Hiv Jurnal
Hiv Jurnal
Hiv Jurnal
Correspondence
[email protected]
In Brief
The cure for HIV is not achievable due to
HIV reservoirs, mostly in resting CD4+
T cells. Monel et al. show that CD8+ T cells
from HIV controllers are able to establish
immunological synapses with HIV+
resting CD4+ T cells, leading to IFN-g,
MIP1-b production, degranulation, and
the elimination of the target cells.
Highlights
d Non-activated CD4+ T cells are permissive for HIV entry but
not productively infected
Article
*Correspondence: [email protected]
https://doi.org/10.1016/j.celrep.2019.03.016
142 Cell Reports 27, 142–153, April 2, 2019 ª 2019 The Authors.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
that resting cells may also be targeted (Buckheit et al., 2013), production was detected by means of HIV long terminal repeat
although the role of antigen processing, restricting human (LTR)-driven GFP expression (Cavrois et al., 2002; Tilton et al.,
leukocyte antigen (HLA) alleles, and synapse formation in the 2014).
observed elimination of infected cells was not evaluated. We Using this system, we assessed viral entry and levels of produc-
tested the hypothesis that HIV-specific CD8+ T cells from HIV tive infection, comparing activated to non-activated CD4+ T cells
controllers, a very small proportion of the HIV-infected popula- from healthy donors. The activation status of live CD3+CD4+
tion who manage to spontaneously control viral replication T cells in whole peripheral blood mononuclear cells (PBMCs)
and maintain stable CD4+ T cell counts without the need for an- was assessed ex vivo by flow cytometry by analyzing the expres-
tiretroviral therapy (Carrington and Walker, 2012; Deeks and sion of CD25 and CD69, inducible cell surface glycoproteins
Walker, 2007; Walker and Yu, 2013), could recognize and kill acquired during lymphocyte activation. In the absence of exoge-
non-activated, infected CD4+ T cells due to the recognition of nous stimulation, CD4+ T cells within the PBMCs were quiescent,
processed viral proteins following viral entry, without requiring but were readily activated by incubation with CD3/CD28 beads
productive infection. We used a combination reporter virus sys- for 2 days. A representative experiment is shown in Figure S1A.
tem that allowed us to sensitively and specifically track the ki- Of note, the activation status was similar when CD4+ T cells
netics of infection of resting CD4+ T cells after viral entry and were first isolated from PBMCs (data not shown).
before any de novo viral protein production. We show that Two hours following infection, activated and non-activated
CD8+ T cells from HIV controllers readily establish functional CD4+ T cells were assessed for viral entry, as evidenced by
synapses with non-activated infected CD4+ T cells, leading to b-lactamase-mediated cleavage and fluorescence of the cyto-
HLA class I-restricted degranulation, cytokine production, and plasmic substrate. Non-activated (CD25 , CD69 ) CD4+ T cells
target cell death, and does not require reverse transcription, were highly permissive to entry by X4-tropic HIV (Figure 1A),
indicating that de novo viral protein production is not needed. with viral entry detected in 65% ± 11% of resting CD4+ T cells
Moreover, we show that cell-cell transmission also sensitized at the multiplicity of infection used (Figure 1B, top). The entry
cells to HIV-specific CD8+ T cell recognition, before viral reverse of R5 tropic viruses was also detected, but to a lesser extent
transcription occurs. This response is significantly more potent (5% ± 1% of resting CD4+ T cells), consistent with lower C-C
in HIV controllers than in progressors, suggesting a mechanism chemokine receptor type 5 (CCR5) expression on the resting
whereby the immune response may influence the size of the HIV CD4+ T cells (Figures 1B, bottom, and S1B). Similar levels
reservoir. of infection were observed when non-activated CD4+ T cells
were first isolated from PBMCs (data not shown). To be certain
RESULTS that the cleaved substrate corresponded to viral entry, a virus
missing the envelope (HIV DEnv) and a fusion-defective virus
HIV Infection of Primary Non-activated CD4+ T Cells (HIV X4 Env-F522Y) were used as controls (Figure S2). Quan-
Direct HIV infection of non-activated CD4+ T cells leads predom- tification of GFP expression in CD4+ T cells 2 days later re-
inantly to abortive infection and to a lesser extent, latent infec- vealed that most of the non-activated HIV-exposed CD4+
tion, which renders cells largely invisible to HIV-specific CD8+ T cells remained non-productively infected, contrary to activated
T cells (Pan et al., 2013; Tilton et al., 2014). Since incoming CD4+ T cells (Figure 1C). These results are consistent with previ-
virions can sensitize activated CD4+ T cells for recognition by ous reports (Haqqani et al., 2015; Tilton et al., 2014) and further
CD8+ T cells (Buseyne et al., 2001; Kløverpris et al., 2013; Payne suggest that most of the directly infected non-activated CD4+
et al., 2010), we first sought to confirm whether resting CD4+ T cells remain non-productively infected during the period
T cells would likewise be permissive for HIV entry, as previously observed.
shown (Tilton et al., 2014), and to determine whether these cells
could be recognized by CD8+ T cells pre-integration and thus Recognition of HIV+ Non-activated CD4+ T Cells by HIV-
before possible abortive infection or establishment of latent Specific CD8+ T Cells
infection. We next tested whether an HIV Gag-specific CD8+ T cell line
To assess the ability of non-activated CD4+ T cells to become derived by stimulating PBMCs from an HIV controller with a
infected with HIV, we used a combination reporter virus system Gag peptide pool could recognize these non-productively in-
that allowed for discrimination between viral entry into the fected CD4+ T cells. We first examined the formation of func-
cytoplasm and subsequent de novo virion production in the tional immunological synapses by fluorescence microscopy. A
infected cell (Tilton et al., 2014). Resting CD4+ T cells were in- cytolytic immunological synapse requires recognition of the viral
fected with HIV containing b-lactamase fused to HIV Vpr (Vpr- peptide:major histocompatibility complex (MHC) by the T cell
blam). Viral entry was detected by pre-labeling cells with a receptor (TCR) and polarization of adhesion molecules including
fluorescence resonance energy transfer (FRET) cytoplasmic LFA-1 and actin toward the point of contact to form a ring with
substrate (coumarin cephalosporin fluorescein, a fluorescent central clearance, through which cytolytic granules are released
beta-lactamase substrate [CCF2-AM]) composed of a hydroxy- (Ritter et al., 2015; Stinchcombe and Griffiths, 2003). To deter-
coumarin donor conjugated to a fluorescein acceptor via a mine whether immune synapse formation was occurring, HIV-
b-lactam ring. Cleavage of the b-lactam ring is mediated via infected non-activated CD4+ T cells (defined by the expression
the b-lactamase protein carried by the incoming virus, inducing of the b-lactamase cleaved substrate) were co-cultured for
an emission shift that allows for the colorimetric detection of 30 min with an autologous HIV Gag-specific CD8+ T cell line
viral entry into the cell by flow cytometry. De novo HIV protein 3 h post-HIV entry, a time point shown above to lack viral protein
non-productively HIV-infected CD4+ T cells, before viral protein I, the main adhesion molecule involved in immunological synapse
production, elicit similar CD8+ T cell responses to autologous formation (Dustin and Shaw, 1999; Grakoui et al., 1999), resulted
cells loaded with pooled Gag peptides. in decreased degranulation of CD8+ T cells (Figure 2I). Blocking
The above results suggest that autologous CD8+ T cells can LFA-3 or ICAM-2 did not reduce the CD8+ T cell response, sug-
form immunological synapses and degranulate and release gesting that these molecules may be less important to the recog-
cytokines in response to non-productively HIV-infected resting nition of infected, non-activated CD4+ T cells.
CD4+ T cells. To assess the relation between synapse formation
and degranulation, we inhibited the formation of immunological CD8+ T Cell Recognition of Non-activated C4+ T Cells
synapses using antibodies against adhesion molecules involved following Cell-Cell Transmission
in this interaction (LFA-1, intercellular adhesion molecule 2 The above results were obtained using cell-free infection with
[ICAM-2], LFA-3). Consistent with our hypothesis, blocking LFA- a high viral inoculum. To further confirm our results under
30
studied for diverse viruses in antigen-presenting cells (APCs), B HIV cell-to-cell transmission
B Chromium release assay
mostly in dendritic cells and macrophages in the context of the
exogenous antigens’ pathway to prime the CD8+ T cells. Howev- B Quantification of HIV DNA reservoir by digital droplet
er, CD4+ T cells are not known as efficient APCs, and thus the PCR
presentation of viral antigens from incoming particles through d QUANTIFICATION AND STATISTICAL ANALYSIS
HLA-I that does not rely on de novo protein production is not
completely expected and is not well characterized in CD4+ SUPPLEMENTAL INFORMATION
T cells. Furthermore, activated CD4+ T cells and resting CD4+
Supplemental Information can be found online at https://doi.org/10.1016/j.
T cells possess different enzymatic activities regarding the anti-
celrep.2019.03.016.
gen-processing pathway (J.B., unpublished data), and due to
their activation status, resting CD4+ T cells are not expected to ACKNOWLEDGMENTS
efficiently present antigen from incoming particles. Therefore,
our study confirms that resting CD4+ T cells are able to present The authors thank Thomas Murooka and Thorsten Mempel for the HIV proviral
antigens from incoming particles through HLA-I and provides constructs, the Ragon Cell Processing Lab, the clinical staff at the Massachu-
some mechanistic insight regarding this antigen-presentation setts General Hospital, and all of the study participants, as well as the Ragon
Microscopy Core (particularly Thomas J. Diefenbach) and the Ragon Flow
pathway.
Cytometry Core (particularly Michael Waring). The authors thank also Thomas
Additional studies will be required to understand why CD8+
Obadia from Institut Pasteur (Paris) for help with the biostatistics analysis. This
T cells from HIV controllers are more efficient in recognizing study was supported in part by grants from the Bill and Melinda Gates
non-activated infected cells and to determine factors that modu- Foundation (OPP1066973 and OPP1146433), the NIAID (R37AI067073 and
late this recognition. It will be important also to determine R01AI118544), and the Harvard University Center for AIDS Research (CFAR;
whether this is unique to the protective alleles HLA-B*27 and P30 AI060354), which is supported by the following institutes and centers
B*57 or can be seen in the context of other alleles. Nevertheless, co-funded by and participating with the NIH: NIAID, NCI, NICHD, NHLBI,
NIDA, NIMH, NIA, FIC, and OAR.
these data indicate that non-productively infected cells can be
targeted by CD8+ T cells, suggesting a path forward to reducing
AUTHOR CONTRIBUTIONS
the viral reservoir. Furthermore, our study could have implica-
tions for T cell-based prophylactic vaccines as an adjunct to anti- B.M. designed the study, performed the experiments, analyzed the data, and
body-mediated vaccines, as the presence of these cells at the wrote the manuscript; P.L-M. provided intellectual input and helped to write
time of infection may limit the establishment of the reservoir. the manuscript; A.M. and P.J. performed the experiments; Y.P. performed
Figure 6. The CD8+ T Cell Response to HIV+ Non-activated CD4+ T Cells Participates in HIV Control
(A–D) CD8+ T cell responses from HIV controller (red) or progressor (blue) patients (all HLA-B*27 or B*57, and n = 10 patients for both groups) in the context of
whole non-activated PBMCs were measured by (A) CD107a expression, (B) intracellular IFN-g expression, (C) intracellular MIP-1b expression, and (D) intracellular
TNF-a expression after 2 h of HIV infection and 5 h of incubation. Results are shown following background subtraction. Statistical differences were assessed
using PRISM software and an unpaired t test. *p < 0.05, **p < 0.01, and ns, not statistically significant.
(E) CD8+ T cell response to Gag pool peptides (final concentration 1 mg/mL/peptide) was measured for each patient by intracellular IFN-g staining. A repre-
sentative experiment is shown for an HIV progressor (top) and for an HIV controller (bottom).
(F) The CD8+ T cell response to HIV+ non-activated CD4+ T cells expressed relative to the Gag pool peptide response.
(G) The HIV DNA reservoir was measured by droplet PCR, and the Spearman correlation between the log10 of Gag copies/106 non-activated CD4+ T cells and the
CD8+ T cell IFN-g response is shown for all of the patients (left, n = 12 patients) and by sub-groups (right, n = 6 patients for each group).
See also Table 1 and Figures S5 and S6.
Further information and requests for reagents may be directed to and will be fulfilled by the Lead Contact, Bruce Walker (bwalker@
mgh.harvard.edu).
Study subjects
PBMC from HIV-1-infected individuals were used for this study according to protocols approved by the Institutional Review Board of
the Massachusetts General Hospital. Peripheral blood mononuclear cells (PBMC) were isolated using Ficoll-Hypaque density
gradient centrifugation
HIV Controllers include Elite Controllers who were defined as having HIV-1 RNA below the level of detection for the respective
available ultrasensitive assay (e.g., < 75 RNA copies/ml by cDNA or < 50 copies by ultrasensitive PCR) without antiretroviral therapy
and Viremic Controllers as having HIV-1 RNA between 50 and 2000 RNA copies/mL. CD4+ T cell counts, viral loads and HLA types
were determined as described (Pereyra et al., 2008). HIV Progressors were treated with Anti-retroviral (ARV) therapy for 1 to 4 years.
Characteristics of the study subjects are shown in Table 1.
Cell lines
HeLa cells and HEK293T cells (both from ATCC) were cultured at 37 C in Dulbecco Modified Eagle Medium (DMEM) supplemented
with 10% of Fetal Bovine Serum and 1% of penicillin/streptomycin.
METHOD DETAILS
Cells
Peripheral blood mononuclear cells (PBMC) were isolated using Ficoll-Hypaque density gradient centrifugation, rested a few hours at
37 C before being infected, checked for activity status or activated with anti-CD3/anti-CD28 beads. CD4+ T cells were isolated from
PBMC by negative selection (StemCell Technologies, Vancouver, Canada). HeLa cells and HEK293T cells (both from ATCC)
were cultured at 37 C in Dulbecco Modified Eagle Medium (DMEM) supplemented with 10% of Fetal Bovine Serum and 1% of
penicillin/streptomycin.
Most of the data are represented as Mean ± standard deviation and statistical analyses were performed using GraphPad PRISM soft-
ware. t tests were used to compare two groups of samples, one-way ANOVA multiple comparison tests were used to compare each
column to a control column or to every other columns and two-way ANOVA multiple comparison tests were used when grouped data
were analyzed and compare to a control column (Figures 5B and 5C). Regarding Figure 6G we used a Spearman correlation to assess
monotonic relationships between the Log10 of Gag copies/10^6 non-activated CD4+ T cells and the CD8+ T cell IFN-g response. The
Spearman’s rank correlation coefficient (rs) is indicated in the Figure. The corresponding tests are indicated for each figure in the
legend. In Figure 6, the n indicated in the legend represents the number of patients studied for each group (n = 10 Controllers and
n = 10 Progressors for Figures 6A–D and 6F; n = 12 for Figure 6 left panel; n = 6 Controllers and n = 6 Progressors for Figure 6G right
panel. No methods were used to determine whether the data met assumptions of the statistical approach.