1565
Frequency and Phenotyping of Human Immunodeficiency Virus
(HIV)–Specific CD8+ T Cells in HIV-Infected Children, Using
Major Histocompatibility Complex Class I Peptide Tetramers
Daniel Scott-Algara,1 Florence Buseyne,2
Stephane Blanche,3 Christine Rouzioux,4 C. Jouanne,1
F. Romagné,5 and Yves Rivière2,a
1
Unité d’Immunohématologie-Immunopathologie and 2Laboratoire
d’Immunopathologie Virale, Centre National de la Recherche
Scientifique Unité de Recherche Associée 1930, Institut Pasteur,
and 3Fédération de Pédiatrie et 4Laboratoire de Virologie, Hôpital
Necker, Paris, and 5Beckman Coulter Immunotech, Marseille, France
HLA-A*02 tetramers complexed to human immunodeficiency virus (HIV) Gag SLYNTVATL and HIV Pol ILKEPVHGV peptides were used to characterize HLA class I–restricted
CD8+ T cells in 41 HIV-infected children. The frequencies and the phenotype of specific
circulating CD8+ T cells were determined in whole-blood samples by means of cytometric
analysis. Background staining of 13 HLA-A*02–negative patients showed that the frequency
of CD8+ T cells was !0.01%. Of the 28 HLA-A*02–positive patients, blood samples from 26
stained positive at least once the Gag tetramer (mean CD8+ T cells, 0.87%; range, 0.1%–3.9%),
and blood samples from 21 stained positive for the Pol tetramer (mean CD8+ T cells, 0.59%;
range, 0.1%–5.5%). The tetramer-binding cells were CD28⫺, CD45RA⫺, CD45RO+, HLADR+, and CD69⫺ T lymphocytes. HIV-specific CD8+ T cells can be detected easily in peripheral
blood of HIV-infected children, using HLA tetramers combined with HIV peptides. These
cells are memory activated CD28⫺CD8+ T lymphocytes.
Cytotoxic T lymphocytes (CTL) are important in the control
of viral infections [1, 2]. A strong CTL response has been implicated in the control of disease progression and in protection
against both human immunodeficiency virus (HIV) type 1 infection in humans and simian immunodeficiency virus infection
in macaques [3–9]. An inverse correlation between virus load
and HIV-1–specific CTL precursor frequencies has been reported for adult patients with primary HIV-1 infection [10, 11].
A selective reduction of HIV-1–specific CTL frequency was
observed in patients in the advanced stages of HIV-1 infection
[6, 12]. More recently, an inverse correlation between viremia
and the number of circulating CD8⫹ T cells stained with HLAA*02 HIV-1 Gag and Pol tetramers has been described for adult
patients with chronic HIV infection [13, 14]. Tetramer staining
allows for the direct visualization of CD8⫹ effector cells by
means of flow cytometry and provides a highly sensitive and
specific tool for the quantitation of the frequency of cytotoxic
Received 4 December 2000; revised 26 February 2001; electronically published 27 April 2001.
Consent was obtained from parents or legal guardians of the study subjects, and the institutional review ethical board approved the study.
Financial support: “Agence Nationale de la Recherche sur le SIDA”:
Centre National de la Recherche Scientifique; Institut Pasteur; Pediatric
AIDS Foundation.
a
Y.R. is an Elisabeth Glaser scientist.
Reprints or correspondence: Dr. Y. Rivière, Laboratoire d’Immunopathologie Virale, URA CNRS 1930, Institut Pasteur, 28 rue du Dr Roux,
75015 Paris, France ([email protected]).
The Journal of Infectious Diseases 2001; 183:1565–73
䉷 2001 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2001/18311-0003$02.00
CD8⫹ cells, without use of in vitro culture. These new tools for
the detection of circulating virus-specific CD8⫹ T cells can be
combined with phenotypic characterization of the cells, which
will allow for a better understanding of the dynamics of differentiation and activation of these cells.
Previous studies have suggested that HIV-1–specific cytotoxic
CD8⫹ T cells in HIV-infected children have similar frequency
and specificity to those found in HIV-infected adults, after in
vitro stimulation and expansion [3, 15, 16]. However, only a
small percentage of infants born to infected mothers had detectable circulating ex vivo–activated CTL, compared with that
in adults, as measured by using a conventional chromium release assay [3]. A large proportion of circulating CD8⫹ T cells
in HIV-infected adult donors are activated, as evidenced by
high levels of expression of CD38, HLA-DR, and CD57 antigens, as well as cytolytic serine esterase granzyme A. In HIV
infection, there is an increased proportion of activated memory
T cells that do not express CD28. The circulating HIV-specific
cytotoxicity is mediated by the CD28⫺CD8⫹ T cell subpopulation [17]. After activation by T cell–receptor engagement, the
CD28⫺CD8⫹ T cells produce interferon (IFN)–g and have reduced proliferative capacity [18], which suggests that these cells
are terminally differentiated effector CTL.
We used complex of peptide and major histocompatibility
complex (MHC) class I complex tetramers to study CD8⫹ T
cells that recognize 2 HLA-A*02–restricted HIV-1 CTL Gag
and Pol epitopes in HIV-infected children. Our data show that
HIV-specific CD8⫹ T cells can be detected easily in peripheral
lymphocytes in whole-blood samples obtained from HIV-infected children, using HLA tetramers combined with HIV pep-
1566
Scott-Algara et al.
tides. Furthermore, CD8⫹ T cells that bind tetramers are memory activated CD8⫹ T cells (CD28⫺, CD45RA⫺, CD69⫺, and
HLA-DR⫹), and their frequencies are comparable with those
in HIV-infected adults. Finally, in a longitudinal study of a
child treated with potent antiretroviral therapy (ART), the rela
tively high frequency of these cells appeared to be maintained
by the presence of HIV-1 replication, since the detection of
epitope-specific CD8⫹ T cells declined only when viral replication is reduced markedly by antiretroviral drugs.
Patients and Methods
Patients. All patients were followed prospectively at the Hôpital
Necker (Paris). Most patients were born to HIV-infected mothers.
Four were infected by blood transfusion (patients EM50, EM66,
EM67, and EM76). Of the 41 patients in the study, 28 were positive
for HLA-A*02, and 13 were negative for HLA-A*02. The patients’
mean age was 13.3 years. Twenty-three (79%) of the 28 HLAA*02–positive patients were studied at least twice. For each patient,
biological data, the Centers for Disease Control and Prevention
(CDC) disease stage designation [19], and the ART at the time when
the patient was tested for tetramer staining are shown in table 1. For
the HLA-A*02–positive patients, mean values were as follows: CD4⫹
T cells, 27% (range, 3%–53%); CD8⫹ T cells, 49.6% (range, 27%–
79%), and log10 HIV-1 RNA level, 3.6 copies/mL (range, !1.3–
5.2 copies/mL). The biological characteristics of HLA-A*02–
positive and –negative patients were not significantly different.
Tetrameric peptide–MHC class I complexes. The HLA-A*02
tetrameric complexes (Tetra) used in this study were synthesized
as described elsewhere [20, 21]. The peptide–MHC class I complexes were formed with 2 HIV-1 immunodominant epitopes, one
in the p17 region of Gag (SLYNTVATL or SL9) and the other in
reverse transcriptase (RT; Pol, ILKEPVHGV or IV9).
Flow cytometry analyses. Absolute CD4 and CD8 cell counts
were performed on fresh whole-blood samples by means of 4-color
flow cytometry analysis of cells positive for CD45, CD3, CD4, or
CD8 T cells, using fluorescent beads as an internal standard (flowcount beads; Beckman Coulter Immunotech). CD8⫹ T cell populations were studied in the same whole-blood samples by means of
4-color flow cytometry analysis and standard lysis procedure of
red blood cells. After lysis, cells were washed once in PBS, were
adjusted to 500 mL in PBS, and were stored at 4⬚C until analysis.
The following associations were used: 1, CD45–fluorescein isothiocyanate (FITC), CD3–phycoerythrin-cyanin 5.1 (PCy5), CD4–
phycoerythrin (RD1), and CD8–phycoerythrin–Texas Red (ECD);
2, CD3-ECD, CD8-PyC5, Tetra-phycoerythrin (PE), HLA-DRFITC; 3, CD3-ECD, CD8-PyC5, Tetra-PE, and CD28-FITC; 4,
CD3-ECD, CD8-PCy5, Tetra-PE, and CD45RA-FITC; 5, CD4ECD, CD8-PCy5, Tetra-PE, and CD16-FITC; 6, CD3-ECD, CD8PCy5, Tetra-PE, and CD45RO-FITC; 7, CD3-ECD, CD8-PCy5,
Tetra-PE, and CD69-FITC. Naive cells were defined by the expression of CD45RA. All cells that were not coexpressing CD45RA
but were expressing CD45RO were considered to be memory cells.
Cells expressing HLA-DR and/or CD69 were considered to be
activated T cells. Labeled cells were analyzed on an EPICS-XL
MCL flow cytometer (Coulter Electronics). Event accumulations
JID 2001;183 (1 June)
were followed until reaching 50,000 living cells on the lymphocyte
gate that was set up by using both forward and right-angle scatters.
All monoclonal antibodies were purchased from Beckman Coulter
Immunotech.
Plasma HIV-1 RNA monitoring. The Amplicor HIV Monitor
test (Roche) was used to quantitate HIV-1 RNA levels in plasma.
This technique is based on reverse transcription of the RNA that
corresponds to the HIV gag gene, which is followed by amplification of the cDNA by polymerase chain reaction (PCR) assay in
the presence of a quantitation standard [22]. The cut-off value
varied from 1.3 to 2.3 log copies/mL.
HLA class I typing. HLA genotyping was performed in the
Department of Immunology at the Chelsea and Westminster Hospital
(London), using amplification refractory mutation system PCR [23].
Statistical analysis.
The biological parameters of HLAA*02–positive or –negative patients were analyzed by the 2-tailed
Student’s t test and by the x2 test. The correlation among the
frequency of tetramer-binding cells, plasma virus load, and the
percentage of CD4⫹ and CD8⫹ T cells was determined by using
Spearman’s rank test. Virus load and lymphocyte subsets from the
different groups of HLA-A*02–positive children were compared
by use of the Mann-Whitney U test. P ! .05 was considered to be
statistically significant.
Results
Frequency of Gag- and Pol-specific CD8⫹ T cells. Our first
objective was to determine the specificity and the frequency of
epitope-specific CD8⫹ T cells in our cohort of HIV-infected
children. Background binding in samples from the 13 HLAA*02–negative patients was !0.01% of CD8⫹ T cells (not
shown). Specific binding of peptide–MHC class I Gag and Pol
tetramers was observed only in the CD3⫹CD8⫹ T cell subpopulation in samples from HLA-A*02–positive patients, and no
specific binding was detected in the CD3⫹CD4⫹ subpopulation
(figure 1). Of the 28 HLA-A*02–positive patients tested, 26 had
samples that stained positive at least once for the Gag tetramer
(mean CD8⫹ T cells, 0.87%; range, 0.1%–3.9%), and 21 had
samples that stained positive for the Pol tetramer (mean of
CD8⫹ T cells, 0.59%; range 0.1%–5.5%; figure 2). Twenty-three
patients were tested 2–6 times. Of these 23 patients, 15 (65%)
were found repeatedly to be positive for the Gag tetramer–
specific CD8⫹ T cells; 1 (4%) patient was found repeatedly to
be negative, and the remaining 7 (30%) were found to be positive in some, but not all, samples. Nine (39%) of the 23 patients
were found repeatedly to be positive for Pol tetramer–specific
CD8⫹ T cells, 6 (26%) were found repeatedly to be negative,
and the remaining 8 (35%) were found to be positive in some,
but not all, samples. Of the 20 patients who displayed binding
to both Gag and Pol tetramers, 13 had higher frequencies of
CD8⫹ cells that recognize the Gag epitope.
Among the 28 HLA-A*02–positive children, a cross-sectional
analysis of the frequency of either Gag tetramer– or Pol tetramer–binding cells failed to show any correlation with plasma
CD8⫹ Tetramer-Positive T Cells in HIV-Infected Children
JID 2001;183 (1 June)
Table 1.
Characteristics of patients.
HLA-A*02 status,
patient
Positive
EM003
EM018
EM020
EM023
EM025
EM026
EM028
EM030
EM031
EM040
EM042
EM045
EM050
EM054
EM063
EM064
EM066
EM067
EM071
EM076
EM078
EM080
EM083
EM086
EM089
EM096
EM102
EM103
Negative
EM002
EM060
EM068
EM072
EM074
EM079
EM082
EM087
EM093
EM094
EM095
EM097
EM106
1567
Age,
years
ART
CD3⫹ T
b
cells, %
CD3⫹CD4⫹
b
T cells, %
CD3⫹CD8⫹
a,b
T cells, %
11.6
13.6
16.6
14.3
14.5
12.6
11.8
13.1
17
14.4
9.9
9.3
16.8
8.2
7.9
15.3
10
16
12.1
12
8.8
12.6
12
10.4
3.9
15
2.3
2.9
2
2
2
2
0
3
0
0
0
0
3
1
2
3
2
2
1
1
2
0
1
3
3
1
1
2
3
0
86
72
84
63
76
79
90
88
86
88
80
82
82
70
79
92
87
90
79
81
86
66
71
89
92
89
84
62
28
31
25
24
14
28
53
12
13
41
20
26
18
37
23
2
19
52
34
21
28
29
19
42
28
21
39
28
54
38
56
35
60
45
36
73
65
44
56
49
61
27
53
81
64
36
40
55
54
33
35
43
59
67
42
30
12
7.2
14.2
5.6
12
13.6
13.2
14
12.2
3.2
5.2
10.8
7.4
2
3
0
3
0
2
3
3
0
3
3
3
3
46
74
76
85
70
68
80
77
80
77
67
76
84
9
28
19
38
19
3
31
37
28
30
34
19
27
31
39
53
39
47
59
46
38
44
45
26
49
48
a
Virus load,
log copies/mL
3.6
c
Stage
3.4
4.1
4.9
2.3
!2.2
5.0
4.8
2.6
3.8
3.4
4.0
!1.3
5.1
5.3
4.3
2.7
!2.2
5.2
3.6
4.1
3.0
3.8
4.2
2.3
4.7
4.1
B
B
C
B
N
A
B
A
N
N
A
N
B
N
C
C
A
N
C
B
B
A
A
A
N
A
C
N
3.2
5.0
5.0
4.2
4.1
2.7
3.1
3.4
3.6
4.5
4.2
3.6
2.9
A
B
B
C
B
A
A
A
A
N
A
A
A
!1.3
NOTE. ART, antiretroviral therapy; HIV, human immunodeficiency virus; RT, reverse transcriptase.
a
ART groups: 0, no treatment; 1, 2 RT inhibitors; 2, tritherapy with a nonnucleoside RT inhibitor; and 3,
tritherapy with an HIV protease inhibitor.
b
Determined by flow cytometry.
c
Stage of HIV disease, as defined by the Centers for Disease Control and Prevention [19].
virus load or with the percentage of CD4⫹ or CD8⫹ T cells.
No difference in the frequency of tetramer-binding cells was
observed according to the disease stage. When the groups of
children (group 0, untreated; group 1, 2 RT inhibitors; group
2, tritherapy with an nonnucleoside RT inhibitor (NNRTI); and
group 3, tritherapy with an HIV protease inhibitor [PI]) were
paired and compared, no differences were observed for virus
load, percentage of CD4⫹ or CD8⫹ T cells, or frequency of Pol
tetramer–binding cells. For Gag tetramer–binding cells, group
1 had a significantly higher frequency than did group 2 (P !
.03, Mann-Whitney U test), and there was a trend for higher
frequency group 1 than in groups 0 or 3 (P ! .11 and P ! .06,
respectively, Mann-Whitney U test). No differences were observed between group 0 and group 2 or 3 and between group
2 and group 3.
Phenotype of tetramer-binding cells.
The phenotype of
tetramer-binding cells was studied for 20 HLA-A*2–positive
patients. The results for a representative example are shown in
figure 3. The median (range) values for the percentage of expression of the various surface antigens on gated CD8⫹ cells
that bind Gag tetramers were 12% (0%–27%) for CD45RA,
94% (81%–99%) for CD45RO, 8% (0%–12%) for CD28, 91%
1568
Scott-Algara et al.
JID 2001;183 (1 June)
Figure 1. Tetramer staining of peripheral lymphocytes in whole-blood sample from an HLA-A*02–positive human immunodeficiency virus
(HIV)–infected child. Whole blood was incubated with a combination of CD4, CD8 specific monoclonal antibodies and tetramer molecules. After
washing, cells were analyzed by means of flow cytometry. Percentages of CD8⫹ and CD4⫹ tetramer-binding and tetramer-negative cells are shown.
(87%–95%) for HLA-DR, and 0% for CD69. A similar result
was obtained with the Pol tetramer–binding CD8⫹ cells. More
than 98% of tetramer-binding CD8⫹ cells expressed the CD3
antigen. The majority of tetramer-binding CD8⫹ cells lacked
expression of CD28 antigen and expressed CD45RO memory
cells but not CD45RA naive cell markers. However, CD45RApositive cells with intermediate fluorescence intensity were detected in a few of the patients. The occurrence of CD45RAbrightpositive cells was very rare and accounted for !0.1% of cells
in our study, as were CD45RA⫺/CD45RO⫺ double-negative
cells and CD45ROlow cells. This indicates that HIV-specific tetramer-binding cells were predominantly memory cells. The majority of tetramer-binding CD8⫹ cells expressed the HLA-DR
activation marker. The lack of expression of the activation
marker CD69 indicates that fresh Gag tetramer– or Pol tetramer–positive cells recently had not been activated in vivo in
these patients. Thus, the phenotype of peptide–MHC class I
Gag tetramer– and Pol tetramer–positive CD8⫹ cells in this
group of HIV-infected children corresponded to antigen-experienced memory CD8⫹ T cells.
Longitudinal study of tetramer-binding T cells during ART.
We were able to follow the changes in the frequency of Gag
tetramer– or Pol tetramer–binding CD8⫹ cells in a 13-year-old
child receiving a new potent ART that was initiated 2 months
after the interruption of the previous ART. The new potent
ART was a combination of stavudine, abacavir, and efavirenz
that was given from week 0 to week 29, which then was changed
to stavudine, abacavir, and nevirapin that was given from week
29 to week 33. The plasma RNA copy numbers declined from
week 0 to week 21 (figure 4A). After several weeks of the new
potent ART, this child showed a significant decline in the frequency of Gag-specific CD8⫹ T cells, which was concomitant
with the decrease in virus load. Later, during follow-up, an
increase of Gag-specific CD8⫹ T cells was associated with an
increase in plasma virus load (figure 4A). The reduction in the
frequency of tetramer-binding cells in this patient was associated with CD4⫹ and CD8⫹ T cell changes (figure 4B). The
decrease in the frequency of tetramer-positive cells correlated
positively with the percentage of circulating CD8⫹ T cells
(r p .99; P ! .001), correlated negatively the percentage of cir-
JID 2001;183 (1 June)
CD8⫹ Tetramer-Positive T Cells in HIV-Infected Children
1569
though the plasma virus load was below the threshold of detection (!1.3 copies/mL) for patient EM54, the levels of Gag
tetramer– and Pol tetramer–binding cells were low but detectable. However, from week 36 to week 84 of ART, there was a
2-fold reduction in the total number of Gag tetramer– and Pol
tetramer–binding cells expressed in the percentage of CD8⫹ T
cells (0.7%–0.4 %) and in the absolute T8 number (5175 to 2528
cells/mL). In the 3 other patients, the virus load was not controlled. Despite high levels of plasma virus load during followup, patient EM080 had a very low frequency of both the Gagtetramer or Pol-tetramer CD8⫹ populations. The frequencies
of the Gag-tetramer and Pol-tetramer CD8⫹ populations were
stable for patient EM050, as was the patient’s virus load. The
changes in the frequency of peptide-MHC Gag-tetramer– or
Pol-tetramer–positive CD8⫹ T cells appeared to be parallel to
changes in the plasma virus load for patient EM067 (figure 5).
Among the 5 patients studied, there was no change in the
phenotype of Gag tetramer– or Pol tetramer–binding CD8⫹ T
cells that were consistently CD3⫹, HLA-DR⫹, CD45RO⫹,
CD45RA⫺, CD28⫺, or CD69⫺ (data not shown), despite the
variation in their frequencies during follow-up.
Discussion
Figure 2. Frequencies of CD3⫹CD8⫹ tetramer-binding cells, which
are expressed as percentages of total CD8⫹ lymphocytes for the 28
HLA-A*02–positive human immunodeficiency virus (HIV)–infected
children in this study. Analysis was performed as described in figure
1. Antiretroviral therapy (ART) groups are as follows: 0, no treatment;
1, 2 reverse-transcriptase (RT) inhibitors; 2, triple-drug combination
therapy with a nonnucleoside RT inhibitor; 3, triple-drug combination
therapy with an HIV protease inhibitor. Virus load is expressed in log10
copies/mL.
culating CD4⫹ T cells (r p ⫺.93; P ! .01), and correlated positively to virus load (r p .97; P ! .001). After week 21, increase
of the percentage of Gag tetramer-positive CD8⫹ cells correlated positively with an increase in the percentage of total CD8⫹
T cells and virus load.
Four other HIV-infected children who were receiving ART
were tested several times (figure 5). The first assay for the detection of tetramer-binding cells was performed between 32 and
116 months after initiation of ART. As shown in figure 5, al-
In the present study of 28 HIV-1–infected HLAA*02–positive children, we found that a large proportion of
these children had circulating CD3⫹CD8⫹ T cells that were
specific to 1 or 2 viral epitopes. The HIV-infected children in
our study had 93% and 75% of CD8⫹ cells that recognized,
at least once, the Gag (SLYNTVATL) and the Pol (ILKEPVHGV) epitopes, respectively. The lower proportion of
CD8⫹ cells that recognized the Pol epitope, compared with the
proportion that recognized the Gag epitope, may reflect the
fact that fewer Pol epitopes are expressed by infected cells [24].
Despite the lower Pol frequencies, 71% of the patients had
circulating CD8⫹ T cells that recognized both epitopes. Our
findings differ from the exclusivity of epitope recognition that
was reported for HIV-1–infected adult patients [25]. The frequencies detected in children are comparable to those of infected adult patients and support the immunodominance of
these 2 epitopes [13, 14, 25]. In HIV infection, evidence for viral
escape from HIV-specific CTL recognition has been reported
elsewhere [26, 27]. In our study, the lack of binding of the 2
immunodominant epitopes could be due to viral escape that is
generated in vivo. The high frequency of children with peripheral blood mononuclear cells (PBMC) that bind tetramers is
consistent with an absence of immune selection pressure by
immunodominant HLA-A*0201–restricted cytotoxic response,
as reported elsewhere for adults with chronic HIV-1 infection
[28]. There is a discrepancy between our results, which showed
that the CD8⫹ T cells in nearly all the HLA-A*2.01–positive
children recognize the SL9 epitope, and those reported recently
by Betts et al. [30], who found that the response to the SL9
1570
Scott-Algara et al.
JID 2001;183 (1 June)
Figure 3. Surface marker expression of CD8⫹ tetramer-binding cells: representative flow cytometric analysis of an infected child. Results are
expressed as the percentage of positive cells among tetramer-binding CD8⫹ lymphocytes (gated CD8⫹ tetramer-positive cells). The coexpression
of CD3, CD28, CD45RO, CD45RA, HLA-DR, and CD69 in fresh Gag- and Pol-specific CD8⫹ cells is shown.
epitope was detected in only 4 of 10 HIV-1–infected subjects.
However, in their study, the detection of peptide-specific T cells
was not performed by the measure of binding of PBMC with
HLA-A02 tetramers but by the measure of intracellular IFNg production by CD8⫹ T cells after peptide stimulation [29].
Thus, it is possible that this difference in the frequencies of
detection of peptide-specific CD8⫹ T cells is due to distinct
technical approaches.
The tetramer technology allows for not only the estimation
of the frequency of epitope-specific CD8⫹ T cells but also the
possibility for the characterization of the phenotypes that correspond to various differentiation stages of reacting T cells.
CD45RO is considered to be a marker for activated and memory T cells [18], whereas CD45RA is considered to be a marker
for naive cells. We observed that the majority of Gag tetramer–
and Pol tetramer–binding CD8⫹ cells were CD45RO⫹, whereas
intermediate levels of CD45RA were expressed on a significant
minority of the tetramer-positive cells in a few of the children
JID 2001;183 (1 June)
CD8⫹ Tetramer-Positive T Cells in HIV-Infected Children
Figure 4. Longitudinal follow-up of patient EM30. Treatment history was as follows: monotherapy (zidovudine or didanosine), bitherapy (zidovudine/zalcitabine or didanosine/nelfinavir), tritherapy (stavudine/lamivudine/indinavir), and a 2-month period without any
treatment before starting the new therapy (stavudine/abacavir/efavirenz) from week 0 to week 29, which was changed to stavudine/abacavir/nevirapin from week 29 to week 33. Plasma virus loads are shown
(cross). A, Frequency of CD8⫹ tetramer-binding T cells (Tetra), which
is expressed as the percentage of total CD8⫹ lymphocytes. B, Frequency
of circulating CD4 (T4) and CD8 (T8) T cells, which is expressed as
the percentage of total lymphocytes.
in our study. However, it has been reported that a subpopulation of CD45RA⫹ cells has high cytolytic activity and may
be an important effector population [18]. The different expression of CD45 isoforms probably corresponds to different cellular activation status. Markers of activation and late differentiation, as well as costimulatory molecules, were expressed
on a majority of Gag tetramer– and Pol tetramer–binding
CD8⫹ cells. The HLA-DR marker was expressed at a high level
on a very high proportion (⭓90%) of the cells, which shows
that most tetramer-binding cells were activated in vivo. To get
more insight into the functional markers of tetramer-binding
cells, we checked the expression of CD69, which is considered
to be a very early T cell activation marker. CD69 is rapidly
up-regulated after T cell receptor (TCR)–mediated signaling,
but its expression is transient and depends on continuous TCR
activation. We were not able to detect tetramer-binding CD8⫹
T cells that coexpressed CD69. Thus, the lack of substantial
CD69 antigen expression on fresh circulating epitope-specific
CD8⫹ T cells (figure 3) probably is more compatible with the
presence of memory T cells that have not been activated
recently. Moreover, recently activated cells may have down-
1571
regulated their TCRs. In our group of patients, Gag tetramer–
and Pol tetramer–binding CD8⫹ cells were detected easily in
whole blood, and we did not observe a down-regulation of the
tetramer expression in fresh CD8⫹ lymphocytes. Overall, these
results support the hypothesis that these cells have not been
activated recently in vivo. The tetramer-positive cells were in
the CD28⫺ T cell compartment. In HIV-infected adults, circulating CD28⫺CD8⫹ T cells have been shown to be effector
CTL that present a direct cytolytic activity, as measured by use
of a chromium-release assay [17]. Down-regulation of CD28
seen on peptide–MHC class I Gag tetramer– and Pol tetramer–positive CD8⫹ cells may reflect recurrent restimulation and
perhaps terminal differentiation of subsets of the CTL in vivo.
In the follow-up of ex vivo–activated CTL from the same
pediatric cohort studied in 1997 and 1998, we found that ex
vivo–activated cytotoxic activity (fresh PBMC) that is specific
to Gag- or Pol-expressing targets could be detected in !10% of
children, using a conventional chromium-release assay (authors’ unpublished data). Although the studies of ex vivo cytotoxic activity and tetramer-binding cells were not performed at
the same time with the same samples of PBMC, the difference
in the frequencies of tetramer-positive cells and ex vivo CTL
activities may reflect the differences in the threshold of detection
of these 2 assays and/or the nature of the subpopulation of
CD8⫹ T cells measured by these 2 assays. Furthermore, tetramer-binding cells may be nonfunctional and thus undetectable
in a chromium-release assay performed with fresh PBMC, as
was shown in mice [29]. In addition to the tetramer-binding
assay, PBMC from 22 of the 28 HLA-A*02–positive patients
were studied for CTL activity after expansion in vitro, using a
chromium-release assay. A majority of the T cell lines (63%)
derived from children with Gag tetramer–binding CD8⫹ T cells
also had a p55Gag-specific cytotoxic activity (data not shown).
A similar percentage of patients (59%) were positive for Polspecific activities. These results suggest that CD8⫹ T cells need
either maturation or expansion in vitro to obtain significant
CTL activity, as measured by use of a functional chromiumrelease assay.
Recent controversies have been raised concerning the function of tetramer-binding cells [30–34]. Goulder et al. [32] have
reported that most tetramer-binding CD8⫹ T cells produced
IFN-g after peptide stimulation, which suggests that functionally inert CTL are not present in chronic HIV infection. On
the other hand, Goepfert et al. [31] have reported that a significant number of tetramer-binding CD8⫹ T cells do not produce IFN-g after stimulation with the cognate peptide in vitro.
Others [33, 34] (authors’ unpublished data) have observed similar results. The correlation between the number of reactive
CD8⫹ T cells detected by the tetramer assay and the ELISPOT
assay (secretion of IFN-g) has been documented by different
groups [31, 32] (authors’ unpublished data). Thus, there is a
need for more-extensive study to clarify the existence and sig-
1572
Scott-Algara et al.
JID 2001;183 (1 June)
Figure 5. Longitudinal follow-up of the tetramer-binding (Tetra) cells in 4 human immunodeficiency virus (HIV)–infected children receiving
antiretroviral therapy. Time is indicated in weeks of therapy. Antiretroviral therapies were as follows: patient EM050, zidovudine/lamivudine/
efavirenz; patient EM054, stavudine/lamivudine/nelfinavir; patient EM067, zidovudine/didanosine; and patient EM080, stavudine/lamivudine/
nelfinavir. Plasma virus loads are shown for each patient (cross). *For patient EM054, the virus load was below the threshold of detection (!1.3
/mL).
nificance of nonfunctional HIV-specific CD8⫹ T cells in the
pathogenesis of HIV infection.
In the cross-sectional study of 28 HLA-A*02–positive children,
no difference in the frequency of tetramer-binding cells or in virus
load was observed in untreated and treated children. When the
treated children were divided into different groups according to
the different therapies, group 1 (combination therapy) had a
significantly higher frequency of Gag tetramer–binding cells than
did group 2 (tritherapy with an NNRTI), and there was a trend
for higher frequency in group 1, compared with group 0 (untreated) or group 3 (tritherapy with a PI). This unexpected result
might reflect an influence of combination therapy on CD8 responses, but additional data are needed to confirm this observation. Analyzing the frequency of either Gag or Pol tetramerbinding cells failed to show any correlation with plasma virus
load. This is in contrast to a previous study in which a highly
significant inverse correlation was found between virus load and
the frequency of Gag tetramer–binding cells [25]. In that study,
the 14 patients were untreated adults, whereas most patients
(75%) in this study were treated children. The frequency of epitope-specific cells also was followed longitudinally in patients
receiving potent ART. One child showed a significant loss of
tetramer-positive CD8⫹ cells and a suppression of replicating
HIV, which were associated with new ART (figure 4A). The decrease of tetramer-binding cells was correlated positively with the
percentage of circulating CD8⫹ T cells and negatively correlated
with the percentage of total circulating CD4⫹ T cells (figure 4B).
This observation suggests that the frequency of HIV-1–specific
clones could depend on the persistence of replicating virus, which
has been suggested in other studies [35–38]. However, tetramerbinding cells still can be detected, despite undetectable virus load,
as was shown for patient EM54 (figure 5).
Our data further illustrate that the inhibition of HIV replication leads to a reduction of the magnitude of immune responses. For the maintenance of T cell memory and thus a
potential protective immune response, consistent antigen restimulation might be crucial. Although there are contradicting
points of view for the maintenance of T cell memory [39], our
findings are consistent with the notion that a reduction of viral
antigenic stimulus is associated with a decreased number of
antigen-specific experienced memory CD8⫹ T cells, and, without constant stimulation, the frequency of antigen-specific
CD8⫹ T cells decline but do not disappear.
Acknowledgments
We thank Frances Gotch (Chelsea and Westminster Hospital, London)
for HLA genotyping, Béatrice Corre and Françoise Porrot (Institut Pasteur, Paris) for expert technical assistance with the chromium-release
assay, Eida Bui (Institut Pasteur) for collecting clinical information, and
Ara Hovanessian (Institut Pasteur) for critical review of the manuscript.
References
1. Riddell SR, Gilbert MJ, Greenberg PD. CD8⫹ cytotoxic T cell therapy of
cytomegalovirus and HIV infection. Curr Opin Immunol 1993; 5:484–91.
2. Walker CM, Erickson AL, Hsueh FC, Levy JA. Inhibition of human im-
JID 2001;183 (1 June)
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
CD8⫹ Tetramer-Positive T Cells in HIV-Infected Children
munodeficiency virus replication in acutely infected CD4⫹ cells by CD8⫹
cells involves a noncytotoxic mechanism. J Virol 1991; 65:5921–7.
Buseyne F, Burgard M, Teglas JP, et al. Early HIV-specific cytotoxic T lymphocytes and disease progression in children born to HIV-infected mothers. AIDS Res Hum Retroviruses 1998; 14:1435–44.
Gotch FM, Nixon DF, Alp N, McMichael AJ, Borysiewicz LK. High frequency of memory and effector gag-specific cytotoxic T lymphocytes in
HIV-seropositive individuals. Int Immunol 1990; 2:707–12.
Kaul R, Plummer FA, Kimani J, et al. HIV-1–specific mucosal CD8⫹ lymphocyte responses in the cervix of HIV-1–resistant prostitutes in Nairobi.
J Immunol 2000; 164:1602–11.
Klein MR, van Baalen CA, Holwerda AM, et al. Kinetics of Gag-specific
cytotoxic T lymphocyte responses during the clinical course of HIV-1
infection: a longitudinal analysis of rapid progressors and long-term
asymptomatics. J Exp Med 1995; 181:1365–72.
Gea-Banacloche JC, Migueles SA, Martino L, et al. Maintenance of large
numbers of virus-specific CD8⫹ T cells in HIV-infected progressors and
long-term nonprogressors. J Immunol 2000; 165:1082–92.
Rivière Y, McChesney MB, Porrot F, et al. Gag-specific cytotoxic responses
to HIV-1 are associated with a decreased risk of progression to ARC or
AIDS. AIDS Res Hum Retroviruses 1995; 11:903–7.
Donahoe SM, Moretto WJ, Samuel RV, et al. Direct measurement of CD8⫹
T cell responses in macaques infected with simian immunodeficiency virus.
Virology 2000; 272:347–56.
Borrow P, Lewicki H, Hahn BH, Shaw GM, Oldstone MB. Virus-specific
CD8⫹ cytotoxic T-lymphocyte activity associated with control of viremia
in primary human immunodeficiency virus type 1 infection. J Virol 1994;
68:6103–10.
Koup RA, Safrit JT, Cao Y, et al. Temporal association of cellular immune
responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Virol 1994; 68:4650–5.
Greenough TC, Brettler DB, Somasundaran M, Panicali DL, Sullivan JL.
Human immunodeficiency virus type 1–specific cytotoxic T lymphocytes
(CTL), virus load, and CD4 T cell loss: evidence supporting a protective
role for CTL in vivo. J Infect Dis 1997; 176:118–25.
Ogg GS, Kostense S, Klein MR, et al. Longitudinal phenotypic analysis of
human immunodeficiency virus type 1–specific cytotoxic T lymphocytes:
correlation with disease progression. J Virol 1999; 73:9153–60.
Kalams SA, Goulder PJ, Shea AK, et al. Levels of human immunodeficiency
virus type 1–specific cytotoxic T-lymphocyte effector and memory responses decline after suppression of viremia with highly active antiretroviral therapy. J Virol 1999; 73:6721–8.
Aldhous MC, Watret KC, Mok JY, Bird AG, Froebel KS. Cytotoxic T
lymphocyte activity and CD8 subpopulations in children at risk of HIV
infection. Clin Exp Immunol 1994; 97:61–7.
McFarland EJ, Harding PA, Luckey D, Conway B, Young RK, Kuritzkes
DR. High frequency of Gag- and envelope-specific cytotoxic T lymphocyte precursors in children with vertically acquired human immunodeficiency virus type 1 infection. J Infect Dis 1994; 170:766–74.
Fiorentino S, Dalod M, Olive D, Guillet JG, Gomard E. Predominant involvement of CD8⫹CD28⫺ lymphocytes in human immunodeficiency virus–specific cytotoxic activity. J Virol 1996; 70:2022–6.
Hamann D, Baars PA, Rep MH, et al. Phenotypic and functional separation
of memory and effector human CD8⫹ T cells. J Exp Med 1997; 186:1407–14.
Centers for Disease Control and Prevention. 1994 Revised classification system for human immunodeficiency virus infection in children less than 13
years of age. MMWR Morb Mortal Wkly Rep 1994; 43(RR-12):1–10.
Altman JD, Moss PAH, Goulder PJR, et al. Phenotypic analysis of antigenspecific T lymphocytes. Science 1996; 274:94–6.
Bodinier M, Peyrat MA, Tournay C, et al. Efficient detection and immuno-
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
1573
magnetic sorting of specific T cells using multimers of MHC class I and
peptide with reduced CD8 binding. Nat Med 2000; 6:707–10.
Mulder J, Resnick R, Saget B, et al. A rapid and simple method for extracting
human immunodeficiency virus type 1 RNA from plasma: enhanced sensitivity. J Clin Microbiol 1997; 35:1278–80.
Krausa P, Brywka MI, Savage D, et al. Genetic polymorphism within HLAA*02: significant allelic variation revealed in different populations. Tissue
Antigens 1995; 45:223–31
Tsomides TJ, Aldovini A, Johnson RP, Walker BD, Young RA, Eisen HN.
Naturally processed viral peptides recognized by cytotoxic T lymphocytes
on cells chronically infected by human immunodeficiency virus type 1. J
Exp Med 1994; 180:1283–93.
Ogg GS, Jin X, Bonhoeffer S, et al. Quantitation of HIV-1–specific cytotoxic
T lymphocytes and plasma load of viral RNA. Science 1998; 279:2103–6.
Phillips RE, Rowland-Jones S, Nixon DF, et al. Human immunodeficiency virus
genetic variation that can escape cytotoxic T cell recognition. Nature
1991;354:453–9.
Goulder PJ, Phillips RE, Colbert RA, et al. Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression
to AIDS. Nat Med 1997; 3:212–7.
Brander C, Hartman KE, Trocha AK, et al. Lack of strong immune selection
pressure by the immunodominant, HLA-A*201–restricted cytotoxic response in chronic human immunodeficiency virus–1 infection. J Clin Invest
1998;101:2559–66.
Zajac AJ, Blattman JN, Murali-Krishna K, et al. Viral immune evasion due to
persistence of activated T cells without effector function. J Exp Med
1998;188:2205–13.
Betts MR, Casazza JP, Patterson BA, et al. Putative immunodominant human
immunodeficiency virus–specific CD8⫹ T-cell responses cannot be predicted by major histocompatibility complex class I haplotype. J Virol
2000; 74:9144–51.
Goepfert PA, Bansal A, Edwards BH, et al. A significant number of human
immunodeficiency virus epitope-specific cytotoxic T lymphocytes detected
by tetramer binding do not produce gamma interferon. J Virol 2000;
74:10249–55.
Goulder PJR, Tang Y, Brander C, et al. Functionally inert HIV-specific cytotoxic T lymphocytes do not play a major role in chronically infected adults
and children. J Exp Med 2000; 192:1819–31.
Appay V, Nixon DF, Donahoe SM, et al. HIV-specific CD8⫹ T cells produce
antiviral cytokines but are impaired in cytolytic function. J Exp Med 2000;
192:63–76.
Shankar P, Russo M, Harnish B, Patterson M, Skolnik P, Lieberman J.
Impaired function of circulating HIV-specific CD8⫹ T cells in chronic
human immunodeficiency virus infection. Blood 2000; 96:3094–101.
Pantaleo G, Soudeyns H, Demarest JF, et al. Accumulation of human immunodeficiency virus–specific cytotoxic T lymphocytes away from the
predominant site of virus replication during primary infection. Eur J Immunol 1997; 27:3166–73.
Wong JK, Gunthard HF, Havlir DV, et al. Reduction of HIV-1 in blood and
lymph nodes following potent antiretroviral therapy and the virologic
correlates of treatment failure. Proc Natl Acad Sci USA 1997; 94:12574–9.
Spiegel HM, Chandwani R, Sheehy ME, et al. The impact of early initiation
of highly active antiretroviral therapy on the human immunodeficiency
virus type 1–specific CD8 T cell response in children. J Infect Dis 2000;
182:88–95.
Luzuriaga K, McManus M, Catalina M, Mayack S, Sharkey M, Stevenson
M. Early therapy of vertical human immunodeficiency virus type 1 (HIV1) infection: control of viral replication and absence of persistent HIV1–specific immune responses. J Virol 2000; 74:6984–91.
Bevan MJ, Braciale TJ. Why can’t cytotoxic T cells handle HIV? Proc Natl
Acad Sci USA 1995; 92:5765–7.