The TRIM5 gene modulates penile mucosal acquisition of simian immunodeficiency virus in rhesus monkeys

Abstract

There is considerable variability in host susceptibility to human immunodeficiency virus type 1 (HIV-1) infection, but the host genetic determinants of that variability are not well understood. In addition to serving as a block for cross-species retroviral infection, TRIM5 was recently shown to play a central role in limiting primate immunodeficiency virus replication. We hypothesized that TRIM5 may also contribute to susceptibility to mucosal acquisition of simian immunodeficiency virus (SIV) in rhesus monkeys. We explored this hypothesis by establishing 3 cohorts of Indian-origin rhesus monkeys with different TRIM5 genotypes: homozygous restrictive, heterozygous permissive, and homozygous permissive. We then evaluated the effect of TRIM5 genotype on the penile transmission of SIVsmE660. We observed a significant effect of TRIM5 genotype on mucosal SIVsmE660 acquisition in that no SIV transmission occurred in monkeys with only restrictive TRIM5 alleles. In contrast, systemic SIV infections were initiated after preputial pocket exposures in monkeys that had at least one permissive TRIM5 allele. These data demonstrate that host genetic factors can play a critical role in restricting mucosal transmission of a primate immunodeficiency virus. In addition, we used our understanding of TRIM5 to establish a novel nonhuman primate penile transmission model for AIDS mucosal pathogenesis and vaccine research.

Fig. 1.

Partial amino acid sequence alignment of SIV Gag. SIVmac251 and SIVsmE660 sequences were obtained by single-genome amplification of viral challenge stocks that were expanded in human PBMCs stimulated with concanavalin A and interleukin-2, while Gag sequences of SIVmac239 and SIVsmE543 were obtained from the Los Alamos HIV Sequence Database. Residues 89, 90, and 97 (SIVmac239 numbering) in the SIV capsid that are postulated to mediate host TRIM5 restriction are highlighted.

Fig. 2.

Systemic infections following repeated penile exposures of rhesus monkeys to SIVsmE660. Rhesus monkeys (n = 6/group) were exposed to SIVsmE660 in a 0.5-ml volume via the penile mucosa for 12 weeks. The first exposure was on week 0, and the last exposure was on week 11. Monkeys were monitored for systemic infection weekly by assessing plasma SIV gag RNA levels by qRT-PCR. The lower limit of detection using qRT-PCR was 500 SIV RNA copies/ml. SIV RNA levels are shown for the first 20 weeks following exposures in TRIM5 homozygous restrictive (A), heterozygous permissive (B), and homozygous permissive (C) monkeys. Viral RNA levels are shown as log₁₀ copies of plasma viral RNA/ml of plasma at each time point. Monkeys that developed a positive plasma virus RNA level were not given further virus exposures. The numbers of monkeys that became infected in each group were statistically different by a 3 × 2 Fisher's exact test (P = 0.005).

Fig. 3.

Effect of TRIM5 genotype on number of exposures before penile infection with SIVsmE660 in rhesus monkeys. The effects of restrictive and permissive TRIM5 alleles on the number of exposures before infection are shown by Kaplan-Meier curves. There were 3 groups of monkeys (n = 6/group): the first group had only restrictive TRIM5 alleles (red), the second group had one permissive allele (blue), and the third group had only permissive alleles (black). The statistical comparison of the numbers of exposures before infection between the three groups of monkeys with different TRIM5 genotypes was made using the log rank test (P = 0.017).

Fig. 4.

Clinical immune status after SIV penile exposures. The percentages of CD4⁺ T cells (A) and CCR7⁺ CCR5⁻ CD28⁺ CD95⁺ central memory CD4⁺ T cells (B) were monitored in all 3 cohorts of rhesus monkeys. Asterisks indicate monkeys that resisted mucosal infection in the TRIM5 heterozygous and homozygous permissive groups. (A) There was a decline in the percentage of CD4⁺ T cells in monkeys in the TRIM5 heterozygous (P = 0.04) and homozygous (P = 0.02) permissive groups. The two-tailed Wilcoxon signed-rank test was used to compare the changes in percentage of CD4⁺ T cells after 20 weeks. (B) There was no significant decline in the percentage of central memory CD4⁺ T cells within all three groups. (C) Comparison of differences in percentages of CD4⁺ T cells after infection between the three groups. Bars representing the median values are shown for each group. The Kruskal-Wallis test was used to evaluate the differences in the percentages of CD4⁺ T cells between groups (P = 0.01).

Fig. 5.

Phylogenetic analyses of T/F viruses in TRIM5 heterozygous permissive monkeys. A maximum likelihood tree of env sequences at ramp-up (open squares) and peak (closed squares) viremia in TRIM5 heterozygous permissive monkeys and of SIVsmE660 inoculum env sequences (black circles) is shown. env sequences from each monkey are displayed as similarly colored symbols. Asterisks indicate monkeys that were infected with homogeneous T/F viruses. Bootstrap values are shown for nodes with at least 50% support. The long branch lengths relative to the T/F virus in some of the monkeys (e.g., AV33X and the single long branch in 9E9) were entirely due to APOBEC-mediated G-to-A hypermutation. This is commonly observed in acute HIV infection and adjusted for in the Poisson modeling of viral diversity.

Fig. 6.

Phylogenetic analyses of T/F viruses in TRIM5 homozygous permissive monkeys. A maximum likelihood tree of env sequences at ramp-up (open circles) and peak (closed circles) viremia in TRIM5 homozygous permissive monkeys and of SIVsmE660 inoculum env sequences (black circles) is shown. env sequences from each monkey are displayed as similarly colored symbols. Asterisks indicate monkeys that were infected with homogeneous T/F viruses. Bootstrap values are shown for nodes with at least 50% support. The long branch lengths observed for some of the animals, such as DBME, resulted from APOBEC-mediated hypermutation.

Fig. 7.

Partial amino acid sequence alignment of SIVsmE660 Gag sequences in acutely infected monkeys. Gag sequences were obtained by bulk sequencing of plasmas of 10 infected monkeys at ramp-up viremia (RU) and were aligned with the Gag sequence from the SIVsmE660 challenge stock. Residues 89, 90, and 97 (SIVmac239 numbering) in the SIV capsid that are postulated to mediate host TRIM5 restriction are highlighted.