Cell surface CCR5 density determines the postentry efficiency of R5 HIV-1 infection

Abstract

We have recently reported that the mean number of CCR5 coreceptors at the surface of CD4(+) T cells (CCR5 density) correlates with viral load and disease progression in HIV-1-infected persons. Here, we definitively establish that CCR5 density determines the level of virus production and identify the stages of HIV-1 replicative cycle modulated by this effect. We show, by transducing the CCR5 gene into CCR5(+) cells, that CCR5 overexpression resulted in an HIV-1 overinfectability. We sorted HOS-CD4(+)-CCR5(+) cells into two subpopulations, HOS(high) and HOS(low), the former expressing seven times more cell surface CCR5 molecules than the latter. Virus production was 30-80 times higher in HOS(high) cells than in HOS(low) cells after a single round of infection. In contrast, only twice as many viral particles entered the cytosol of HOS(high) cells as compared with the cytosol of HOS(low) cells. Yet, seven times as many early, and 24 times as many late, reverse transcription products were found in HOS(high) cells as compared with HOS(low) cells. Moreover, a 24- to 30-fold difference in the number of copies of integrated HIV-1 DNA was observed. No difference in HIV-1 LTR activation between the two cell lines was evident. Finally, we show that the higher virus production observed in HOS(high) cells is inhibited by pertussis toxin, a Galphai protein inhibitor. Thus, CCR5 density mainly modulates postentry steps of the virus life cycle, particularly the reverse transcription. These data explain why CCR5 density influences HIV-1 disease progression and underline the therapeutic interest of lowering CCR5 expression.

Fig 1.

CCR5 (A and B) and CD4 (C and D) expression on HOS^low (A and C) and HOS^high (B and D) cells. Cells were exposed to an anti-CCR5 or anti-CD4 antibody (full line) or to a negative control antibody (dotted line), labeled with an FITC-conjugated anti-mouse Ig probe, and analyzed by flow cytometry.

Fig 2.

(A) Effect of CCR5 overexpression on HIV-1 R5 production. HOS^high cells (▵), HOS^low cells either nontransduced (♦) or transduced with V_CCR5/EGFP (•), V_Δ32/EGFP (○), or V_EGFP (▪), were exposed to the HIV-1 R5 strain AD8 and cultured. Infection was monitored by measuring p24 concentration in the cell supernatant. (B) Structure of the HIV-1 vectors used for gene transfer.

Fig 3.

Efficiency of one round of HIV-1 R5 infection correlates with CCR5 density on the target cell. HOS^low (▪) and HOS^high (•) cells were exposed to env-defective HIV-1 R5 virus harboring the luciferase marker gene and pseudotyped with an R5 envelope at various quantities (A) or pseudotyped with the G protein of the VSV at the same quantities (B) or at 150-fold lower quantities to result in equal viral copy number entering the cells (C). Luciferase activity was measured in cell lysates 72 h later.

Fig 4.

Efficiency of HIV-1 R5 entry into HOS^low (▪) and HOS^high (•) cells. HOS^low and HOS^high cells were exposed to various amounts of R5 HIV-1 strains for 3 h. Cytosolic and endosomic fractions were then prepared, and p24 concentration was measured in each fraction.

Fig 5.

Efficiency of HIV-1 R5 early reverse transcription in HOS^low (▪) and HOS^high (•) cells. HOS^low and HOS^high cells were exposed to various amounts of HIV-1 R5 for 3 h, and the quantity of early transcripts was determined by quantitative PCR.

Fig 6.

Effect of exposure of HOS^low and HOS^high cells to HIV-1 R5 on LTR expression. (A) HOS^low and HOS^high cells were transfected with a luciferase gene driven by a minimal promoter containing three NF-κB boxes, and exposed or not (empty bars) to 100 ng/ml tumor necrosis factor α (filled bars), or to 100 ng of p24 of the Ada-M strain (hashed bars). (B) HOS^low and HOS^high cells were transfected with an HIV-1 LTR-driven luciferase gene and exposed or not (empty bars) to 100 ng of p24 of the Ada-M strain (hatched bars); as a positive control, cells were cotransfected with a tat-expressing plasmid (filled bars). Luciferase activity was measured, and the induction of gene expression was calculated.

Fig 7.

PTX reduces the efficiency of a single round of R5-enveloped, but not VSV-pseudotyped, HIV-1 virions. HOS^high cells treated with PTX (A) or herbimycin A (B) were exposed to R5-enveloped (▪) or VSV-pseudotyped (•) env-defective HIV-1 R5 virions harboring the luciferase marker gene, and luciferase activity was measured in cell lysates 72 h later. Comparable results were obtained in other experiments (n = 4).