Polymorphisms in Rhesus Macaque Tetherin Are Associated with Differences in Acute Viremia in Simian Immunodeficiency Virus Δnef-Infected Animals

As a consequence of ongoing evolutionary conflict with viral pathogens, tetherin has accumulated numerous species-specific differences that represent important barriers to the transmission of viruses between species. This study reveals extensive polymorphism in rhesus macaque tetherin and identifies specific alleles that are associated with lower viral loads during the first few weeks after infection with nef-deleted SIV. These observations suggest that the variable selective pressure of viral pathogens, in addition to driving the diversification of tetherin among species, also operates within certain species to maintain sequence variation in tetherin.

ABSTRACT

Tetherin (BST-2 or CD317) is an interferon-inducible transmembrane protein that inhibits virus release from infected cells. To determine the extent of sequence variation and the impact of polymorphisms in rhesus macaque tetherin on simian immunodeficiency virus (SIV) infection, tetherin alleles were sequenced from 146 rhesus macaques, including 68 animals infected with wild-type SIV_mac239 and 47 animals infected with SIV_mac239Δnef. Since Nef is the viral gene product of SIV that counteracts restriction by tetherin, these groups afford a comparison of the effects of tetherin polymorphisms on SIV strains that are, and are not, resistant to tetherin. We identified 15 alleles of rhesus macaque tetherin with dimorphic residues at 9 positions. The relationship between these alleles and plasma viral loads was compared during acute infection, prior to the onset of adaptive immunity. Acute viremia did not differ significantly among the wild-type SIV-infected animals; however, differences in acute viral loads were associated with polymorphisms in tetherin among the animals infected with SIVΔnef. In particular, polymorphisms at positions 43 and 111 (P₄₃ and H₁₁₁) were associated with lower acute-phase viral loads for SIVΔnef infection. These observations reveal extensive polymorphism in rhesus macaque tetherin, maintained perhaps as a consequence of variability in the selective pressure of diverse viral pathogens, and identify tetherin alleles that may have an inherently greater capacity to restrict SIV replication in the absence of Nef.

IMPORTANCE As a consequence of ongoing evolutionary conflict with viral pathogens, tetherin has accumulated numerous species-specific differences that represent important barriers to the transmission of viruses between species. This study reveals extensive polymorphism in rhesus macaque tetherin and identifies specific alleles that are associated with lower viral loads during the first few weeks after infection with nef-deleted SIV. These observations suggest that the variable selective pressure of viral pathogens, in addition to driving the diversification of tetherin among species, also operates within certain species to maintain sequence variation in tetherin.

INTRODUCTION

Tetherin (BST-2 or CD317) is an interferon-stimulated gene product that inhibits the detachment of enveloped viruses from infected cells. Although initially identified as the cellular factor that accounts for a defect in the release of vpu-deleted human immunodeficiency virus type 1 (HIV-1) from restrictive cells (1, 2), tetherin has since been shown to have broad antiviral activity against diverse families of enveloped viruses (3–14). The antiviral activity of tetherin reflects unique structural features of the protein, including an N-terminal transmembrane domain and a C-terminal glycosylphosphatidylinositol (GPI) anchor (15) that enable opposite ends of tetherin dimers to become incorporated into viral and cellular membranes and to thereby link nascent virions to the surface of infected cells (16–18). In addition to physically impairing virus release, secondary effects of tetherin may also contribute to antiviral activity. Virion-induced cross-linking of tetherin results in NF-κB activation and the release of proinflammatory cytokines (19–21), and the accumulation of captured virions at the plasma membrane increases the susceptibility of infected cells to antibody-dependent killing (22, 23). Hence, the impact of tetherin on virus replication in vivo may be amplified by the combined effects of innate and adaptive immunity.

Many viruses have acquired mechanisms to overcome restriction by tetherin. Among the primate lentiviruses, at least three different viral proteins have evolved to counteract tetherin. Whereas most simian immunodeficiency viruses (SIVs) use Nef to counteract the tetherin proteins of their nonhuman primate hosts, HIV-1 (group M) and HIV-2 use Vpu and Env, respectively, to counteract human tetherin due to the absence of a five-amino-acid sequence in the cytoplasmic domain of human tetherin that confers susceptibility to Nef (1, 2, 24–27). Instances of lentiviral adaptation to tetherin have also been observed in nonhuman primate models, including compensatory changes in the gp41 cytoplasmic tail of a nef-deleted strain of SIV that regained a pathogenic phenotype in rhesus macaques (28), changes in Nef that restore the ability to counteract tetherin in HIV-1-infected chimpanzees (29), and changes acquired by a simian-adapted strain of HIV-1 that enable Vpu to antagonize macaque tetherin (30). Tetherin has therefore had a significant impact on the course of lentiviral evolution in primates.

Viruses have in turn shaped the evolution of tetherin in primates. A comparison of tetherin coding sequences among various primate species revealed signatures of strong positive selection and a concentration of nonsynonymous (amino acid replacement) changes in regions of the gene that encode the cytoplasmic and transmembrane domains (31, 32), which coincide with the surfaces of the protein known to interact with lentiviral Nef and Vpu proteins (1, 2, 9, 24–27, 33). Thus, similar to the TRIM5 and APOBEC3 proteins (34–36), tetherin has accumulated species-specific differences at sites that interact with viral proteins. In addition to these interspecies differences, sequence comparisons within the same species have revealed allelic variation for certain nonhuman primates, including rhesus macaques (31). Since the rhesus macaque is an important animal model for HIV-1 pathogenesis and AIDS vaccine development (37), we investigated the extent of sequence variation in rhesus macaque tetherin and associations between polymorphisms in tetherin and viral loads in SIV-infected animals.

RESULTS

Polymorphisms in rhesus macaque tetherin.

To determine the extent of polymorphism in rhesus macaque tetherin and the potential impact of allelic differences on SIV replication, tetherin cDNA was amplified and sequenced from the cryopreserved peripheral blood mononuclear cells (PBMCs) of 146 rhesus macaques, including 68 animals infected with wild-type SIV_mac239 and 47 animals infected with SIV_mac239Δnef, from studies previously performed at the New England Primate Research Center and at the Frederick National Laboratory for Cancer Research (38–42). Fifteen tetherin alleles were identified with unique combinations of 10 single-nucleotide polymorphisms, of which nine encode amino acid differences and one corresponds to a silent change at nucleotide 129 (G129C) (Fig. 1). Four of the nine amino acid polymorphisms are located in the cytoplasmic and transmembrane domains, corresponding to regions of tetherin targeted by lentiviral Nef and Vpu proteins (1, 2, 9, 24, 25, 33). The frequency of each of the tetherin alleles among the animals in this study is shown in Table 1. In order of frequency, the four most common variants were rBST-2.1 (36%), rBST-2.5 (17.8%), rBST-2.2 (11.6%), and rBST-2.8 (9.9%) (Table 1).

FIG 1.

Amino acid sequence alignment of 15 alleles of rhesus macaque tetherin reveals dimorphic residues at nine positions throughout the protein. Shaded sequences correspond to the transmembrane domain (red), conserved cysteine residues (gray), potential N-linked glycosylation sites (magenta), and predicted site of GPI addition (green). Positions of amino acid identity are indicated by a period.

TABLE 1.

Rhesus macaque tetherin alleles

Tetherin allele	No. of animalsa	Allele frequencyb (%)	GenBank accession numberc
rBST-2.1	67	36.0	MG816214
rBST-2.2	24	11.6	HM775182
rBST-2.3	15	6.8	MG816215
rBST-2.4	1	0.3	MG816216
rBST-2.5	36	17.8	MG817217
rBST-2.6	7	2.7	MG816218
rBST-2.7	13	5.8	MG816219
rBST-2.8	20	9.9	HQ596987
rBST-2.9	9	3.8	MG816221
rBST-2.10	2	1.0	MG816222
rBST-2.11	2	1.4	MG816223
rBST-2.12	5	1.7	MG816224
rBST-2.13	1	0.3	MG816225
rBST-2.14	1	0.3	MG816226
rBST-2.15	1	0.3	MG816227

^aThe number of animals that carry at least one copy of a particular tetherin allele is indicated.

^bThe frequency of occurrence of individual rBST-2 alleles in rhesus macaques is calculated as a ratio of the number of alleles of a particular variant in 146 diploid animals.

^cThe GenBank accession numbers for the respective rhesus macaque tetherin alleles are shown.

Restriction of wild-type versus nef-deleted SIV by each allotype of rhesus macaque tetherin.

To assess differences in antiviral activity, we compared virus release for wild-type versus nef-deleted SIV_mac239 from cells expressing each allotype of rhesus macaque tetherin. 293T cells were cotransfected with full-length infectious molecular clones of SIV_mac239 or SIV_mac239Δnef together with expression constructs for each tetherin allele, and virus accumulation in the cell culture supernatant was quantified by SIV p27 antigen capture enzyme-linked immunosorbent assay (ELISA) (Fig. 2A). Differences in virus release and the expression of tetherin were also confirmed by Western blot analysis of cell lysates and supernatant from the transfected cells (Fig. 2B). Comparisons of virus release for SIV_mac239 versus SIV_mac239Δnef demonstrated that all 15 allotypes of rhesus macaque tetherin restrict virus release for nef-deleted SIV to a greater extent than that for wild-type SIV (Fig. 2A); however, one allele (rBST-2.4), which was only identified in a single animal, was less efficient than the others. This deficiency was attributable to a unique L₁₇₅Q polymorphism at the C terminus of rBST-2.4, because this residue was the only difference between rBST-2.1 and rBST-2.4 and the introduction of Q₁₇₅ into rBST-2.10 recapitulated the impaired restriction observed for rBST-2.4 (Fig. 2C and D). These results confirm that 14 of the 15 allotypes of rhesus macaque tetherin are capable of restricting SIV Δnef and identify a rare C-terminal polymorphism that significantly impairs this activity.

FIG 2.

Wild-type versus nef-deleted SIV were tested for sensitivity to restriction by variants of rhesus macaque tetherin. (A) The amount of virus released into the cell culture supernatant was measured by SIV p27 antigen-capture ELISA from 293T cells cotransfected with full-length infectious molecular clones of SIV_mac239 or SIV_mac239Δnef (100 ng) and the indicated tetherin expression constructs (20 ng). Percent maximal release was calculated relative to the amount of virus released in control transfections with an empty vector that does not express tetherin. Differences in virus release for wild-type (WT) and nef-deleted (Δnef) strains were compared using an unpaired t test (ns, nonsignificant; *, P < 0.05; **, P < 0.01). (B) The effects of each of the tetherin variants on virus release for SIV_mac239 versus SIV_mac239Δnef were confirmed by Western blot analysis. Viruses recovered from the culture supernatant and cell lysates were separated by SDS-PAGE, transferred to a PVDF membrane, probed with antibodies to tetherin, p27/p55 Gag, Hsp90 or actin followed by an HRP-conjugated goat anti-mouse secondary antibody, developed in chemiluminescent substrate, and visualized using an ImageQuant LAS-4000 image reader. (C) The effect of the L₁₇₅Q polymorphism on virus release was tested by introducing the Q₁₇₅ change into rBST-2.10. Differences in virus release for SIV_mac239 versus SIV_mac239Δnef were assessed by ELISA (C) and Western blot analysis (D) using 20 ng of the indicated tetherin expression constructs and 100 ng of proviral DNA as described above.

Variation in acute-phase viral loads for macaques infected with nef-deleted SIV is associated with allelic differences in tetherin.

To determine if polymorphisms in rhesus macaque tetherin are associated with differences in SIV replication in animals, differences in peak viral loads were compared for animals infected with wild-type SIV_mac239 versus SIV_mac239Δnef, which affords an opportunity to determine the effects of tetherin polymorphisms on viruses that are, and are not, resistant to this restriction factor. We reasoned that the impact of tetherin on SIV replication would be most apparent during acute infection (weeks 1 to 4 postinfection [p.i.]) (43, 44), since the peak of tetherin upregulation on CD4⁺ lymphocytes coincides with the peak of detectable IFN-α in plasma on day 10 postinfection (45), and adaptive immune responses have not yet obscured the effects of innate immunity. Although viral loads did not vary significantly among animals infected with SIV_mac239 (Fig. 3A), significant variation in peak viremia was observed among animals infected with SIV_mac239Δnef (P = 0.025 by Kruskal-Wallis test) (Fig. 3B). In most cases, peak viremia occurred on day 14 postinfection; however, due to differences in sampling schedules, viral load data were not available at every time point for all of the animals (see Table S1 in the supplemental material). Therefore, to be sure that our results were not biased by any particular time point, we also calculated and compared area under the curve (AUC) values for acute-phase viral loads in SIV_mac239- and SIV_mac239Δnef-infected animals. By integrating measurements over multiple time points, AUC values provide a measure of total viremia during acute infection. Similar to comparisons of peak viral loads, significant variation in AUC values as a function of tetherin was detectable for SIV_mac239Δnef-infected animals (P = 0.027 by Kruskal-Wallis test) but not for SIV_mac239-infected animals (Fig. 3C and D).

FIG 3.

Variation in viral loads for rhesus macaques infected with wild-type versus nef-deleted SIV was analyzed according to allelic differences in tetherin. Peak viral RNA loads in plasma are shown for animals infected with SIV_mac239 (A) or SIV_mac239Δnef (B), and AUC values representing acute viremia are shown for animals infected with SIV_mac239 (C) or SIV_mac239Δnef (D) that are either homozygous (black) or heterozygous (color-coded) for the indicated alleles of tetherin. Viral load data from macaques with tetherin alleles present in fewer than three animals were omitted from these analyses. Animals for which viral load data were not available at weeks 1, 2, or 4 were also omitted from AUC comparisons. Variation in viral loads as a function of allelic differences in tetherin was assessed by the Kruskal-Wallis test. The horizontal bars represent median values.

Differences in acute viremia for macaques infected with SIV_mac239Δnef are associated with specific amino acid polymorphisms in tetherin.

To assess the effects of individual polymorphisms in tetherin, differences in peak and total viremia during acute infection with SIV_mac239Δnef were compared on the basis of amino acid differences at each position. Dimorphisms at positions 43 (L₄₃P) and 111 (Q₁₁₁H), but not at positions 9 (C₉R), 14 (D₁₄G), 29 (V₂₉I), or 159 (S₁₅₉P), were associated with differences in peak and total (AUC) viral load measurements (Fig. 4). Viral load differences were most significant for L₄₃P, where the P₄₃ polymorphism in the transmembrane domain was associated with lower peak and total viremia in either the homozygous or heterozygous state (Fig. 4D and J).

FIG 4.

Differences in peak and total viremia during acute infection with SIV_mac239Δnef were compared with respect to individual amino acid differences in tetherin. Peak viral RNA loads in plasma (A to F) and AUC values reflecting total viremia (G to L) were compared with respect to amino acid polymorphisms at positions 9, 14, 29, 43, 111, and 159 using the Kruskal-Wallis test followed by an uncorrected Dunn's test.

P₄₃ and H₁₁₁ are both present in rBST-2.5 and rBST-2.6. We therefore compared acute-phase viral loads for animals with at least one copy of either of these alleles to animals with neither of these alleles. The increased power of this aggregate analysis revealed a greater difference in viral loads. Both peak and total viremia were significantly lower (P < 0.0001 and P = 0.0002 by Mann-Whitney U test) for animals expressing rBST-2.5 or rBST-2.6 compared to animals with neither of these alleles (Fig. 5A and B). This suggests that rBST-2.5 and rBST-2.6 have an inherently greater capacity to inhibit the replication of viruses like SIV_mac239Δnef that are unable to counteract tetherin.

FIG 5.

Differences in peak (A) and total (B) viral loads during acute SIV_mac239Δnef infection were compared for animals expressing at least one copy of rBST-2.5 or rBST-2.6 to animals lacking both of these alleles using the Mann-Whitney U test. The horizontal lines indicate median values.

DISCUSSION

Tetherin has broad antiviral activity against diverse families of enveloped viruses and, as a consequence of ongoing evolutionary conflict with viral pathogens, has accumulated numerous species-specific differences (7, 31, 46). Some of these differences have imposed barriers to the cross-species transmission of simian immunodeficiency viruses that have shaped the course of lentiviral evolution in primates (4, 24, 25, 47, 48). Initial comparisons of tetherin sequences from nonhuman primates also revealed variation among members of the same species, including rhesus macaques (24, 25, 31, 48). Since SIV infection of rhesus macaques is an important animal model for lentiviral pathogenesis and AIDS vaccine development (37) and polymorphisms in another restriction factor (TRIM5) can profoundly affect the outcome of SIV infection in this model (36, 49), we determined the extent of sequence variation in tetherin and potential associations with peak viral loads in SIV-infected animals. From an analysis of 146 rhesus macaques, including 68 animals infected with wild-type SIV_mac239 and 47 animals infected with SIV_mac239Δnef, we identified 15 tetherin alleles with amino acid differences at 9 positions. While these differences were distributed throughout the protein, 4 of the 9 polymorphic residues were located in sequences corresponding to the cytoplasmic and transmembrane domains, which are known to interact with viral antagonists and exhibit strong signatures of positive selection (1, 2, 9, 24, 25, 33). Although total acute-phase viremia did not vary significantly among animals infected with wild-type SIV, significant variation was observed among animals infected with nef-deleted SIV. Moreover, two alleles (rBST-2.5 and rBST-2.6) were associated with lower viremia in SIVΔnef-infected animals, suggesting that these variants have an inherently greater capacity to inhibit virus replication in the absence of antagonism by Nef.

Notwithstanding polymorphic differences, all of the tetherin alleles were functional and inhibited the release of nef-deleted SIV more efficiently than wild-type SIV in cell culture assays; however, one allele appeared to be less restrictive than the others. Further characterization of this variant (rBST-2.4) identified a unique polymorphism at position 175 (Q₁₇₅) that accounted for its diminished capacity to inhibit virus release. While the C-terminal location of this polymorphism suggests that it could interfere with cleavage and addition of the GPI anchor, immunoblots did not reveal a shift in molecular weight that would indicate a defect in posttranslational modification. The reason for the functional impairment of rBST-2.4 is unclear but may reflect effects of the Q₁₇₅ polymorphism on subcellular trafficking and/or partial effects on GPI addition that are not apparent from the heterogeneous banding of tetherin observed by Western blot analysis.

A few of the tetherin alleles also appeared to restrict virus release better than the others. Although differences in virus release among all of the alleles except rBST-2.4 were all less than 3-fold, five alleles (rBST-2.2, rBST-2.7, rBST-2.9, rBST-2.13, and rBST-2.14) reproducibly inhibited the detachment of nef-deleted SIV more efficiently than the others (Fig. 2A). Interestingly, these alleles were distinguished by the presence of glycine at position 14, suggesting that the G₁₄ residue contributed to the increased activity of these variants; however, the D₁₄G polymorphism was not associated with differences in acute viremia in SIV_mac239Δnef-infected animals. The reason for this disparity is not immediately apparent, but since the D₁₄G polymorphism coincides with the five-amino-acid sequence that is missing from human tetherin, it is conceivable that this polymorphism differentially affects virus release from human 293T cells used for cell culture assays but not in SIV Δnef-infected macaques.

Conversely, although rBST-2.5 and rBST-2.6 were associated with lower viremia during the acute phase of SIV_mac239Δnef infection, these alleles did not restrict the release of nef-deleted SIV more efficiently in cell culture. This may reflect the amplification of small differences in virus release over multiple cycles of replication in animals that were not detectable in single-round cell culture assays. Differences in the efficiency of virus restriction may also be magnified in vivo by the induction of proinflammatory cytokines and by an increased sensitivity of virus-infected cells to antibody-dependent cellular cytotoxicity (19–23). Furthermore, whereas cell culture assays measure the effects of individual tetherin alleles on virus release in isolation from other factors, the cumulative effects of tetherin on acute viremia in animals are complicated by additional components of innate immunity, including in most animals a second allele of tetherin and polymorphisms in other restriction factors, such as the TRIM5 and APOBEC3 proteins (35, 36). Thus, it is perhaps not surprising that differences in virus release measured using an in vitro assay subject to experimental variation in transfection efficiency did not correspond to the 0.5- to 1.0-log differences in peak viremia observed in SIVΔnef-infected animals.

The fact that differences in acute viremia were associated with tetherin in animals infected with SIVΔnef, but not wild-type SIV, suggests that these differences reflect polymorphisms in tetherin or possibly the product of another closely linked gene counteracted by Nef. Nevertheless, while these associations are intriguing, they are not definitive. Since our analyses are based on coding sequences, we cannot exclude the possibility that polymorphisms in untranslated regions of the tetherin gene that affect protein expression may have contributed to the differences in viremia among SIV_mac239Δnef-infected animals.

Whereas deleterious mutations in a gene are usually rapidly eliminated by purifying selection, mutations that do not affect fitness are eventually either eliminated or driven to fixation through the random process of genetic drift. Thus, the extensive polymorphism observed in rhesus macaque tetherin implies that balancing selection has acted to maintain tetherin diversity in this species. The nature of this selective pressure, however, remains unclear. Although more than 40 species of African apes and Old-World monkeys are naturally infected with simian immunodeficiency viruses, these viruses are not found in Asian species of macaques (50). It is therefore highly unlikely that allelic variation in rhesus macaque tetherin reflects the selective pressure of a lentiviral pathogen, at least not an extant lentivirus; however, macaques are natural hosts for other types of retroviruses, including simian type-D retroviruses and simian T-cell leukemia viruses, as well as many other families of enveloped viruses that may be subject to restriction by tetherin. Hence, any number of diverse viral pathogens may account for the selective pressure to maintain polymorphism in rhesus macaque tetherin.

In contrast to the extensive polymorphism of rhesus macaque tetherin, sequence variation in human tetherin is limited to a few nucleotide differences in noncoding regions and low-frequency variants in coding regions. An insertion/deletion polymorphism in the promoter and a single-nucleotide polymorphism (SNP) in the 3′-untranslated region of human tetherin were associated with higher rates of HIV-1 disease progression (51). SNPs in noncoding regions were also associated with modest differences in HIV-1 acquisition among intravenous drug users and a nominal decrease in the risk of HIV-1 transmission by breast feeding (52, 53). Of the few rare variants in the coding region, one particularly interesting variant in the cytoplasmic domain (R₁₉H) was found to disrupt the signaling activity of human tetherin (54, 55). Thus, while the extent of variation in human tetherin is more limited, there is evidence that certain polymorphisms affect the biology and clinical course of HIV-1 infection.

In summary, we show that tetherin is highly polymorphic in the rhesus macaque, a species that has become an especially valuable animal model for HIV-1/AIDS and other viral diseases. This implies that positive selection is not only driving the diversification of tetherin among primate species but also operating in the form of balancing selection to maintain tetherin polymorphism within certain species. Significant variation in acute viremia for animals infected with nef-deleted SIV, but not wild-type SIV, and the identification of two alleles of tetherin associated with lower viral loads during acute SIVΔnef infection further suggest that some tetherin alleles have an inherently greater capacity to restrict virus replication in the absence of a specific viral antagonist.

MATERIALS AND METHODS

Rhesus macaque BST-2 sequencing.

Peripheral blood mononuclear cells (PBMCs) and plasma samples from rhesus macaques infected with SIV_mac239 or SIV_mac239Δnef were archived during the course of studies conducted with IACUC approval by laboratories at the New England Primate Research Center, Southborough, MA, and the AIDS and Cancer Virus Program (ACVP), Frederick, MD (38–42). Plasma viral load measurements for all samples were performed by the ACVP, Frederick, MD. To determine the tetherin alleles present in the animals, total RNA was extracted from the respective PBMCs and cDNA was synthesized by reverse transcription. Full-length tetherin sequences were amplified from cDNA using primers 5′-GACAGGTACCGTCGCCACCATGGCACCTATTTTGTATGAC-3′ and 5′-CACACTCGAGTCACAGCAGCAGAGCGCTCAAG-3′ and then T/A cloned into pGEM-T easy vector (Promega) and analyzed by Sanger sequencing. The tetherin alleles were confirmed by independent PCRs, and at least 10 to 20 cDNA clones were sequenced to determine tetherin genotypes.

Cell lines.

293T cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM l-glutamine, and 1× antibiotic antimycotic solution. PBMCs from rhesus macaques were isolated using a Ficoll-Hypaque gradient and cultured in RPMI 1640 supplemented with 10% FBS, 25 mM HEPES, 2 mM l-glutamine, and 1× antibiotic-antimycotic solution.

Plasmids.

Tetherin cDNA clones were subcloned into KpnI and XhoI sites of pcDNA3 as previously described (24) to produce expression constructs for each of the rhesus tetherin alleles. A single-nucleotide change corresponding to the L₁₇₅Q substitution was introduced into the pcDNA3 construct encoding rBST-2.10 by PCR overlap extension. All constructs were sequence confirmed.

Virus release assay.

293T cells were transfected with wild-type or nef-deleted SIV proviral DNA (100 ng) and pcDNA3-tetherin expression constructs (4 or 20 ng). To assess the effect of the L₁₇₅Q polymorphism on virus restriction, 20 ng of the tetherin expression constructs was cotransfected with 100 ng of SIV_mac239 or SIV_mac239Δnef proviral DNA (100 ng). Differences in the amount of plasmid DNA in each transfection were offset by the addition of empty pcDNA3 vector. All transfections were performed in 24-well plates using GenJet lipid transfection reagents (SignaGen Laboratories, Gaithersburg, MD) according to the manufacturer's instructions. Forty-eight hours posttransfection, the amount of virus released into the cell culture supernatant was measured by SIV p27 antigen-capture ELISA (Advanced Bioscience Laboratories, Inc., Kensington, MD), and virus release was expressed as the percentage of maximal particle release relative to control transfections in the absence of tetherin, as previously described (24, 28). The data presented are averages from at least three independent experiments.

Immunoblotting.

Forty-eight hours posttransfection, 293T cell lysates were harvested in radioimmunoprecipitation assay (RIPA) buffer (Thermo Scientific) supplemented with Halt protease inhibitor (Thermo Scientific). The lysates were clarified by centrifugation at 20,000 × g for 15 min at 4°C, and protein sample loading buffer (0.25 M Tris, pH 6.8, 10% sodium dodecyl sulfate, 50% glycerol, 5% beta-mercaptoethanol, 0.25% bromophenol blue) was added. Virions were recovered from the cell culture supernatant by centrifugation at 20,000 × g for 2 h at 4°C and resuspended in 2× SDS sample buffer (Bio-Rad, Hercules, CA). Samples were heated to 95°C for 10 min, resolved by electrophoresis on 12% polyacrylamide gels, and transferred to polyvinylidene difluoride (PVDF) membranes. The membranes were blocked with 5% nonfat dry milk in phosphate-buffered saline (PBS) containing 0.05% Tween 20 for 1 h and probed for 1 h at room temperature with one of the following antibodies. Rhesus macaque tetherin alleles and related mutants were detected with a mouse polyclonal antibody (ab88523; Abcam, Cambridge, MA) at a dilution of 1:1,000. The SIV Gag proteins p27 and p55 were detected with the mouse monoclonal antibody 183-H12-5C (AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH) at a dilution of 1:1,000. SIV Nef was detected using the mouse monoclonal antibody 17.2 (AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH) at a dilution of 1:2,000. Endogenous β-actin was detected with the monoclonal antibody C4 (Chemicon, Billerica, MA) at a dilution of 1:1,000. Hsp90 was detected with the monoclonal antibody F-8 (Santa Cruz Biotechnology, Dallas, TX). The PVDF membranes were then washed three times for 5 min in PBS–0.05% Tween 20 and were probed with a horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibody (Bio-Rad, Hercules, CA) at a dilution of 1:1,000 for 1 h at room temperature. The blots were then rinsed three more times in PBS–0.05% Tween 20, treated with Clarity western ECL substrate (Bio-Rad, Hercules, CA), and imaged using an ImageQuant LAS 4000 reader (GE Health Care).

Statistical analyses.

Peak and area under the curve differences in acute viremia were compared as a function of polymorphic differences in tetherin. Peak viral loads were defined as the highest plasma viral RNA load measurement during acute infection (weeks 1 to 4 p.i.), and AUC values were calculated from plasma viral RNA load measurements at weeks 1, 2, and 4 p.i., which were available for most of the animals. Animals that were missing viral load measurements at any of these time points were excluded from the analysis. The Kruskal-Wallis test was used to compare variation in peak and total (AUC) viremia among SIV_mac239- and SIV_mac239Δnef-infected animals. The Kruskal-Wallis test, followed by an uncorrected Dunn's test, was used to compare differences in peak and total viremia as a function of individual amino acid polymorphisms in tetherin for the SIV_mac239Δnef-infected animals. The Mann-Whitney U test was used to compare differences in peak and total viremia for SIV_mac239Δnef-infected animals with at least one copy of rBST-2.5 or rBST-2.6 to viral loads for the other animals that lack both of these alleles.

Accession number(s).

Sequences were deposited in GenBank under accession numbers MG816214 to MG816219 and MG816221 to MG816227 (Table 1).