The efficiency of Vpx-mediated SAMHD1 antagonism does not correlate with the potency of viral control in HIV-2-infected individuals

Abstract

Background: The presence of a vpx gene distinguishes HIV-2 from HIV-1, the main causative agent of AIDS. Vpx degrades the restriction factor SAMHD1 to boost HIV-2 infection of macrophages and dendritic cells and it has been suggested that the activation of antiviral innate immune responses after Vpx-dependent infection of myeloid cells may explain why most HIV-2-infected individuals efficiently control viral replication and become long-term survivors. However, the role of Vpx-mediated SAMHD1 antagonism in the virological and clinical outcome of HIV-2 infection remained to be investigated.

Results: Here, we analyzed the anti-SAMHD1 activity of vpx alleles derived from seven viremic and four long-term aviremic HIV-2-infected individuals. We found that effective Vpx-mediated SAMHD1 degradation and enhancement of myeloid cell infection was preserved in most HIV-2-infected individuals including all seven that failed to control the virus and developed AIDS. The only exception were vpx alleles from an aviremic individual that predicted a M68K change in a highly conserved nuclear localization signal which disrupted the ability of Vpx to counteract SAMHD1. We also found that HIV-2 is less effective than HIV-1 in inducing innate immune activation in dendritic cells.

Conclusions: Effective immune control of viral replication in HIV-2-infected individuals is not associated with increased Vpx-mediated degradation of SAMHD1.

Figure 1

Evolutionary relationships among HIV-2 and SIV Vpx amino acid sequences. HIV-2 Vpx sequences newly analyzed in the present study are highlighted in green (ECs) and red (NCs). Numbers on branches are percentage estimated posterior probabilities.

Figure 2

Alignment of HIV-2 and SIV Vpx sequences. The HIV-2 ROD Vpx sequence is shown on top for comparison. The three putative α-helical regions, the nuclear localization signal (NLS), residue Q76 involved in DCAF1 binding, and the C-terminal proline-rich region are indicated. Dots indicate amino acid identity.

Figure 3

Most HIV-2 Vpx proteins from controllers and non-controllers of virus replication degrade SAMHD1. (A) SAMHD1, ß-actin, eGFP control and Vpx-AU-1 levels in HeLa cells stable expressing Flag-SAMHD1 with pCGCG constructs expressing the indicated HIV-2 Vpx proteins, the control HIV-2 ROD or SIVmac239 Vpx alleles or only eGFP. Protein expression levels were determined by western blot two days post-transfection. (B) Average levels of SAMHD1 in the presence of the indicated HIV-2 Vpx proteins. The SAMHD1 signals were obtained by western blot analysis as described in panel A, quantified by Licor Odyssey software and normalized to the ß-actin signals. The graph shows mean values and standard deviation (SD) of normalized SAMHD1 expression levels from four independent experiments. HIV-2 vpx genes were grouped based on the viral loads of the patients and are color coded green (ECs) or red (NCs). (C) The levels of SAMHD1 expression were determined in stable transfected HeLa cells in the presence of vpx alleles from HIV-2-infected ECs and NCs and are shown relative to those measured in the absence of Vpx (100%).

Figure 4

HIV-2 Vpx-mediated enhancement of virus infection of macrophages. (A) Macrophages were transduced with the VSV-G pseudotyped wild-type HIV-1 NL4-3 and SIVmac239 IRES-eGFP virions (panels 1 and 2) or a VSV-G pseudotyped vpx-defective SIVmac239 IRES-eGFP construct produced in the presence of pCGCG vectors expressing eGFP alone (panel 3) or together with the indicated Vpx proteins (panels 4–10). Virus stocks were produced by transient transfection of 293T cells. High infection rates were associated with increased rates of apoptosis and thus reduced numbers of cells analyzed in some experiments. (B) Average percentages and SDs of virally infected GFP+ cell levels detected in macrophages derived from four different donors quantified by flow cytometric analysis at four days post-transduction. Virus infectivity was normalized to infection of TZM-bl indicator cells. (C) Enhancement of macrophages infection by vpx alleles derived from HIV-2-infected individuals grouped based on their viral loads and the infecting viruses. See legend to figure 3 for further detail.

Figure 5

Effect of naturally occurring mutations in the NLS of HIV-2 Vpx proteins on SAMHD1 degradation and macrophage infection. (A) HeLa cells stably expressing FLAG tagged SAMHD1 were mock transfected or transfected with constructs expressing the indicated AU-1 tagged Vpx proteins for 48 h before whole-cell extraction and analysis by western blot using the indicated antibodies. (B) Quantitation of Vpx-mediated degradation of SAMHD1. Shown are the average levels and SEM of SAMHD1 expression in Hela cells in the presence of the indicated Vpx alleles. Results were derived from four independent experiments. (C) MDM isolated from four donors were transduced by VSV-G-pseudotyped SIVmac239 eGFP reporter viruses produced in the presence of the indicated Vpx proteins and GFP⁺ MDM were quantified four days later.

Figure 6

Effect of HIV-1 and HIV-2 on DC activation. (A) CD86 surface expression in DCs after infection with HIV-1 and HIV-2 IRES-eGFP constructs alone or in combination with VSV-G-pseudotyped SIVmac239 particles. The levels of CD86 expression by virally infected (eGFP+) cells were measured at 3 days post-infection. Panels A and B show mean values (±SEM) derived from four experiments. HIV-1 alone did not infect DCs at detectable levels. (B) Levels of IFN-γ in the supernatant of the infected DC cultures.