Analysis of Mycobacterium tuberculosis-specific CD8 T-cells in patients with active tuberculosis and in individuals with latent infection.

Abstract

CD8 T-cells contribute to control of Mycobacterium tuberculosis infection, but little is known about the quality of the CD8 T-cell response in subjects with latent infection and in patients with active tuberculosis disease. CD8 T-cells recognizing epitopes from 6 different proteins of Mycobacterium tuberculosis were detected by tetramer staining. Intracellular cytokines staining for specific production of IFN-gamma and IL-2 was performed, complemented by phenotyping of memory markers on antigen-specific CD8 T-cells. The ex-vivo frequencies of tetramer-specific CD8 T-cells in tuberculous patients before therapy were lower than in subjects with latent infection, but increased at four months after therapy to comparable percentages detected in subjects with latent infection. The majority of CD8 T-cells from subjects with latent infection expressed a terminally-differentiated phenotype (CD45RA+CCR7(-)). In contrast, tuberculous patients had only 35% of antigen-specific CD8 T-cells expressing this phenotype, while containing higher proportions of cells with an effector memory- and a central memory-like phenotype, and which did not change significantly after therapy. CD8 T-cells from subjects with latent infection showed a codominance of IL-2+/IFN-gamma+ and IL-2(-)/IFN-gamma+ T-cell populations; interestingly, only the IL-2+/IFN-gamma+ population was reduced or absent in tuberculous patients, highly suggestive of a restricted functional profile of Mycobacterium tuberculosis-specific CD8 T-cells during active disease. These results suggest distinct Mycobacterium tuberculosis specific CD8 T-cell phenotypic and functional signatures between subjects which control infection (subjects with latent infection) and those who do not (patients with active disease).

Full text

Page 1

Analysis of Mycobacterium tuberculosis-Specific CD8 T-
Cells in Patients with Active Tuberculosis and in
Individuals with Latent Infection
Nadia Caccamo1*, Giuliana Guggino1, Serena Meraviglia1, Giuseppe Gelsomino1, Paola Di Carlo2, Lucina
Titone2, Marialuisa Bocchino3, Domenico Galati3, Alessandro Matarese3, Jan Nouta4, Michel R. Klein5,
Alfredo Salerno1, Alessandro Sanduzzi3, Francesco Dieli1., Tom H. M. Ottenhoff4.
1Dipartimento di Biopatologia e Metodologie Biomediche, Universita ` di Palermo, Palermo, Italy, 2Dipartimento di Medicina Clinica e delle Patologie Emergenti,
Universita ` di Palermo, Palermo, Italy, 3TB Infection Screening Unit, Department of Clinical and Experimental Medicine, University of Naples ‘‘Federico II’’, Monaldi Hospital,
Naples, Italy, 4Department of Immunohematology & Blood Transfusion and Department of Infectious Diseases, Leiden University Medical Center, Leiden, The
Netherlands, 5National Institute of Public Health and the Environment, Bilthoven, The Netherlands
Abstract
CD8 T-cells contribute to control of Mycobacterium tuberculosis infection, but little is known about the quality of the CD8 T-
cell response in subjects with latent infection and in patients with active tuberculosis disease. CD8 T-cells recognizing
epitopes from 6 different proteins of Mycobacterium tuberculosis were detected by tetramer staining. Intracellular cytokines
staining for specific production of IFN-c and IL-2 was performed, complemented by phenotyping of memory markers on
antigen-specific CD8 T-cells. The ex-vivo frequencies of tetramer-specific CD8 T-cells in tuberculous patients before therapy
were lower than in subjects with latent infection, but increased at four months after therapy to comparable percentages
detected in subjects with latent infection. The majority of CD8 T-cells from subjects with latent infection expressed a
terminally-differentiated phenotype (CD45RA+CCR72). In contrast, tuberculous patients had only 35% of antigen-specific
CD8 T-cells expressing this phenotype, while containing higher proportions of cells with an effector memory- and a central
memory-like phenotype, and which did not change significantly after therapy. CD8 T-cells from subjects with latent
infection showed a codominance of IL-2+/IFN-c+and IL-22/IFN-c+T-cell populations; interestingly, only the IL-2+/IFN-c+
population was reduced or absent in tuberculous patients, highly suggestive of a restricted functional profile of
Mycobacterium tuberculosis-specific CD8 T-cells during active disease. These results suggest distinct Mycobacterium
tuberculosis specific CD8 T-cell phenotypic and functional signatures between subjects which control infection (subjects
with latent infection) and those who do not (patients with active disease).
Citation: Caccamo N, Guggino G, Meraviglia S, Gelsomino G, Di Carlo P, et al. (2009) Analysis of Mycobacterium tuberculosis-Specific CD8 T-Cells in Patients with
Active Tuberculosis and in Individuals with Latent Infection. PLoS ONE 4(5): e5528. doi:10.1371/journal.pone.0005528
Editor: Madhukar Pai, McGill University, Canada
Received January 26, 2009; Accepted April 10, 2009; Published May 13, 2009
Copyright: ? 2009 Caccamo et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work has been supported by grants from the Ministry for Instruction, University and Research (MIUR-PRIN to FD), the University of Palermo (60% to
FD and NC) and the European Commission European Commission within the 6th Framework Programme, contract no. LSHP-CT-2003-503367 to FD and THMO,
(the text represents the authors’ views and does not necessarily represent a position of the Commission who will not be liable for the use made of such
information) and the Bill and Melinda Gates Foundation, Grand Challenges in Global Health (GC6#74, GC12#82). The funders had no role in study design, data
collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: caccamo@unipa.it.
. These authors contributed equally to this work.
Introduction
Globally, Tuberculosis (TB) accounts for approximately nine
million new cases of disease and around two million deaths every
year [1]. TB is presenting new challenges as a global health
problem, especially with new threats of HIV coinfection and
multidrug-resistant and extensively drug-resistant strains of
Mycobacterium tuberculosis (Mtb). TB is transmitted directly from
human to human and the control of the infection depends on early
identification and proper treatment of individuals with active
disease. However, the lack of accurate diagnostic techniques has
contributed to the emergence of TB as a threat to global health.
To date, there is no simple, rapid, sensitive and specific test that
can differentiate active TB from latent infection, and slowly
progressive TB.
T-cells, T-cell derived cytokines and cytotoxic molecules are
crucial for protection against TB. Although a role for CD4 T-cells
in protection against Mtb is well documented, there is also a large
body of evidence derived from human and non human models
that suggests an involvement of CD8 T-cells [2–5]. CD8 T-cells
contribute to control of Mtb infection by mediating specific
effector functions, including IFN-c and TNF-a production upon
recognition of mycobacterial antigens [6–8], lysis of infected host
cells [6–9], and direct killing of mycobacteria [5,10,11]. A limited
number of studies focused on the T-cell repertoire in Mtb
infection, demonstrating clonal T-cell expansion in granulomas
from subjects with LTBI [12] and changes in the peripheral blood
and pleural fluid T-cell repertoire from TB patients [13].
Furthermore, CD8 T-cells specific for numerous mycobacterial
antigens can be isolated at high frequency from human and mouse
PLoS ONE | www.plosone.org1 May 2009 | Volume 4 | Issue 5 | e5528

Page 2

models, consistent with the hypothesis that CD8 T lymphocytes
are constantly being stimulated with antigen [9,10].
However, there are few studies which have compared the
frequency, phenotype and function of antigen-specific CD8 T-cells
in TB patients and subjects with latent infection (LTBI). Among
them, we ourselves previously found that the frequency of Ag85A
peptide-specific CD8 T-cells was reduced in tuberculous children
before therapy, but increased after therapy to levels similar to
those detected in healthy tuberculin skin test positive children.
Ag85A epitope-specific CD8 T-cells during active TB were mainly
present among central memory cells and produced low levels of
IFN-c and perforin, which recovered after therapy [14]. In a
parallel study, Kaufmann and colleagues found clonal expansion
of effector-memory CD8 T-cells in older children with TB, with
potential impact on course and severity of disease [15]. However,
the CD8 repertoire of children could well be different from that in
adult individuals given the different clinical manifestation of TB in
children and adults; moreover, little is known about the size,
quality and specificity of Mtb-specific CD8 T-cell responses in
adult patients with active TB disease compared to treated TB and
subjects with LTBI.
To start addressing these issues, we have in this study
determined the ex-vivo frequencies, phenotype and functional
properties of HLA-A*0201 CD8 T-cells specific for different
peptides of Mtb proteins in adult subjects with LTBI and adult TB
patients with active disease, both before and following four months
of anti-mycobacterial therapy.
Results
Ex vivo analysis of circulating epitope-specific CD8 T-cells
To determine the ex vivo frequency of peptide-specific CD8 T-
cells, PBMC from HLA-A*0201 patients with active TB before
(T0) and after four months of chemotherapy (T4) and individuals
with LTBI were stained with HLA-A*0201/tetramers and anti-
CD8 antibody and analysed by FACS.
Four out of the six selected epitopes (Ag85B p5–13, Esat-6 p82–
90, Hsp65 p362–370 and 16 kDa p120–128) were previously
identified as CD8 T cell epitopes, while Rv1490 p325–333 and
Rv1614 p197–205 are newly identified in this study as candidate
epitopes, based upon whole Mtb genome screening for 9-mer
peptides sequences with high/intermediate HLA-A*0201 binding
affinity (see also Table 1). Moreover, a minimum of three and a
maximum of all six tetramers marked the CD8 T-cell response in
each group of individuals (data not shown). In all instancies,
specificity of tetramer staining was confirmed by the negative data
obtained both using tetramer of an irrelevant specificity (the HLA-
A*0201/HIV-1 gag peptide p76–84) and staining PBMC from
normal, uninfected donors with Mtb tetramers (see Table 2).
Although there was considerable variability in the proportions of
CD8 T-cells that bound to single tetramers, an immunodominance
Table 2. Ex vivo analysis of frequency of peptide-specific CD8 T-cells.
TetramerLTBIT0T4 PPD2Healthy Donor
16 kDa0.30.13a
0.22c
,0.01
(0.19–0.55)(0.05–0.34) (0.20–0.72)
Rv14900.54 0.19a
0.3c
,0.01
(0.16–0.80)(0.11–0.65)(0.22–0.36)
Rv16140.650.30b
0.3
,0.01
(0.58–0.85)(0.11–0.59)(0.29–0.45)
Ag85B 0.80.60 0.64
,0.01
(0.56–0.94)(0.31–0.92) (0.24–1.35)
Esat-60.56 0.29b
0.58d
,0.01
(0.33–0.96) (0.17–0.66)(0.18–0.92)
Hsp650.570.41b
0.34
,0.01
(0.42–0.80) (0.31–0.74)(0.22–0.58)
HIV-1 gag
,0.01
,0.01
,0.01
,0.01
Cumulative data on the frequencies of the tetramer-specific CD8 T-cells in peripheral blood of subjects with LTBI, patients with active TB before (T0) and four months
after therapy (T4). Data are presented as median values while interquartile range is shown in brackets.
ap,0.01 andbp,0.05 when compared to values in LTBI subjects.
cp,0.05 anddp,0.02 when compared to values in TB patients before therapy (T0).
doi:10.1371/journal.pone.0005528.t002
Table 1. HLA-A*0201 binding of predicted HLA-A*0201
binders.
Peptide AA sequencerel IC50
Rv1490 p325–333 FLLGLLFFV 2,50
Rv1614p197–205 FLYELIWNV1,79
Ag85B p5–13GLPVEYLQV7
Esat-6p82–90 AMASTEGNV 1,25
Hsp65 p362–370KLQERLAKL 0,25
16 kDa p120–128GILTVSVAV 2,25
HBV core p47–56FLPSDYFPSV1
Peptides were tested for their ability to compete binding of 1.6 mM biotinylated
peptide HBV core p47–56 to HLA-A*0201 molecules. The concentration of
peptide yielding 50% inhibition (IC50) was deduced from the dose-response
curve. One of the six peptides bound to HLA-A*0201 with high affinity
(IC50,1 mM), while the other peptides bound the HLA-A*0201 with
intermediate affinity (IC50 1–10 mM). Each peptide was tested in at least two
separate experiments. Data are expressed as relative (rel) IC50, compared to the
IC50 of the standard peptide HBV core p47–56, which was considered as 1.
Values are the mean of rel IC50 of two independent experiments with SE being
always ,10%.
doi:10.1371/journal.pone.0005528.t001
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org2May 2009 | Volume 4 | Issue 5 | e5528

Page 3

hierarchy in epitope-specific CD8 T-cell response was found both in
LTBI subjects and in TB patients at T0 and T4 (Figure 1A and
Table 2). In LTBI subjects, the mean ex-vivo frequency of peptide-
specific CD8+tetramer+T-cells was 0.8% for Ag85B, 0.66% for
Hsp65, 0.65% Rv1614, 0.57% for Esat-6, 0.49% for Rv1490 and
0.43% for 16 kDa. The ex-vivo frequency of tetramer-specific CD8 T-
cells was higher in LTBI subjects than in TB patients (Table 2) and
this difference attained statistical significance with most of the studied
epitopes (i.e, Esat-6, Rv1614, Rv1490, Hsp65 and 16 kDa). Ag85B
peptidewasthemostimmunodominantalsoinTBpatientsatT0and
at T4, as estimated by enumerating the frequencies of tetramer-
specific CD8 T-cells; however significant differences in frequencies of
epitope-specificCD8T-cells were observed in TBpatientsbefore and
after chemotherapy. In three instances (Esat-6, Rv1490 and 16 kDa
antigens), mean frequencies of epitope-specific CD8 T-cells signifi-
cantly increased after therapy: the mean frequency of Esat-6-specific
CD8 T-cells was 0.39% in patientsat T0 and 0.55%in patients at T4
(p,0.02), the mean frequency of Rv1490-specific CD8 T-cells was
0.35% at T0 and raised to 0.41% at T4 (p,0.05), the mean
frequency of 16 kDa-specific CD8 T-cells was 0.24% at T0 and
0.38% at T4 (p,0.05) and the mean frequency Ag85B-specific CD8
T cells was 0.65% at T0 and 0.78% at T4. However, the frequencies
of CD8 T-cells specific for the two other studied epitopes remained
virtually unchanged before and four months after therapy: in fact, the
mean frequency of Rv1614-specific CD8 T-cells was 0.38% at T0
and 0.41% at T4, and finally, the mean frequency of Hsp65-specific
CD8 T-cells was 0.52% at T0 and 0.55% at T4.
Figure 1B shows FACS analysis of the tetramer+CD8+T-cells
of one representative LTBI subject, one TB patient at T0 and one
TB patient at T4.
Figure 1. Comparison of the frequencies of tetramer+CD8 T-cells in peripheral blood from LTBI subjects and TB patients with active
disease before therapy (T0) and after four months of therapy (T4). (A) In each group tested, LTBI subjects, TB patients at T0 and TB patients
at T4, the median proportion of tetramer+CD8 T-cells was estimated as 100% and the relative percentages of individual tetramer+CD8 T-cells
calculated accordingly. (B) Dot plot analysis of tetramer+CD8+T-cell populations of one representative LTBI subject, one TB patient at T0 and one TB
patient at T4.
doi:10.1371/journal.pone.0005528.g001
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org3May 2009 | Volume 4 | Issue 5 | e5528

Page 4

Altogether, these results indicate that (a) the frequencies of Mtb
epitope-specific CD8 T-cells during active pulmonary TB disease
in adults are lower than in LTBI individuals, but that they increase
after anti-mycobacterial therapy; (b) that between 50 and 100% of
epitopes selected are recognized by individuals with latent or
active Mtb infection, including two new epitopes (Rv1490 and
Rv1614) and (c), that there appears to be an immunodominance
hierarchy in the recognition of different epitopes of Mtb in
individuals with LTBI, as well as in patients with active TB before
and after therapy.
Phenotypic analysis of tetramer-specific CD8 T-cells
CD8 T-cells can be divided into at least four different
populations of naive, central memory, effector memory and
terminally-differentiated effector memory T-cells, based on the
expression of surface markers associated with their maturation
[16]. We have compared the phenotype of circulating tetramer-
specific CD8 T-cell subsets in HLA-A*0201 TB patients at T0 and
T4 and in individuals with LTBI. Representative data are shown
in Figure 2A and cumulative data are shown in Figure 2B.
The mean frequencies of tetramer-specific CD8 T-cells with a
CCR7+CD45RA+naive phenotype were found to be comparable
in TB patients and in subjects with LTBI. However, in the latter
we found that approximately 60% of CD8 T-cells expressed
CD45RA but not CCR7, indicating a terminally-differentiated
phenotype; this was irrespective on their antigen specificity. In
LTBI subjects, about 5% of the specific CD8 T-cells had an
effector memory-like phenotype (CD45RA2CCR72), while less
than 5% had a central memory-like phenotype (CD45RA2
CCR7+).
In TB patients, although cells expressing a terminally-differen-
tiated CD45RA+CCR72phenotype still comprised the predom-
inant subset among specific CD8 T-cells, their mean percentage
was lower than in subjects with LTBI (35% versus 60%) and
remained virtually unchanged before and four months after
therapy (35% versus 42%). However, although lower percentages of
CD45RA+CCR72
cells was detectable within all studied
tetramer+CD8 T-cells, none of the differences between LTBI
subjects and TB patients before or after therapy attained statistical
significance. Conversely, tetramer-specific CD8 T-cells from TB
patients contained higher proportions of cells with an effector
memory-like (15% in TB patients versus 5% in LTBI subjects) and
a central memory-like (10% in TB patients versus 3% in LTBI
subjects) phenotype; however, also the frequencies of these two
memory subsets did not change significantly before and after
therapy. Thus, the data here reported also point to qualitative
differences between TB patients and LTBI subjects in their
antigen-specific CD8 T-cell compartment and suggest that the
pool of terminally-differentiated CD45RA+CCR72epitope-
specific CD8 T cells is reduced in TB patients.
Analysis of cytokine production by peptide-specific CD8
T-cells at the single cell level
IFN-c and IL-2 have been shown to be the most relevant
cytokines to define functional populations of antigen-specific CD4
and CD8 T-cells [16–18]. With regard to CD8 T-cells, two cell
populations can be defined on the basis of the ability to secrete IL-
2 and IFN-c: CD8 T-cells secreting simultaneously IL-2 and IFN-
c (dual IL-2+/IFN-c+), and CD8 T-cells secreting only IFN-c
(single IFN-c). To assess these two profiles, we stimulated PBMC
of HLA-A*0201 LTBI subjects and TB patients with the same
individual peptides as those present in tetramers used in this study
and determined the proportion of tetramer specific CD8 T-cells
that produced IFN-c and/or IL-2 by intracellular FACS analysis,
after short-term stimulation with peptides. Representative data are
shown in Figure 3A and cumulative data are shown in Figure 3B.
Irrespective of the peptide, in subjects with LTBI 35%–45% of
CD8 T-cells secreted both IL-2 and IFN-c. The remainder of
CD8 T-cells (55% to 65%) secreted only IFN-c. In contrast, TB
patients had a lower frequency of single IFN-c-secreting cells
against all tested peptides, but attained statistical significance only
for Hsp65 peptide-specific CD8 T-cells. However, the most
impressive finding was the consistent reduction of dual IL-2+/IFN-
c+cytokine-secreting CD8 T-cells in TB patients, highly suggestive
of a more restricted functional profile of Mtb-specific CD8 T-cells
during active disease. Because of limited blood sample volume, it
was possible to do intracellular cytokine staining after therapy only
in five out of the patients at the first study time point. The results
obtained showed, that four months after therapy all these five
patients had still a dominance of IFN-c-only secreting cells for all
peptides tested except for the 16 kDa peptide. However, the
limited number of patients after therapy did not allow statistical
analysis of data.
These results indicate for the first time that the percentage of
double IFNc+/IL-2+producing CD8 T-cells is significantly higher
in LTBI subjects than in TB patients before therapy, suggesting a
protective role of the two cytokines jointly in association with
antigen specific CD8+T-cell responses towards Mtb in latently
infected healthy subjects. Figure 4 shows peptide-specific CD8 T-
cell producing single IFN-c+or double IFN-c+/IL2+: each
portions of a pie chart indicates the mean percentage of
peptides-specific T cells that responded with one or two functions.
Discussion
CD8 T-cells play a critical role in chronic viral infection, but
during recent years their role has gained increasing attention also
in Mtb infection. In the present study, we investigated the ex-vivo
frequencies, multifunctional cytokine production and memory
phenotype of circulating CD8 T-cells specific for different peptide-
nonamers of Mtb proteins in adult HLA-A*0201 subjects with
LTBI and in TB patients before (T0) and after four months of anti-
mycobacterial therapy (T4).
The ex-vivo frequencies of circulating tetramer specific CD8 T-
cells in TB patients before therapy was lower than in LTBI
subjects, but increased at four months after therapy to comparable
percentages detected in subjects with LTBI; this pattern was
consistently found for all tested tetramers. Thus, the frequency of
circulating Mtb-specific CD8 T-cells is halved during active TB,
compared to LTBI individuals. The reason of the reduced antigen-
specific CD8 T-cell frequencies in TB patients at the beginning of
therapy and their recovery after four months is not known, but the
simplest explanation is that in TB patients large numbers of CD8
T-cells are sequestered at the site of disease and repopulate the
peripheral blood compartment after successful anti-mycobacterial
therapy. The phenomenon of sequestration of antigen-specific cells
has been widely observed in TB for both CD4 and CD8 T-cells
[14,19,20].
The data here reported also point to qualitative differences
between TB patients and LTBI subjects in their antigen-specific
CD8 T-cell compartment: while approximately 60% of antigen-
specific CD8 T-cells in LTBI expressed a terminally-differentiated
phenotype (CD45RA+CCR72), in TB patients this was only 35%
of antigen-specific CD8 T-cells. Conversely, TB patients had
higher proportions of cells with an effector memory-like and a
central memory-like phenotype. Interestingly, the phenotype of
the CD8 T-cell population did not change significantly four
months after therapy. These findings are somewhat surprising on
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org4 May 2009 | Volume 4 | Issue 5 | e5528

Page 5

CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org5May 2009 | Volume 4 | Issue 5 | e5528

Page 6

the basis of our previous study of CD8 T-cell phenotype in
children with TB, where the antigen-specific CD8 T-cell
distribution pattern consistently changed four months after
therapy, with a significant recovery of terminally differentiated
effector memory T-cells and decreased frequencies of central
memory T-cells [14]. Another study reported expansion of
effector-memory CD8 T-cells in children with TB [15], which
express a CD28 and CD27 double negative phenotype. However,
care should be taken in correlating the phenotype with the
functional properties. For instance, CD45RAhiCD8 T-cells may
accumulate during chronic viral infections in elderly individuals
representing a pool of apoptosis-resistant memory cells that retain
replicative potential [21].
The reason for the reduction of antigen-specific CD8 T-cells
during active TB is unknown. As discussed above, one explanation
we favour is that these cells are sequestered at sites of infection.
Accordingly, a significantly high percentage of Mtb Ag85A
epitope-specific CD8 T-cells was previously reported in the
cerebrospinal fluid (CSF) of a child with TB meningitis [14].
Alternatively, it is possible that the low frequency of specific
CD8 T-cells in active TB could be the consequence of sustained in
vivo mycobacterial stimulation, which causes their apoptosis. For
example, high levels of bacteria (such as occurs in TB patients due
to the inability to contain and prevent their spread) could result in
chronic stimulation of CD8 T-cells and induce their apoptosis.
A final aim of our study was to assess the capability of Mtb-
specific CD8 T-cells to coproduce IFN-c and IL-2, which is
thought to be an indication of their multi-functionality and which
has been associated with protective immunity [22]. Our results
show that while in subjects with LTBI there was a high percentage
of IL-2+/IFN-c+and IL-22/IFN-c+peptide-specific CD8 T-cells,
the IL-2+/IFN-c+population was consistently reduced in TB
patients, highly suggestive of a restricted functional profile of Mtb-
specific CD8 T-cells during active disease. While to our knowledge
there has been no study comparing the cytokine response of CD8
T-cells at a single cell level in LTBI subjects and patients with
active TB, a recent study on the CD4 T-cell response to Esat-6
and CFP-10 reported that there was a shift in the IFN-c and IL-2
cytokine profile, notably from a dominance of IFN-c-only T-cells
in active tuberculosis to a dominance of IFN-c/IL-2-double
secreting T-cells [18]. These results, together with those reported
in this paper, suggest distinct T-cell functional signatures between
subjects which control Mtb infection (LTBI individuals) and those
who do not (active TB disease). Accordingly, studies on CD8 T-
cell responses in chronic infections such as HIV, CMV, EBV and
HCV [23,24] have highlighted signatures of protective antiviral
immunity: poly-functional (i.e. IL-2 and IFN-c secretion) and not
mono-functional (i.e. IFN-c only secreting) CD4 and CD8 T-cell
responses represent correlates of protective antiviral immunity in
chronic viral infections. Furthermore, the levels of antigen load
modulate the phenotypic and functional patterns of the T-cell
response within the same virus infection. Accordingly, Lalvani and
colleagues demonstrated that functional CD4 T-cell heterogeneity
is also associated with changes in Mtb antigen load: in active
disease, in which antigen load if high, IFN-c is secreted from two
functional subsets, namely IFN-c-only and IFN-c/IL-2 dual
secreting T-cells, whereas after therapy when antigen load is
low, IFN-c is predominantly secreted from IFN-c/IL-2 dual
secreting CD4 T-cells [18].
Although more extensive phenotyping of Mtb-specific IFN-c-
and IL-2-secreting T-cells is beyond the scope of this study,
previous studies have identified a relationship between the
function and phenotype of memory CD4 T-cells and proposed
that the IL-2-only secreting cells are typical of central memory T-
cells that persist after antigen clearance while the IFN-c/IL-2- and
IFN-c-only secreting T-cells are typical of effector memory T-cells
[20,22]. Accordingly, a recent paper reported that in children
vaccinated with BCG, specific CD8 T-cells identified by
intracellularIFN-c
secretion,
CD45RA2CCR72effector memory phenotype, while a central
memory population (CD45RA2CCR7+) was the second most
common [25].
Phenotypic and functional signatures of CD8 T-cells could be
used as an immunological marker of mycobacterial load, to
monitor the response to treatment, to evaluate new therapies for
active tuberculosis and the efficacy of new vaccines in clinical trials
where new biomarkers are needed. Moreover, phenotypic and
functional signatures of CD8 T-cells could also be used to monitor
individuals latently infected with Mtb at a high risk of progression
to active tuberculosis, such as those with HIV coinfection or on
anti-TNF therapy.
displayeda predominant
Materials and Methods
Human subject
Peripheral blood was obtained from 13 HLA-A*0201 positive
adults with TB disease (7 men, 6 women, age range 50–58 years)
from the Dipartimento di Medicina Clinica e delle Patologie
Emergenti, University Hospital, Palermo, and Monaldi Hospital,
Naples, Italy, and 9 HLA-A*0201 LTBI subjects (5 men, 4
women, age range 30–45 years) and 5 tuberculin (PPD)2negative
healthy subjects (3 men, 2 women, age range 35–55 years). TB
patients had clinical and radiological findings consistent with
active pulmonary TB [26]. Diagnosis was confirmed by bacteri-
ological isolation of Mtb in 12 patients and 1 further patient was
classified as having highly probable pulmonary TB on the basis of
clinical and radiological features highly suggestive of TB that were
unlikely to be caused by another disease and a decision was made
by the attending physician to initiate anti-tuberculosis chemother-
apy, which resulted in an appropriate response to therapy. All
patients were treated in accordance with italian guidelines and
received therapy for 6 months. Treatment was successful in all
participants as evidenced by no clinical or radiographic evidence
of current disease, the completion of anti-tuberculosis chemother-
apy and sterile mycobacterial cultures. Peripheral blood was
collected before (T0) and 4 months after chemotherapy (T4). The
follow-up time point of four months after starting therapy
following was chosen on the basis of previous studies by our and
other groups [19,27–29] which have demonstrated change in
many different immune responses in TB patients at this time point
after therapy, including the CD8 T cell phenotype in childhood
TB [14]. None of the TB patients had been vaccinated during
Figure 2. Phenotypic analysis of tetramer+Mtb-specific CD8 T-cells. Peripheral blood mononuclear cells (PBMC) were stained with individual
tetramers, and anti-CD8, -CD45RA and -CCR7 mAbs to separate functionally distinct subpopulations. After gating on tetramers+CD8+cells, the
percentage of cells expressing CD45RA and CCR7 was determined. (A) Representative phenotyping data for one subject with LTBI (LTBI), one TB
patient before (T0) and one TB patient 4 months after therapy (T4). Numbers in the corners indicate the percentage of positive cells in each quadrant.
(B) Summary cumulative data of the phenotype of tetramer+Mtb-specific CD8 T-cells. Data are presented with box plot reporting the median values
and the interquartile range. Black columns=LTBI; grey columns=T0; white columns=T4.
doi:10.1371/journal.pone.0005528.g002
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org6May 2009 | Volume 4 | Issue 5 | e5528

Page 7

CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org7May 2009 | Volume 4 | Issue 5 | e5528

Page 8

infancy with BCG, or had evidence of human immunodeficiency
virus (HIV) infection, or was being treated with steroid or other
immunosuppressive or anti-tubercular drugs at the time of their
first sampling. Tuberculin (purified protein derivative, PPD) skin
tests were considered positive when the induration diameter was
larger than 5 mm at 72 hrs since injection of 1 U of PPD (Statens
Seruminstitut, Copenaghen, Denmark). The study was approved
by the Ethical Committee of the Dipartimento di Medicina
Clinica e delle Patologie Emergenti, University Hospital, Palermo,
and Monaldi Hospital, Naples, Italy where the patients were
recruited. Informed consent was written by all participants.
For the identification of LTBI subjects, in the absence of a gold
standard, the most widely used diagnostic test remains the
tuberculin skin test, based on the delayed-type hypersensitivity
reaction that develops in Mtb infected individuals upon intrader-
mal injection of PPD. However, this test suffers from many
limitations, including false-negative results, especially in some
high-risk groups, and false-positive results in BCG-vaccinated
individuals or in subjects exposed to non-tuberculous mycobacte-
ria [30]. Moreover, in vitro release of IFN-c by T lymphocytes
upon stimulation with the Mtb-specific antigens Esat-6 and CFP-
10 (the T-SPOT.TB test-Oxford Immunotec, Oxford, United
Kingdom and the QuantiFERON-TB Gold test-QFT-G, Cellestis,
Victoria, Australia), was performed for all patients, for all LTBI
subjects that resulted positive, and for PPD2healthy donors
included in the present study [31,32].
Individuals with LTBI were defined as healthy people with a
positive tuberculin skin test and no symptoms and signs of active
TB. All of the LTBI subjects were health care workers, and thus
very likely to be close contacts of TB index cases. Moreover, all the
LTBI subjects included in this groups have not been vaccinated
with BCG.
All the subjects were HLA typed serologically. The HLA
subtype, A*0201, was confirmed by PCR amplification technique
using sequence-specific oligonucleotide primers.
HLA-A*0201 and b2-microglobulin
Recombinant HLA-A*0201 was over expressed in E. coli, purified
as described [33] and dissolved in 8M urea. The integrity of the
protein was confirmed by TOF-MALDI mass spectrometry using
insulin as an internal reference. Human b2-micro-globulin was
purchasedfromSigma(St.Louis,MO)anddissolvedinH2O.Heavy
chain (50 mM) stock solutions were stored at 220uC until use.
Synthetic peptides
A total of 6 nonamer peptides derived from the sequence of the
proteins of Mtb and containing HLA-A*0201-binding motifs [34]
were prepared using solid-phase/Fmoc chemistry, as described in
detail elsewhere [19,35,36]. The peptides were of 90% purity, and
their homogeneity was confirmed by analytical reverse-phase high-
performance liquid chromatography, mass spectroscopy, and amino
acid composition analysis. The sequences of the peptides were:
Ag85B p5–13 (GLPVEYLQV), Esat-6 p82–90 (AMASTEGNV),
Hsp65 p362–370 (KLQERLAKL), 16 kDa p120–128 (GILTVS-
VAV), Rv1490 p325–333 (FLLGLLFFV) and Rv1614 p197–205
(FLYELIWNV).Whereasthe former4peptideshave been identified
previously, the latter two are newly identified in this study as
candidate epitopes, based upon whole Mtb genome screening for 9-
mer peptides sequences with the high and intermediate predicted
HLA-A*0201 binding affinity as indicated in Table 1.
The two peptides derived from Rv1490 and Rv1614 were
identified after a full genome-wide screening with the HLA_BIND
algorithm. The two selected peptides had the highest predicted
score for binding to HLA-A*0201, and there are reported strong
associations between peptide binding HLA-A*0201 affinity and
epitope recognition for CD8 T-cells in tuberculosis [37]. The in
silico approach matched the strategy previously used to identify
CD8 T-cell epitopes across the entire Mtb genome [17,38].
Candidate HLA-A*0201 binding peptides in Hsp65 were
selected using the MOTIFS software described previously [39].
Figure 3. Polyfunctional cytokine production analysis of tetramer+Mtb-specific CD8 T-cells. Peripheral blood mononuclear cells (PBMC)
were stimulated with the same individual peptides as those present in tetramers and were stained with mAbs to CD8, IFN-c and IL-2, or with isotype-
control mAbs. After gating on CD8+cells, the percentage of cells expressing IFN-c and IL-2 was determined. (A) Representative intracellular cytokine
staining data in one subject with LTBI, one TB patient before therapy and one PPD2healthy donor. Numbers in the corners indicate the percentage
of CD8+cytokine-positive cells in each quadrant. (B) Summary cumulative data of the IFN-c and IL-2 secretion capability of tetramer+Mtb-specific
CD8 T-cells in LTBI subjects (white bars) and TB patients with active disease before therapy (black bars). The data are expressed as the percentage of
CD8+T-cells that are IFN-c+/IL-22or IFN-c+/IL-2+. The values reported are the mean percentage of the different subset analysed for each group tested
6 standard deviations (SD). *p,0.001 and **p,0.01 when compared to values in LTBI subjects.
doi:10.1371/journal.pone.0005528.g003
Figure 4. Peptides-specific CD8 T cell responses in LTBI subjects, in TB patients at T0 and T4. Peptides-specific CD8 T cell responses are
shown as a pie chart. Each portion of a pie chart indicates the percentage of peptides-specific T cells that responded with one or two functions, i.e.
producing IFN-c alone or the combination of IFN-c and IL-2 (see legend).
doi:10.1371/journal.pone.0005528.g004
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org8 May 2009 | Volume 4 | Issue 5 | e5528

Page 9

Positive scores were given for each potential anchor residue found
in the peptide, and negative scores were given to inhibitory
residues. The overall peptide score was the sum of the scores for
individual anchor and inhibitor residues.
HLA-A*0201-peptide binding assay
HLA-A*0201 was titered in the presence of 100 fmol standard
peptide to determine the HLA concentration necessary to bind
20–50% of the total fluorescent signal [35]. All subsequent
inhibition assays were then performed at this concentration. HLA-
A*0201 was incubated in 96-well plates (polypropylene, seroclus-
ter, Costar) at RT (pH 7) for 24 h with 0.5 ml b2M (15 pmol) and
1 ml (100 fmol) fluorescent labelled peptide in 92.5 ml assay buffer
(100 mM Na-phosphate, 75 mM NaCl, 1 mM CHAPS), 2 ml
protease inhibitor mixture (1 mM chymostatin, 5 mM leupeptin,
10 mM pepstatin A, 1 mM EDTA, 200 mM pefabloc) and 2 ml of
the peptides of which HLA-binding capacity was to be
determined. As a standard peptide we used FLPSDC(Fl)FPSV.
The HLA-peptide complexes were separated from free peptide by
gelfiltrationonaSynchropak
(250 mm64.6 mm; Synchrom, Inc., Lafayette, Indiana). Fluores-
cent emission was measured at 528 nm on a Jasco FP-920
fluorescence detector (B&L Systems, Maarssen, The Netherlands).
As HPLC running buffer, assay buffer containing 5% CH3CN was
used. The percentage of labelled peptide bound was calculated as
the amount of fluorescence bound to MHC divided by total
fluorescence. The concentration of peptide inhibitor yielding 50%
inhibition (IC50) was deduced from the dose-response curve. Each
peptide was tested in at least two separate experiments. Data are
expressed as relative (rel) IC50, compared to the IC50 of the
standard peptide HBV core p47–56, which was considered as 1.
Binding affinities of peptides to the HLA-A*0201 molecule were
defined as high (IC50,1 mM), intermediate (IC50 1–10 mM) and
low (IC50.10 mM). Note that, although the binding affinity of the
Ag85 peptide is intermediate, it has been shown that is able to
induce potent CD8 CTL activity [35].
GPC 100column
Generation of HLA-A*0201-Peptide Tetramers
Tetrameric HLA-A2-peptide complexes were prepared as
follows: recombinant HLA-A*0201 and human b2-microglobulin,
produced in Escherichia coli, were solubilized in urea and injected
together with each synthetic peptide into a refolding buffer
consisting of 100 mM Tris (pH 8.0), 400 mM arginine, 2 mM
EDTA, 5 mM reduced glutathione, and 0.5 mM oxidized
glutathione. Refolded complexes were purified by anion exchange
chromatography using DE52 resin (Whatman) followed by gel
filtration on a Superdex 75 column (Amersham Pharmacia
Biotech). The refolded HLA-A*0201-peptide complexes were
biotinylated by incubation for 16 hrs at 30uC with BirA enzyme
(Avidity, Denver). Tetrameric HLA-peptide complexes were
produced by the stepwise addition of extravidin-conjugated
phycoerythrin (PE) (Sigma) to achieve a 1:4 molar ratio
(extravidin-PE/biotinylated HLA class I). The PE-labelled HLA-
A*0201 tetramer complexed with the HIV-1 gag peptide p76–84
(SLYNTVATL was obtained from Proimmune Ltd. (Oxford, UK)
and used as a negative control of tetramer staining.
Tetramer staining and immunophenotyping
PE-labelled HLA-A*0201 tetramer complexes loaded with the
Mtb peptides Ag85B p5–13 (GLPVEYLQV), Esat-6 p82–90
(AMASTEGNV), Hsp65 p362–370 (KLQERLAKL), 16 kDa
p120–128 (GILTVSVAV), Rv1490 p325–333 (FLLGLLFFV)
and Rv1614 p197–205 (FLYELIWNV) were used throughout.
Tetramer staining was carried out as described in detail previously:
peripheral blood mononuclear cells (PBMC) were isolated from
heparinized blood samples by Ficoll-Hypaque (Sigma) density
centrifugation. PBMC were incubated in U-bottom 96-well plates,
washed twice in phosphate buffered saline (PBS, Euroclone,
Milan, Italy) containing 1% fetal calf serum (FCS, Sigma) and
stained for 30 min at 4uC with PE-labelled tetramers (3 ml),
washed and subsequently stained with FITC-labelled anti-CD8
mAb (clone HIT8a, BD Biosciences, San Jose `, CA) and analyzed
by flow cytometry on a FACS Calibur analyzer with the use of the
CellQuest software (BD Biosciences). Viable lymphocytes were
gated by forward and side scatter and the analysis was performed
on 100.000 acquired events for each sample.
To assess the phenotype of tetramer+T-cells, cells were stained
with FITC-labelled anti-CD8 mAb, APC-labelled anti-CCR7
(clone 3D12) mAb and PE-Cy5-labelled anti-CD45RA mAb
(clone HI100) all from BD Biosciences in incubation buffer (PBS-
1% FCS-0,1% Na azide) for 30 min at 4uC. Cells were then
washed twice in PBS 1% FCS and analyzed by flow cytometry as
previously described. Analysis was performed on 100.000 acquired
events for each sample.
Intracellular cytokine staining
PBMC (106/ml) were stimulated with peptides (1 mg/ml, final
concentration), in the presence of monensin for 6 hrs at 37uC in
5% CO2. The cells were harvested, washed and stained with APC-
conjugated anti-CD8 mAb (BD) in incubation buffer (PBS-1%
FCS-0.1% Na azide) for 30 min at 4uC. The cells were washed
twice in PBS-1% FCS and fixed with PBS-4% paraformaldehyde
overnight at 4uC. Fixation was followed by permeabilization with
PBS–1% FCS–0.3% saponin–0.1% Na azide for 15 min at 4uC.
Staining of intracellular cytokines was performed by incubation of
fixed permeabilized cells with PE-labelled anti-IFN-c (clone B27)
and FITC-labelled IL-2 antibody (clone MQ1-17H12) or an
isotype-matched control mAb. All mAbs were from BD Biosci-
ences. Cells were acquired and analysed by FACS as described
above.
Statistical considerations
Negative control (background) values for cytokine staining were
not subtracted from peptide-induced responses. For phenotype
distribution analysis we used a cut off of a minimum number of 50
events, with a mean of 250 events for all tetramers tested and for
each group of individuals tested.
Nonparametric Mann-Whitney U test was used to determine
statistical differences in the distribution of the results. Values of
p,0.05 were considered significant. Data were analyzed using
statistical software SYSTAT 11 (Systat Software).
Acknowledgments
We would like to thank Sarah Klein who actually did the MTB genome
scan for potential HLA-A*0201 peptide-binding sequences. We thank
Annemieke Geluk for making available results from peptide/HLA-A2
binding experiments.
Author Contributions
Conceived and designed the experiments: NC AS FD THMO. Performed
the experiments: GG SM GG. Analyzed the data: NC SM. Contributed
reagents/materials/analysis tools: GG GG PDC LT MLB DG AM JN
MK AS. Wrote the paper: NC FD THMO.
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org9 May 2009 | Volume 4 | Issue 5 | e5528

Page 10

References
1. WHO (2008) Global Tuberculosis Control: Surveillance, Planning, Financing.
Geneva: World Health Organization, Available at http://www.who.int/entity/
tb/publications/global_report/2008/pdf/fullreport.pdf.
2. Flynn JL, Goldstein MM, Triebold KJ, Koller B, Bloom BR (1992) Major
histocompatibility complex class I-restricted T cells are required for resistance to
Mycobacterium tuberculosis infection. Proc Natl Acad Sci USA 89: 12013–12017.
3. Liebana E, Girvin RM, Welsh M, Neill SD, Pollock JM (1999) Generation of
CD8+T-cell responses to Mycobacterium bovis and mycobacterial antigen in
experimental bovine tuberculosis. Infect Immun 67: 1034–1044.
4. Koga T, Wand-Wurttenberger A, DeBruyn J, Munk ME, Schoel B, et al. (1989)
T cells against a bacterial heat shock protein recognize stressed macrophages.
Science 4922: 1112–1115.
5. Stenger S, Mazzaccaro RJ, Uyemura K, Cho S, Barnes PF, et al. (1997)
Differential effects of cytolytic T cell subsets on intracellular infection. Science
276: 1684–1687.
6. Cho S, Mehra V, Thoma-Uszynski S, Stenger S, Serbina N, et al. (2000)
Antimicrobial activity of MHC class I-restricted CD8+T cells in human
tuberculosis. Proc Natl Acad Sci U S A 97: 12210–5.
7. Kaufmann SH, McMichael AJ (2005) Annulling a dangerous liaison: vaccination
strategies against AIDS and tuberculosis. Nat Med 11: S33–44.
8. Flynn JL, Chan J (2001) Tuberculosis: latency and reactivation. Infect Immun
69: 4195–201.
9. Lalvani A, Brookes R, Wilkinson RJ, Malin AS, Pathan AA, et al. (1998) Human
cytolytic and interferon gamma-secreting CD8+T lymphocytes specific for
Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 95: 270–275.
10. Ottenhoff THM, Lewinsohn DA, Lewinsohn DM (2008) Human CD4 and CD8
T cell responses to Mycobacterium tuberculosis: antigen specificity, function,
implications and applications. Handbook of Tuberculosis. Immunology and
Cell Biology. Kaufmann SHE, Britton WJ, eds. Weinheim, Germany: Wiley-
VCH Verlag GmbH & Co.KGaA.
11. Klein MR, Fox A (2001) Mycobacterium-specific human CD8 T cell responses.
Arch Immunol Ther Exp (Warsz) 49: 379–89.
12. Tully G, Kortsik C, Ho ¨hn H, Zehbe I, Hitzler WE, et al. (2005) Highly focused
T cell responses in latent human pulmonary Mycobacterium tuberculosis infection.
J Immunol 174: 2174–84.
13. Gambon-Deza F, Pacheco Carracedo F, Cerda Mota T, Montes Santiago J
(1995) Lymphocyte populations during tuberculosis infection: V repertoires.
Infect Immun 63: 1235–40.
14. Caccamo N, Meraviglia S, La Mendola C, Guggino G, Dieli F, et al. (2006)
Phenotypical and functional analysis of memory and effector human CD8 T
cells specific for mycobacterial antigens. J Immunol 177(3): 1780–5.
15. Jacobsen M, Detjen AK, Mueller H, Gutschmidt A, Leitner S, et al. (2007)
Clonal expansion of CD8+effector T cells in childhood tuberculosis. J Immunol
179: 1331–9.
16. Geginat J, Lanzavecchia A, Sallusto F (2003) Proliferation and differentiation
potential of human CD8+memory T-cell subsets in response to antigen or
homeostatic cytokines. Blood 101: 4260–66.
17. Hammond AS, Klein MR, Corrah T, Fox A, Jaye A, et al. (2005) Mycobacterium
tuberculosis genome-wide screen exposes multiple CD8 T cell epitopes. Clin Exp
Immunol 140: 109–16.
18. Millington KA, Innes JA, Hackforth S, Hinks TS, Deeks JJ, et al. (2007)
Dynamic relationship between IFN-c and IL-2 profile of Mycobacterium
tuberculosis-specific T cells and antigen load. J Immunol 178: 5217–26.
19. Dieli F, Singh M, Spallek R, Romano A, Titone L, et al. (2000) Change of Th0
to Th1 cell-cytokine profile following tuberculosis chemotherapy.
Scand J Immunol 52: 96–102.
20. Dieli F, Friscia G, Di Sano C, Ivanyi J, Singh M, et al. (1999) Sequestration of T
lymphocytes to body fluids in tuberculosis: reversal of anergy following
chemotherapy. J Infect Dis 180: 225–28.
21. Dunne PJ, Belaramani L, Fletcher JM, Fernandez de Mattos S, Lawrenz M, et
al. (2005) Quiescence and functional reprogramming of Epstein-Barr virus
(EBV)-specific CD8+T cells during persistent infection. Blood 15;106(2):
558–65.
22. Seder RA, Darrah PA, Roederer M (2008) T-cell quality in memory and
protection: implications for vaccine design. Nat Rev Immunol 8(4): 247–58.
23. Harari A, Dutoit V, Cellerai C, Bart PA, Du Pasquier RA, et al. (2006)
Functional signatures of protective antiviral T-cell immunity in human virus
infections. Immunol Rev 211: 236–54.
24. Pantaleo G, Koup RA (2004) Correlates of immune protection in HIV-1
infection: what we know, what we don’t know, what we should know. Nat Med
10: 806–10.
25. Soares AP, Scriba TJ, Joseph S, Harbacheuski R, Murray RA, et al. (2008)
Bacillus Calmette-Gue ´rin vaccination of human newborns induces T cells with
complex cytokine and phenotypic profiles. J Immunol 180: 3569–3577.
26. Diagnostic standards and classification of tuberculosis in adults and children: this
official statement of the American Thoracic Society and the Centers for Disease
Control and Prevention was adopted by the ATS Board of Directors, July (1999)
This statement was endorsed by the Council of the Infectious Disease Society of
America, September (1999). Am J Respir Crit Care Med 161: 20001376–1395.
27. Carrara S, Vincenti D, Petrosillo N, Amicosante M, Girardi E, et al. (2004) Use
of a T cell-based assay for monitoring efficacy of antituberculosis therapy. Clin
Infect Dis 38: 754–756.
28. Di Liberto D, Buccheri S, Caccamo N, Meraviglia S, Romano A, et al. (2007)
Decreased serum granulysin levels in childhood tuberculosis which reverse after
therapy. Tuberculosis 87: 322–328.
29. Dieli F, Sireci G, Caccamo N, Di Sano C, Titone L, et al. (2002) Selective
depression of interferon-gamma and granulysin production with increase of
proliferative response by Vgamma9/Vdelta2 T cells in children with
tuberculosis. J Infect Dis 186: 1835–1839.
30. Menzies D (1999) Interpretation of repeated tuberculin tests: boosting,
conversion and reversion. Am J Respir Crit Care Med 159: 15–21.
31. Richeldi L (2006) An update on the diagnosis of tuberculosis infection.
Am J Respir Crit Care Med 174: 736–742.
32. Menzies D, Pai M, Comstock G (2007) Meta-analysis: new tests for the diagnosis
of latent tuberculosis infection: areas of uncertainty and recommendations for
research. Ann Intern Med 146: 340–354.
33. Garboczi DN, Hung DT, Wiley DC (1992) HLA-A2 peptide complexes:
Refolding and crystallization of molecules expressed in E. coli and complexed
with single antigenic peptides. Proc Natl Acad Sci USA 89: 3429.
34. Rammensee HG, Bachmann J, Emmerich NPN, Bachor OA, Stevanovic S
(1999) SYFPEITHI: database for MHC ligands and peptide motifs. Immuno-
genetics 50: 213–9.
35. Geluk KE, van Meijgaarden KLMC, Franken JW, Drijfhout S, D’Souza A, et al.
(2000) Identification of a major epitope of Mycobacterium tuberculosis Ag85B that is
recognized by HLA-A*0201-restricted CD8+T cells in HLA-transgenic mice
and humans. J Immunol 165: 6463–6471.
36. Geluk A, van Meijgaarden KE, Drijfhout JW, Ottenhoff TH (1995) Clip binds
to HLA class II using methionine-based, allele-dependent motifs as well as allele-
independent supermotifs. Mol Immunol 32: 975–981.
37. Teixeira A, Benckhuijsen WE, de Koning PE, Valentijn AR, Drijfhout JW
(2002) The use of DODT as a non-malodorous scavenger in Fmoc-based
peptide synthesis. Protein Pept Lett 9: 379–385.
38. Klein MR, Hammond AS, Smith SM, Jaye A, Lukey PT, et al. (2002) HLA-
B*35-restricted CD8(+)-T-cell epitope in Mycobacterium tuberculosis Rv2903c.
Infect Immun 70: 981–4.
39. D’Amaro J, Houbiers JG, Drijfhout JW, Brandt RM, Schipper R, et al. (1995) A
computer program for predicting possible cytotoxic T lymphocyte epitopes
based on HLA class I peptide binding motifs. Hum Immunol 43: 13–8.
CD8 T-Cell in Tuberculosis
PLoS ONE | www.plosone.org10May 2009 | Volume 4 | Issue 5 | e5528