HLA-C increases HIV-1 infectivity and is associated with gp120

Abstract

Background: A recently identified genetic polymorphism located in the 5' region of the HLA-C gene is associated with individual variations in HIV-1 viral load and with differences in HLA-C expression levels. HLA-C has the potential to restrict HIV-1 by presenting epitopes to cytotoxic T cells but it is also a potent inhibitor of NK cells. In addition, HLA-C molecules incorporated within the HIV-1 envelope have been shown to bind to the envelope glycoprotein gp120 and enhance viral infectivity. We investigated this last property in cell fusion assays where the expression of HLA-C was silenced by small interfering RNA sequences. Syncytia formation was analyzed by co-cultivating cell lines expressing HIV-1 gp120/gp41 from different laboratory and primary isolates with target cells expressing different HIV-1 co-receptors. Virus infectivity was analyzed using pseudoviruses. Molecular complexes generated during cell fusion (fusion complexes) were purified and analyzed for their HLA-C content.

Results: HLA-C positive cells co-expressing HIV-1 gp120/gp41 fused more rapidly and produced larger syncytia than HLA-C negative cells. Transient transfection of gp120/gp41 from different primary isolates in HLA-C positive cells resulted in a significant cell fusion increase. Fusion efficiency was reduced in HLA-C silenced cells compared to non-silenced cells when co-cultivated with different target cell lines expressing HIV-1 co-receptors. Similarly, pseudoviruses produced from HLA-C silenced cells were significantly less infectious. HLA-C was co-purified with gp120 from cells before and after fusion and was associated with the fusion complex.

Conclusion: Virionic HLA-C molecules associate to Env and increase the infectivity of both R5 and X4 viruses. Genetic polymorphisms associated to variations in HLA-C expression levels may therefore influence the individual viral set point not only by means of a regulation of the virus-specific immune response but also via a direct effect on the virus replicative capacity. These findings have implications for the understanding of the HIV-1 entry mechanism and of the role of Env conformational modifications induced by virion-associated host proteins.

Figure 1

Fusion efficiency of CHO cells expressing HLA-C and HIV-1 Env. Panel A: Syncytia formation after co-cultivation of effector CHO cells expressing gp120/gp41 and HLA-C, or CHO cells expressing only gp120/gp41, with target CHO-CD4-CCR5 cells. The number and the extent of syncytia is significantly higher (p < 0.05) when effector cells express HLA-C. Panel B: ELISA analysis of Env expression. CHO, negative control; CHO-gp120, cells stably expressing the Env gene of the R5 tropic HIV-1 isolate 91US005; CHO-gp120-HLA-C: CHO-gp120 cells stably expressing HLA-Cw4; gp120: positive control, consisting of a mixture of five different gp120s. The higher fusion efficiency of CHO-gp120-HLA-C cells is not due to an increased level of Env expression, since they express 27% less gp120 than CHO-gp120 cells.

Figure 2

Transient transfections of CHO cells expressing human HLA-C with different env sequences. CHO (-, grey bars) and CHO-HLA-C (+, black bars) cells transiently transfected with plasmids encoding Tat, Rev and Env from different primary and laboratory HIV-1 isolates and co-cultivated for 6 hours with TZM-bl target cells. After Tat driven transactivation of firefly luciferase expression, fusion efficiency was quantified and expressed as counts per second (CPS). Each value represents the average of four replicates. The gp120/gp41 of primary isolates 93MW965 (R5), 91US005 (R5) and 92UG024 (X4) are HLA-C sensitive (p < 0.05) while isolates J500 (X4) and NDK (X4) are less sensitive to the presence of HLA-C (p not significant).

Figure 3

Specific silencing of HLA-C in human cell lines. Panel A: off-target effect analysis by RT-PCR in HLA-C silenced (+) and non-silenced (-) HeLa cells expressing HIV-1 gp120/gp41 (ADA). PCR was performed with primers specific for HLA (A, B, C), gp120, β₂-microglobulin and GAPDH. M: molecular weight marker. No off-target effect due to HLA-C mRNA silencing is affecting the mRNA levels of the other MHC class I genes, as well as β₂-microglobulin, HIV-1 gp120 or the housekeeping control gene GAPDH. Panel B: western-blot analysis of HLA-C protein expression. After 72 hours from siRNAs transfection, HLA-C is undetectable both in HeLa-ADA and in 293T cells.

Figure 4

Cell fusion of HLA-C silenced HeLa-Env cells with HeLa-P4.2 target cells. Analysis of syncytia formation by co-cultivating HLA-C silenced (+) and non-silenced (-) HeLa-LAI and HeLa-NDK cells with target HeLa-P4.2 cells, expressing CD4 and CXCR4. The number of syncytia formed is lower (p < 0.01) using HLA-C silenced HeLa-LAI cells. Fusion efficiency of HeLa-NDK cells is not significantly affected by HLA-C silencing.

Figure 5

Comparison of the fusion efficiency of HLA-C silenced HeLa-Env cells with 3T3.T4.CCR5 and 3T3.T4.CXCR4 cells. HLA-C silenced (+, grey bars) and non-silenced (-, black bars) HeLa cells expressing gp120/gp41 of different HIV-1 isolates (ADA, LAI, NDK) co-cultivated with NIH 3T3.T4.CXCR4 and NIH 3T3.T4.CCR5 cells. Fusion efficiency of X4 tropic gp120 LAI is significantly lower (p < 0.01) in HLA-C silenced cells when fusing with CXCR4 target cells. Similarly, fusion efficiency of the R5 tropic gp120 ADA is lower (p < 0.01) in HLA-C silenced cells when fusing with CCR5 target cells. The fusion of ADA gp120 in HLA-C silenced cells with cells expressing CXCR4 is significantly (p < 0.01) less efficient, while that of LAI gp120 with cells expressing CCR5 is similar, irrespective of HLA-C silencing. The NDK gp120 is HLA-C insensitive, when using either the CXCR4 or the CCR5 co-receptor.

Figure 6

Transduction efficiency of pseudoviruses produced in HLA-C silenced cells. Panel A: luciferase reporter gene assay analysis after transduction with pseudoviruses expressing subtype B HIV-1 env (6535.3 and pRHPA4259.7) or subtype D HIV-1 env (NDK), produced in HLA-C silenced (dashed line, open circles) and non silenced (continuous line, close squares) 293T cells. Each point (expressed as counts per second, CPS) represents average and standard deviation of four replicates. HLA-C sensitive pseudoviruses 6535.3 and pRHPA4259.7 show a significant lower infectivity (p < 0.0001) when produced on HLA-C silenced cells. The NDK pseudovirus as well as a virus pseudotyped with the VSV-G envelope protein, do not show significant differences in infectivity when produced in HLA-C silenced or non silenced 293T cells. Panel B: analysis of the relation between pseudovirus infectious dose and HLA-C sensitivity. 1×, pseudovirus infectious titer giving a luciferase signal (expressed as counts per second, CPS) of 1000 at 16 hours post infection. When the HLA-C insensitive NDK pseudovirus was analyzed at lower infectious titers (0.3× and 0.1×), its infectivity was significantly increased by HLA-C. When the HLA-C sensitive pseudovirus pRHPA4259.7 was analyzed at higher infectious doses (3.3×, 10×), it remained sensitive to HLA-C presence.

Figure 7

Co-purification of fusion complexes containing HLA-C molecules. Panel A: dot-blot analysis of purified fusion complexes for the presence of HLA-C. Lanes a, b, c and d: cell lysates before purification. Lanes e, f, g and h: cell lysates purified on Galanthus nivalis (GN) lectin columns. Panel B: western blot analysis to detect the presence of HLA-C in purified fusion complexes. Cells were treated with DTSSP, which fixes only proteins present on the cell membrane, and lysates purified on GN lectin columns. PC: positive control (HeLa cells expressing HLA-C); the arrow indicates HLA-C.