Cytokines Elevated in HIV Elite Controllers Reduce HIV Replication In Vitro and Modulate HIV Restriction Factor Expression

Abstract

A subset of HIV-infected individuals termed elite controllers (ECs) maintain CD4⁺ T cell counts and control viral replication in the absence of antiretroviral therapy (ART). Systemic cytokine responses may differentiate ECs from subjects with uncontrolled viral replication or from those who require ART to suppress viral replication. We measured 87 cytokines in four groups of women: 73 ECs, 42 with pharmacologically suppressed viremia (ART), 42 with uncontrolled viral replication (noncontrollers [NCs]), and 48 HIV-uninfected (NEG) subjects. Four cytokines were elevated in ECs but not NCs or ART subjects: CCL14, CCL21, CCL27, and XCL1. In addition, median stromal cell-derived factor-1 (SDF-1) levels were 43% higher in ECs than in NCs. The combination of the five cytokines suppressed R5 and X4 virus replication in resting CD4⁺ T cells, and individually SDF-1β, CCL14, and CCL27 suppressed R5 virus replication, while SDF-1β, CCL21, and CCL14 suppressed X4 virus replication. Functional studies revealed that the combination of the five cytokines upregulated CD69 and CCR5 and downregulated CXCR4 and CCR7 on CD4⁺ T cells. The CD69 and CXCR4 effects were driven by SDF-1, while CCL21 downregulated CCR7. The combination of the EC-associated cytokines induced expression of the anti-HIV host restriction factors IFITM1 and IFITM2 and suppressed expression of RNase L and SAMHD1. These results identify a set of cytokines that are elevated in ECs and define their effects on cellular activation, HIV coreceptor expression, and innate restriction factor expression. This cytokine pattern may be a signature characteristic of HIV-1 elite control, potentially important for HIV therapeutic and curative strategies.IMPORTANCE Approximately 1% of people infected with HIV control virus replication without taking antiviral medications. These subjects, termed elite controllers (ECs), are known to have stronger immune responses targeting HIV than the typical HIV-infected subject, but the exact mechanisms of how their immune responses control infection are not known. In this study, we identified five soluble immune signaling molecules (cytokines) in the blood that were higher in ECs than in subjects with typical chronic HIV infection. We demonstrated that these cytokines can activate CD4⁺ T cells, the target cells for HIV infection. Furthermore, these five EC-associated cytokines could change expression levels of intrinsic resistance factors, or molecules inside the target cell that fight HIV infection. This study is significant in that it identified cytokines elevated in subjects with a good immune response against HIV and defined potential mechanisms as to how these cytokines could induce resistance to the virus in target cells.

Keywords: HIV; chemokine receptors; cytokines; elite control; restriction factor.

FIG 1

Correlation of cytokines with elite control of HIV. Multiplex assays were used to measure 87 cytokines in samples from patients in the WIHS and SCOPE cohorts. Sample groups included 74 EC, 42 ART, 42 viremic, and 49 HIV-negative patients. Bars represent median values. Asterisks denote statistical significance as determined by a Mann-Whitney test. Corrections for multiple comparisons were made using the method of Benjamini-Hochberg; an FDR of <0.1 and P value of <0.05 were considered significant (*, P < 0.05; **, P < 0.01).

FIG 2

In vitro suppression of HIV by individual and combined cytokines. (A and B) Pooled CD8-depleted PBMCs from three donors (mixed-lymphocyte reaction stimulated) were infected with 81-A or NL4-3 virus at an MOI of 10⁻² and cocultured with the individual indicated cytokines. (C and D) Infections were cocultured with SDF-1α/β, CCL21, XCL1, CCL14, and CCL27 (Combo), IL-2, or medium alone. Supernatants were measured for p24 by ELISA on day 6. Data from four experiments were combined for analysis. *, P < 0.05; **, P < 0.01; ***, P < 0.001 (2-by-3 ANOVA of cytokine results versus those of the medium control).

FIG 3

In vitro suppression of HIV in individual and pooled donor PBMCs. Infections with 81-A and NL4-3 viruses were performed as previously described in pooled (mixed-lymphocyte reaction-stimulated) or nonpooled (resting) PBMCs and cocultured with combined SDF-1α/β, CCL21, XCL1, CCL14, and CCL27 (Combo), IL-2 alone, or medium alone. Culture supernatants were measured for p24 on day 6. Data were combined for analysis from two experiments.

FIG 4

Cytokine stimulation of PBMCs increases CD69 and decreases CCR7 and CXCR4 expression. PBMCs from three donors were separately stimulated with individual cytokines or combined SDF-1α/β, CCL21, XCL1, CCL14, and CCL27 (Combo) for the indicated times. Following incubation, surface expression of CXCR4, CCR5, CCR7, and CD69, as indicated, was measured by flow cytometry. Means and standard errors of the means for three donors are shown. *, P < 0.05; **, P < 0.01 (2-by-3 ANOVA of cytokine results versus those of the unstimulated condition). MFI, mean fluorescence intensity.

FIG 5

HIV-infected cultures cocultured with cytokines increases CD69 and decreases CXCR4 and CCR7 expression. Resting CD8-depleted PBMCs from three donors were infected with HIV NL4-3 and cocultured with the indicated cytokines. Combo indicates coculture with SDF-1α/β, CCL21, XCL1, CCL14, and CCL27. (A to D) Following infection for 1 day cells were measured by flow cytometry for expression of CD69, CCR5, CXCR4, and CCR7. (E to J) Cells from 6-day infections were measured by flow cytometry for expression of CXCR4, CCR5, and CCR7, as indicated. Individual and combined cytokine experiments were performed separately. Means and standard errors of the means for three donors are shown. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 (2-by-3 ANOVA of cytokine results versus those of the medium control).

FIG 6

Gene expression profile of HIV restriction factors. CD4⁺ T cells from three donors were negatively selected and stimulated overnight with the indicated cytokines. Combo indicates SDF-1α/β, CCL21, XCL1, CCL14, and CCL27. (A) Thirty-one HIV innate restriction factors were measured on a custom qPCR array and normalized to values for the housekeeping genes. (B) Elevated expression of IFITM1 and IFITM2 mRNA (top) and decreased expression of RNase L and SAMHD1 mRNA (bottom). Fold induction was determined using the C_T method. Means and standard errors of the means of three donors are shown, and conditions were compared using repeated-measures ANOVA. *, P < 0.05; **, P < 0. 01. Unstim, unstimulated.

FIG 7

Elevated mRNA expression of IFITM1 and IFITM2. CD4⁺ T cells from 10 donors were negatively selected and stimulated for 24 or 72 h with IFN-α, combined cytokines (SDF-1α/β, CCL21, XCL1, CCL14, and CCL27), or medium (Unstim, unstimulated) alone. Lysates were then measured by qPCR for IFITM1 and IFITM2, as indicated. Means and standard errors of the means are shown, with comparisons between groups made using repeated-measures ANOVA. *, P < 0.05; ****, P < 0.0001.

FIG 8

Cytokine-induced protein expression of IFITM1 and IFITM2. CD4⁺ T cells were negatively selected and stimulated with IFN-α, combined cytokines (SDF-1α/β, CCL21, XCL1, CCL14, and CCL27), or medium (Unstim) alone for 72 h. (A) Total cell lysates were transferred to PVDF membranes and probed for expression of IFITM1 (left) or IFITM2 (right). GAPDH was used as a loading control. Images from two representative experiments are shown. (B) Plots summarize the expression ratio as a percentage of IFITM1 or IFITM2 compared to GAPDH expression after each treatment. Means and standard errors of the means from 10 donors are shown; comparisons were made using repeated-measures ANOVA of log-transformed data. *, P < 0.05; **, P < 0.01.