mTOR signaling mediates effects of common gamma-chain cytokines on T cell proliferation and exhaustion: implications for HIV-1 persistence and cure research

Abstract

Chronic elevation of plasma cytokines is a key feature of HIV infection. The physiological consequences of this response to infection and its role in HIV persistence are not fully understood. Here, we show that common gamma chain (γc)-cytokines induce both proliferation and expression of T cell exhaustion markers in a mammalian target of rapamycin (mTOR)-dependent fashion, suggesting a possible therapeutic target that, if inhibited, could diminish HIV reservoir expansion, persistence, and resistance to immune surveillance.

Introduction

Deregulation of the cytokine network is a hallmark of acute HIV and SIV infections that impacts pathogenesis [1,2]. Common gamma chain (γc)-cytokines (e.g. IL-7, IL-15) are elevated in plasma throughout infection [3,4], contributing to chronic immune activation and exhaustion, as well as the latent viral reservoir and viral set-point [4–8]. Better understanding of mechanisms by which these cytokines mediate their effects will inform strategies to counter immunopathogenesis not reversed by current antiretroviral therapeutics. As γc-cytokines have been reported to trigger mTOR activity in natural killer and CD8⁺ T cells [9,10], we assessed the role of mTOR in mechanisms underlying γc-cytokine-mediated CD4⁺ T-cell proliferation, activation, and induction of immune checkpoint receptors (ICR).

First, we studied cell proliferation because γc-cytokine-stimulated homeostatic CD4⁺ T cell proliferation has been hypothesized to contribute to persistence of HIV-1 via clonal expansion of the HIV-1 latent reservoir [6]. Notably, during HIV-1 infection and other lymphopenic conditions, homeostatic proliferation of CD4⁺ T cells is associated with elevated serum concentrations of γc-cytokines IL-7 and IL-15 [3,11,12]. On their own, IL-7 and IL-15 are poorly mitogenic, but synergize together or in concert with myeloid antigen-presenting cells (APC) to stimulate CD4⁺ T cell proliferation [13,14]. This is consistent with studies determining that homeostatic proliferation of memory CD4⁺ T cells is mediated by a mixture of IL-7 and IL-15 in vivo[15]. After finding that combined exposure to both IL-7 and IL-15 (IL-7/IL-15) activated mTOR to a greater degree than either alone (data not shown), we tested whether cell-cycle entry stimulated by IL-7 with IL-15 was mTOR-dependent. Resting primary CD4⁺ T cells were exposed to IL-7 with IL-15 for 5 days, with or without treatment with a potent, highly specific catalytic mTOR inhibitor (AZD2014). Cells were then stained intracellularly for Ki-67, a cellular marker that reliably assesses T cell proliferation in vitro[16], and phosphorylated ribosomal protein S6 (p-S6), a prototypical downstream target of mTOR [17]. In addition, we monitored cell size by assessing forward side scatter area (FSC-A), as mTOR-dependent proliferation is linked to its role in regulating cellular growth [18]. AZD2014 completely blocked cell-cycle entry, cell growth, and mTOR activity induced by IL-7/IL-15, as reflected by ameliorated increases in Ki-67, p-S6, and FSC-A, respectively (Fig. 1a). Surface staining for activation markers demonstrates that IL-7/IL-15 also increased surface expression of CD38 and, to a lesser extent, CD69; increases in these activation markers that precede cell proliferation caused by the γc-cytokines were also blocked by AZD2014 pretreatment (Fig. 1b). Of note, catalytic mTOR inhibition did not perturb increases in these early activation markers after anti-CD3⁺/CD28⁺ antibody-coated bead stimulation (Fig. 1b). This observation is consistent with earlier reports that mTOR inhibition does not impair antigen-specific T cell function [19,20].

Fig. 1.

mTOR signaling drives homeostatic proliferation and expression of markers of T cell exhaustion in response to cytokines, and is abrogated by catalytic mTOR inhibition.

Isolated resting CD4⁺ T cells from blood were untreated (grey-filled histogram) or stimulated in the absence (black line) or presence of a catalytic mTOR inhibitor, AZD2014 (red line). AZD2014 was started 24 h prior to the following stimuli, and continued for duration of stimulation. Stimuli were IL-7 with IL-15 (100 ng/ml each) for 5 days or anti-CD3/CD28 beads for 72 h, as indicated. Exposure to IL-7/IL-15 increased CD4⁺ T cell size, proliferation, and mTOR activity as determined by FSC-A, Ki-67 expression, and p-S6 abundance, respectively (black line in a). AZD2014 blocked these increases (red line in a). T cell activation markers CD38 and CD69 each increased after both stimuli (black lines in b) with AZD2014 blunting this after IL-7/IL-15 exposure (red lines in b, upper panel). T cell exhaustion was evaluated by surface expression of co-inhibitory receptors and ligands (TIGIT, PD-1, and PD-L1) (c–e). (c) Demonstrates gating strategy to identify PD-1⁺, TIGIT⁺, PD-1⁺/TIGIT⁺ double positive, and TIGIT hi/dim/neg CD4⁺ T cell populations. The gated cell population frequencies and MFI of p-S6 in TIGIT subpopulations are indicated in black and red, respectively. Fluorescence minus one (FMO) controls for TIGIT and PD-1 are also shown. PD-1, TIGIT, and PD-L1 increases seen after either IL-7/IL-15 or anti-CD3/CD28 beads (black lines in d) were decreased by AZD2014 (red lines in d). In (e), expression of PD-1, TIGIT, and abundance of PD-1⁺/TIGIT⁺ double-positive CD4⁺ T cells were determined by gating from within a CD45⁺/CD3⁺ population to exclude other cell types present in mononuclear cells isolated from colon biopsies after 5 days of exposure to IL-7/IL-15 in the absence (black bar), or presence (red bar) of AZD2014. Statistical significance of differences was determined by two-tailed Student's t-tests ^∗P < 0.05, ^∗∗P < 0.01; ^∗∗∗P < 0.005.

Secondly, we assessed if mTOR contributes to induction of immune checkpoint receptors (ICR). We studied the co-inhibitory receptor PD-1, previously reported to be induced after γc-cytokine stimulation [21] and another receptor-mediating immune exhaustion, T cell immunoreceptor with Ig and ITIM domains (TIGIT). We found that PD-1 and TIGIT were each induced in blood CD4⁺ T cells by γc-cytokines. We determined whether induction of PD-1 and TIGIT is associated with mTOR activity by assessing p-S6 levels in subpopulations (neg/dim/hi) of untreated and γc-cytokine-stimulated CD4⁺ T cells defined by relative mean fluorescence intensity (MFI) of PD-1 (data not shown) or TIGIT (Fig. 1c, right bottom). The IL-7/IL-15-induced TIGIT^hi CD4⁺ T cell subpopulation had the highest level of mTOR activity (p-S6) and IL-7/IL-15-exposed TIGIT^neg CD4⁺ T cells had the lowest, suggesting mTOR activity may cause increased TIGIT expression (Fig. 1c, right bottom). Of note, mTOR activity (p-S6) did not vary significantly among the untreated TIGIT^neg/dim/hi subpopulations (Fig. 1c, right bottom). We also found an association between p-S6 and PD-1 expression, albeit less robust (not shown). To directly test the role of mTOR in driving expression of PD-1 and TIGIT, we treated cells with AZD2014 prior to stimulation with IL-7/IL-15 and observed a block in the upregulation of both PD-1 and TIGIT on blood CD4⁺ T cells in response to IL-7/IL-15 exposure (Fig. 1d). This is relevant because both PD-1 and TIGIT are enriched on blood CD4⁺ T cells either actively replicating HIV or harboring latent provirus [22–25]. We also determined that induction of the ligand for PD-1 (PD-L1) in CD4⁺ T cells by IL-7/IL-15 was also mTOR-dependent (Fig. 1d). Next, as PD-1 expression on CD4⁺ T cells in gastrointestinal tract mucosal tissue of antiretroviral therapy (ART)-treated individuals is associated with HIV-1 persistence [24], we also evaluated the effects of γc-cytokine exposure and mTOR signaling on PD-1 and TIGIT expression on CD4⁺ T cells in normal human colon tissue treated ex vivo. Colon biopsies obtained from healthy adult volunteers during routine screening colonoscopies [after Institutional Review Board (IRB)-approved consent was obtained] were cultured in the presence or absence of IL-7/IL-15 for 5 days. AZD2014-treated tissue cultured in the presence of γc-cytokines was compared with control-treated (DMSO) tissue under the same conditions. Similar to our results with peripheral blood CD4⁺ T cells, IL-7/IL-15 significantly increased abundance of PD-1⁺ (P = 0.0186), TIGIT⁺ (P = 0.0016), and TIGIT⁺PD-1⁺ double positive (P = 0.007) CD4⁺ T cells in colon tissue explants ex vivo (Fig. 1e, black bars). Importantly, pretreatment with the catalytic mTOR inhibitor blocked the IL-7/IL-15-mediated increases in each of these ICRs in colon mucosal CD4⁺ T cells ex vivo (Fig. 1e, red bars).

These results confirm that catalytic mTOR inhibition blocks γc-cytokine-mediated homeostatic proliferation of CD4⁺ T cells. Notably, TIGIT⁺ memory CD4⁺ T (Tm) cells from HIV-infected individuals receiving suppressive ART are significantly enriched for integrated HIV DNA when compared with their TIGIT⁻ counterparts [22]. Evidence also demonstrates that PD-1-mediated signaling enhances the establishment of latency by APCs [26]. Taken together with those earlier studies, the current results raise the possibility that catalytic mTOR inhibition could mitigate HIV-1 persistence resulting from γc-cytokine-mediated homeostatic proliferation of latently-infected CD4 T cells. The physiological relevance of this γc-cytokine/ICR axis in HIV pathogenesis is underscored by the fact that myeloid dendritic cells and monocytes, APCs known to promote HIV latency in CD4⁺ Tm cells, potently induce polyclonal expansion of these cells in the presence of IL-7/IL-15 [13,14]. Further, these APCs highly express both IL-15 receptor alpha (IL-15Rα), which potentiates the effects of IL-15 via trans-presentation, and the poliovirus receptor (PVR), the ligand for TIGIT [27]. In contrast, plasmacytoid dendritic cells express low levels IL-15Rα and consequently fail to potently induce HIV latency or T_M-cell proliferation in the presence of IL-7/IL-15 [14,27].

In addition, PD-L1 expression has been recently described as mediating immune escape of retrovirus-infected cells from killing by cytotoxic CD8⁺ T cells (CTL) [28]. Therefore, prevention of γc-cytokine induction of PD-L1 expression through catalytic mTOR inhibition may enhance immune surveillance, by enabling CTL-mediated killing of CD4⁺ T cells harboring replication-competent HIV. In particular, these results raise the testable hypothesis that γc-cytokine-mediated increases in CD4⁺ T cell PD-L1 expression in a novel ex-vivo culture system [29] causes the observed resistance to CTL-mediated killing of CD4⁺ T cells reactivating HIV-1 after treatment with potent HIV-1 latency-reversing agents, and that mTOR inhibition may enhance immune-mediated elimination of HIV-infected CD4⁺ T cells under those ex-vivo conditions. If so, catalytic mTOR inhibitors may bear clinical evaluation as an adjunct to ‘kick and kill’ cure strategies.

Materials and methods

CD4⁺ T cell isolation and culture

Isolation of resting CD4⁺ T cells from uninfected subjects’ PBMC (Lifesource, Rosemont, Illinois, USA) and culture were performed as described elsewhere [30]. Wherever indicated, cells were treated with cytokines (Peprotech, Rocky Hill, New Jersey, USA) at 100 ng/ml, or pretreated with AZD2014 (5 μmol/l, Selleckchem, Houston, Texas, USA).

Colon tissue culture and isolation of mucosal single cell suspensions

Northwestern University IRB-approved consent for colon biopsies was obtained from unidentifiable, healthy adults undergoing routine screening colonoscopy. Colon tissue biopsies were washed in RPMI (Cellgro, Fisher Scientific, Northern America, USA) and cultured in RPMI containing 10% FBS, Penicillin/Streptomycin/l-glutamine (Gibco, ThermoFisher Scientific, Waltham, Massachusetts, USA), and 500 μg/ml piperacillin/tazobactam (Zosyn; Wyeth, Madison, New York, USA). Where indicated, biopsies were treated with cytokines at 10 ng/ml or pretreated with AZD2014 (5 μmol/l) for 24 h. Mucosal mononuclear single cell suspensions were isolated from colon biopsies by digesting tissue in RPMI culture medium containing collagenase IV (5 mg/ml) and DNAse I (200 U/ml) for 45 min at 37 °C with gentle rotation. Liberated cells were passed through a 70 μm cell strainer and washed in culture medium before use.

Flow cytometry

Isolated CD4⁺ T cells or colon-derived mononuclear cells were stained with LIVE/DEAD Fixable Blue Dead Cell Stain Kit (ThermoFisher Scientific, Waltham, Massachusetts, USA) to exclude nonviable cells and surface phenotyped by staining with fluorochrome-conjugated antibodies at 4 °C for 30 min. For intracellular staining, cells were subsequently fixed and permeabilized using Cytofix/Cytoperm solution (BD Biosciences, San Jose, California, USA). Flow cytometric data was obtained on a LSRFortessa (Becton Dickinson, Franklin Lakes, New Jersey, USA) and analyzed with FlowJo software (Tree Star, Ashland, Oregon, USA) and whenever applicable, fluorescence minus one (FMO) gating strategies were used to set the manual gates. The following antibodies were used: CD3-BV510 (BD Biosciences, #563109), CD4-PE-Cy7 (Biolegend, San Diego, California, USA, #344612), CD45-BV421 (BD Bioscience, #563879), CD69-FITC (BD Bioscience, #555530), CD38-AF700 (BD Bioscience, #560676), TIGIT-PE (eBiosciences, ThermoFisher Scientific, Waltham, Massachusetts, USA, #12-9500-41), PD-1-PE/Dazzle (Biolegend, #329939), PD-L1-APC (eBioscience, #17-5983-41), and Ki-67-PE (Biolegend, #350503).

Acknowledgements

This work was supported by P01 AI 131346 and a Developmental Core Pilot Project award to HET from the Third Coast Center for AIDS Research (CFAR), an NIH-funded center (P30 AI117943). Flow cytometry was conducted at the Northwestern University – Flow Cytometry Core Facility supported by Cancer Center Support Grant (NCI CA060553). Also appreciated are Third Coast CFAR Viral Pathogenesis and Clinical Sciences Core services (P30 AI117943).

Author contributions

Conceptualization, H.E.T.; methodology and experimental design, H.E.T. and R.T.D.; performed experiments, H.E.T.; formal analysis, H.E.T., N.C., and R.T.D.; writing – original draft, H.E.T. and R.T.D.; writing – review and editing, H.E.T., N.C., and R.T.D.; funding acquisition, H.E.T. and R.T.D.

Conflicts of interest

There are no conflicts of interest.