TRIM5α Resistance Escape Mutations in the Capsid Are Transferable between Simian Immunodeficiency Virus Strains

ABSTRACT

TRIM5α polymorphism limits and complicates the use of simian immunodeficiency virus (SIV) for evaluation of human immunodeficiency virus (HIV) vaccine strategies in rhesus macaques. We previously reported that the TRIM5α-sensitive SIV from sooty mangabeys (SIVsm) clone SIVsmE543-3 acquired amino acid substitutions in the capsid that overcame TRIM5α restriction when it was passaged in rhesus macaques expressing restrictive TRIM5α alleles. Here we generated TRIM5α-resistant clones of the related SIVsmE660 strain without animal passage by introducing the same amino acid capsid substitutions. We evaluated one of the variants in rhesus macaques expressing permissive and restrictive TRIM5α alleles. The SIVsmE660 variant infected and replicated in macaques with restrictive TRIM5α genotypes as efficiently as in macaques with permissive TRIM5α genotypes. These results demonstrated that mutations in the SIV capsid can confer SIV resistance to TRIM5α restriction without animal passage, suggesting an applicable method to generate more diverse SIV strains for HIV vaccine studies.

IMPORTANCE Many strains of SIV from sooty mangabey monkeys are susceptible to resistance by common rhesus macaque TRIM5α alleles and result in reduced virus acquisition and replication in macaques that express these restrictive alleles. We previously observed that spontaneous variations in the capsid gene were associated with improved replication in macaques, and the introduction of two amino acid changes in the capsid transfers this improved replication to the parent clone. In the present study, we introduced these mutations into a related but distinct strain of SIV that is commonly used for challenge studies for vaccine trials. These mutations also improved the replication of this strain in macaques with the restrictive TRIM5α genotype and thus will eliminate the confounding effects of TRIM5α in vaccine studies.

INTRODUCTION

Simian immunodeficiency virus (SIV)-infected rhesus macaques are widely used as an animal model to evaluate the efficacy of vaccine strategies against human immunodeficiency virus (HIV)/AIDS (1–3). Macaques infected with SIV develop disease progression remarkably similar to that with human AIDS, including acute robust and chronic persistent viral replication, progressive loss of CD4 T cells, immunodeficiency, and opportunistic infections (4). Similarly to HIV infections in humans, the clinical outcomes for SIV-infected rhesus macaques are variable. For SIV-infected rhesus macaques, this variability is at least partially due to host genetic differences, including major histocompatibility complex class I (MHC-I) and TRIM5α protein polymorphisms (5–12). TRIM5α was first identified as a restriction protein responsible for the inhibition of HIV-1 replication in macaque cell lines (13, 14). It is found in all mammals and acts as a cross-species restriction factor that inhibits retroviral infection (13–20). Recently, we and several other groups reported that TRIM5α polymorphisms common among Indian rhesus macaques affected virus replication and clinical outcomes following SIV infection. Based on an insertion/deletion polymorphism at amino acids 339 to 341 in the B30.1/SPRY domain of TRIM5α and associated differences in inhibitory activity against SIV strains, three functional types of TRIM5 alleles can be identified, TRIM5^Q, TRIM^TFP, and TRIM^CypA (21, 22). Of the three alleles, TRIM5^Q is permissive and TRIM^TFP and TRIM^CypA are restrictive for SIVsm replication. In SIVsmE543-3-infected rhesus macaque cohorts, viral loads in animals with restrictive TRIM5^TFP/TFP and TRIM5^TFP/CypA genotypes were 2 to 3 logs lower than those in macaques with at least one permissive TRIM5^Q allele (7–9). A modest association of viral loads and TRIM5 genotypes (excluding TRIM5^CypA-expressing macaques) was observed in a cohort of rhesus macaques infected with SIVmac251, but the magnitude was considerably lower than what we observed in SIVsmE543-3-infected cohorts (11). This TRIM5 effect on SIVmac251 has since been questioned by conflicting results from other vaccine studies using this virus (23). The presence of restrictive TRIM5 genotypes also decreased the mucosal transmission of uncloned SIVsmE660 (E660) but had no effect on SIVmac239 in repeated low-dose inoculations in rhesus macaques (24, 25). These results indicated that rhesus TRIM5 polymorphisms may confound vaccine study results, especially when uncloned SIVsmE660 is used as the challenge strain. The current approach to avoiding confounding TRIM5 effects has been to exclude macaques with restrictive TRIM5 genotypes by genetic screening or balancing groups for TRIM5 genotype. However, this greatly increases the cost and logistics of these studies since more than 50% of rhesus macaques have restrictive TRIM5 genotypes (7).

In the present study, we generated TRIM5-resistant SIVsmE660 variants that can overcome TRIM5 restriction. We previously reported that the closely related infectious molecular clone SIVsmE543-3 escaped from TRIM5 restriction when passaged in rhesus macaques expressing two restrictive TRIM5α alleles. Two single amino acid substitutions (P37S and R98S) and amino acid substitutions 87 to 91 in the CypA binding loop (GPLPA) in the capsid region were identified as being associated with escape from restriction by the TRIM5^TFP and TRIM5^Cyp alleles, respectively (8). We introduced these mutations into SIVsmE660 clones and evaluated one clone in rhesus macaques with different TRIM5 genotypes.

MATERIALS AND METHODS

Animal care.

This study was carried out in strict accordance with recommendations described in the Guide for the Care and Use of Laboratory Animals (26) of the National Institutes of Health, the Office of Animal Welfare, and the U.S. Department of Agriculture. Colony-bred rhesus macaques of Indian origin were obtained from the Morgan Island, SC, rhesus monkey breeding colony. All animal work was approved by the NIAID Division of Intramural Research Animal Care and Use Committees (IACUC), Bethesda, MD (animal study protocol ASP-LMM-6). The animal facility is accredited by the American Association for Accreditation of Laboratory Animal Care. All procedures were carried out under ketamine anesthesia by trained personnel under the supervision of veterinary staff, and all efforts were made to ameliorate welfare and to minimize animal suffering in accordance with recommendations of the Weatherall report on the use of nonhuman primates (27). Animals were housed in adjoining individual primate cages, allowing social interactions, under controlled conditions of humidity, temperature, and light (12-h-light/12-h-dark cycles). Food and water were available ad libitum. Animals were monitored twice daily (pre- and postchallenge) and fed commercial monkey chow, treats, and fruit twice daily by trained personnel. Early endpoint criteria, as specified by IACUC-approved score parameters, were used to determine when animals should be humanely euthanized.

Viruses.

Infectious SIVsmE660 clone FL6 (SIVsmE660-FL6), SIVsmE660-FL10, and SIVsmE660-FL14 were isolated and constructed as previously described (28). Env from the tier 2 SIVsmE660 clone H807-16wk-6, with moderate sensitivity to neutralizing antibodies (NAbs), was amplified from plasma collected from SIVsmE660-FL14-infected rhesus macaque H807 at 16 weeks postinfection (wpi) as previously described (28). An infectious tier 2 SIVsmE660 clone was constructed by replacing SIVsmE660-FL14 Env regions with H807-16wk-6 Env by restriction digestion with BsmI and BglII. Amino acid substitutions P37S and R98S were introduced into the capsid of tier 2 SIVsmE660 to create a TRIM5α-resistant mutant by using QuikChange II site-directed mutagenesis kits (Agilent) according to the manufacturer's instructions. This virus is referred to as SIVsmE660-SS (E660-SS). The following primer sets were used to introduce mutations: forward primer 5′-GGC AGA GGT AGT GTC AGG ATT TCA GGC-3′ and reverse primer 5′-GCC TGA AAT CCT GAC ACT ACC TCT GCC-3′ for the P37S mutation and forward primer 5′-GCA ACT TAG AGA GCC ATC AGG ATC AGA CAT TGC AG-3′ and reverse primer 5′-CTG CAA TGT CTG ATC CTG ATG GCT CTC TAA GTT GC-3′ for the R98S mutation. The presence of the substitutions was confirmed by sequencing of the Gag region of the mutant by using an Applied Biosystems 3130XL genetic analyzer as previously described (8).

Single-cycle infectivity assay.

SIV-based retroviral vector carrying the enhanced jellyfish green fluorescent protein inserted into the nef region (V1EGFP-SIV) and Crandell-Rees feline kidney (CRFK) cell lines (ATCC) stably expressing common rhesus TRIM5α alleles, including TRIM5^TFP, TRIM5^Q, and TRIM5^CypA alleles, were described previously (7). The gag-pol-vif genes of SIVsmE660 clones and mutants were subcloned into V1EGFP-SIV by cutting with restriction enzymes NarI and BstBI. Single-cycle SIVs were produced in HEK293T cells by cotransfection of V1EGFP-SIV and vesicular stomatitis virus G protein expressing vector pVSV-G, and titers were determined and infectivity was measured with CRFK cell lines (ATCC) stably expressing TRIM5 alleles as previously described (7).

Preparation of virus stocks.

Virus stocks were made by transfection of 293T cells. 293T cells were maintained in complete Gibco GlutaMAX Dulbecco's modified Eagle medium (DMEM) and transfected with 10 μg of the plasmid by using the FuGENE 6 transfection reagent (Roche Diagnostics, Indianapolis, IN) as previously described (28). Virus stocks were collected from the supernatant of transfected cells after 48 h and filtered with a 0.22-μm filter. The 50% tissue culture infectivity dose (TCID₅₀) of virus stocks was tested on TZM-bl cells.

Viruses were expanded on rhesus peripheral blood mononuclear cells (PBMCs) collected from a permissive TRIM5^Q/Q macaque, RhDCCW, before inoculation into rhesus macaques. PBMCs were activated by phytohemagglutinin (PHA) stimulation and infected with viruses at a multiplicity of infection (MOI) of 0.001 as previously described (28). Virus stocks were collected at day 6 and filtered, followed by titration on TZM-bl cells and quantitation by real-time reverse transcription-PCR (RT-PCR).

Animal inoculation.

Twenty-four simian T-lymphotropic virus (STLV)-, simian retrovirus (SRV)-, and SIV-seronegative rhesus macaques were divided into four groups according to their TRIM5 genotypes. The TRIM5 genotypes of rhesus macaques were determined as previously described (7), and MHC-I genotypes were determined by the Rhesus Macaque MHC Typing Core facility at the University of Miami Miller School of Medicine. Each macaque was inoculated intrarectally (i.r.) with a 1:50 dilution of SIVsmE660 or SIVsmE660-SS PBMC virus stocks (1,000 TCID₅₀). After inoculation, viral RNA levels in plasma were determined by quantitative RT-PCR. Four weeks later, any of the macaques that remained uninfected were inoculated intrarectally on a weekly schedule with same amount of virus until viral RNAs became detectable in plasma. After infection, blood and plasma were collected, and plasma viral RNA levels were determined.

Statistical analyses.

All statistical analyses and graphic analyses were performed by using GraphPad Prism5 (GraphPad Prism Software, La Jolla, CA). The rates of acquisition of infection were plotted as Kaplan-Meier curves based on the number of inoculations required to achieve infection and compared by a log rank test. Two-way analysis of variance (ANOVA) was used for comparison of virus replication between groups during the acute phase of infection. The nonparametric Kruskal-Wallis test was used for comparison of peak viral loads. The cumulative survival rates of infected macaques were plotted as Kaplan-Meier curves and were compared by a log rank test.

RESULTS

We previously isolated three infectious SIV clones, SIVsmE660-FL6, -FL10, and -FL14, from the SIVsmE660 stock that has been commonly used for vaccine challenge studies (29–31). Two clones, SIVsmE660-FL6 and -FL14, were evaluated in rhesus macaques with a heterozygous (permissive and restrictive) TRIM5^TFP/Q genotype. Both of the clones were highly sensitive to neutralization by antibody, analogous to the tier 1 designation used for HIV-1. However, both clones replicated efficiently and induced AIDS in rhesus macaques (28). To evaluate the sensitivity of these clones to TRIM5 restriction, we measured their infectivity on cell lines stably expressing different rhesus TRIM5 alleles by single-cycle infectivity assays. As shown in Fig. 1, virus clone SIVsmE660-FL6 replicated in the cell lines expressing the TRIM5^Q allele but was restricted by the TRIM5^TFP and TRIM5^CypA alleles. The other two clones, SIVsmE660-FL10 and SIVsmE660-FL14, were able to replicate in cell lines expressing the TRIM5^Q and TRIM5^CypA alleles but were still restricted by the TRIM5^TFP allele. The capsid sequences of the three SIVsmE660 clones were aligned and compared to SIVsmE543-3, which was sensitive to both the TRIM5^TFP and TRIM5^Cyp alleles in our previous studies. Only a few amino acid substitutions in the capsid were observed in SIVsmE660 clones compared to SIVsmE543-3 (Fig. 2). We previously observed that amino acid substitutions (P37S and R98S) in the capsid conferred resistance to TRIM5^TFP restriction when introduced into SIVsmE543-3. The clones SIVsmE660-FL10 and SIVsmE660-FL14 had amino acid substitutions in the CypA binding loop GPLPA87-91 region, consistent with their resistance to TRIM5^Cyp restriction.

FIG 1.

TRIM5α polymorphism restricts replication of SIVsmE660 clones. Single-cycle infectivity of three SIVsmE660 clones was measured in triplicate on a panel of cell lines stably expressing rhesus TRIM5^TFP, TRIM5^Q, and TRIM5^CypA alleles. Infectivity was measured as the percentage of green fluorescent protein-positive (GFP⁺) cells. Black bars show the negative vector without TRIM5 expression as a control.

FIG 2.

Capsid sequences of SIVsmE660 clones. The amino acid sequences of the Gag capsids of SIVsmE660 clones were aligned to the capsid sequence of the TRIM5α-sensitive clone SIVSmE543-3. Identical amino acids are shown as dots. Amino acid residues 37 and 98 and the region of the cyclophilin A binding loop, which are associated with TRIM5α sensitivity, are highlighted in yellow.

We chose the virus clone SIVsmE660-FL14, which was pathogenic in macaques with TRIM5^TFP/Q genotypes and had already acquired resistance to TRIM5^Cyp restriction. We generated a clone that was fully resistant to the effects of rhesus TRIM5 by introducing two amino acid substitutions, P37S and R98S, into the capsid of SIVsmE660-FL14 by site-directed mutagenesis, generating clone SIVsmE660-FL14SS (FL14SS). The infectivity of SIVsmE660-FL14SS was evaluated by utilizing cell lines stably expressing different rhesus TRIM5 alleles with a single-cycle infectivity assay. As shown in Fig. 3, the introduction of these two amino acid substitutions into the FL14 capsid greatly improved its replication (>10-fold) in cells expressing the TRIM5^TFP allele. Compared to the original clone, SIVsmE660-FL14, the mutant clone FL14SS was resistant to not only the TRIM5^Q and TRIM5^CypA alleles but also the TRIM5^TFP allele. The replication of the FL14SS mutant in cells expressing different TRIM5 alleles was similar to that of the rhesus macaque-adapted clone SIVmac239, making it a potential improvement in comparison to the TRIM5-sensitive SIVsmE660 clones as a challenge strain.

FIG 3.

Introduction of amino acid substitutions into the SIVsmE660 clone confers resistance to TRIM5α restriction. Amino acid substitutions P37S and R98S were introduced into the capsid of the virus clone SIVsmE660-FL14 to generate SIVsmE660-FL14SS. (A to C) The single-cycle infectivity of SIVsmE660-FL14SS (B) was measured in triplicate on a panel of cell lines stably expressing rhesus TRIM5^TFP, TRIM5^Q, and TRIM5^CypA alleles and compared with those of wild-type SIVsmE660-FL14 (A) and SIVmac239 (C). (D) The infectivity of HIV-1 NL4-3 was measured as a negative control.

An ideal SIV model in rhesus macaques not only would be unaffected by all TRIM genotypes but also would mirror the NAb sensitivity of primary HIV-1 isolates, which are generally characterized as tier 2, or moderately resistant to NAb. In contrast, SIVsmE660-FL14 was extremely sensitive (tier 1) to neutralizing antibodies. Therefore, prior to initiating in vivo studies, we chose to introduce an envelope variant with tier 2 sensitivity into our TRIM5-resistant E660-FL14SS variant. This was possible since we had previously amplified several env variants with diverse neutralization sensitivities from SIVsmE660-FL6- and -FL14-infected rhesus macaques and constructed infectious chimeric SIVsmE660 variants by replacing the env regions of SIVsmE660-FL14 (28). Recently, we evaluated three clones representative of the full range of neutralization sensitivity, from extremely sensitive (tier 1) to moderately sensitive (tier 2) to highly resistant (tier 3), in TRIM5^TFP/Q rhesus macaques. These clones replicated efficiently and induced AIDS-like symptoms in rhesus macaques (unpublished data). To generate a TRIM5-resistant, tier 2 SIVsmE660 clone, we introduced the P37S and R98S capsid substitutions into the tier 2 SIVsmE660 variant described in Materials and Methods. This clone contained the backbone of FL14 and a tier 2 envelope cloned from the 16-week plasma of macaque H807 that had been inoculated with FL14 (H807-16wk-6) (28). This variant, referred to as E660-SS, was evaluated in macaques with different TRIM5 genotypes.

Twenty-four rhesus macaques were divided into four groups of six macaques each, according to the TRIM5 genotypes TRIM5^Q/Q, TRIM5^TFP/TFP, TRIM5^TFP/CypA, and TRIM5^TFP/Q (Table 1). Macaques of the TRIM5^CypA/CypA genotype were not evaluated since this is a quite rare genotype in our population of rhesus macaques. In addition, MHC-I genotypes known to have an effect on SIVmac239 viremia (A01 and B08) were equally distributed among the groups. The exception for this distribution was in the TRIM5^Q/Q group, which included three macaques with the restrictive MHC A01 genotype due to the scarcity of this genotype. Macaques in the TRIM5^Q/Q, TRIM5^TFP/TFP, and TRIM5^TFP/CypA groups were inoculated with the tier 2 neutralization-sensitive SIVsmE660 clone carrying the P37S and R98S amino acid substitutions (designated E660-SS). Macaques in the TRIM5^TFP/Q group were inoculated with the wild-type (WT) tier 2 SIVsmE660 clone (designated E660) as controls predicted to be relatively resistant to TRIM5 restriction. Before inoculation, the two viruses were expanded in PBMCs from a macaque with a permissive TRIM5^Q/Q genotype. gag sequences of the virus stocks were evaluated after expansion in PBMCs, and no additional mutations were found in either of the two stocks (data not shown).

TABLE 1.

TRIM5 and MHC genotypes and sexes of study animals

Inoculated virus	TRIM5α genotype	Animal	MHC genotype(s)a	Gender of animalb
E660-SS	Q/Q	Rh862	A01, B01	M
		Rh863	A01	M
		Rh864	A08	M
		Rh865	A08, B01	M
		Rh866	Negative	M
		Rh867	A01	F
	TFP/TFP	Rh868	Negative	M
		Rh869	A02	M
		Rh870	A11	M
		Rh871	A01, B01	M
		Rh872	A02	M
		Rh873	A08	M
	TFP/CypA	Rh874	B01	M
		Rh875	Negative	M
		Rh876	A02	M
		Rh877	B01	M
		Rh878	A01, A02	M
		Rh879	A08, B01	M
E660	TFP/Q	Rh850	Negative	M
		Rh851	A02, B01	M
		Rh852	B01	M
		Rh853	A02	M
		Rh854	Negative	F
		Rh855	A02	F

^aNine rhesus MHC class I alleles, including Mamu-A*001, -A*002, -A*008, -A*011, -B*001, -B*003, -B*004, -B*008, and -B*017, were tested and listed for each macaque. Negative, none of the 9 alleles observed. MHC-I genotypes known to have an effect on SIVmac239 viremia are indicated in bold.

^bM, male; F, female.

Since our previous studies of in vivo TRIM resistance of SIVsmE543-3 demonstrated a significant delay in acquisition in an intrarectal challenge model (8), we chose to evaluate virus acquisition in the present study. Each macaque was inoculated i.r. with 1,000 TCID₅₀, and infection was evaluated by monitoring plasma viral RNA loads. Four weeks later, any of the macaques that remained uninfected were inoculated intrarectally on a weekly schedule with the same amount of virus until they became infected. The rate of acquisition of infection in each group is shown by Kaplan-Meier curves in Fig. 4A. In the control TRIM5^TFP/Q group inoculated with wild-type E660, five of six macaques were infected after the first inoculation, and the last macaque was infected after the third inoculation. The rates of acquisition of infection between the latter group and the three groups inoculated with the E660-SS variant were compared, and no significant difference was observed between any of the groups (P = 0.319, as determined by a log rank test). In the “permissive” TRIM5^Q/Q group inoculated with E660-SS, four of six macaques were infected after the first inoculation, and the other two macaques in this group were infected after the third and fifth inoculations, respectively. In the “restrictive” TRIM5^TFP/TFP group, two macaques were infected after the first inoculation of E660-SS, two were infected after the third inoculation, and the last two were infected after the fourth inoculation. Finally, in the restrictive TRIM5^TFP/CypA group, four macaques were infected after the first inoculation of E660-SS, and the last two were infected after the third inoculation. Therefore, rates of acquisition were similar among the four cohorts, consistent with the lack of an apparent effect of TRIM genotype on this parameter.

FIG 4.

Acquisition and replication of SIVsmE660 and SIVsmE660-SS clones during the acute phase of infection in rhesus macaques. (A) Each macaque was inoculated i.r. with 1,000 TCID₅₀ of virus, and infection was monitored by measuring plasma viral RNA loads. Four weeks later, any of the macaques that remained uninfected were inoculated intrarectally on a weekly schedule with same amount of virus until they became infected. The acquisition of infection in each group is shown as a percentage of uninfected macaques after each inoculation, and groups were compared by a log rank test (P = 0.3193). (B) Median plasma viral RNA copy numbers during the acute phase of infection (up to 8 weeks postinfection) in each group are shown and were compared by two-way ANOVA (P > 0.05). (C) Peak plasma viral RNA copy numbers in each group are shown and were compared by a Kruskal-Wallis one-way ANOVA (P = 0.3523). (D) Plasma viral RNA copy numbers at set points are shown and were compared by a Kruskal-Wallis one-way ANOVA (P = 0.3413).

We also compared plasma viral loads of the four groups after infection. The median plasma viral loads during the acute phase of infection (0 to 8 wpi) for each group are shown in Fig. 4B. The mutant virus E660-SS replicated as efficiently as the WT E660 virus, indicated by the similar kinetics of plasma viremia during the acute phase of infection. Furthermore, significant differences were not observed in comparisons of viral replication in macaques with different TRIM5 genotypes (P > 0.05, as determined by two-way ANOVA). Compared to the macaques in the TRIM5^TFP/Q group infected with the WT E660 virus, median peak viral loads in macaques infected with E660-SS in any of the three genotype groups, TRIM5^Q/Q, TRIM5^TFP/TFP, and TRIM5^TFP/CypA, were not significantly different (P = 0.3523, as determined by a Kruskal-Wallis one-way ANOVA) (Fig. 4C), although these values were slightly lower (differences of 7.8-fold, 1.4-fold, and 4.6-fold, respectively). The plasma viral loads at the set point in each group were also compared, and no significant difference was observed (P = 0.3413, as determined by a Kruskal-Wallis one-way ANOVA) (Fig. 4D).

Despite the similarity in acute-phase viremia, infected macaques in each group showed variable plasma viral loads and disease progression during the chronic phase of infection. Sequential plasma viral loads for each macaque in each group up to 80 weeks after infection are shown in Fig. 5A to D. Two macaques (Rh852 and Rh854) infected with the WT E660 virus in the TRIM5^TFP/Q group suppressed viremia below 10³ copies/ml. In macaques infected with E660-SS, four macaques (Rh863, Rh865, Rh866, and Rh867) in theTRIM5^Q/Q group, two macaques (Rh869 and Rh871) in the TRIM5^TFP/TFP group, and one macaque (Rh874) in the TRIM5^TFP/CypA group suppressed viremia below 10³ copies/ml during the chronic phase of infection. Three macaques infected with the WT E660 virus in the TRIM5^TFP/Q group progressed to AIDS. For macaques infected with E660-SS, one macaque in the TRIM5^Q/Q group, three macaques in the TRIM5^TFP/TFP group, and five macaques in the TRIM5^TFP/CypA group progressed to AIDS. The survival rates of these three groups were compared by a log rank test and are shown as a Kaplan-Meier plot in Fig. 5E. Paradoxically, the TRIM5^Q/Q macaques infected with E660-SS, which we had anticipated would be the most susceptible group, had prolonged survival rates compared to those of the other three groups, but this difference did not reach statistical significance (P = 0.134, as determined by a log rank test). Overall, the lack of differences in virus acquisition, peak viremia, and disease progression, regardless of TRIM5 genotype, suggested that the introduction of the P37S and R98S amino acid substitutions conferred resistance to TRIM5 restriction to SIVsmE660 clones in vivo.

FIG 5.

Viral loads and survival curves of SIVsmE660- and SIVsmE660-SS-infected rhesus macaques. Viral loads were quantified and are shown as RNA copy numbers in rhesus macaques. (A) SIVsmE660-SS-infected macaques with the TRIM5^Q/Q genotype. (B) SIVsmE660-SS-infected macaques with the TRIM5^TFP/TFP genotype. (C) SIVsmE660-SS-infected macaques with the TRIM5^TFP/CypA genotype. (D) Wild-type SIVsmE660-infected macaques with the TRIM5^TFP/Q genotype. (E) Cumulative survival of macaques in each group are shown as Kaplan-Meier curves and were compared by a log rank test (P = 0.134). Animals expressing an MHC allele known to be restrictive for SIVmac239 are identified by an asterisk.

DISCUSSION

SIV/simian-human immunodeficiency virus (SHIV)-infected rhesus macaques have been extensively used as an animal model for studying HIV pathogenesis and evaluating HIV vaccine strategies. Compared to the high diversity of circulating HIV-1 primary strains, the most widely used SIV challenge strains are all derived from limited lineages, including SIVmac239, SIVmac251, SIVsmE660, and their derivatives (32, 33). The use of new SIV isolates in the rhesus macaque model is limited at least partially by the restriction of cross-species restriction factors, including TRIM5, APOBEC3G, and BST2, etc. (8, 34, 35). We and others previously reported that the polymorphism of rhesus TRIM5 affected SIVsm replication in rhesus macaques (7, 10, 24, 25). We also found that SIVsm can overcome TRIM5 restriction by gaining amino acid substitutions in its capsid region through passage in macaques expressing restrictive TRIM5 alleles (8, 9). Here we demonstrated that adaptation to restrictive TRIM5 alleles can be transferred to other SIVsm clones without passage in rhesus macaques. The TRIM5-sensitive SIVsmE660 clones acquired resistance to TRIM5 restriction after the introduction of the same capsid amino acid substitutions P37S and R98S, which were found in the escape variants during passage of SIVsmE543-3 in rhesus macaques with restrictive TFP alleles. The mutant SIVsmE660 clone replicated efficiently in cell lines expressing restrictive TRIM5 alleles. When evaluated in rhesus macaques, the mutant infected and replicated in macaques with restrictive TRIM5 genotypes as efficiently as in macaques with permissive TRIM5 genotypes. Indeed, replication was more consistent in macaques with restrictive TRIM genotypes than in TRIM5^Q/Q macaques for reasons that are presently unclear. We speculate that this was due to the inclusion of more macaques with restrictive MHC genotypes in this group due to the limited availability of macaques with this genotype. These results indicate that assisted mutations in the SIV capsid can confer resistance to TRIM5 restriction without animal passage, suggesting an applicable method to generate more diverse SIV strains for SIV-rhesus macaque models. It is important to note that the amino acid sequences of the SIVsmE660 and SIVsmE543-3 capsids are very similar. Whether the same amino acid substitutions, P37S and R98S, can confer TRIM5 resistance to other more divergent SIVsm lineages will need to be evaluated at least in vitro before application in rhesus macaques.

We previously reported that TRIM5 restriction has a long-term effect on disease progression and survival rates during SIV infection. Macaques with restrictive TRIM5^TFP/TFP genotypes infected with a TRIM5-resistant mutant survived longer than did those infected with wild-type SIVsmE543-3 (9). Similar results were also observed in a SIVmac251-infected rhesus macaque cohort by another group (11). In the present study, we also observed diverse survival rates among infected macaques. However, the difference in survival rates among animals was not segregated by study group and thus did not appear to be attributable to TRIM5 restriction in this study. In most reports, TRIM5 restriction affects virus acquisition and replication during the acute phase of infection. In contrast, such differences were not observed in this study when macaques were infected with the TRIM5-resistant clone. Furthermore, diversity in survival rates of infected macaques was observed in all groups. The intergroup difference, which would reflect the impact of TRIM5 restriction, was not statistically different. The majority of differences in viremia in this study were observed during the chronic phase of infection, which could be affected not only by restriction factors but also by the adaptive immune response, including neutralizing-antibody and cytotoxic T lymphocyte (CTL) responses. A similar variability was observed for TRIM5^TFP/Q animals inoculated with our wild-type tier 2 SIVsmE660 clone, where the presence of the permissive TRIM^Q allele generally allowed unrestricted replication. The degree of variability in survival in this study was higher than what we observed in our previous study using the E543-SS capsid mutant. In that study, persistent viremia was observed in all but one of the six animals inoculated with E543-SS. One potential reason for the higher degree of variability among animals in this study may be the relative neutralization sensitivity of the viruses. SIVsmE543-3, used in the previous study, is highly resistant to neutralization (tier 3), whereas the E660 clone in the present study was only moderately resistant to neutralizing antibody (tier 2). We hypothesize that the neutralizing-antibody resistance of the inoculum may play a subsequent role in disease progression. Furthermore, other restriction factors, including APOBEC3G and BST2, may also impact disease progression in SIV-infected macaques. Recently, Krupp et al. described a rhesus APOBEC3G polymorphism that impacts SIV infectivity and selects a Gly¹⁷-to-Glu¹⁷ adaptation of the Vif protein during SIVsmE543 passage in rhesus macaques, suggesting that APOBEC3G-mediated restriction is also a barrier for the generation of SIV challenge strains in macaque models (34).

In summary, the E660-SS variant described in the present study would be an improved alternative to uncloned TRIM-sensitive SIVsmE660, as it would not be confounded by TRIM5 genotype differences. Artificial quasispecies mixtures of tier 1, 2, and 3 variants that are not susceptible to TRIM genotypes could also be used to create molecularly characterized virus stocks for challenge studies in vaccine trials using SIV immunogens.