Role of common human TRIM5α variants in HIV-1 disease progression
© Goldschmidt et al. 2006
Received: 10 April 2006
Accepted: 22 August 2006
Published: 22 August 2006
Skip to main content
© Goldschmidt et al. 2006
Received: 10 April 2006
Accepted: 22 August 2006
Published: 22 August 2006
The retroviral restriction factor tripartite motif protein (TRIM)5α, is characterized by marked amino acid diversity among primates, including specific clusters of residues under positive selection. The identification of multiple non-synonymous changes in humans suggests that TRIM5α variants might be relevant to retroviral pathogenesis. Previous studies have shown that such variants are unlikely to modify susceptibility to HIV-1 infection, or the course of early infection. However, the longterm effect of carrying Trim5α variants on disease progression in individuals infected with HIV-1 has not previously been investigated.
In a cohort of 979 untreated individuals infected with HIV-1 with median follow up 3.2 years and 9,828 CD4 T cell measurements, we analysed common amino acid variations: H43Y, V112F, R136Q, G249D, and H419Y. The rate of CD4 T cell decline before treatment was used as the phenotype. In addition, we extended previous work on the in vitro susceptibility of purified donor CD4 T cells (n = 125) to HIV-1 infection, and on the susceptibility of HeLa cells that were stably transduced with the different TRIM5 variants. Haplotypes were analysed according to the most parsimonious evolutionary structure, where two main human TRIM5α groups can be defined according to the residue at amino acid 136. Humans present both Q136 and R136 at similar frequency, and additional TRIM5α amino acid variants are almost exclusively derived from R136-carrying haplotypes.
We observed modest differences in disease progression for evolutionary branches carrying R136-derived haplotypes, and with the non-synonymous polymorphisms G249D and H419Y. In vitro analysis of susceptibility of donor CD4 T cells, and of the various transduced HeLa cell lines supported the absence of significant differential restriction of HIV-1 infection by the various huTRIM5α alleles.
Common human variants of TRIM5α have no effect or modest effect on HIV-1 disease progression. These variants occur at sites conserved throughout evolution, and are remote from clusters of positive selection in the primate lineage. The evolutionary value of the substitutions remains unclear.
The tripartite motif (TRIM) family is a well conserved family of proteins characterized by a structure comprising a RING domain, one or two B-boxes and a predicted coiled-coil region . In addition, most TRIM proteins have additional C-terminal domains. Members of the TRIM protein family are involved in various cellular processes, including cell proliferation, differentiation, development, oncogenesis and apoptosis (for recent review [2, 3]). Some TRIM proteins display antiviral properties, targeting retroviruses in particular .
TRIM5α is a retroviral restriction factor targeting the early steps of cellular infection . TRIM5α restricts retroviral infection by specifically recognizing the capsid and promotes its premature disassembly . Human TRIM5α (huTRIM5α) has limited efficacy against HIV-1, while some primate TRIM5α orthologues can potently restrict this particular lentivirus (for review see [2, 3]). Considerable inter-species sequence diversity characterizes TRIM5α and might underlie differences in the pattern and breadth of restriction of multiple lentiviruses. Evolutionary analysis reveals that up to 2% of codons of TRIM5α are predicted to be under positive selection with high confidence [6, 7]. Residues under positive selection cluster in the C-terminal B30.2 domain. A first cluster resides between amino acids 322 to 340 in the variable region v1 [7, 8], a region previously described as a "patch" of positive selection . Replacement of the v1 region, or of specific amino acids within v1, modifies the restriction pattern of TRIM5α [9, 10]. The second cluster, localized between amino acids 381 to 389 , corresponds to the previously described variable region v2 of the B30.2 domain . Substitution of the human v2 region by a v2 from Rhesus monkeys exhibits no inhibitory activity against HIV-1 [9, 10]. However, v2 variants are thought to result in species-specific lentiviral restriction patterns . An additional region of considerable variation among Sooty mangabeys and Rhesus monkeys has been mapped to the coiled-coil motif .
Recently, two studies have addressed the potential role of huTRIM5α variants in modulating susceptibility to HIV-1 [13, 14]. Sawyer et al. identified several non-synonymous SNPs in huTRIM5α, but only one of these (H43Y) was found to have a functional consequence . H43Y lies in the RING domain of TRIM5α and may negatively affect its putative E3 ubiquitin ligase activity. Although huTRIM5α weakly restricts HIV-1, H43Y might further reduce viral restriction to a level similar to that of cells expressing no exogenous huTRIM5α . To assess whether the impaired retroviral restriction seen with exogenously expressed huTRIM5α H43Y resulted in altered susceptibility in human cells, Sawyer et al. tested B-lymphocytes from four individuals: one homozygous for H43, two homozygous for 43Y, and one heterozygous at this residue. Challenge with HIV-1 failed to demonstrate a significant effect of the H43Y change, although 43Y homozygous cells could be infected with N-MLV about 100-fold more efficiently than cells with the common H43 allele . In a second study, Speelmon et al. assessed the association of various non-synonymous variants with susceptibility to HIV-1 among 110 HIV-1 infected subjects and 96 exposed seronegative persons . This study identified possible associations between specific haplotypes and alleles and susceptibility to infection and viral setpoint after acute infection.
In our study, we analysed data from a large cohort of subjects infected with HIV-1 to explore whether different huTRIM5α variants are associated with long-term disease evolution. The study is complemented by analysis of CD4 cell susceptibility to HIV-1, and the in vitro functionality of selected huTRIM5α variants. We conclude from our study that there is no impact for some and negligible to modest impact for other common human variants of TRIM5α on disease progression.
Speelmon et al.  reported on the permissiveness of purified CD4 T cells from 77 seronegative donors. Analysis included assessment of H43Y, R136Q, H419Y, and a series of less common variants. We performed similar experiments by infecting purified CD4 T cells from 125 Caucasian healthy blood donors with replicating HIV-1. Alleles tested included H43Y, R136Q, H419Y, and the common variant G249D (not tested in the above study). There was no significant association of specific variants or haplotypes with in vitro p24 production after 7 days (Additional file 3). None of the additional SNPs investigated in vitro were associated with differences in cell permissiveness (Additional file 1).
The reports by Sawyer et al and Speelmon et al. [13, 14] suggested that some of the alleles could indeed have an impact on HIV-1 susceptibility in vivo. We extended their analyses by investigating the association of the various huTRIM5α variants with long-term disease progression in a large cohort of individuals infected with HIV-1. The clinical phenotype was defined as the patient-specific rate of CD4 T cell decline, a recognized marker of disease progression . Analysis excluded any CD4 T cell values after initiation of treatment. The median follow up time was 3.2 years, during which 979 cohort participants, not receiving antiretroviral treatment, contributed 9,828 CD4 T cell determinations to the analysis (median 7 CD4 T cell determinations per participant). We first tested for associations of individual non-synonymous polymorphisms with differences in the natural history of disease progression. H43Y and R136Q had no effect on disease progression. Participants who had one or two copies of G249D or H419Y had slower progression although confidence intervals were wide due to small numbers. Compared to non-carriers who had a square root transformed CD4 gradient of -2.02, participants who were carriers of G249D and H419Y had gradients of -1.74 (95% CI -1.39 to -2.09, p = 0.11) and -1.64 (95% CI -1.33 to -1.95, p = 0.02) respectively.
Whilst our study was being completed, Sawyer et al.  and Speelmon et al  reported on an additional common variant V112F (allelic frequency of 7%). We re-genotyped the cohort and identified the presence of 112F in R136-carrying haplotypes. We confirm the absence of significant effect of this particular amino acid on disease progression [CD4 T cell gradient: -2.21 (95% CI -1.89 to -2.534) compared to mean reference slope -2.01]. In addition, Speelmon et al. included in their analysis the 5'UTR -2C>G SNP . The -2C represents the ancestral allele and is present in high linkage disequilibrium with Q136. Speelmon et al. identified a rare haplotype, where individuals carried -2G in the context of Q136, possibly associated with susceptibility to HIV infection (OR 5.49, p = 0.02) . We also identified this rare haplotype (frequency of 1.4%), but did not observe any association with disease progression [CD4 T cell gradient: -2.04 (95% CI -1.20 to -2.88)].
We have extended previously reported findings on huTRIM5α by showing that the amino acid variant H43Y which results in loss of restriction of N-MLV  is not associated with significant differences in HIV-1 disease progression in a large human cohort. The present study also underscores the modest or negligible effect of human variants G249D and H419Y and of some haplotypes on HIV-1 susceptibility and disease progression. Despite the conserved nature of these residues in primates, the evolutionary relevance of the variants in humans is uncertain. However, it is possible that the polymorphisms found in TRIM5α might have been selected in past epidemics by viruses unrelated to HIV-1.
CD4 T cells from 125 healthy Caucasian blood donors were isolated by anti-CD4 magnetic beads (Miltenyi Biotech) and cultured ex vivo in RPMI-1640 (Gibco-Invitrogen), supplemented with 20% fetal calf serum (FCS), 20 U/ml human IL-2 (Roche) and 50 μg/ml gentamicin, following stimulation with 2 μg/ml phytohaemagglutinin (PHA) during two days. CD4 T cells (106 cells) were infected with R5 clone HIV-1 NL4-3BaL env (1000 pg p24 antigen) for 2 hours at 37°C, 5% CO2, in 1 ml final volume. Cells were washed and cultured for 7 days. Virus-containing supernatant was harvested and p24 antigen production was monitored by ELISA (Abbott). Permissiveness was defined as the ability of cells to be infected and sustain replication of HIV-1 . The ex vivo viral replication for each genotype was represented by the median p24 antigen production at day 7.
Single nucleotide polymorphism (SNP) discovery used single strand conformation polymorphism and sequencing of 94 chromosomes (47 Caucasian blood donors). For this, a total of 21 PCR reactions were designed to cover exons, putative promoter regions, and intron-exon boundaries (6771 bp/subject). SNPs resulting in non-synonymous substitutions were then genotyped by using TaqMan allelic discrimination (Additional file 4).
The pLPCX oncoretroviral vector containing the human and Rhesus TRIM5α gene with an HA epitope tag was obtained from the NIH AIDS Reagent Program (donated by J. Sodroski). Variants of huTRIM5α were made by using the QuikChange protocol (Stratagene). Retroviral vectors were packaged by co-transfecting the various pLPCX constructs with the pNB-tropic MLV Gag-Pol and pVSV-G packaging plasmids . Supernatants were concentrated and used to transduce HeLa cells. Seventy two hours after transduction, cells were selected in 0.5 mg/ml puromycin. Expression of HA-tagged TRIM5α proteins were determined by Western blotting using an anti-HA antibody (Roche). Tubulin was detected with the anti-α tubulin antibody (Sigma). Single-cycle infectivity assays in HeLa cells used the VSV-pseudotyped recombinant viruses HIV.1-GFP and N-MLV-GFP at various m.o.i. Cells were analysed by fluorescence-activated cell sorter (FACS) 48 h after transduction.
Study participants (n = 979) were recruited within the genetics project of the Swiss HIV Cohort Study . The ethics committees of all participant centers approved the study. Patients gave written informed consent for genetic testing. DNA from PBMCs was used for genotyping. Their characteristics are shown in Additional file 5. The rate of decline in CD4 T cell count during the natural history of disease progression was defined as study phenotype as previously reported . The CD4 T cell trajectories were modeled using a repeated measures hierarchical approach using Mlwin software . Square root transformed CD4 T cell counts were modeled as a linear function of time since estimated date of seroconversion with random effects for both the intercept and the gradient with additional terms for sex, age, and risk group . For each genotype, the average square root CD4 decline per year was estimated in dominant and recessive models. Haplotypes were attributed using PHASE .
This study has been financed in the framework of the Swiss HIV Cohort Study, supported by the Swiss National Science Foundation. (Grant no 3345-062041), and by Swiss National Science Foundation grant no. 310000-110012/1 to A.T.
The members of the Swiss HIV Cohort Study are S. Bachmann, M. Battegay, E. Bernasconi, H. Bucher, Ph. Bürgisser, M. Egger, P. Erb, W. Fierz, M. Fischer, M. Flepp (Chairman of the Clinical and Laboratory Committee), P. Francioli (President of the SHCS, Centre Hospitalier Universitaire Vaudois, CH-1011- Lausanne), H.J. Furrer, M. Gorgievski, H. Günthard, P. Grob, B. Hirschel, L. Kaiser, C. Kind, Th. Klimkait, B. Ledergerber, U. Lauper, M. Opravil, F. Paccaud, G. Pantaleo, L. Perrin, J.-C. Piffaretti, M. Rickenbach (Head of Data Center), C. Rudin (Chairman of the Mother & Child Substudy), J. Schupbach, R. Speck, A. Telenti, A. Trkola, P. Vernazza (Chairman of the Scientific Board), R. Weber, S. Yerly.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.