Level of double negative T cells, which produce TGF-β and IL-10, predicts CD8 T-cell activation in primary HIV-1 infection

Article (PDF Available)inAIDS 26(2):139–148 · January 2012with97 Reads
DOI: 10.1097/QAD.0b013e32834e1484
Abstract
Objective: Persistent immune activation plays a central role in the pathogenesis of HIV disease. Besides natural regulatory T cells (nTregs), ‘double negative’ T cells shown to exhibit regulatory properties could be involved in the control of harmful immune activation. The aim of this study was to analyze, in patients with primary HIV infection (PHI), the relationship between CD4+CD25+CD127lowFoxP3+ nTregs or CD3+CD4−CD8− double negative T cells and systemic immune activation. Design: A prospective longitudinal study of patients with early PHI. Methods: Twenty-five patients were included. Relationships between frequency of Treg subsets and T-cell activation, assessed on fresh peripheral blood mononuclear cells, were analyzed using nonparametric tests. Cytokine production by double negative T cells was assessed following anti-CD3/anti-CD28 stimulation. Results: No relationship was found between T-cell activation and frequencies of nTregs. In contrast, a strong negative relationship was found at baseline between the proportion of double negative T cells and the proportion of activated CD8 T cells coexpressing CD38 and HLA-DR (P = 0.005) or expressing Ki-67 (P = 0.002). In addition, the frequency of double negative T cells at baseline negatively correlated with the frequency of HLA-DR+CD38+CD8+ T cells at month 6, defining the immune activation set point (P = 0.031). High proportions of stimulated double negative T cells were found to produce the immunosuppressive cytokines transforming growth factor-β1 and/or IL-10. Conclusion: The proportion of double negative T cells at baseline was found to be predictive of the immune activation set point. Our data strongly suggest that double negative T cells may control immune activation in PHI. This effect might be mediated through the production of TGF-β1/IL-10 known to downmodulate immune activation.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Level of double negative T cells, which produce
TGF-b and IL-10, predicts CD8 T-cell activation in
primary HIV-1 infection
Gae
¨
l Petitjean
a,M
, Mathieu F. Chevalier
a,b,M
, Feriel Tibaoui
c
,
Ce
´
line Didier
a
, Maria Elena Manea
d
, Anne-Sophie Liovat
a
,
Pauline Campa
e
, Michaela Mu¨ller-Trutwin
a
, Pierre-Marie Girard
e
,
Laurence Meyer
c
, Franc¸oise Barre
´
-Sinoussi
a
, Daniel Scott-Algara
a
and
Laurence Weiss
a,d,f
Objective: Persistent immune activation plays a central role in the pathogenesis
of HIV disease. Besides natural regulatory T cells (nTregs), ‘double negative’ T cells
shown to exhibit regulatory properties could be involved in the control of harmful
immune activation. The aim of this study was to analyze, in patients with primary HIV
infection (PHI), the relationship between CD4
þ
CD25
þ
CD127
low
FoxP3
þ
nTregs or
CD3
þ
CD4
CD8
double negative T cells and systemic immune activation.
Design: A prospective longitudinal study of patients with early PHI.
Methods: Twenty-five patients were included. Relationships between frequency of
Treg subsets and T-cell activation, assessed on fresh peripheral blood mononuclear
cells, were analyzed using nonparametric tests. Cytokine production by double nega-
tive T cells was assessed following anti-CD3/anti-CD28 stimulation.
Results: No relationship was found between T-cell activation and frequencies of
nTregs. In contrast, a strong negative relationship was found at baseline between
the proportion of double negative T cells and the proportion of activated CD8 T cells
coexpressing CD38 and HLA-DR (P ¼ 0.005) or expressing Ki-67 (P ¼ 0.002). In
addition, the frequency of double negative T cells at baseline negatively correlated
with the frequency of HLA-DR
þ
CD38
þ
CD8
þ
T cells at month 6, defining the immune
activation set point (P ¼ 0.031). High proportions of stimulated double negative T cells
were found to produce the immunosuppressive cytokines transforming growth factor-
b1 and/or IL-10.
Conclusion: The proportion of double negative T cells at baseline was found to be
predictive of the immune activation set point. Our data strongly suggest that double
negative T cells may control immune activation in PHI. This effect might be mediated
through the production of TGF-b1/IL-10 known to downmodulate immune activation.
ß 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins
AIDS 2012, 26:139148
Keywords: double negative T cells, immune activation, primary/acute HIV
infection, regulatory T cells, transforming growth factor-b
a
Institut Pasteur, Re
´
gulation des infections re
´
trovirales, De
´
partement de Virologie, Paris,
b
Universite
´
Paris Diderot, Sorbonne Paris
Cite
´
, Paris,
c
INSERM U 1018, AP-HP, Universite
´
Paris Sud, Paris,
d
AP-HP, Ho
ˆ
pital Europe
´
en Georges Pompidou, Paris,
e
AP-HP,
Ho
ˆ
pital Saint-Antoine, Paris, and
f
Universite
´
Paris Descartes, Sorbonne Paris Cite
´
, Paris, France.
Correspondence to Laurence Weiss, MD, PhD, Institut Pasteur, 25 rue du Dr Roux, 75015, Paris, France.
Tel: +33 156 093 297; fax: +33 156 093 026; e-mail: laurence.weiss@egp.aphp.fr.
Gae
¨
l Petitjean and Mathieu F. Chevalier contributed equally to the writing of this article.
Received: 28 July 2011; revised: 29 September 2011; accepted: 12 October 2011.
DOI:10.1097/QAD.0b013e32834e1484
ISSN 0269-9370 Q 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins
139
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Introduction
HIV infection is associated with a progressive depletion of
CD4 T lymphocytes and defective HIV-specific T-cell
responses. Persistent immune activation plays a central
role in driving CD4 T-cell depletion and progression to
AIDS [1,2]. High levels of immune activation occur early
in primary HIV infection (PHI) and may be related to
activation of innate and adaptive immune responses,
microbial translocation, activation by HIV viral proteins
and reactivation of other viruses (e.g. cytomegalovirus,
hepatitis viruses) [3]. During PHI, immune activation set
point was demonstrated to predict the subsequent CD4
T-cell decrease independently of viral load [4]. Thus, the
early control of immune activation could be a goal in
clinical management of PHI patients.
CD4
þ
CD25
þ
natural regulatory T cells (nTregs) are able
to suppress antigen-specific T-cell responses against a
variety of pathogens and may also control inappropriate or
exaggerated immune activation, thus, limiting immune-
mediated tissue damage [5]. Thus, nTregs may influence
the outcome of various infections [6]. In HIV/SIV
infection, Tregs, capable of suppressing HIV/SIV-specific
immune responses, are detected in per ipheral blood
and lymphoid tissues and may contribute to immune
deficiency [79]. Whether these cells are harmful by
suppressing HIV-specific immune responses and/or
beneficial through a decrease in immune activation
remains debatable [8,1014] and may depend on the stage
of the disease. Indeed, our previous data suggested that
nTregs may be efficient in controlling residual immune
activation in patients with antiretroviral-mediated viral
suppression but failed to control immune activation
associated with viral rebound following treatment
inter ruption [15].
We previously demonstrated that HIV-specific nTregs
were present in PHI and exhibited suppressive activity
against HIV-specific CD4 T-cell responses [16]. In PHI,
however, only sparse data are available regarding the
relationship between the proportion of nTregs and
immune activation, and no longitudinal study has
definitely elucidated the impact of nTregs on generalized
immune activation in the post-PHI phase in humans. In
addition, studies on regulatory T cells in HIV infection
were mainly focused on natural CD4
þ
CD25
high
FoxP3
þ
Tregs. However, other T-cell subsets including
CD8
þ
FoxP3
þ
regulatory T cells [17,18] and ‘double
negative’ CD3
þ
CD4
CD8
T cells (double negative
T cells) [19,20] have also been shown to possess the ability
to down-regulate specific immune responses in viral
infections and/or transplantation. Neither the presence
nor the role of double negative T cells have been studied
yet in HIV infection.
In the present study, we investigated whether nTregs and
double negative T cells in PHI may participate in the
control of generalized T-cell activation and, therefore,
influence the immunologic and virologic outcome of
HIV infection. To assess the impact of these regulatory
T-cell subsets on immune activation in PHI, we analyzed
the relationship between the activation level of peripheral
CD4 and CD8 T cells and the frequency and/or the
number of peripheral regulatory T-cell subsets.
Methods
Study population
This is a prospective multicenter study conducted in four
clinical sites in France. To be enrolled, individuals must
have shown evidence of acute HIV infection as defined by
a negative or weakly positive ELISA, and a known
contamination date by one of these criteria: less than three
bands on HIV Western Blot, a positive p24 antigenaemia,
and/or a detectable plasma HIV-RNA (patients coin-
cluded in the CO6-PRIMO ANRS cohort). Individuals
were offered to participate in this study before initiation
of combination antiretroviral therapy (ART). Some of the
patients started ART during the follow-up, based on
clinical symptoms, CD4 cell counts (below 500 cells/ml
according to French recommendations) and the decision
of both physicians and patients. Written informed
consent was provided by study participants according
to French ethical laws. The ethical committee of Ile de
France II approved the study. Blood from patients was
collected at day 0 of enrollment (baseline), day 15, month
1 (M1), month 3 (M3) and month 6 (M6). Blood was also
collected from healthy volunteer s (n ¼ 3).
Cell isolation and flow cytometric analysis
Fresh peripheral blood mononuclear cells (PBMC) were
purified by density gradient centrifugation (Isopaque-
Ficoll, PAA, Austria) within 24 h after blood sampling.
After washings, cells were stained using multicolor panels
and analyzed by flow cytometry (LSRII, Becton
Dickinson). The following monoclonal antibodies (mAbs)
conjugated to Phycoerythrin Texas Red (ECD), peridinin
chlorophyll proteincyanin 5.5 (PerCPCy5.5), Alexa
Fluor 488, Alexa Fluor 700, allophycocyanin (APC),
allophycocyaninHilite7 (APCH7), phycoerythrin
cyanin 7 (PEC7), phycoerythrin (PE), fluorescein
isothiocyanate (FITC), allophycocyanineFluor 780, or
V450 were used: anti-CD4PerCPCy5.5, anti-CD4
APCH7, anti-CD8Alexa Fluor 488, anti-CD8
Alexa Fluor 700, anti-CD25APC, anti-CD25PE
Cy7, anti-HLADRPerCPCy5.5 anti-CD38APC,
anti-CD152(CTLA-4)PE and anti-perforinFITC
(BD Biosciences); anti-CD8APCeFluor 780, anti-
CD45RAV450, anti-CD127PECy7, anti-CD39
PECy7 and anti-FoxP3Alexa Fluor 700, (eBiosciences,
San Diego, California, USA); anti-CD3ECD and anti-
gd T-cell receptorPE (Beckman Coulter, Brea,
California, USA); and anti-Ki-67FITC (Dako, Glostrup,
140 AIDS 2012, Vol 26 No 2
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Denmark). For intracellular staining of FoxP3, CTLA-4
and Ki-67, cells were fixed and permeabilized using FoxP3
Staining Buffer Set (eBioscience) according to the manu-
facturer’s recommendations. Analyses were performed
using FlowJo software (TreeStar, Ashland, Oregon, USA).
Cytokine assay
Intracellular cytokine production was assessed by flow
cytometry on fresh PBMC stimulated with plate-bound
anti-CD3 and soluble anti-CD28 mAbs (1 mg/ml of each;
Invitrogen, Carlsbad, California, USA) for 24 h at 378C.
After 2 h of culture, brefeldin A (5 mg/ml; Sigma,
St Louis, Missouri, USA) was added. Anti-CD3ECD
(Beckman Coulter), anti-CD4APCH7, anti-CD8
PE-Cy7, anti-IL-4FITC, anti-IFN-gAlexa Fluor
700, anti-IL-17Alexa Fluor 700 (BD Biosciences),
anti-IL-10eFluor 450 (eBiosciences) and anti-TGFb
PE (IQ Products) mAbs were used for immunostaining.
Statistical analysis
Data were described by medians and interquartile ranges
(IQR) for continuous variables. Nonparametric tests were
used to avoid the impact of potential outlier values in a
small study. All patients at baseline and only untreated
patients at M6 were considered for the analyses.
Comparisons between treated and untreated patients were
performed using the MannWhitney test. The Wilcoxon-
matched pairs test was used to estimate the changes in the
level of T-cell activation, HIV-RNA levels and CD4 T-cell
count throughout the follow-up. The Spearman’s non-
parametric correlation was used to estimate the association
of two continuous variables of interest. P values below 0.05
were considered statistically significant.
Results
Patients’ characteristics
Twenty-five patients diagnosed early during PHI [median
(IQR) of estimated time postinfection: 42 (3051) days]
were prospectively enrolled in the study. Patients’ clinical
characteristics at baseline and at M6 are depicted in
Table 1. Twelve patients remained untreated during the
study period. Ten patients were treated with cART just
after baseline sampling; one patient was treated between
M3 and M6. Treated patients had significant lower CD4
cell counts and higher plasma HIV-RNA levels at baseline
(P ¼ 0.0003 and P ¼ 0.03; respectively). CD8 and CD4
T-cell activation levels were determined by the pro-
portion of cells that expressed the CD38, CD25, HLA-
DR and/or Ki-67 activation markers at all time points of
follow-up. Treated and untreated patients did not differ
for CD4 and CD8 T-cell activation at baseline. As
expected, at baseline, the proportion of CD8 T cells
coexpressing CD38 and HLA-DR and of CD8 T cells
expressing Ki-67 positively correlated with viral
load (r ¼ 0.72, P ¼ 0.0001 and r ¼ 0.80, P < 0.0001,
Double negative T cells predict immune set point Petitjean et al. 141
Table 1. Patients’ characteristics.
Time
points Patients
HIV-1 RNA
(log
10
/mm
3
) CD3%
CD3 cell count
(cells/ml) CD4%
CD4 cell count
(cells/ml) CD8%
CD8 cell count
(cells/ml) CD4/CD8 ratio
Baseline n ¼ 25 5.61 (4.366.42) 79 (7486) 1512 (10272368) 26 (1636) 490 (334624) 54 (4268) 1073 (5481695) 0.48 (0.260.98)
M6 untreated n ¼ 12
a
4.50 (3.365.04) 82 (7785) 1422 (12742278) 35 (2938) 677 (456735) 45 (3950) 725 (6031146) 0.75 (0.581.11)
M6 ART-treated n ¼ 10
b
1.15 (11.34) 78 (7884) 1366 (8681652) 48 (3850) 751 (466901) 35 (2738) 595 (319690) 1.39 (1.141.85)
Data are expressed as median (IQR).
a
For one patient experiencing an intercurrent episode associated with high level of inflammation at M6, clinical data from M3 were used.
b
Nine patients received ART before day 15 and one before M6.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
respectively). The proportion of HLA-DR
þ
CD4
þ
T
cells and Ki-67
þ
CD4
þ
T cells also positively correlated
with viral load (r ¼ 0.60, P ¼ 0.002 and r ¼ 0.53,
P ¼ 0.006, respectively). During the 6 months of
follow-up (Fig. 1), the proportion of double positive
HLA-DR/CD38 among CD8 T cells decreased from a
median (IQR) of 25.5% (12.560.8) to 10.1% (5.812.1)
in untreated patients (P ¼ 0.0005). The proportion of Ki-
67-expressing CD8 T cells (Fig. 1) and of Ki-67-
expressing CD4 T cells (not shown) also significantly
decreased. In contrast, HIV-RNA plasma levels did not
significantly decrease between baseline and M6, as
previously reported in the French ANRS primo cohort
[21]. CD8 T-cell activation remained stable between M3
and M6. At 6 months, the proportion of CD8 T cells that
expressed CD38, the density of CD38 expression and the
proportion of Ki-67
þ
CD8
þ
T cells were positively
correlated to viral load (r ¼ 0.67, P ¼ 0.017; r ¼ 0.85,
P ¼ 0.0004 and r ¼ 0.84, P ¼ 0.0006, respectively).
Double negative T cells but not natural
regulatory T cells negatively correlated with CD8
T-cell activation level
We put forward the hypothesis that regulatory T cells
including nTregs and/or double negative T cells could
impact the level of immune activation during PHI. We
investigated the relationship between CD4 and CD8 T-
cell activation levels and the proportion or absolute
numbers of CD3
þ
CD4
þ
CD25
þ
CD127
low
FoxP3
þ
nTregs and of CD3
þ
CD4
CD8
double negative
T cells during the follow-up (Fig. 2). Longitudinal data
of these regulatory T-cell subsets are depicted in
Supplemental Digital Content Table 1, http://links.lww.
com/QAD/A187. We did not observe any significant
variation over the time in the frequencies and absolute
numbers of neither nTregs nor double negative T cells.
As illustrated in Fig. 2a, neither CD4 nor CD8 T-cell
activation at baseline was associated with the frequencies
or absolute numbers of nTregs. In contrast, we found a
strong negative relationship between the proportion of
double negative T cells and the level of CD8 T cells that
coexpressed CD38 and HLA-DR at baseline (r ¼0.56,
P ¼ 0.005; Fig. 2b) as well as with the proportion of
Ki-67-expressing CD8 T cells (r ¼0.58, P ¼ 0.002). In
addition, the proportion of double negative T cells but
not that of nTregs negatively correlated with HIV-RNA
plasma levels at baseline (r ¼0.42, P ¼ 0.034; Fig. 2).
No relationship was found between the proportion of
CD4 T cells expressing CD25, HLA-DR and/or Ki-67
142 AIDS 2012, Vol 26 No 2
80
(a) (b)
(c) (d)
P = 0.0005
P = 0.003
P = n.s.
60
40
% of CD38
+
HLA-DR
+
cells
among CD8 T cells
20
0
Baseline 15 days M1 M3 M6
8
7
6
5
4
Plasma HIV-1 RNA
log
10
copies/ml
3
2
Baseline 15 da
y
sM1 M3 M6
1200
1000
800
600
400
CD4 T cell count (cells/µl
3
)
200
0
Baseline 15 days M1 M3 M6
100
P = 0.0005
P = 0.001
P = n.s.
60
80
40
% of Ki-67
+
cells
among CD8 T cells
20
0
Baseline 15 days M1 M3 M6
Fig. 1. Individual longitudinal follow-up in untreated patients with acute HIV infection. CD8 T-cell activation was assessed in
untreated patients (n ¼ 12) by measuring the frequency of (a) CD38
þ
HLA-DR
þ
cells and (b) K-i67
þ
cells among CD8 T cells at
baseline, day 15, month 1 (M1), month 3 (M3) and month 6 (M6). (c) Plasma HIV-1 RNA levels and (d) CD4 T cells counts were
plotted as a function of time during the 6 months of follow-up. Wilcoxon matched pairs tests were performed and P values,
considered as significant when less than 0.05, are indicated.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
and the proportion or the number of double negative
T cells at baseline. The absolute numbers of double
negative T cells were not related to CD8 and CD4 T-cell
activation or to HIV-RNA plasma levels.
We next performed similar analyses in untreated patients
at M6. We found that the proportion of double negative
T cells still negatively correlated with the proportion of
CD38
þ
HLA-DR
þ
CD8 T lymphocytes (r ¼0.59,
P ¼ 0.041) but not anymore with viral load (Supple-
mental Digital Content Fig. 1, http://links.lww.com/
QAD/A187). We also observed a trend to a negative
relationship between the proportion of double negative
T cells and CD4 T-cell activation (CD25
þ
HLA-DR
þ
CD4 T cells) at M6 (r ¼0.57, P ¼ 0.055); in addition,
the absolute numbers of double negative T cells
were negatively correlated with the proportion of
CD25
þ
HLA-DR
þ
activated CD4 T cells (r ¼0.62,
P ¼ 0.033). Therefore, double negative T cells, but not
nTregs, were found to be strongly negatively correlated
with the level of CD8 T-cell activation during early
PHI and at month 6 of follow-up.
The proportion of double negative T cells at
baseline predicted the level of T-cell activation at
month 6
We next assessed the impact of double negative T cells at
baseline on T-cell activation and viral set point after
6 months of follow-up. As shown in Fig. 3, the
proportion of double negative T cells at baseline was
negatively related to the proportion of activated CD8
T cells at M6 whether assessed by the expression of
HLA-DR (r ¼0.76, P ¼ 0.004) or by the coexpression
of HLA-DR and CD38 (r ¼0.62, P ¼ 0.031) on CD8
T cells. The proportion of double negative T cells at
baseline also negatively correlated with the proportion of
HLA-DR-expressing CD4 T cells (r ¼0.72, P ¼ 0.008)
and of CD25
þ
HLA-DR
þ
CD4 T cells (r ¼0.78,
P ¼ 0.003) at M6. We also found that the absolute
numbers of double negative T cells at baseline negatively
correlated with T-cell activation at M6 (r ¼0.60,
P ¼ 0.035 for HLA-DR
þ
CD8 T cells and r ¼0.64,
P ¼ 0.024 for HLA-DR
þ
CD25
þ
CD4 T cells). In
contrast, the proportion or the absolute number of double
negative T cells at baseline did not influence the viral set
point. Thus, the proportion of double negative T cells at
baseline was found to be predictive of T-cell activation
level but not of viral load at M6.
Double negative T cells display an
anti-inflammatory cytokine profile
In order to gain insight into the mechanisms by which
double negative T cells could exert immunomodulatory
properties, we then characterized their phenotypic and
functional characteristics. We assessed the expression of
regulatory T-cell markers (CD25, CD127, FoxP3,
Double negative T cells predict immune set point Petitjean et al. 143
100
R = –0.11
P = 0.610
R = 0.13
P = 0.530
80
60
40
20
% CD38
+
HLA-DR
+
of CD8 T cells
at baseline
Natural Tregs
CD8 activation
(a)
(b)
CD4 activation Viral load
DN T cells
0
02
% Natural Tregs among CD4 T cells
at baseline
4 6 8 10 12
100
R = –0.56
P = 0.005
80
60
40
20
% CD38
+
HLA-DR
+
of CD8 T cells
at baseline
0
024681012
100
80
60
40
20
% Ki-67
+
of CD8 T cells
at baseline
0
02
% Natural Tregs among CD4 T cells
at baseline
4681012
R = –0.58
P = 0.002
100
80
60
40
20
% Ki-67
+
of CD8 T cells
at baseline
0
02
% CD4
CD8
among CD3 T cells
at baseline
% CD4
CD8
among CD3 T cells
at baseline
% CD4
CD8
among CD3 T cells
at baseline
% CD4
CD8
among CD3 T cells
at baseline
4681012
R = 0.01
P = 0.960
10
8
6
4
2
% CD25
+
HLA-DR
+
of CD4 T cells
at baseline
0
02
% Natural Tregs among CD4 T cells
at baseline
4681012
R = –0.14
P = 0.487
10
8
6
4
2
% CD25
+
HLA-DR
+
of CD4 T cells
at baseline
0
024681012
R = 0.13
P = 0.520
8
6
4
Viral load (log)
at baseline
2
02
% Natural Tregs among CD4 T cells
at baseline
4681012
R = –0.42
P = 0.034
8
6
4
Viral load (log)
at baseline
2
024681012
Fig. 2. Double negative T cells but not natural regulatory T cells negatively correlated with CD8 T-cell activation level and viral
load at baseline. At day of enrollment, frequencies of (a) CD4
þ
CD25
þ
CD127
low
FoxP3
þ
natural Tregs and (b) CD3
þ
CD4
CD8
double negative T cells were illustrated as function of T-cell activation and viral load in PHI patients (n ¼ 25). CD8 T-cell activation
was defined by the percentage of CD8 T cells coexpressing CD38 and HLA-DR or CD8 T cells expressing Ki-67 (left panels). CD4
T-cell activation was defined by the percentage of CD4
þ
T cells coexpressing CD25 and HLA-DR (middle panels). Viral load is
expressed as plasma HIV-1 RNA log
10
copies/ml (right panels). Spearman’s rank correlation coefficients ‘R and corresponding P
values are indicated on each panel.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
CTLA-4) as well as activation and proliferation markers
(CD38, HLA-DR and Ki-67). As shown in Supplemental
Digital Content Table 2, http://links.lww.com/QAD/
A187, double negative T cells did not express CD25 and
FoxP3 but exhibited slight expression of CTLA-4
(median 5.3%). Double negative T-cell activation level
was positively cor related with activation of CD4 and
CD8 T cells (data not shown). Interestingly, a median of
34% of double negative T cells expressed high levels
of perforin and a median of 53% expressed the gd
T-cell receptor.
We then studied the cytokine profile of double negative T
cells. Fresh PBMCs were stimulated for 24 h with anti-
CD3/anti-CD28 mAbs and assessed for IL-4, IL-10, IL-
17, TGF-b1 and IFN-g production. As illustrated in
Fig. 4, in contrast to CD4 T cells, double negative T cells
from PHI patients and controls displayed an anti-
inflammatory cytokine profile expressing IL-10 and/or
TGF-b1
˙
Expression of TGF-b1 was detected in 30
40% of stimulated double negative T cells (vs. 28% of
unstimulated cells) from PHI patients. In healthy donors,
the proportion of TGF-b1 expressing double negative
T cells tended to be lower increasing from 24% in
unstimulated cells to 810% in stimulated conditions. In
patients, expression of IL-10 was found in 330% of
stimulated double negative T cells (vs. 13% of
unstimulated cells). In healthy donors, the proportion
of IL-10 expressing double negative T cells was less than
5% even in stimulated conditions. Double negative T cells
from healthy donors and PHI patients did not upregulate
IL-4, IFN-g nor IL-17 in the same experimental
conditions.
Discussion
Our data showed that, in contrast to nTregs, double
negative T cells strongly negatively correlated with CD4
and CD8 T-cell activation. Interestingly, the level of
double negative T cells at baseline predicted the immune
activation set point reached by M6 [4]. In our cohort,
levels of viral load and T-cell activation remained stable
between M3 and M6 in untreated patients indicating that
the viral and immunologic set points were reached by M3.
The role of nTregs in controlling harmful T-cell
activation was still questionable. We previously investi-
gated the role of nTregs in the pathogenesis of HIV
144 AIDS 2012, Vol 26 No 2
30
(a)
(b)
R = 0.76
P = 0.004
R =0.78
P = 0.003
R = 0.60
P = 0.035
R = 0.64
P = 0.024
25
20
15
% HLA-DR
+
of CD8 T cells
at M6
% HLA-DR
+
of CD8 T cells
at M6
% CD25
+
HLA-DR
+
of CD4 T cells
at M6
% CD25
+
HLA-DR
+
of CD4 T cells
at M6
10
5
0
024
% CD4
CD8
among CD3 T cells
at baseline
CD4
CD8
T cells/µl
3
CD4
CD8
T cells/µl
3
% CD4
CD8
among CD3 T cells
at baseline
6810
6
4
2
0
0246810
6
4
2
0
0 50 100 150 200 250
30
25
20
15
10
5
0
0 50 100 150 200 250
Fig. 3. The proportion of double negative T cells at baseline predicted the level of T-cell activation at month 6. Panels depict
relationships between (a) the proportion or (b) absolute number of double negative T cells at baseline and the percentage of CD8
þ
T cells expressing HLA-DR (left panels) or the percentage of CD4
þ
T cells coexpressing CD25 and HLA-DR (right panels) at
6 month of follow-up. Spearman’s rank correlation coefficients R and corresponding P values are indicated on each panel.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
infection by studying their suppressive effect on HIV-
specific T-cell responses in acute and chronic infection
[7,15,16]. Although nTregs were found to negatively
correlate with residual immune activation in patients with
ART-mediated viral suppression, they were incapable
of downmodulating immune activation resulting from
ART interruption [15]. We previously demonstrated that
nTregs from PHI patients were functional on the basis of
suppressive activity on antigen-specific CD4 T-cell
proliferation [16]. In the present study, nTregs defined
as CD3
þ
CD4
þ
CD25
þ
CD127
low
Foxp3
þ
were found to
express CTLA-4 and CD39 (data not shown), markers
associated with a suppressive function. As previous reports
[22] suggested a reduction in CD25 expression in nTregs
from HIV-infected patients, we performed the same
analyses after excluding CD25 from the Treg phenotype
and still found a lack of relationship between frequency of
Tregs and T-cell activation. Data from a small cross
sectional study in eight patients also suggested that
CD4
þ
CD25
þ
CD127
low
FoxP3
þ
cells were not effective
in reducing pathogenic immune activation [13]. In most
studies of acute FIV or SIV infection, nTregs were not
likely to play a protective role. Indeed, CD4
þ
CD25
þ
Treg cell depletion did not significantly alter viral load,
CD4 cell count or immune activation in the acute phase
of FIV infection [23]. Furthermore, following acute SIV
infection, the frequency of nTregs was found to increase
in Rhesus Macaques but not in Sooty Mangabeys despite
a lower activation in the Sooty Mangabeys nonpatho-
genic model [24]. Taken together, data from previous and
present studies strongly suggest that nTregs are not able to
decrease T-cell activation in patients with untreated PHI.
Here, we studied double negative T cells and report on a
strong negative correlation between their frequency and
the proportion of CD8 T cells coexpressing HLA-DR
and CD38 at baseline and M6. This suggests a role for
double negative T cells in the control of immune
activation, although we cannot exclude that double
negative T cells relocate to lymphoid organs in patients
with high T-cell activation. Concomitant HLA-DR and
CD38 expression on CD8 T cells, referred to as
generalized immune activation, was demonstrated to
be a strong predictor of disease progression [2]. Also
Double negative T cells predict immune set point Petitjean et al. 145
Count
CD3
CD8
DN T cells
CD4
CD4 T cells
(a) (c)
(b)
Healthy donor
Unstimulated
Anti-CD3/CD28
Unstimulated
Anti-CD3/CD28
Unstimulated
2.7% 24%
6.1%0.6%
0.01% 0.11%
0.09%
0.01%
PHI patient
Anti-CD3/CD28
DN T cells
CD4 T cells
DN T cells
CD4 T cells
DN T cells
CD4 T cells
IL-10
4.6%1.2%
2.8%
IL-10IFN-γ
0.08%
0.06%
0.04%
0.03%
2.43%
0.12%
<0.01%
IL-17
0.13%
1.11%
6.47%
0.14%
0.03%
0.17%
2.76% 0.05%
0.10%
TGF-β
TGF-β
4.8%
Fig. 4. Double negative T cells express anti-inflammatory cytokines TGF-b and IL-10. PBMCs were cultured in medium only
(unstimulated) or in the presence of plate bound anti-CD3 and soluble anti-CD28 mAbs for 24 h. Cytokine expression was assessed
in double negative T cells and CD4 T cells by flow-cytometry. (a) Shows gating strategy. Cytokine profiles obtained with cells (b)
from a patient with primary-HIV infection and (c) from a healthy donor are illustrated. Data are representative of independent
experiments in three patients with PHI and three healthy donors.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
noteworthy is that the proportion of double negative T
cells at baseline predicted the level of CD8 T-cell
activation at M6 whether assessed by the coexpression of
HLA-DR and CD38 or the expression of HLA-DR
alone. In contrast, double negative T cells at baseline did
not impact CD38 expression on CD8 T cells at M6.
Although double negative T cells were not related to
CD4 T-cell activation at baseline, their proportion was
found to predict the level of HLA-DR-expressing CD4 T
cells at M6. Thus, double negative T cells seem to
principally impact the frequency of T cells that express
HLA-DR, which, it has been suggested, reflects the
capacity of cells to proliferate [25], rather than CD38.
Expression of CD38 on CD8 T cells is known to be
closely associated with HIV plasma viral load [26].
Accordingly, we did not find any relationship between
double negative T cells, whether measured at baseline or
at M6, and plasma viral load at M6.
Double negative T cells detected in PHI patients could
originate from CD4 T cells that had downmodulated the
membrane CD4 protein following activation. However,
several findings here argue against this hypothesis
including the higher double negative T-cell frequencies
observed in patients with lower T-cell activation and the
proportion and number of double negative T cells that
remained stable within the first months of PHI (data not
shown). Although double negative T cells were found to
be target cells of HIV [27], mature double negative T cells
were reported to be highly resistant to apoptosis [28].
Thus, in the context of PHI in which immune activation
is particularly high, double negative T cells could possibly
resist activation-induced cell death.
Zhang et al. [19] were the first to characterize the
suppressive function of antigen-specific double negative
regulatory T cells and their role in the control of allograft
rejection in mice. They showed that double negative T
cells can specifically downmodulate proliferative T-cell
responses against alloantigens allowing significant pro-
longation of skin graft survival. Double negative T cells
were demonstrated to inhibit proliferation and cytotoxicity
of CD8 T cells carrying the same TCR specificity. Fischer
et al. [20] first characterized in humans, double negative T
cells with similar functional properties to those described in
mice. Interestingly, an inverse relationship between
frequency of double negative T cells and graft-versus-host
disease severity has been reported in patients receiving
haematopoietic stem-cell transplantation [29], suggesting
that double negative T cells participate in peripheral
tolerance when expanded after allogeneic stimulation.
Consistent with previous phenotypic analyses of double
negative T cells in humans [20,29,30], those from PHI
patients did not express markers of conventional nTregs
such as FoxP3 or CD25. They consisted of both
naive CD45RA
þ
and antigen-experienced CD45RA
neg
T cells (see Supplemental Digital Content Table 2,
http://links.lww.com/QAD/A187). We also found that
they expressed high levels of perforin. A slight proportion
of double negative T cells exhibited CTLA-4 expression.
There is some evidence that double negative T cells
mediate direct T-cell suppression through cell-to-cell
contact [19,20,30] and, in mice, through a perforin/
granzyme-dependent pathway [31], a mechanism that
might be involved in HIV infection. However, we did not
find any correlation between perforin expression by
double negative T cells and their ability to control
immune activation. This is consistent with a recent study
suggesting that, in contrast to mice, human double
negative T cells do not eliminate effector T cells but
rather suppress their proliferation [30]. In addition to
cellcell contact, other mechanisms could be involved in
double negative-mediated suppression. A high pro-
portion of double negative T cells were found here to
produce the immunosuppressive cytokines TGF-b and
IL-10. Among CD3
þ
CD4
CD8
cells, human T cells
bearing the gd TCR were demonstrated to produce these
cytokines [32,33]. Approximately half of double negative
T cells from patients with PHI expressed the gd TCR.
These cells were described as potential effectors playing a
direct role in HIV infection by killing infected cells
through cytotoxic NK-like mechanisms and/or by
producing chemokines [34]. However, several studies
also suggested that gd T cells exhibited immunoregu-
latory properties [32,3538], and that their suppressive
function might be mediated through secretion of TGF-b
and/or IL-10 [33,3840]. Although higher TGF-b
plasma levels were found in patients with advanced stages
of HIV infection [41,42], it was shown to inhibit viral
replication of HIV [43] and hepatitis B virus [44]. TGF-b
also inhibited DC-SIGN expression on dendritic cells
[45] and, thus, could indirectly reduce transinfection of T
cells caused by DC-SIGN/HIV binding. There is
increasing evidence that HIV infection disrupts gastro-
intestinal mucosal barrier integrity leading t o m icrobial
translocation [46]. Incre ased li popolysaccharide (LPS)
plasma levels were found to correlate with innate and
adaptive immune activation [47]. Microbial translocation
might, thus, be one of the major mechanisms involved in
chronic generalized immune activat ion during HIV
infection. Interestingly, mucosal TGF-b and IL-10 have
bee n shown to p revent LPS-driven epithel ial damages
in the human colon [48]. Therefore, double negative
T cells producing TGF-b and IL-10 may participate in
the control of immune activation both by reducing
inflammatory responses and by preserving mucosal
bar rier integrity and, therein, preventing microbial
translocation.
Conclusion
Altogether, data from the present study strongly supports
a role for double negative T cells in the control of harmful
146 AIDS 2012, Vol 26 No 2
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
generalized T-cell activation during PHI. Because the
frequency of double negative T cells in PHI patients was
found to predict the immune activation set point,
reported to impact disease progression, this marker should
be considered for use in clinical practice. It can be easily
measured in hospitals without any additional set up using
the regular four-color CD4 protocol. This needs to
be evaluated in larger prospective cohort studies.
Whether double negative T cells are able to inhibit the
development of HIV-associated immunopathology with-
out compromising specific immune responses remains to
be determined.
Acknowledgements
We thank all patients involved in this study and we also
thank Nelly Desplanques, Nadia Valin, Laurent Fon-
querine (Ho
ˆ
pital Saint Antoine, Paris); Marina Kar-
mochkine, Pascale Kousignian, Jean Derouineau, Martin
Buisson, Isabelle Pierre, Dominique Batisse (Ho
ˆ
pital
Europe
´
en Georges Pompidou, Paris), Gilles Pialoux,
Laurence Slama, Thomas Lyavanc, Laura Iordache
(Ho
ˆ
pital Tenon, Paris), Christine Katlama and Marc-
Antoine Valantin (Ho
ˆ
pital Pitie
´
-Salpe
´
trie
`
re, Paris) for
including patients in the study.
We thank members of the ANRS PRIMO Cohort study
group and especially Christiane Deveau.
L.W designed and supervised the study; G.P, M.F.C,
M.M-T, F.B.S, D.S-A and L.W contributed to the
experimental design and provided intellectual input; G.P,
M.F.C, C.D and A-S.L performed experiments; P.C,
P-M.G, M.E.M included patients and reviewed the
manuscript; G.P, M.F.C, F.T, L.M and L.W. analyzed data;
M.M-T reviewed the manuscript and G.P, M.F.C, D.S-A
and L.W wrote the paper.
G.P. and M.F.C. were recipients of fellowships from the
Agence Nationale de Recherches sur le SIDA et les
He
´
patites Virales (ANRS). This work was supported by
the ANRS and the Assistance Publique Ho
ˆ
pitaux de
Paris (AP-HP, Paris).
Conflicts of interest
There are no conflicts of interest.
References
1. Bentwich Z, Kalinkovich A, Weisman Z, Grossman Z. Immune
activation in the context of HIV infection. Clin Exp Immunol
1998; 111:1–2.
2. Hazenberg MD, Otto SA, van Benthem BH, Roos MT, Coutinho
RA, Lange JM, et al. Persistent immune activation in HIV-1
infection is associated with progression to AIDS. AIDS 2003;
17:1881–1888.
3. Appay V, Sauce D. Immune activation and inflammation
in HIV-1 infection: causes and consequences. J Pathol 2008;
214:231–241.
4. Deeks SG, Kitchen CM, Liu L, Guo H, Gascon R, Narvaez AB,
et al. Immune activation set point during early HIV infection
predicts subsequent CD4R T-cell changes independent of viral
load. Blood 2004; 104:942–947.
5. Belkaid Y, Tarbell K. Regulatory T cells in the control of
host-microorganism interactions. Annu Rev Immunol 2009;
27:551–589.
6. Rouse BT, Sarangi PP, Suvas S. Regulatory T cells in virus
infections. Immunol Rev 2006; 212:272–286.
7. Weiss L, Donkova-Petrini V, Caccavelli L, Balbo M, Carbonneil
C, Levy Y. Human immunodeficiency virus-driven expansion of
CD4RCD25R regulatory T cells, which suppress HIV-specific
CD4 T-cell responses in HIV-infected patients. Blood 2004;
104:3249–3256.
8. Kinter AL, Hennessey M, Bell A, Kern S, Lin Y, Daucher M, et al.
CD25RCD4R regulatory T cells from the peripheral blood of
asymptomatic HIV-infected individuals regulate CD4R and
CD8R HIV-specific T cell immune responses in vitro and
are associated with favorable clinical markers of disease status.
J Exp Med 2004; 200:331–343.
9. Estes JD, Li Q, Reynolds MR, Wietgrefe S, Duan L, Schacker T,
et al. Premature induction of an immunosuppressive regulatory
T cell response during acute simian immunodeficiency virus
infection. J Infect Dis 2006; 193:703–712.
10. Epple HJ, Loddenkemper C, Kunkel D, Troger H, Maul J, Moos
V, et al. Mucosal but not peripheral FOXP3R regulatory T cells
are highly increased in untreated HIV infection and normalize
after suppressive HAART. Blood 2006; 108:3072–3078.
11. Tsunemi S, Iwasaki T, Imado T, Higasa S, Kakishita E, Shirasaka
T, et al. Relationship of CD4RCD25R regulatory T cells to
immune status in HIV-infected patients. Aids 2005; 19:879–
886.
12. Eggena MP, Barugahare B, Jones N, Okello M, Mutalya S, Kityo
C, et al. Depletion of regulatory T cells in HIV infection is
associated with immune activation. J Immunol 2005; 174:
4407–4414.
13. Ndhlovu LC, Loo CP, Spotts G, Nixon DF, Hecht FM. FOXP3
expressing CD127lo CD4R T cells inversely correlate
with CD38R CD8R T cell activation levels in primary HIV-
1 infection. J Leukoc Biol 2008; 83:254–262.
14. Lim A, Tan D, Price P, Kamarulzaman A, Tan HY, James I, et al.
Proportions of circulating T cells with a regulatory cell
phenotype increase with HIV-associated immune activation
and remain high on antiretroviral therapy. Aids 2007; 21:
1525–1534.
15. Weiss L, Piketty C, Assoumou L, Didier C, Caccavelli L,
Donkova-Petrini V, et al. Relationship between regulatory
T cells and immune activation in human immunodeficiency
virus-infected patients interrupting antiretroviral therapy.
PLoS One 2010; 5:e11659.
16. Kared H, Lelievre JD, Donkova-Petrini V, Aouba A, Melica G,
Balbo M, et al. HIV-specific regulatory T cells are associated
with higher CD4 cell counts in primary infection. AIDS 2008;
22:2451–2460.
17. Karlsson I, Malleret B, Brochard P, Delache B, Calvo J, Le Grand
R, et al. FoxP3R CD25R CD8
R T-cell induction during primary
simian immunodeficiency virus infection in cynomolgus
macaques correlates with low CD4R T-cell activation and
high viral load. J Virol 2007; 81:13444–13455.
18. Smith TR, Kumar V. Revival of CD8R Treg-mediated suppres-
sion. Trends Immunol 2008; 29:337–342.
19. Zhang ZX, Yang L, Young KJ, DuTemple B, Zhang L. Identifica-
tion of a previously unknown antigen-specific regulatory T cell
and its mechanism of suppression. Nat Med 2000; 6:782–
789.
20. Fischer K, Voelkl S, Heymann J, Przybylski GK, Mondal K,
Laumer M, et al. Isolation and characterization of human
antigen-specific TCR alpha betaR CD4(S)CD8S double-
negative regulatory T cells. Blood 2005; 105:2828–2835.
21. Ghosn J, Deveau C, Chaix ML, Goujard C, Galimand J, Zitoun Y,
et al. Despite being highly diverse, immunovirological status
strongly correlates with clinical symptoms during primary
HIV-1 infection: a cross-sectional study based on 674 patients
enrolled in the ANRS CO 06 PRIMO cohort. J Antimicrob
Chemother 2010; 65:741–748.
Double negative T cells predict immune set point Petitjean et al. 147
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
22. Nilsson J, Boasso A, Velilla PA, Zhang R, Vaccari M, Franchini
G, et al. HIV-1-driven regulatory T-cell accumulation in
lymphoid tissues is associated with disease progression in
HIV/AIDS. Blood 2006; 108:3808–3817.
23. Mikkelsen SR, Long JM, Zhang L, Galemore ER, VandeWoude S,
Dean GA. Partial regulatory T cell depletion prior to acute
feline immunodeficiency virus infection does not alter disease
pathogenesis. PLoS One 2011; 6:e17183.
24. Pereira LE, Villinger F, Onlamoon N, Bryan P, Cardona A,
Pattanapanysat K, et al. Simian immunodeficiency virus (SIV)
infection influences the level and function of regulatory T cells
in SIV-infected rhesus macaques but not SIV-infected sooty
mangabeys. J Virol 2007; 81:4445–4456.
25. Saez-Cirion A, Lacabaratz C, Lambotte O, Versmisse P, Urrutia
A, Boufassa F, et al. HIV controllers exhibit potent CD8 T cell
capacity to suppress HIV infection ex vivo and peculiar
cytotoxic T lymphocyte activation phenotype. Proc Natl Acad
Sci U S A 2007; 104:6776–6781.
26. Benito JM, Lopez M, Lozano S, Martinez P, Gonzalez-Lahoz J,
Soriano V. CD38 expression on CD8 T lymphocytes as a marker
of residual virus replication in chronically HIV-infected
patients receiving antiretroviral therapy. AIDS Res Hum Retro-
viruses 2004; 20:227–233.
27. Marodon G, Warren D, Filomio MC, Posnett DN. Productive
infection of double-negative T cells with HIV in vivo. Proc Natl
Acad Sci U S A 1999; 96:11958–11963.
28. Khan Q, Penninger JM, Yang L, Marra LE, Kozieradzki I, Zhang
L. Regulation of apoptosis in mature alphabetaRCD4SCD8S
antigen-specific suppressor T cell clones. J Immunol 1999;
162:5860–5867.
29. McIver Z, Serio B, Dunbar A, O’Keefe CL, Powers J, Wlodarski
M, et al. Double-negative regulatory T cells induce allotoler-
ance when expanded after allogeneic haematopoietic stem cell
transplantation. Br J Haematol 2008; 141:170–178.
30. Voelkl S, Gary R, Mackensen A. Characterization of the immu-
noregulatory function of human TCR-alphabetaR CD4S CD8S
double-negative T cells. Eur J Immunol 2011; 41:739–748.
31. Zhang ZX, Ma Y, Wang H, Arp J, Jiang J, Huang X, et al. Double-
negative T cells, activated by xenoantigen, lyse autologous B
and T cells using a perforin/granzyme-dependent, Fas-Fas
ligand-independent pathway. J Immunol 2006; 177:6920–
6929.
32. Kuhl AA, Pawlowski NN, Grollich K, Blessenohl M,
Westermann J, Zeitz M, et al. Human peripheral gammadelta
T cells possess regulatory potential. Immunology 2009; 128:
580–588.
33. Bhagat G, Naiyer AJ, Shah JG, Harper J, Jabri B, Wang TC, et al.
Small intestinal CD8RTCRgammadeltaRNKG2AR intrae-
pithelial lymphocytes have attributes of regulatory cells in
patients with celiac disease. J Clin Invest 2008; 118:281–293.
34. Agrati C, D’Offizi G, Gougeon ML, Malkovsky M, Sacchi A,
Casetti R, et al. Innate gamma/delta T-cells during HIV infec-
tion: Terra relatively incognita in novel vaccination strategies?
AIDS Rev 2011; 13:3–12.
35. Locke NR, Stankovic S, Funda DP, Harrison LC. TCR gamma
delta intraepithelial lymphocytes are required for self-
tolerance. J Immunol 2006; 176:6553–6559.
36. Peng G, Wang HY, Peng W, Kiniwa Y, Seo KH, Wang RF.
Tumor-infiltrating gammadelta T cells suppress T and dendri-
tic cell function via mechanisms controlled by a unique
toll-like receptor signaling pathway. Immunity 2007; 27:
334–348.
37. Drobyski WR, Vodanovic-Jankovic S, Klein J. Adoptively trans-
ferred gamma delta T cells indirectly regulate murine graft-
versus-host reactivity following donor leukocyte infusion ther-
apy in mice. J Immunol 2000; 165:1634–1640.
38. Han GC, Wang RX, Chen GJ, Wang JN, Xu RN, Wang LY, et al.
Interleukin-17-producing gamma delta plus T cells protect
NOD mice from type 1 diabetes through a mechanism invol-
ving transforming growth factor-beta. Immunology 2010;
129:197–206.
39. Nagaeva O, Jonsson L, Mincheva-Nilsson L. Dominant IL-10
and TGF-beta mRNA expression in gammadeltaT cells of
human early pregnancy decidua suggests immunoregulatory
potential.
Am J Reprod