To generate CD1c-restricted T cells, human T cells were isolated from healthy donors and stimulated for several weeks with autologous DCs loaded with the endogenous lipid phosphatidylcholine (PC). These cells were then tested for functionality, CD1c-restriction, and phenotyped. The analysis indicated that these CD8⁺, αβ TCR, non-cytotoxic (data not shown) T cells are PC-reactive and CD1c-restricted (Figure 1).

HIV-1 infection induces the production of cholesterol [²⁹,³⁰], which may be presented by CD1c/CD1d molecules to initiate an immune response. To address this issue, we determined whether increased lipid production results in increased CD1c/CD1d expression. To accomplish this, Jurkat cells were cultured for 10 days in the presence of mevalonate, a precursor and known cholesterol inducer. The measurement of cholesterol production revealed that the treated cells produced significantly more cholesterol than untreated controls (Table 1). However, despite increased cholesterol production, the CD1c/CD1d levels remained unchanged (Figure 2A), suggesting that increased lipid synthesis does not result in increased CD1c/CD1d expression.

We next wanted to know whether cells producing excess cholesterol were more capable of inducing CD1c-restricted T cells to secrete interferon (IFN)-γ compared to cells not producing excess lipid. Jurkat cells were treated with 200 μM of mevalonate and used to stimulate the secretion of IFN-γ from CD1c-restricted T cells. The cells treated with mevalonate had an enhanced capacity to stimulate CD1c-restricted T cells compared to untreated controls (Figure 2B), suggesting that CD1c-restricted T cells may be able to recognize increased lipid levels in the context of CD1c molecules.

Based on our findings that an increase in cholesterol production resulted in increased stimulation of CD1c-restricted T cells and studies by other groups showing that HIV induces cholesterol production [²⁸,²⁹], we next assessed the level of CD1c expression during HIV-1 infection. For this purpose, we infected Jurkat cells with HIV-1 IIIb and measured CD1c/CD1d expression by FACS on day 10. We found that CD1c was downregulated 40% compared to uninfected controls, while CD1d was downregulated by 53% on average from the cell surface (Figure 3A). Similar levels of CD1c/CD1d modulation were found with varying viral titrations, suggesting that the viral titer does not significantly affect the levels of CD1c and CD1d downregulation (data not shown).

The mechanism of CD1 molecule modulation during HIV-1 infection is poorly understood. While some have shown that CD1d is downregulated by Nef [²³,²⁴], one study showed that Nef does not alter the expression of CD1d [²⁶], and another showed that CD1d expression is downregulated by Vpu [²⁷]. When we infected Jurkat cells with wild-type (WT), Nef-deleted (ΔNef) or Vpu-deleted (ΔVpu) viruses, we found a significant difference between the WT and ΔVpu-infected cells, but not the WT and ΔNef infected cells on days 2 (Figure 4A) and 10 (Figure 4B) post-infection. The viral RNA copy numbers among different viruses were similar (Figure 4C). In fact, the viral load of ΔVpu virus appeared to be slightly higher than that of ΔNef virus (Figure 4C), ensuring that the CD1 down-modulation was not due to the amounts of viruses used to infect Jurkat cells.

In order to confirm whether Vpu itself is sufficient to accomplish this function, we transfected Jurkat cells with a Vpu-expressing plasmid and determined the expression of CD1c/CD1d molecules by the transfected Jurkat cells. As shown in Figure 5, the transfection of a plasmid encoding only Vpu could significantly downregulated both CD1c and CD1d albeit to a lesser degree than that caused by WT plasmid encoding a whole HIV-1 sequence.

These results altogether indicate that the levels of CD1c and CD1d expression were downregulated in a Vpu-dependent and Nef-independent fashion.

The measurement of cholesterol production in the HIV-1-infected cells showed a significant 2-fold increase in cholesterol production compared to uninfected controls (Table 1). Thus, these data suggest that, during HIV-1 infection, CD1c is downregulated and cholesterol production is increased. To rationalize why CD1c/CD1d modulation is not more significantly enhanced during HIV-1 infection and understand the role cholesterol production may play in CD1c/CD1d expression during infection, we blocked cholesterol production in the HIV-1-infected cells by incubating the cells with 10 μM simvastatin, an enzyme that inhibits the cholesterol biosynthesis pathway. We measured a dramatic decrease in CD1c and CD1d expression in the presence of the inhibitor (Figure 3B), with CD1c expression decreasing by 17% compared to the HIV infected, untreated controls. Simvastatin treatment alone had minimal (less than 1% modulation) effects on CD1c expression (data not shown). These data suggest that cholesterol production during HIV-1 infection may aid in preventing further decreases in CD1c and CD1d expression.

Because an increased production of cholesterol resulted in increased stimulation of CD1c-restricted T cells, and HIV-1 infection induced the downregulation of CD1c expression, we hypothesized that the HIV-1-mediated downregulation of CD1c and elevated cholesterol production would counteract each other and lead to the overall level of the CD1c-restricted T-cell response. To test this, we cultured HIV-1-infected Jurkat cells with CD1c-restricted T cells and measured IFN-γ production by the CD1c-restricted T cells. Compared to uninfected controls, the HIV-1 infected cells showed a decreased capacity to activate CD1c-restricted T cells (Figure 6A). Thus, the decrease in CD1c expression during HIV-1 infection appears to affect the ability of CD1c-restricted T cells to recognize infected cells. This decrease in the CD1c-restricted T-cell response was even more pronounced in the presence of the cholesterol inhibitor simvastatin, where CD1c expression was further decreased.

The HIV-1-mediated decrease in CD1c expression and the subsequent decrease in IFN-γ secretion by CD1c-restricted T cells suggested that HIV-1 downregulates CD1c in an effort to evade the host immune response. To test this, we cultured HIV-1-infected cells with CD1c-restricted T cells for 24 hours and measured viral protein p24 production. In the presence of CD1c-restricted T cells, less viral p24 was produced than by infected cells alone (Figure 6B). Thus, while the HIV-1-mediated decrease in CD1c expression appears to be an immune evasion mechanism used by the virus to decrease the CD1c-restricted T-cell reactivity, these attempts appear to be ineffective for suppressing the anti-viral response mediated by the CD1c-restricted T cells.

For the previous studies, we used Jurkat cells and CD1c-restricted T cells that were stimulated for several weeks prior to use. Both cell lines are not ideal as a physiological model. Thus, we assessed CD1c and CD1d expression and endogenous lipid reactivity in primary cells. First, peripheral blood mononuclear cells (PBMCs) isolated from healthy donors were stimulated with 0.5 μg/ml phytohaemagglutinin (PHA) for several days and infected with HIV-1 IIIb, as described in the methods section. The levels of CD1c and CD1d expression were measured 10 days later by FACS. We found that, similar to Jurkat cells, there was a decrease in CD1c and CD1d expression on the HIV-infected cells compared to the uninfected controls. However, unlike the Jurkat cells, this decrease was not significant (Figure 7).

We next sought to determine whether endogenous lipid-reactive, CD1c-restricted T cells are present in the peripheral blood. When we stimulated naïve T cells with PC-loaded HeLa-CD1c cells, we found that 1.7% of these naïve cells produced high levels of intracellular IFN-γ (Figure 8). This suggests that endogenous lipid-reactive T cells do circulate at low levels in the periphery and therefore may contribute to anti-HIV-1 immunity during natural infection.

Anti-CD1c antibody was purchased from Ancell (Bayport, MN) and Biolegend (San Diego, CA). Monoclonal antibodies against CD1d and HLA-A, B, and C were purchased from BD Biosciences (San Diego, CA). The anti-human IFN-γ antibody was purchased from Abcam (Cambridge, MA). The human Jurkat cell line was obtained from ATCC (Manassas, VA). Cells were cultured in complete RPMI (CRPMI) medium consisting of RPMI 1640 containing 10% fetal calf serum (FCS), penicillin (50 U/ml), streptomycin (50 μg/ml), 2 mM L-glutamine, 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES, and 1.0 mM sodium pyruvate. The cells were incubated in 5% CO₂ at 37°C. HIV-1 IIIB was obtained from the National Institutes of Health AIDS Research and Reference Reagent Program. A WT plasmid encoding an entire HIV-1 sequence and a PCR3.1 plasmid encoding only HIV-1 Vpu under the CMV promoter were provided by Dr. Paul Bieniasz (Aaron Diamond AIDS Research Center and Rockefeller University). HIV-1 NL4-3 delta Nef (ΔNef), HIV-1 delta Vpu (ΔVpu) and WT viruses were kindly provided by Dr. Derya Unutmaz (New York University School of Medicine), Dr. Paul Bieniasz and Dr. Hiroshi Mori (Aaron Diamond AIDS Research Center), respectively.

PC was purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). Mevalonate and simvastatin were purchased from Sigma Aldrich (St. Louis, MO).

As for the transfection, Jurkat cells were resuspended to 2 x 10⁵ cells/ml in CRPMI medium and placed in a 12-well plate prior to the transfection in a final volume of 1 ml per well. Transfection was performed using plasmid encoding for GFP and the full-length HIV genome (WT) or PCR3.1 vector encoding only Vpu of HIV-1. One μg of DNA was diluted in 200 μl of Optimem serum free media and incubated at room temperature for 15 minutes. After the incubation, 4 μl of Lipofectamine (Invitrogen) was added to the diluted DNA and incubated for 25 more minutes. The DNA-Lipofectamine complex was added to the cells dropwise and incubated for 24 hours at 37°C with 5% CO₂. After the incubation, cells were harvested and checked for the expression of CD1c and CD1d by a flow cytometric analysis. As for the infection, 2 x 10⁶ Jurkat cells in 10 ml CRPMI were incubated with virus (50TCID₅₀) for 2 hours in 10-ml cell culture flasks with gentle mixing every 15–30 minutes. The cells were then incubated for the appropriate number of days at 37°C with 5% CO₂.

After isolating total RNA from Jurkat cells infected with corresponding HIV-1 viruses using RNAzol reagent (Invitrogen), the RNA was used for RT-PCR reactions. Ten-fold serial dilutions of plasmid containing the fragment of the HIV-1 genome, ranging from 1 to 10⁷ HIV-1 copies per tube, were used to construct the HIV-1 RNA calibration curve. Ten-fold serial dilutions of plasmid containing the fragment of the human GAPDH gene, ranging from 1 to 10⁷ copies per tube, were used to construct the GAPDH calibration curve. The number of RNA copies used in the reaction was then normalized based on the number of GAPDH copies. The quantitative PCR performed with primers and a probe specific for HIV-1 gag were; Forward primer; 5^′ATC AAG CAG CCA TGC AAA TGT T3^′(578–599); Reverse primer; 5^′CTG AAG GGT ACT AGT AGT TCC TGC TAT ATC3^′(722–752), and Probe; 5^′FAM-ACC ATC AAT GAG GAA GCT GCA GAA TGG GA-TAMRA3^′(607–636). GAPDH quantitative PCR was performed with Forward primer; 5^′GAA GAT GGT GAT GGG ATT TC3^′, Reverse primer; 5^′GAA GGT GAA GGT CGG AGT C3^′, and Probe; 5^′VIC-CAA GCT TCC CGT TCT CAG CC-TAMRA3^′.

Jurkat cells were harvested and stained for CD1c/CD1d expression for 45 minutes on ice with APC-Cy7-labeled anti-human CD1c antibody (Biolegend) and with PE-labeled anti-human CD1d antibody (BD Biosciences), according to the manufacturer’s instructions. The cells were then washed twice in FACS buffer containing 1× phosphate buffered saline, 2% fetal bovine serum, and fixed in 1% paraformaldehyde. The Mean Fluorescence Intensity (MFI) of both CD1c and CD1d on the GFP-positive Jurkat cells was evaluated on the transfected cells and compared to that on the untransfected controls. Samples were acquired and analyzed using DIVA software on a LSR II (BD, San Jose, CA, USA).

To generate the CD1c-restricted T cells, CD14⁺ cells were isolated from PBMCs, and immature dendritic cells (iDCs) were generated from the CD14⁺ cells after a 3-day incubation in the presence of 300 U/ml granulocyte-macrophage colony-stimulating factor and 100 U/ml interleukin (IL)-4. The iDCs were treated with mitomycin C (50 μg/ml) for 1 hour followed by four washes with culture media. The iDCs were preloaded with PC for 1 hour followed by the addition of the isolated T cells. Six days later, 10 U/ml IL-2 was added, and 30 U/ml IL-2 was added 3 days later or as needed. The cells were cultured for a total of 15 days, re-stimulated for an additional 7–10 days, tested for functionality, and used in experiments.

HeLa cells were transfected with expression vectors encoding CD1c and CD1d glycoproteins (kindly provided by Dr. Shiratsuchi at Aaron Diamond AIDS Research Center) using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. These HeLa-CD1c cells were then loaded with PC (50 μg/ml) and used to stimulate the T cells overnight. The next day, supernatants were collected, and IFN-γ levels were assessed by enzyme-linked immunosorbent assay (ELISA). For the experiments, CD1-restricted T cells (5 × 10⁵) were co-cultured with 5 × 10⁵ Jurkat cells, which were either pre-loaded for 1 hour with the indicated amounts of lipid, unloaded, or infected with HIV. The cells were incubated with or without treatment with an inducer or inhibitor of cholesterol synthesis, as described. After culture for 24 hours, supernatants were collected, and IFN-γ secretion was assessed by ELISA. Where applicable, supernatants were also assessed for p24 antigen production by ELISA according to the manufacturer’s instructions (Beckman Coulter, Brea, CA).

Two million uninfected or HIV-1-infected cells were lysed with 0.1% Triton X solution and used to assess cholesterol production using the Amplex Red Cholesterol Assay Kit (Molecular Probes, Eugene, OR) according to the manufacturer’s instructions.

PBMCs were isolated from healthy donors and stimulated with 0.5 μg/ml PHA for 3 days. Two million of these cells were then infected with HIV-1 IIIb, HIV-1 NL4-3, HIV-1 NL4-3 ΔNef, or HIV-1 ΔVpu, as described above, and cultured for 2 or 10 days. The level of CD1c/CD1d expression was then assessed on days 2 or 10 post-infection, as described above.

T cells were isolated from PBMCs from healthy donors and stimulated with HeLa-CD1c cells that were preloaded with PC (100 μg/ml). After 1 hour, Brefeldin was added, and the cells were cultured for an additional 16 hours. The next day the cells were gated for the CD8 positive population and stained for intracellular IFN-γ, as assessed by a flow cytometric analysis.

The Student’s t-test was used to compare differences among the different experimental conditions tested.

* p < 0.05, ^# p < 0.05.

Total cell cholesterol was measured in Jurkat cells untreated or treated with 200 μM mevalonate or infected with HIV IIIB using the Amplex Red Cholesterol Kit, as described. * Indicates a significant difference between the untreated and the mevalonate treated cells. ^# Indicates a significant difference between untreated cells and HIV infected cells. Data represent results from one of three independent experiments.