN2
SGPL1
F13A1
CTSL
GAA
FCGR1A
MGAT1
LRPAP1
SULT1A1
CYP27A1
LRP1
DPYD
NAIP
ATP6AP1CD300EhCG_14925SLC27A3ARHGEF10LLAMTOR1NCEH1AGTRAPSYNE3C6orf120
LDHDRHOT1
GHDC
LRRC25
PNKD
ATP6V0D2
FAM45A
CDC6
NPL
SLC12A9
TBC1D2CLPB
PLEK
HA4
TLR8
HEBP
1
STAB1
PYCA
RDIRAK3
COMM
D10
SFXN3
TM9SF4
ATP6V1G1
IDH1ND
UFB8ALD
H1A1S10
0A6ANX
A2CD1
4CYC1
ANXA5LGALS1HMOX1
IFI30
POR
TYMP
RAB5A
ATP6V1B2
UQCRC2
CES1
S100A4
BLVRB
UQCRC1
APOBEC3A
HADHA
CRATGPD2
UQCRFS1HNMT
SERPINB8ALDH3A2PLO
D1CKAP4
AHNAKNAPRT
COLG
ALT1
CD14+_Monocytes
CD33+_Myeloid
WholeBlood
CD71+_EarlyErythroid
Liver
0
100
200
300
400
500
CD56+_NKCells
Adrenalgland
AdrenalCortex
ParietalLobe
Amygdala
0
200
400
600
BDCA4+_DendriticCells
CD33+_Myeloid
WholeBlood
CD14+_Monocytes
Thymus0
200
400
600
Enrichment score Enrichment score Enrichment scoreEnrichment score
(A)
(B)
(C)
(D)
(E)
(F)
(G)
(H)
(I)
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
26
Figure 2: Use of proteomics data to segregate immune cell phenotypes. (A) A nonmetric multidimensional scaling (NMDS) plot was used
to visualize the separation of immune cell phenotypes at the major phenotype level based on protein expression data from the
deconvolution data set. We observed that expression profile of certain cell types including neutrophils, monocytes and basophils, resulted
in a clear separation from other immune cell types. To identify specific proteins that could be used to distinguish phenotypes, we used
random forest feature selection. Using this approach, we identified protein features for 33 phenotypes that could be used to disaggregate
individual phenotypes from a complex mixture (Supplementary table 1). We then examined airway proteome data from children with
pneumonia to determine whether any of the proteins identified were detectable in the airway. An example of the results is shown in (B,D,F
& H), where the phenotype classification features for NK cells, DCs, Neutrophils and Monocytes are shown labelled on the circular tile
plot. The green tiles are the phenotype-specific feature proteins that were identified in the random forest model and that were also
detected in the airway. We used enrichment analysis to determine whether the phenotype-classifier proteins identified by the random
forest model for specific phenotypes could be used to recapitulate those the phenotypes in an unsupervised phenotype prediction analysis.
The clustergrams shown in (C,E,G & I) depict a co-expression matrix of feature proteins and their known expression in different cell types.
For 20 of the 33 phenotypes, (including the four examples shown here) the random forest phenotype classification, correctly predicted the
original phenotype in the unsupervised prediction analysis. Each row on the clustergram represents a feature protein, while each colored
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
27
box represents the known expression profile in a particular cell types. Horizonal panels at the bottom indicate enrichment scores from the
unsupervised analysis.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
28
SERP
INB2
●●●● ●●●● ●● ●● ●●●● ●●●● ●●●●●
●●●
● ●●● ●● ●●
●
●
●
●●●● ●●●● ●●●● ●● ● ●●●● ●●●● ●●● ●●●● ●●● ●●●● ●●● ●●●● ●●●● ●●●● ●●●● ●●●● ●● ●● ●●●● ●●●● ●●●● ●●●● ●●●● ●●● ● ●●●● ●●●● ●●●● ●● ●● ●●●● ●●●● ●●●● ●● ●● ●●●● ●● ●●0
5,000,000
10,000,000
15,000,000
B.m
emor
y_ac
tivat
edB.
mem
ory_
stea
dy.s
tate
B.na
ive_a
ctiva
ted
B.na
ive_s
tead
y.st
ate
B.pl
asm
a_st
eady
.sta
teBa
soph
il_st
eady
.sta
teEo
sinop
hil_
stea
dy.s
tate
mDC
_act
ivate
dm
DC_s
tead
y.st
ate
MO
.cla
ssica
l_ac
tivat
edM
O.c
lass
ical_
stea
dy.s
tate
MO
.inte
rmed
iate
_ste
ady.
stat
eM
O.n
oncla
ssica
l_st
eady
.sta
tem
Treg
s_ac
tivat
edm
Treg
s_st
eady
.sta
teNe
utro
phil_
stea
dy.s
tate
NK.b
right
_act
ivate
dNK
.brig
ht_s
tead
y.st
ate
NK.d
im_a
ctiva
ted
NK.d
im_s
tead
y.st
ate
nTre
gs_a
ctiva
ted
nTre
gs_s
tead
y.st
ate
pDC_
activ
ated
pDC_
stea
dy.s
tate
T4.C
M_a
ctiva
ted
T4.C
M_s
tead
y.st
ate
T4.E
M_a
ctiva
ted
T4.E
M_s
tead
y.st
ate
T4.E
MRA
_act
ivate
dT4
.EM
RA_s
tead
y.st
ate
T4.n
aive
_act
ivate
dT4
.nai
ve_s
tead
y.st
ate
T8.C
M_a
ctiva
ted
T8.C
M_s
tead
y.st
ate
T8.E
M_a
ctiva
ted
T8.E
M_s
tead
y.st
ate
T8.E
MRA
_act
ivate
dT8
.EM
RA_s
tead
y.st
ate
T8.n
aive
_act
ivate
dT8
.nai
ve_s
tead
y.st
ate
Th1_
stea
dy.s
tate
Th17
_ste
ady.
stat
eTh
2_st
eady
.sta
te
SERPINB2
Prot
ein
copy
num
ber
ACTG1AGA
AGTRAPATG13ATG2ACCR1CD14
CD163CD36CD37
CDCA7LCEBPD
CEP152CES1
CKAP4CLEC4E
CPMCXCL5CXCL8
CYP27A1DYNLT1ELOVL7
FABP4FAM129B
FAM160A2FCAR
FCGR1AFPR1FPR2FTH1
GABARAPGNAQGUSBHLTFIL1B
IQGAP3ITGA5
LACC1LAMTOR4
LTB4RLYZ
METTL7BMPP1
MTFMTNCSTN
NINJ1OBFC1
PAMPLAC8RAD9A
RALGAPA1RBM47
RIT1RPL6
SERPINB1SERPINB13SERPINB2
SLC44A1SLC7A6STEAP3
SUN1TMEM50A
VCANZNF185
Airway expression level (log2 reporter intensity)
4.54
3.53
210.750.25
NA
SERPINB2
ABHD
10AB
HD11
ACAD
SBAC
OX1
ACSL
1AH
CYL2
AKR1
C1AL
OX5
ANPE
PAN
XA6
APAF
1AP
MAP
ARCN
1AR
PC5
AZU1
BAG3
BAZ1
BBPHL
BPI
CA4CAMK2D
CCAR2
CCDC47
CD14
CKBCMAS
CNN3CORO1B
CPNE1
CRIP1CTNNB1
CTSBCYB5BCYBBDCDDCXRDDX1DHX15DNAJA1DNAJB1DYSFEIF3AEIF3BEIF4G1EMC7EPXFTH1GCAGFM2GFPT1GNSGPD1LGRHPR
GSTO1HADHHIBADH
HIST1H1C
HIST1H3A
HNRNPAB
HNRNPR
HP1BP3
HPRT1
HSD17B11
HSP90AB1
IFI30IGHG4
ILF3IPO9ITG
AMKHDRBS1
KPNB1KRT19LCP1LM
AN2M
APRE1M
E1M
GAMM
LLT4
MM
P9M
NDA
MO
GS
MRP
L24
MRP
S22
MTP
NM
X1M
YH9
MYO
1FNA
CANCF4
NDUF
A11
NDUF
A12
NDUF
A2
NDUFA5
NMES1
NRAS
NUDT21
OLFM4
OPTNPADI4
PAFAH1B1PCPCMT1PDLIM5PEPDPGRMC2PLBD1PMPCBPOLDIP2PON2PRPF19PRRC1PRTN3PSMA3PSMB3
PSMD11PSMD8PSME1PSME2RAB10RAB7ARAP1BRBM39
RETNRPL10RPL17RPL29
RPS14
RPS16
RPS25
RPS29
RRBP1
RUVBL2
S100A12
SAP18
SBDS
SERPINA3
SERPINB10
SERPINB2SH3GLB1
SNX6SRP54
SRSF3SUGP2
SUMF2
TCEB2TNFSF13BTO
MM
40TPPP3
TRAP1TUBB8UCHL3VNN1VRK1
YWHAQ
Fold change
4.543.532.5210.750.50.250NA
SERPINB2
MO classical activated vs Th2 steady stateMO classical activated vs Th17 steady stateMO classical activated vs Th1 steady stateMO classical activated vs T8 naive steady stateMO classical activated vs T8 EMRA steady stateMO classical activated vs T8 EM steady stateMO classical activated vs T8 CM steady stateMO classical activated vs T4 naive steady stateMO classical activated vs T4 EMRA steady stateMO classical activated vs T4 EM steady stateMO classical activated vs T4 CM steady stateMO classical activated vs pDC steady stateMO classical activated vs nTregs steady stateMO classical activated vs NK dim steady stateMO classical activated vs NK bright steady stateMO classical activated vs Neutrophil steady stateMO classical activated vs mTregs steady stateMO classical activated vs MO nonclassical steady stateMO classical activated vs MO intermediate steady stateMO classical activated vs MO classical steady stateMO classical activated vs mDC steady stateMO classical activated vs Eosinophil steady stateMO classical activated vs Basophil steady stateMO classical activated vs B plasma steady stateMO classical activated vs B naive steady stateMO classical activated vs B memory steady state
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●●
●
●●
●
●●
●
●
●
● ●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
● ●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●
●
●
●
●
●●
● ●
●●●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●●
●
●●
●
●
●
●● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●●
●
●
●
● ●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
● ●●
●
●
●●● ●
●
●
●
●
●
●
●
●
● ●
●
●
●
● ●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●●●
●
●
●
●
●
●
●●●●
●
●
●
● ●
●
●
●
●●
●●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●● ●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
● ●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
● ●
●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
● ●●●●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
● ●●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●● ●
●
●
●
●
●
●
●●
●●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
● ●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
● ●●●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●● ●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●● ●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●●
● ●●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●● ●●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
● ●
●●● ●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
● ●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
● ●
●
●
●
●
●
●
●
●
●
●● ●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●● ●●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●●
●
●
● ●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
● ●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●●
●
●●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●● ●
●●
●
●
●
●
●
●
●
●
●● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
● ●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
● ●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●●
●
●
●
●●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●●
●●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●●●
●
● ●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●● ●●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●● ●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●● ●●●
●
●
●●
●
●●●
●
●
●
●
●●
●
●● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●●
●
●
●●
●
●
●
●
●● ●●
●
●●
●
●
●
●
●
●● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
● ●
●
●
●
●
●
●●
●
●
●●
●
●
●
●● ●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
● ●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
● ●
●●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
● ●●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●
● ●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
● ●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
● ●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●●
●
●
●●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●●
●
● ●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
● ●
●
●
● ●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
● ● ●● ●
● ●
●
● ● ● ● ●
●●
●
●● ● ●
●
●●
● ●●
●
●●
●●
0
10
20
B.m
emor
y_ac
tivat
edB.
mem
ory_
stea
dy.s
tate
B.na
ive_a
ctiva
ted
B.na
ive_s
tead
y.st
ate
B.pl
asm
a_st
eady
.sta
teBa
soph
il_st
eady
.sta
teEo
sinop
hil_
stea
dy.s
tate
mDC
_act
ivate
dm
DC_s
tead
y.st
ate
MO
.cla
ssica
l_ac
tivat
edM
O.c
lass
ical_
stea
dy.s
tate
MO
.inte
rmed
iate
_ste
ady.
stat
eM
O.n
oncla
ssica
l_st
eady
.sta
tem
Treg
s_ac
tivat
edm
Treg
s_st
eady
.sta
teNe
utro
phil_
stea
dy.s
tate
NK.b
right
_act
ivate
dNK
.brig
ht_s
tead
y.st
ate
NK.d
im_a
ctiva
ted
NK.d
im_s
tead
y.st
ate
nTre
gs_a
ctiva
ted
nTre
gs_s
tead
y.st
ate
pDC_
activ
ated
pDC_
stea
dy.s
tate
T4.C
M_a
ctiva
ted
T4.C
M_s
tead
y.st
ate
T4.E
M_a
ctiva
ted
T4.E
M_s
tead
y.st
ate
T4.E
MRA
_act
ivate
dT4
.EM
RA_s
tead
y.st
ate
T4.n
aive
_act
ivate
dT4
.nai
ve_s
tead
y.st
ate
T8.C
M_a
ctiva
ted
T8.C
M_s
tead
y.st
ate
T8.E
M_a
ctiva
ted
T8.E
M_s
tead
y.st
ate
T8.E
MRA
_act
ivate
dT8
.EM
RA_s
tead
y.st
ate
T8.n
aive
_act
ivate
dT8
.nai
ve_s
tead
y.st
ate
Th1_
stea
dy.s
tate
Th17
_ste
ady.
stat
eTh
2_st
eady
.sta
te
Prot
ein
copy
num
ber (
log2
)
MO
(A)
(B)
(C)
(D)
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
29
Figure 3: Detailed analysis of the phenotype-specific feature proteins. (A) We used a circular heat map to visualize differential expression
of phenotype-specific proteins in a selection of immune cell phenotypes. In this example, the fold difference in protein expression between
classical activated monocytes and other immune cell subsets is shown. Each spoke of the wheel represents a single protein, and each
segment on a spoke represents the log10 fold difference between protein’s expression level in classical activated monocytes and each of
the listed cell types. An increase in the intensity of the red hue indicates greater expression in classical activated monocytes, relative to
the comparator phenotype. (B) Protein expression across different phenotypes: in this example, the expression level of a feature protein
for activated monocytes (SERPINB2 – identified from the random forest model) was compared between different phenotypes and was
seen to be expressed almost exclusively by monocytes. (C) The airway proteome data from paediatric pneumonia admissions was
examined to determine whether any of the phenotype classification features for classical activated monocytes were also detected in the
airways of children. Each segment represents a single feature, with proteins that were also detected in the airway depicted in a green hue.
An increase in hue intensity, indicates increased airway expression for a particular protein. SERPINB2 (red arrow), was expressed at
relatively high levels in the infant airway. (D) The expression level of all monocyte-specific features was plotted for monocytes and other
cell types. Each black open circle denotes the expression level of a particular feature in other phenotypes, while red markers indicate their
expression in monocytes. The dashed black and red lines indicated median expression of all features in monocytes and alternate cell types
respectively.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
30
Figure 4: (A) Performance of all phenotype specific protein features in segregating different immune phenotypes and sub phenotypes was
visualized using a t-SNE plot. Most phenotype classification features were able to disaggregate different phenotype hierarchies in the 2D
t-SNE space. Neutrophils, eosinophils, monocytes, NK cells, dendritic cells, basophils among others were clearly separated from all other
B.memory
B.memoryB.memory
B.naiveB.naiveB.naive
B.naive
B.plasmaB.memory
B.memory
B.memory
B.memory B.naive
B.naive
B.naive
B.naive
B.plasmaB.plasma
B.plasmaBasophil
Basophil
BasophilBasophil
EosinophilEosinophil
EosinophilEosinophil
mDCmDC
mDC
mDC
mDC
mDC
mDC
mDC
MO.classical
MO.classical
MO.classical
MO.intermediateMO.intermediate
MO.nonclassical
MO.intermediate
MO.classicalMO.classical
MO.intermediate
MO.nonclassicalMO.nonclassical
MO.nonclassical
MO.classical
MO.classical
mTregs
mTregs
mTregs
mTregs
mTregs
mTregs
mTregs
Neutrophil
Neutrophil
NeutrophilNeutrophil
NK.bright
NK.brightNK.dim
NK.bright
NK.dimNK.dim
NK.bright
NK.bright
NK.dim
NK.dim
NK.dim
NK.dim
NK.bright
NK.bright
nTregsnTregs
nTregs
nTregs
nTregs
nTregsnTregs
pDC
pDC
pDC
pDC pDC
pDCpDC
pDC
T4.naive
T4.naive
T4.EMRAT4.EMRA
T4.CM
T4.EMRA
T4.CM
T4.CM
T4.EM
T4.CM
T4.naive
T4.naive
T4.naive
T4.naive
T4.naive
T4.naive
T4.EMRA
T4.CM
T4.EMRA
T4.CM T4.CM
T4.EM
T4.CM
T4.EM
T4.EM
T4.EM
T4.EM
T4.EMRA
T4.EMRA
T4.EMT4.EM
T4.EMRA
T8.CM
T8.naiveT8.naive
T8.CM
T8.naive
T8.naive
T8.EMRA
T8.EMRA
T8.naive
T8.naive
T8.CMT8.CM
T8.EM
T8.EM
T8.EM
T8.EM
T8.EMRA
T8.EMRA
T8.EMRA
T8.naive
T8.naiveT8.CM
T8.CM
T8.CMT8.CM
T8.EM
T8.EMT8.EM
T8.EM
T8.EMRA
T8.EMRAT8.EMRA
Th1
Th1 Th1
Th1
Th17
Th17
Th17
Th17
Th2
Th2
Th2
Th2
−20
−10
0
10
20
−20 −10 0 10 20
labelsB
Basophil
Eosinophil
mDC
MO
mTregs
Neutrophil
NK
nTregs
pDC
T4
T8
Th1
Th17
Th2
(A)
(B)
(C)
P<0.01 n.s.
P<0.05
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
31
phenotypes. There was less clear segregation for T-cell sub phenotypes, including CD8, CD4 and regulatory T-cell phenotypes. (B) the
expression of eosinophil- and neutrophil-specific features was compared between children with different survival outcomes of pneumonia
as well as in well controls. Eosinophil features were expressed at significantly higher levels in survivors compared with non-survivors and
well controls. In contrast, there was no difference in the expression of neutrophil features between these groups. (C) Feature classifiers
for other immune cell phenotypes were also compared by survival status. No significant difference in expression by survival status was
observed, with the exception of the CD8 (T8) phenotype, whose features were significantly upregulated in the airways of survivors relative
to non-survivors and well controls.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
32
Figure 5: Validation of the phenotype associations with pneumonia mortality using a retrospective pneumonia cohort. (A, first panel)
Eosinophils counts were analysed in a retrospective cohort of >10,000 children who had been admitted to hospital with clinical pneumonia
0.0
0.5
1.0
1.5
2.0
2.5
Survived Died
Eosin
ophil
s x10
3 cells
/uL
N=10,859 N=1,604
P = 0.0004
Eosinophils(A)
0
10
20
30
40
50
Survived Died
Neut
roph
ils x1
03 cells
/uL
N=10,871 N=1,606
P = 0.26
Neutrophils(B)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+++++++++++ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++
++++++++++++++
+
+++
0.00
0.25
0.50
0.75
1.00
0 1 2 3 4 5 6 7 8 9 10
Surv
ival
pro
babi
lity
++
High
Low
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++++++++++++++++++++ +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++
0.00
0.25
0.50
0.75
1.00
0 1 2 3 4 5 6 7 8 9 10
Surv
ival
pro
babi
lity
++
High
Low
Surv
ival p
roba
bilit
ySu
rviva
l pro
babi
lity
Days of admission
Days of admission
MUAC < 11.5cm
MUAC < 11.5cm
P=0.2
P<0.0001
0.0
0.1
0.2
0.3
0.4
Eosi
noph
ils x
103 c
ells
/uL
Eosin
ophi
ls x1
03ce
lls/µ
l
High
Low
High
Low
0.0
2.5
5.0
7.5
10.0
Eosi
noph
ils x
103 c
ells
/uL
Neut
roph
ils x
103
cells
/µl
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
33
and who either survived and were discharged or died during admission. Pneumonia survivors had significantly greater eosinophil counts
at admission compared to non-survivors. (A, second panel) In a subset of children who were acutely undernourished at admission (defined
as by a mid-upper arm (MUAC) circumference <11.5cm), children were stratified into two groups on the basis of the median eosinophil
count. In the first 10 days after admission, children whose eosinophil counts were below the median were more likely to die relative to
those with above-median counts. (B) In contrast, there was no association between neutrophil counts and survival status.
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
34
References
1 Causes of severe pneumonia requiring hospital admission in children without HIV infection from Africa and Asia: the PERCH multi-country case-control study. Lancet (London, England) 394, 757-779, doi:10.1016/s0140-6736(19)30721-4 (2019).
2 Shi, T. et al. Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in young children in 2015: a systematic review and modelling study. Lancet (London, England), doi:10.1016/s0140-6736(17)30938-8 (2017).
3 McNally, L. M. et al. Effect of age, polymicrobial disease, and maternal HIV status on treatment response and cause of severe pneumonia in South African children: a prospective descriptive study. Lancet (London, England) 369, 1440-1451, doi:10.1016/s0140-6736(07)60670-9 (2007).
4 Chisti, M. J., Tebruegge, M., La Vincente, S., Graham, S. M. & Duke, T. Pneumonia in severely malnourished children in developing countries - mortality risk, aetiology and validity of WHO clinical signs: a systematic review. Tropical medicine & international health : TM & IH 14, 1173-1189, doi:10.1111/j.1365-3156.2009.02364.x (2009).
5 Moisi, J. C. et al. Sensitivity of hospital-based surveillance for severe disease: a geographic information system analysis of access to care in Kilifi district, Kenya. Bulletin of the World Health Organization 89, 102-111, doi:10.2471/blt.10.080796 (2011).
6 Mejias, A. et al. Whole blood gene expression profiles to assess pathogenesis and disease severity in infants with respiratory syncytial virus infection. PLoS medicine 10, e1001549, doi:10.1371/journal.pmed.1001549 (2013).
7 Johnson, J. E., Gonzales, R. A., Olson, S. J., Wright, P. F. & Graham, B. S. The histopathology of fatal untreated human respiratory syncytial virus infection. Mod Pathol 20, 108-119, doi:10.1038/modpathol.3800725 (2007).
8 Jochems, S. P. et al. Novel Analysis of Immune Cells from Nasal Microbiopsy Demonstrates Reliable, Reproducible Data for Immune Populations, and Superior Cytokine Detection Compared to Nasal Wash. PloS one 12, e0169805, doi:10.1371/journal.pone.0169805 (2017).
9 Connors, T. J. et al. Developmental Regulation of Effector and Resident Memory T Cell Generation during Pediatric Viral Respiratory Tract Infection. Journal of immunology (Baltimore, Md. : 1950) 201, 432-439, doi:10.4049/jimmunol.1800396 (2018).
10 Sande, C. J. et al. Untargeted analysis of the airway proteomes of children with respiratory infections using mass spectrometry based proteomics. Scientific reports 8, 13814, doi:10.1038/s41598-018-32072-3 (2018).
11 Sande, C. J. et al. Airway response to respiratory syncytial virus has incidental antibacterial effects. Nature communications 10, 2218, doi:10.1038/s41467-019-10222-z (2019).
12 Rieckmann, J. C. et al. Social network architecture of human immune cells unveiled by quantitative proteomics. Nature immunology 18, 583-593, doi:10.1038/ni.3693 (2017).
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
35
13 Kursa, M. B. & Rudnicki, W. R. Feature Selection with the Boruta Package. 2010 36, 13, doi:10.18637/jss.v036.i11 (2010).
14 Chen, E. Y. et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC bioinformatics 14, 128, doi:10.1186/1471-2105-14-128 (2013).
15 Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic acids research 44, W90-97, doi:10.1093/nar/gkw377 (2016).
16 Lindemans, C. A. et al. Systemic eosinophil response induced by respiratory syncytial virus. Clinical and experimental immunology 144, 409-417, doi:10.1111/j.1365-2249.2006.03084.x (2006).
17 Kim, H. H., Lee, M. H. & Lee, J. S. Eosinophil cationic protein and chemokines in nasopharyngeal secretions of infants with respiratory syncytial virus (RSV) bronchiolitis and non-RSV bronchiolitis. Journal of Korean medical science 22, 37-42, doi:10.3346/jkms.2007.22.1.37 (2007).
18 Harrison, A. M., Bonville, C. A., Rosenberg, H. F. & Domachowske, J. B. Respiratory syncytical virus-induced chemokine expression in the lower airways: eosinophil recruitment and degranulation. American journal of respiratory and critical care medicine 159, 1918-1924, doi:10.1164/ajrccm.159.6.9805083 (1999).
19 Dimova-Yaneva, D., Russell, D., Main, M., Brooker, R. J. & Helms, P. J. Eosinophil activation and cysteinyl leukotriene production in infants with respiratory syncytial virus bronchiolitis. Clinical and experimental allergy : journal of the British Society for Allergy and Clinical Immunology 34, 555-558, doi:10.1111/j.1365-2222.2004.1918.x (2004).
20 Everard, M. L. et al. Analysis of cells obtained by bronchial lavage of infants with respiratory syncytial virus infection. Archives of disease in childhood 71, 428-432 (1994).
21 McNamara, P. S., Ritson, P., Selby, A., Hart, C. A. & Smyth, R. L. Bronchoalveolar lavage cellularity in infants with severe respiratory syncytial virus bronchiolitis. Archives of disease in childhood 88, 922-926, doi:10.1136/adc.88.10.922 (2003).
22 Wang, S. Z. et al. Shedding of L-selectin and PECAM-1 and upregulation of Mac-1 and ICAM-1 on neutrophils in RSV bronchiolitis. The American journal of physiology 275, L983-989, doi:10.1152/ajplung.1998.275.5.L983 (1998).
23 Wang, S. Z. et al. The apoptosis of neutrophils is accelerated in respiratory syncytial virus (RSV)-induced bronchiolitis. Clinical and experimental immunology 114, 49-54, doi:10.1046/j.1365-2249.1998.00681.x (1998).
24 Kerrin, A. et al. Differential lower airway dendritic cell patterns may reveal distinct endotypes of RSV bronchiolitis. Thorax 72, 620-627, doi:10.1136/thoraxjnl-2015-207358 (2017).
25 Bem, R. A. et al. Activation of the granzyme pathway in children with severe respiratory syncytial virus infection. Pediatric research 63, 650-655, doi:10.1203/PDR.0b013e31816fdc32 (2008).
26 Gill, M. A. et al. Mobilization of plasmacytoid and myeloid dendritic cells to mucosal sites in children with respiratory syncytial virus and other viral respiratory infections. The Journal of infectious diseases 191, 1105-1115, doi:10.1086/428589 (2005).
27 Heidema, J. et al. CD8+ T cell responses in bronchoalveolar lavage fluid and peripheral blood mononuclear cells of infants with severe primary respiratory syncytial virus infections. Journal of immunology (Baltimore, Md. : 1950) 179, 8410-8417 (2007).
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint
36
.CC-BY-ND 4.0 International licenseavailable under awas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint (whichthis version posted January 19, 2020. ; https://doi.org/10.1101/840090doi: bioRxiv preprint