Macrophages and Myeloid Dendritic Cells Lose T Cell–Stimulating Function in Simian Immunodeficiency Virus Infection Associated with Diminished IL-12 and IFN-α Production

Mononuclear phagocytes, including dendritic cells (DC) and macrophages, are integral components of both innate and adaptive immunity. HIV and SIV infection leads to depletion of CD4 T cells and DC (1–5) and diminished Ag-specific T cell responses (6–8), but the relationship between mononuclear phagocyte function and the T cell response remains ill defined. Many groups have examined the impact of HIV and SIV infection on production of proinflammatory cytokines by isolated DC and macrophages (3, 9–15) as well as the effect of HIV exposure in vitro on the IFN response (16). However, studies exploring the T cell–stimulating function of myeloid DC (mDC) and plasmacytoid DC (pDC) in HIV or SIV infection have been limited (11, 13, 14, 17), and virtually nothing is known about the APC function of macrophages in these infections.

The major site of virus replication and T cell priming by mononuclear phagocytes is the lymph node, and SIV-specific T cell responses in lymph nodes but not blood correlate with vaccine-induced protection from infection (18). SIV infection has profound effects on mononuclear phagocyte subsets in the lymph node. During acute infection, there is increased recruitment and turnover of pDC, mDC, and macrophages (15, 19–23), and pDC and macrophages from SIV-infected lymph nodes have reduced responsiveness to ex vivo stimulation (15). However, the capacity for lymph node DC and macrophages to serve as effective APC in HIV and SIV infection is understudied (24, 25). To address these gaps in knowledge, we performed a comprehensive study of DC and macrophage CD4 T cell–stimulating functions in lymph nodes of rhesus macaques with pathogenic SIV infection.

A total of 30 adult male Indian-origin rhesus macaques were used in this study. All protocols and experiments performed on macaques were reviewed and approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh and were in compliance with the U.S. Department of Health and Human Services Guide for the Care and Use of Laboratory Animals. Five animals were infected by i.v. inoculation with SIVmac251 and sacrificed at acute infection (day 36) when inguinal and axillary lymph nodes were harvested, as previously reported (15). Preinfection lymph node biopsies from these animals were also available for analysis. An additional 10 macaques were infected by i.v. inoculation with either SIVmac251 or SIVΔB670 and sacrificed at the chronic stage of infection (range, day 77–470; median, day 404) when inguinal and axillary lymph nodes were harvested, as previously described (26–28) (Table I). Inguinal and axillary lymph nodes from 15 healthy, SIV-naive macaques were used as controls. Lymph nodes were digested, and single-cell suspensions generated using 1 mg/ml collagenase D (Sigma-Aldrich) and 20 μg/ml DNAse I (Roche) in RPMI 1640 with 2% FBS and 10 mmol HEPES and cryopreserved for later experiments.

For responder cells, we used allogeneic CD4 T cells isolated from an SIV-naive rhesus macaque rather than autologous cells, as SIV infection affects CD4 T cell function and Ag specificity to varying degrees between animals (29), which would produce a confounding variable in our experiments. Given the extent of experiments, we used isolated CD4 T cells from two different animals over the course of the study. Preliminary data indicated there was no discernable difference in CD4 T cell proliferation or IFN-γ production between the two donor animals (data not shown). PBMC were enriched for CD4 T lymphocytes by negative selection through MACS with Ab against MHC class II (MHC-II), CD20, CD14, and CD8 according to the manufacturer’s protocol (Miltenyi Biotec). The resulting responder cell populations were >80% pure as measured through flow cytometric analysis, aliquoted, and cryopreserved for later use.

mAbs used in flow cytometry experiments were purchased from BD Biosciences (unless otherwise noted) and were specific for CD3 (clone SP34-2), CD20 (2H7; eBioscience), MHC-II (L243/G46-6), CD163 (GHI/61), CD11c (S-HCL-3), CD123 (7G3), and CD8 (RPA-T8). An enriched population of lymph node mononuclear phagocytes was generated by sorting CD3⁻CD20⁻CD8⁻MHC-II⁺ cells. To positively select subsets of mononuclear phagocytes from lymph nodes, cells were sorted based on the following phenotype: pDC were CD3⁻CD20⁻MHC-II⁺CD163⁻CD123⁺CD11c⁻, mDC were CD3⁻CD20⁻MHC-II⁺⁺CD163⁻CD123⁻CD11c⁺, and macrophages were CD3⁻CD20⁻MHC-II⁺CD163⁺ (15). All sorting was done using a biocontained FACSAria (BD Biosciences), and data were analyzed using FlowJo 8.8.7 (Tree Star).

Twenty thousand enriched or positively selected mononuclear phagocytes were incubated in complete media (RPMI 1640 containing 10% FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mmol glutamine, 10 mmol HEPES, 0.1 mmol nonessential amino acids, and 0.1 mmol sodium pyruvate) with or without toxic shock syndrome toxin (TSST)-1 at 1, 10, or 100 ng/ml for 1 h at 37°C. Cells were then washed and transferred to a sterile 96-well round-bottom tissue culture–treated plate. Responder CD4 T cells isolated from one donor animal were labeled with CFSE according to the CellTrace CFSE cell proliferation kit protocol (Invitrogen). Two hundred thousand CFSE-labeled responder cells were added to each well and the volume of all of the wells normalized to 200 μl with complete media. The coculture was allowed to incubate for 5 d. The proliferation of T lymphocytes was then measured through flow cytometric analysis by staining for CD3 and CD4 while recording CFSE dilution on the FITC channel. Dead cells were excluded through the use of a Live/Dead viability kit (Invitrogen). All terminal flow cytometric data were collected on an LSR II flow cytometer with BD FACS Diva 6.0 (BD Biosciences) and analyzed with FlowJo 8.8.7 (Tree Star).

Twenty thousand mononuclear phagocytes were placed directly into an anti–IFN-γ ELISPOT plate that had already been coated and blocked according to the manufacturer’s protocol (3420M-2A; Mabtech). The mononuclear phagocytes were either incubated in complete media or exposed to 1 μg SIVmac251-derived ssRNA oligonucleotide Env976 that is a ligand for TLR7/8 and a potent activator of mononuclear phagocytes as previously shown (15). After a 2-h low-volume incubation, 200,000 responder CD4 T cells isolated from one donor animal were added to each well and the volume of all of the wells normalized to 200 μl. In assays measuring recovery of function, exogenous recombinant rhesus IFN-α2 (1000 IU/ml; PBL InterferonSource) and/or recombinant rhesus IL-12 (10 ng/ml; R&D Systems) was added to each well. After 2 d of coculturing, supernatants were harvested and stored at −80°C for IL-12 detection. The plates were then developed to quantify IFN-γ spot-forming units (SFU) according to the manufacturer’s protocol (3420M-2A; Mabtech) using an Immunospot Series 3B Analyser ELISPOT reader.

Paired single-cell suspensions of lymph nodes collected before and postinfection were exposed to 10 μg/ml Env976 for 7 h at 37°C. The samples were then stained for viability, the surface markers mentioned above for cell identification, and CD40 (5C3), CD80 (L307.4), CD86 (2331 [FUN-1]), and CD274 (programmed death ligand-1, 29E.2A3). Paired samples from the same animals were stained and analyzed together to control for potential daily differences in flow cytometer calibration.

Lymph node cell suspensions were maintained in complete media with or without 10 μg/ml Env976 for 7 h at 37°C. The final 5 h of incubation were in the presence of 1 μg/ml brefeldin A. The samples were then stained for viability, surface markers, and the intracellular cytokines IFN-α (225.C) and IL-12 (11C.5) as previously described (15). Supernatants from sorted pDC, mDC, and macrophage cocultures were collected after 48 h as mentioned above. The ELISA assay to detect the p40 subunit of IL-12 was performed according to the manufacturer’s protocol (CT719-10; U-Cytech).

Statistical analyses such as t tests were performed using GraphPad Prism v6 software (GraphPad), whereas ANOVA and repeated-measures analyses were done with R v3.1. The p values <0.05 were considered significant. Comparisons between treatment arms and/or study groups were tested by means of linear contrasts on the ANOVA parameters. The ANOVA assumptions of all models were graphically checked, and data were log-transformed when necessary. Tests used for each experiment are noted in the figure legends.

In order to determine the ability of lymph node mononuclear phagocytes to stimulate T cell proliferation, we cultured enriched mononuclear phagocytes isolated from lymph nodes of five acutely infected macaques (Table I) with CD4 T cells and measured T cell proliferation through CFSE dilution. Mononuclear phagocytes were enriched by sorting cells that expressed MHC-II but not CD3, CD8, or CD20 (Fig. 1A). The mononuclear phagocyte gate encompassed MHC-II^bright autofluorescent cells to include the majority of mDC in lymph nodes as we previously have shown (1). Allogeneic CD4 T cells isolated from one of two different SIV-naive macaques were used as responders to avoid the potential for differences in autologous CD4 T cell function as a result of SIV infection. Postinfection, enriched lymph node mononuclear phagocytes were less able to stimulate T cells to proliferate, as CFSE dilution was significantly diminished, both in the presence and absence of the superantigen TSST (Fig. 1B). We next determined the ability of enriched mononuclear phagocytes to drive IFN-γ production from allogeneic CD4 T cells. In this study, we used the biologically relevant polyuridine-rich ssRNA sequence Env976 encoded by SIVmac251 to stimulate pDC, mDC, and macrophages through TLR7/8 (15). Enriched mononuclear phagocytes from acutely infected lymph nodes were significantly less able to stimulate T lymphocytes to produce IFN-γ both in the absence and presence of TLR7/8-mediated activation as measured by ELISPOT assay (Fig. 1C).

To define which mononuclear phagocyte subset was responsible for this defect in T cell–stimulating function and determine if the defect persisted into chronic infection, we used flow cytometry to purify pDC, mDC, and macrophages from lymph nodes (Fig. 2A) of the acutely infected macaques along with lymph nodes from a group of 10 chronically infected macaques (Table I) and cocultured them with CFSE-labeled CD4 T cells. Lymph node pDC were relatively weak at inducing CD4 T cell proliferation in naive monkeys but did not lose this capacity as a result of acute or chronic SIV infection (Fig. 2B, 2C). Conversely, both mDC and macrophages induced strong CD4 T cell proliferation in the naive state based on CFSE staining, and both responses were significantly impaired in acutely infected macaques. Moreover, macrophage induction of T cell proliferation was further reduced in chronic infection and was not corrected in the presence of TSST superantigen, in contrast to mDC from acute infection (Fig. 2B, 2C).

In naive lymph nodes, pDC, mDC and macrophages all were capable of stimulating IFN-γ release from allogeneic CD4 T cells, with macrophages being the most potent inducer of IFN-γ release both with and without activation with Env976 (Fig. 3A, 3B). No IFN-γ production was detected by APC alone, in the absence of CD4 T cells (Fig. 3A). After acute SIV infection, the capacity for macrophages to induce T cell release of IFN-γ was significantly suppressed, even with TLR7/8 stimulation, but this function was largely restored in chronic infection. There was a trend toward loss of pDC- and mDC-mediated IFN-γ production by CD4 T cells in acute infection, but this did not reach statistical significance (Fig. 3B).

To begin determining the mechanism of defective T cell stimulation by DC and macrophages, we first assayed for the expression of the costimulatory molecules CD40, CD80, and CD86 as well as CD274 (programmed death ligand-1) on paired samples of naive and acutely infected lymph node suspensions stimulated with Env976 (Fig. 4A). There was no change in expression of any of these surface markers when comparing preinfection and postinfection lymph node preparations from the same animals with the exception of macrophages, which had modest increases in expression of CD80 and CD274 postinfection (Fig. 4B).

We next focused on production of IFN-α and IL-12, two key cytokines released by activated mononuclear phagocytes that contribute to T cell responses. As expected, pDC in cell suspensions from SIV-naive lymph nodes stimulated with the TLR7/8 ligand Env976 were the principal producer of IFN-α, with 40% of pDC producing this cytokine, but macrophages were also significant contributors to IFN-α production in the naive state. mDC produced negligible amounts of IFN-α after stimulation with Env976 (Fig. 5A, 5B). SIV infection had variable effects on the capacity of lymph node cells to produce IFN-α, with significant loss of IFN-α production by pDC and macrophages in chronic and acute infection, respectively (Fig. 5B). In contrast to IFN-α, pDC, mDC, and macrophages from naive lymph nodes all were significant producers of IL-12 after stimulation with Env976. pDC production of IL-12 was unaffected by SIV infection, whereas the proportion of both mDC and macrophages that produced IL-12 after stimulation with Env976 was markedly suppressed in acute and chronic infection (Fig. 5B).

To explore if diminished IFN-γ production from T cells was directly related to suppressed cytokine production from mononuclear phagocytes, we measured the concentration of cytokine in supernatants from each CD4 T cell coculture. We focused on IL-12 due to limited sample volume and evidence that IL-12 is more critical than IFN-α in driving T cell immune responses (30, 31). Importantly, IL-12 was never detected in any of the wells that contained CD4 T cell responders alone, in the absence of APC (Fig. 6A). Without TLR7/8 stimulation, macrophages were the only mononuclear phagocyte population that produced significant amounts of IL-12 in coculture with T cells, whereas after stimulation with Env976, pDC were the dominant producer of IL-12, followed by macrophages, and, to a lesser extent, mDC (Fig. 6A). SIV infection had significant impact on the capacity for mononuclear phagocytes to secrete IL-12 in coculture with allogeneic CD4 T cells. Unstimulated macrophages showed progressive loss of IL-12 production, which dropped 10-fold in chronic infection relative to naive lymph nodes (Fig. 6A). In the presence of TLR7/8 stimulation, pDC, mDC, and macrophages from chronically infected lymph nodes had marked suppression of IL-12 production in cocultures with CD4 T cells (Fig. 6A). There was a strong positive relationship between IL-12 production by macrophages isolated from naive and acutely infected lymph nodes and the number of IFN-γ–producing CD4 T cells in the same cultures (Fig. 6B). This relationship was not significant for pDC (p = 0.18; r² = 0.10) and mDC (p = 0.45; r² = 0.03), consistent with the fact that loss of IL-12 from these cells in cocultures was most evident during chronic infection (Fig. 6A), whereas loss of capacity to stimulate T cells to produce IFN-γ was most impacted during acute infection (Fig. 3B). We next performed stepwise supplementation of IFN-α, IL-12, or both cytokines to cocultures of pDC, mDC, or macrophages isolated from lymph nodes at acute infection and allogeneic CD4 T cells in ELIPSOT assays to determine if cytokine addition would restore or enhance T cell function. Again, we focused on acutely infected samples as we observed the greatest stimulatory defect at this time point. Exogenous IL-12 with or without IFN-α enhanced T cell production of IFN-γ induced by mDC, in both unstimulated and Env976-stimulated conditions, whereas IFN-α and IL-12 together recovered the T cell–stimulatory capacity of Env976-stimulated macrophages. Cytokine addition did not significantly impact the T cell–stimulating function of pDC isolated from lymph nodes during infection (Fig. 6C). Exogenous cytokines added to responder cells alone did not drive the production of IFN-γ (data not shown), revealing that mononuclear phagocytes were required for this response.

Our study reveals that an enriched population of mononuclear phagocytes from lymph nodes of SIV-infected macaques has significant impairment in the ability to stimulate CD4 T cell proliferation and IFN-γ release that was not diminished by addition of superantigen or stimulation with virus-encoded TLR7/8 ligand. The loss of T cell–stimulating capacity was mediated primarily by dysfunction of macrophages and mDC, but not pDC, and was linked to deficient production of IFN-α and IL-12, two key cytokines in T cell priming (30, 31).

The most striking defect in lymph node mononuclear phagocyte function that has not previously been described was seen with macrophages, which lost capacity to both drive CD4 T cell proliferation and induce T cell release of IFN-γ in SIV-infected macaques. Macrophages accumulate in lymph nodes and gut mucosa in both SIV and HIV infection (15, 23, 32, 33), and the level of monocyte turnover is predictive of disease progression (21, 22). Our data suggest that loss of macrophage function in stimulating adaptive T cell responses may be a critical component in progressive HIV and SIV infection. Lymph node macrophages lost capacity to produce IFN-α, consistent with previous studies (15), although macrophages in gut mucosa of untreated HIV-infected patients produce proinflammatory molecules, suggesting there may be anatomical differences in macrophage function (32). Indeed, macrophages in humans and rhesus macaques have considerable heterogeneity across tissues (33, 34), and it will be important in future studies to determine if T cell–stimulating function of the different macrophage subsets is uniformly or differentially affected by SIV infection and how this relates to disease progression.

Lymph node mDC in acute SIV infection failed to induce normal CD4 T cell proliferation and in both acute and chronic infection had markedly reduced production of IL-12 after stimulation with virus-encoded TLR7/8 ligand. Although suppression of mDC-mediated stimulation of IFN-γ release from CD4 T cells was not statistically significant, addition of IL-12 strongly enhanced the capacity for mDC to drive cytokine production from T cells. Increased recruitment of mDC to lymphoid tissues is associated with disease progression in SIV-infected macaques (27), and in HIV infection, mDC from the blood of elite controllers but not individuals with chronic progressive infection stimulate CD4 T cell proliferation and IFN-γ release from HIV-specific CTL (17). Collectively, these data are suggestive of a generalized defect in mDC T cell–stimulating function in HIV and SIV infection that reduces immune control of virus, thereby promoting disease development. This may be associated with mDC-driven induction of regulatory T cells in lymph nodes, as demonstrated in both SIV and HIV infection (24, 25).

Our data show that mDC from virus-infected lymph nodes produce negligible amounts of IFN-α in response to stimulation with stimulatory virus RNA, consistent with the notion that HIV can escape intracellular recognition in mDC (16, 35, 36). However, recent studies suggest that mDC function in elite controllers is linked to robust production of type I IFN following exposure to virus (17). Together with the loss of IFN-α production by macrophages and the beneficial effect of IFN-α on macrophage function in our study, these data suggest that the IFN-α response is a critical component in mDC- and macrophage-mediated T cell responses in infection. We were not able to measure production of IFN-β in our study due to lack of cross-reactive reagents, and studies with HIV suggest that subversion of the IFN-β response in macrophages and DC is an important mechanism of immune evasion (16, 37). It is possible that SIV infection impacts IFN-β production by macrophages and DC in a manner that is independent from IFN-α, and analysis of this issue awaits further study.

pDC are known to have robust APC function (38, 39), and in our study, lymph node pDC harvested prior to infection were strong inducers of IFN-γ release from T cells. Notably, however, this function was maintained in SIV infection, despite loss of IFN-α and IL-12 production. These findings suggest that additional factors may be involved in driving pDC function distinct from macrophages and mDC.

Our findings are consistent with previous reports describing a loss of IL-12 production by PBMC from HIV-infected patients (40–43) and the beneficial effects of supplementation of IL-12 to CD4 and CD8 T cell responses in HIV-infected individuals (44). Factors in the plasma of HIV-infected patients have also been shown to inhibit IL-12 production from monocyte-derived DC (45, 46). The findings suggest that direct administration of IL-12 may act to promote DC and macrophage function in stimulating antiviral T cell responses, and this approach has been shown to decrease virus load and prolong survival in SIV-infected macaques (47). However, IL-12 administration could also potentially be detrimental, as the broadly immunostimulatory effects of this cytokine (48) could create an increased pool of activated CD4 T cells that would serve as targets for infection.

We limited our analyses to using allogeneic CD4 T cells to control for differential autologous T cell responses in SIV-infected macaques, as have others in studying HIV-infected individuals (11, 13, 14, 24). Data from the rhesus macaque model suggest that lymph node DC stimulation of autologous T cells may also be dysfunctional in SIV infection (25). Our findings of marked mDC and macrophage dysfunction in acute infection and persistent macrophage dysfunction in chronic infection suggest that both the capacity to stimulate early virus-specific responses as well as responses to opportunistic pathogens in progressive infection may be impacted.

In summary, our findings identify previously unappreciated effects of pathogenic SIV infection on the T cell–stimulating function of macrophages and mDC in lymph nodes, indicating that immunosuppression during progressive infection may be due in part to mononuclear phagocyte dysfunction. Future studies aimed at assessing and reversing immunopathology in SIV and HIV infection should take into account the Ag-presenting and T cell–stimulating function of these cells.

We thank Tim Sturgeon for technical expertise and Karen A. Norris for rhesus macaque blood samples.

The authors have no financial conflicts of interest.