|Home | About | Journals | Submit | Contact Us | Français|
Human immunodeficiency virus type 1 (HIV-1) does not replicate in primary cells of New World primates. To better understand this restriction, we expressed owl monkey (Aotus nancymaae) CD4 and CXCR4 in the owl monkey kidney cell line, OMK. An HIV-1 variant modified to evade the owl monkey restriction factor TRIM-cyp replicated efficiently in these cells but could not replicate in primary A. nancymaae CD4-positive T cells. To understand this difference, we examined APOBEC3G and tetherin orthologs from OMK cells and primary A. nancymaae cells. We observed that OMK cells expressed substantially lower levels of APOBEC3G than did A. nancymaae cells. A. nancymaae, but not marmoset (Callithrix jacchus), APOBEC3G was partially downregulated by HIV-1 vif and reduced but did not abolish HIV-1 replication when stably expressed in OMK cells. The functional difference between A. nancymaae and marmoset APOBEC3Gs mapped to residue 128, previously shown to distinguish African green monkey from human APOBEC3G. We also characterized tetherin orthologs from OMK and A. nancymaae cells. The A. nancymaae tetherin ortholog, but not OMK tetherin, prevented HIV-1 release. Alteration of threonine 181 of OMK tetherin rescued its function and its efficient N glycosylation. All alleles of Aotus lemurinus griseimembra examined, but none of A. nancymaae or Aotus vociferans, encoded this nonfunctional tetherin ortholog. Our data indicate that HIV-1 replication in owl monkeys is not restricted at entry but can be limited by APOBEC3G and tetherin. Further, A. lemurinus griseimembra does not restrict HIV-1 replication via tetherin, a property likely useful for the study of tetherin-restricted viruses.
Replication of retroviruses in species other than their natural host species can be blocked at multiple stages in the viral life cycle (1). In recent years, significant progress has been made in describing the mechanisms underlying these restrictions. Barriers to replication of human immunodeficiency virus type 1 (HIV-1) that function at entry are usually due to species-specific differences in the HIV-1 receptor CD4 or coreceptor CCR5 or CXCR4 (13). Differences in other cellular proteins necessary for replication, for example, murine cyclin T, can also prevent viral replication (7, 8). In addition, there are at least three barriers to HIV-1 replication that do not arise from dependence on a cellular protein and that require species-specific viral evasion strategies: TRIM5α or TRIM-cyp proteins, which interact with and are thought to prematurely destabilize the viral capsid (16, 18, 23); APOBEC3 proteins—in particular, APOBEC3G—which deaminate the negative strand of viral DNA during reverse transcription (21); and tetherin (Bst-2 and CD317), which inhibits release of the virion after budding (15, 24). The HIV-1 protein vif promotes the degradation of human APOBEC3G in the virus-producing cell, preventing subsequent deamination of the viral genome in the target cell (21, 22). Similarly, the viral protein vpu facilitates escape from tetherin and release of the virion from the producing cell (14, 15, 24). The mechanism by which HIV-1 evades human TRIM5α but not, for example, that of rhesus macaques remains unclear (1).
New World primates, such as squirrel monkeys (Saimiri sciureus), marmosets (Callithrix jacchus), and owl monkeys (Aotus nancymaae, Aotus vociferans, and Aotus lemurinus griseimembra) have been used as animal models for a number of viral and nonviral diseases (5), and they would be attractive models for the study of primate lentiviruses. However, to date, no lentiviruses have been isolated from these animals, and neither HIV-1 nor simian immunodeficiency virus replicates in their primary cells. It has been observed that marmoset and squirrel monkey CD4 and CCR5 molecules do not bind HIV-1 gp120 or facilitate viral entry, and it had been suggested that entry is the major restriction to viral replication in most primary New World primate cells (13). Subsequent to these studies, replication of HIV-1 in the owl monkey cell line OMK was shown to be restricted by owl monkey TRIM-cyp, a variant of the rhesus macaque restriction factor TRIM5α in which the TRIM5α B20.4 domain is replaced by a nearly intact cyclophilin domain (16, 18). HIV-1 variants modified in the cyclophilin-binding domain of their capsid protein could readily evade owl monkey TRIM-cyp (9) and could replicate efficiently in OMK cells transfected to express human CD4 and CXCR4 (4).
Because engineered HIV-1 could readily evade TRIM-cyp (4, 9) and because owl monkeys have been used as animal models for a number of infectious diseases, including both major human malarias (5, 25), they are of particular interest in the efforts to generate an animal model of HIV-1 infection. However, study of the owl monkey has been complicated by some confusion as to the origin of the widely used OMK cell line (17). Although these cells have been used extensively in HIV-1 research, they were derived before the extent of owl monkey speciation was fully appreciated, and they are usually incorrectly described as being of Aotus trivigatus origin. There are at least 10 known species of owl monkeys, differing in karyotypes and some physical features, distributed from western Panama to northern Argentina (6).
Here, we show that OMK cells expressing owl monkey orthologs of CD4 and CXCR4 also permit robust replication of an HIV-1 variant modified to escape TRIM-cyp. In contrast, the primary CD4-positive cells of one owl monkey species, A. nancymaae, did not permit the replication of this HIV-1 variant. We explored the differences between these cells and observed critical differences in the HIV-1 restriction factors APOBEC3G and tetherin. OMK cells express substantially less APOBEC3G than primary A. nancymaae cells, and stable expression of A. nancymaae APOBEC3G limited but did not abolish HIV-1 replication. We further observed that the tetherin ortholog found in A. nancymaae efficiently restricted HIV-1 but that the OMK cell-derived tetherin had no effect on HIV-1 replication due to a homozygous mutation at tetherin residue 181. All three of the A. lemurinus griseimembra monkeys examined were found to be homozygous for this same nonfunctional tetherin ortholog, confirming that OMK cells derive from this owl monkey species and indicating that A. lemurinus griseimembra cannot restrict HIV-1 via tetherin. In contrast, all A. nancymaae and A. vociferans tetherin orthologs were observed to be functional, and HIV-1 vpu did not overcome the restriction of these functional orthologs. Thus, HIV-1 replication in A. nancymaae CD4-positive cells is not restricted at entry but can be restricted by both APOBEC3G and tetherin.
OMK cells (American Type Culture Collection) were maintained in minimal essential medium with 20% fetal bovine serum and antibiotics. OMK cells expressing owl monkey CD4 and CXCR4 (omCD4 and omCXCR4) were generated by transduction with the retroviral vector pQCXIX (Clontech) containing omCD4 and then with the same vector containing omCXCR4. The cells were stained with anti-human CD4 (clone Q4120) and anti-human CXCR4 (clone 12G5) antibodies and then sorted for high- and low-level expression of omCD4 (omCD4hi/omCXCR4-OMK and omCD4lo/omCXCR4-OMK). We had previously confirmed that Q4120 and 12G5 cross-react with their respective owl monkey receptors (not shown). omCD4lo/omCXCR4-OMK cells stably expressing owl monkey (A. nancymaae) APOBEC3G were generated by transduction with the lentiviral vector pLenti 6/V5-D-topo (Invitrogen) and selected with blasticidin.
Peripheral blood lymphocytes were isolated from blood by Ficoll (Histopaque, Sigma) gradient, stimulated with 10 μg/ml concanavalin A (Sigma) or 0.25 μg/ml phytohemagglutinin (Murex) for 2 days, and then maintained with 100 U/ml recombinant human interleukin-2 (Roche) in RPMI, 20% fetal bovine serum, and antibiotics. The cells were CD4 enriched by staining them with anti-human CD4 (clone M-T466) microbeads (Miltenyi) and sorted with an automated magnetic sorter (autoMACS) before infection.
Full-length replication-competent pNL4-3 was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH. The pNL4-3.G89V capsid variant was generated by replacing the 60-nucleotide SphI-SpeI fragment of pNL4-3 with a double-stranded oligonucleotide that had the glycine codon, GGG, mutated to valine, GTA, at codon 89 of the region encoding the HIV-1 capsid protein. vpu-negative variants of pNL4-3 and pNL43.G89V (pNL4-3.Δvpu and pNL43.G89V.Δvpu) were generated by deleting nucleotide 6076, resulting in multiple stop codons after the fifth codon of vpu. Viruses were produced in 293T cells by calcium phosphate transfection and assayed for p24 production with HIV-1 enzyme-linked immunosorbent assay (ELISA) kits (Perkin Elmer). Comparative viral infection was performed with p24-normalized amounts of virus. The virus was incubated for 16 h with cells and washed with phosphate-buffered saline (PBS). Viral production was monitored by p24 ELISA.
Total RNA was purified from cells using Qiagen RNeasy and reverse transcribed with the ImProm-II Reverse Transcription System from Promega, primed with oligo(dT)12. New World primate genes encoding CD4, CXCR4, APOBEC3G, and tetherin were amplified using either PfuTurbo (Invitrogen) or Ex Taq (Takara) DNA polymerase with forward (f) and reverse (r) primers as follows: owl monkey CD4 (from A. nancymaae peripheral blood mononuclear cells [PBMC]) with ATGGATGGGGGAATCCCTTT (f) and TCAAATGGGGCTACATGTCTT (r); owl monkey CXCR4 (from A. nancyimaae PBMC) with ATGGAGGGCATCAGTATATAC (f) and TTAGCTGGAGTGAAAACTTGA (r); owl monkey (from A. nancymaae PBMC) and marmoset (from C. jacchus PBMC) APOBEC3G with ATGAAGCCTCAGACCAGGAACACAGTGG (f) and TCAGTTTCCCATGATCTGGAGAATGGCC (r); and owl monkey (A. nancymaae, A. lemurinus griseimembra, and A. vociferans PBMC and OMK cells) tetherin with GAGGGGAGATCTGGATGGC (f) and GATATGCCAGCTTCCTGGGAT (r). Genes encoding owl monkey and marmoset APOBEC3G and owl monkey tetherins were cloned into protein expression plasmids derived from a pUC-based vector with a cytomegalovirus promoter and a hepatitis delta antigen polyadenylation signal. Point mutations were generated by site-directed mutagenesis using the QuikChange method (Stratagene), and coding regions amplified by PCR were sequenced.
HIV-1 vif-mediated degradation of APOBEC3G orthologs was measured by transfecting 106 HEK293T cells per well of a six-well plate with 1 μg of vector expressing carboxy-terminal hemagglutinin (HA)-tagged APOBEC3G and 1 μg of HIV-1 vif-expressing plasmid (NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH; pcDNA-HVif was from Stephan Bour and Klaus Strebel) using Lipofectamine 2000 (Invitrogen). The cells were lysed 3 days posttransfection with protein-loading buffer, resolved on a 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel, transferred onto polyvinylidene difluoride membranes (Invitrogen), and immunoblotted with either anti-HA.11 (Covance) or anti-vif (obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH; HIV-1HXBC2 vif antiserum was from Dana Gabuzda).
Virus was produced in the presence of APOBEC3Gs by calcium phosphate transfection of HEK293T cells with a proviral vector and an HA-tagged APOBEC3G expression vector. Virus was harvested 48 h posttransfection, and infectious viral titers were determined by infecting 2 × 104 GHOST-CD4/CXCR4 cells per well in six-well plates with equal volumes of viral supernatant for 16 h, washing them with PBS, and harvesting them for flow cytometric analysis 5 days postinfection. The cells were lysed and immunoblotted with anti-HA antibody to determine APOBEC3G expression.
Virus release in the presence of tetherin orthologs was performed by Lipofectamine 2000 transfection of 2 × 105 HEK293T cells per well in 24-well plates with 800 ng of pNL4-3 or pNL4-3.Δvpu proviral vector and 80 ng of N-terminally HA-tagged tetherin expression vectors. Forty-eight hours postinfection, virus release was assayed by p24 ELISA, and infectious titers were determined by infecting 2 × 104 GHOST-CD4/CXCR4 cells per well in six-well plates with 400 μl of viral supernatant for 16 h, washing them with PBS, and harvesting them for flow cytometric analysis 5 days postinfection.
293T cells transfected with N-terminally HA-tagged tetherin expression vectors were metabolically labeled with EXPRE35S35S (Perkin Elmer) for 30 h. The cells were lysed in 1% NP-40 and precipitated with anti-HA monoclonal antibody and protein A-Sepharose. The precipitated proteins were denatured and treated with either N-glycosidase F (PNGase F) (New England Biolabs) or buffer alone. Samples were separated on 14% SDS-PAGE and quantified by phosphorimaging.
The GenBank accession numbers for cDNA encoding A. nancymaae CD4, CXCR4, and APOBEC3G and marmoset APOBEC3G are FJ623078, FJ638411, FJ638412, and FJ638413, respectively. The GenBank accession numbers for cDNAs encoding tetherin orthologs are FJ638414 (A. lemurinus griseimembra), FJ638415 and FJ638416 (A. nancymaae alleles 1 and 2), and FJ638417 and FJ638418 (A. vociferans alleles 1 and 2).
A previous study of factors limiting HIV-1 replication in New World primates showed that the CD4 molecules of squirrel monkeys (S. sciureus) and marmosets (C. jacchus) did not support efficient HIV-1 entry. However, alteration of marmoset CD4 residues 48 and 59 to their human counterparts rescued efficient entry of several HIV-1 isolates (13). Arginine 59 of human CD4, in particular, makes a critical salt bridge with aspartic acid 368, highly conserved in the HIV-1 envelope glycoprotein (12). Unlike marmoset CD4, residue 59 of A. nancymaae is a lysine (see Fig. S1 in the supplemental material), as it is in rhesus macaque CD4, which supports HIV-1 entry. Consistent with this similarity, A. nancymaae CD4 (omCD4) could immunoprecipitate 14 out of 18 HIV-1 envelope glycoproteins assayed, including those of KB9, 89.6, HXB2, and NL4-3 isolates, comparably to human CD4 (not shown). Unlike their CD4 orthologs, marmoset and squirrel monkey CXCR4s are homologous to human CXCR4 (13). These CXCR4 orthologs have been shown to function with human CD4 to permit efficient HIV-1 entry. Similarly, A. nancymaae CXCR4 (omCXCR4) differs modestly from human CXCR4 (see Fig. S2 in the supplemental material). Accordingly, we tested whether omCD4 and omCXCR4 together could support efficient replication of the X4 isolate NL4-3.
To do so, we used an NL4-3 variant (NL4-3.G89V) modified at capsid residue 89 to evade the owl monkey restriction factor TRIM-cyp (9). This variant has been shown to replicate efficiently in OMK cells expressing human CD4 and CXCR4 (4). OMK cells were transduced with retroviral vectors expressing omCD4 and omCXCR4. The cells were then sorted into populations expressing relatively high (omCD4hi/omCXCR4-OMK) and low (omCD4lo/omCXCR4-OMK) levels of CD4 (Fig. (Fig.1A).1A). Jurkat, HeLa-CD4, omCD4hi/omCXCR4-OMK, and omCD4lo/omCXCR4-OMK cells were incubated with wild-type HIV-1 NL4-3 or with NL4-3.G89V. As expected, wild-type NL4-3 replicated efficiently in HeLa-CD4 and Jurkat cells but did not replicate in either OMK population (Fig. (Fig.1B).1B). NL4-3.G89V replicated inefficiently in both human cell lines, as previously described (3), but replicated robustly in both omCD4hi/omCXCR4-OMK and omCD4lo/omCXCR4-OMK cells. HIV-1 replicated somewhat more efficiently in cells with less CD4. Consequently, omCD4lo/omCXCR4-OMK cells were used in the remaining experiments. Virus harvested from omCD4lo/omCXCR4-OMK cells on day 11 postinfection could easily infect omCD4lo/omCXCR4-OMK cells, and indeed, replication could be maintained indefinitely in this manner (not shown). We conclude that HIV-1 can efficiently replicate in cells expressing omCD4 and omCXCR4 and therefore that HIV-1 is not restricted by A. nancymaae receptors.
We then compared the abilities of NL4-3.G89V to replicate in omCD4lo/omCXCR4-OMK cells and in A. nancymaae PBMC sorted for CD4 expression. NL4-3.G89V replicated efficiently in the OMK cells expressing A. nancymaae receptors but could not efficiently replicate in primary A. nancymaae CD4-positive cells (Fig. (Fig.1C).1C). Moreover, we were unable to promote replication of NL4-3.G89V in these primary cells despite multiple attempts at adaptation. We therefore sought to define differences between OMK cells and primary A. nancymaae cells that might account for this variation.
We first characterized the expression of the HIV-1 restriction factor APOBEC3G in OMK cells and in primary A. nancymaae PBMC. OMK cells express substantially lower levels of APOBEC3G than PBMC obtained from A. nancymaae, as measured by reverse transcriptase (RT) PCR (Fig. (Fig.2A)2A) or real-time PCR (not shown), and expression in OMK cells was only modestly increased by alpha interferon induction. We also examined the ability of NL4-3 vif to downregulate expression of A. nancymaae APOBEC3G. NL3-4 vif partially downregulated wild-type A. nancymaae APOBEC3G, but less efficiently than it downregulated human APOBEC3G (Fig. (Fig.2B,2B, compare lanes 1 and 2 with lanes 5 and 6). It was previously demonstrated that HIV-1 vif does not downregulate rhesus macaque or African green monkey APOBEC3G but can downregulate a variant of African green monkey APOBEC3G in which an aspartic acid, present in human APOBEC3G, is introduced at lysine 128 (2, 19). Like human APOBEC3G, the A. nancymaae ortholog has an aspartic acid at this position (see Fig. S3 in the supplemental material), and its replacement with a lysine prevents its downregulation by vif (lanes 7 and 8, 13, and 14). Expression of marmoset APOBEC3G was not altered by vif, nor was expression of marmoset APOBEC3G variants in which its residue 128 was altered to aspartic acid or lysine (lanes 3 and 4, 9 and 10, 11 and 12).
The ability of these APOBEC3G orthologs and variants to be downregulated by vif was correlated with the infectious titers of viruses produced in their presence. In particular, A. nancymaae APOBEC3G partially diminished the titers of NL4-3 produced in HEK293T cells, but less so than APOBEC3G variants resistant to NL4-3 vif (Fig. (Fig.2C).2C). This effect was more pronounced over multiple rounds of replication: substantially less virus was produced in omCD4lo/omCXCR4-OMK cells stably expressing A. nancymaae APOBEC3G than in omCD4lo/omCXCR4-OMK cells lacking exogenous APOBEC3G (Fig. (Fig.2D).2D). Thus, HIV-1 vif can downregulate A. nancymaae APOBEC3G more efficiently than marmoset APOBEC3G, in part because it shares with human APOBEC3G an aspartic acid at residue 128. Nonetheless A. nancymaae APOBEC3G expression can substantially limit HIV-1 replication.
We also characterized the expression and function of the HIV-1 restriction factor tetherin in OMK cells and A. nancymaae PBMC. As shown in Fig. Fig.3A,3A, OMK cells and A. nancymaae PBMC expressed comparable levels of tetherin mRNA. Despite the presence of tetherin message in both cells, a vpu-negative variant of NL4-3.G89V replicated in both sets of cells as efficiently as NL4-3.G89V with an intact vpu gene (Fig. (Fig.3B).3B). These data suggest that HIV-1 vpu did not overcome any tetherin-mediated restriction.
We then cloned and sequenced tetherin from OMK cells and from the PBMC of three A. nancymaae owl monkeys. We identified two distinct alleles of A. nancymaae tetherin and a single allele from OMK cells. The OMK cell-derived tetherin differed at 5 amino acids from each A. nancymaae ortholog. These tetherin orthologs were assayed for the ability to limit release of HIV-1 virions by measuring p24 levels in supernatants of virus-producing cells (Fig. (Fig.3C,3C, left) or by measuring the infectious titers of vpu-positive or vpu-negative NL4-3 virus produced in the presence of each tetherin variant (Fig. (Fig.3D,3D, right). As an additional control, vpu was also provided in trans when vpu-negative NL4-3 was assayed. In the absence of vpu, human tetherin and both A. nancymaae tetherin variants prevented virus release, but OMK tetherin did not. vpu provided by the NL4-3 provirus or in trans rescued virion release in cells expressing human tetherin, but not those expressing either of the A. nancymaae tetherin variants. Thus, the tetherin found in A. nancymaae cells, but not that in the OMK cell line, restricts HIV-1 virion release, and this restriction is not affected by vpu.
The inability of OMK cell-derived tetherin to limit HIV-1 release suggested that at least one owl monkey species lacked this restriction factor. Ribeiro et al. showed that the cyclophilin domain of TRIM-cyp cloned from OMK cells most closely resembled that from A. lemurinus griseimembra (17). We accordingly cloned tetherins from three A. lemurinus griseimembra and three A. vociferans owl monkeys (Fig. (Fig.4).4). A. vociferans and A. lemurinus griseimembra are members of the gray-necked group of owl monkeys and are more closely related to each other than to A. nancymaae, a red-necked owl monkey. Two distinct A. vociferans tetherin alleles and a single A. lemurinus griseimbra tetherin allele were identified. The A. lemurinus griseimembra tetherin gene, present in all three animals, was identical at the nucleotide level to that cloned from OMK cells, confirming that this cell line derives from A. lemurinus griseimembra. In addition, two of these animals expressed an additional allele that differed from the OMK tetherin gene at 2 nucleotides (not shown) but encoded a protein identical to that expressed in OMK cells. Both A. vociferens tetherin orthologs, like those from A. nancymaae, efficiently limited HIV-1 release and were not affected by expression of HIV-1 vpu (Fig. (Fig.5A).5A). In contrast, virion release was efficient in the presence of A. lemurinus griseimembra/OMK tetherin.
Four residues distinguish A. nancymaae and A. lemurinus griseimembra tetherin molecules (Fig. (Fig.4),4), but one of these, isoleucine 131, is common to A. vociferans and A. lemurinus griseimembra and therefore is unlikely to account for the functional defect in the latter tetherin. Accordingly, we altered the remaining 3 residues of A. lemurinus griseimembra/OMK tetherin to their A. nancymaae analogs. Two of these changes had no effect on A. lemurinus griseimembra/OMK tetherin function, but replacement of threonine 181 with an isoleucine present in all of the A. nancymaae and A. vociferans tetherin orthologs rescued tetherin restriction of HIV-1 (Fig. (Fig.5A).5A). Conversely, an A. nancymaae tetherin variant in which isoleucine 181 was modified to threonine could not restrict HIV-1. Thus, threonine 181 alone determines the inability of A. lemurinus griseimembra/OMK tetherin to restrict HIV-1.
In order to correlate biochemical properties of these owl monkey-tetherin orthologs with their ability to restrict, we compared OMK and A. nancymaae tetherins, as well as the gain-of-function T181I variant of OMK tetherin and the loss-of-function I181T variant of A. nancymaae tetherin. Immunoprecipitation of N-terminally tagged forms of these tetherin molecules revealed two bands: a sharp high-mobility form and a diffuse form that migrated more slowly (Fig. (Fig.5B).5B). Strikingly, the ratio of these bands was correlated with the ability of each of these tetherins to restrict HIV-1. Specifically, wild-type OMK tetherin was primarily found in the lower band, whereas A. nancymaae tetherin was observed mostly in the higher, diffuse band. Treatment of these tetherin molecules with an N-glycosidase increased the mobility of the higher band, consistent with its modification by N glycosylation. The markedly lower efficiency with which tetherins bearing a threonine at residue 181 were modified by N glycosylation suggests that these tetherin variants had a different topology or conformation within the membrane of the endoplasmic reticulum, thereby preventing their subsequent maturation and ability to restrict HIV-1. We finally sought to determine if the residual glycosylated form of OMK tetherin could restrict HIV-1 at high levels of expression. Some restriction of HIV-1 was observed when OMK tetherin was expressed at high levels, although substantially less than that observed with the T181I variant (Fig. (Fig.5C).5C). In addition, high levels of OMK tetherin resulted in some cytotoxicity that was not observed with the T181I variant (Fig. (Fig.5C,5C, bottom). We conclude that a maturation defect in OMK/A. lemurinus griseimembra tetherin impairs the ability of this tetherin ortholog to restrict HIV-1.
The development of an animal model of HIV-1 infection remains an important but elusive goal. To develop such a model, the virus will likely be engineered and adapted to more efficiently use animal orthologs of human proteins necessary for replication and to evade factors of intrinsic immunity that restrict viral replication. These factors present an especially difficult hurdle, because they have undergone positive selection that makes them highly diverse and they are typically refractory to viral evasion strategies evolved in the host species of the virus. As part of an effort to understand these factors in New World primates, we sought to define the underlying differences between an owl monkey kidney cell line, OMK, which permits efficient replication of an HIV-1 variant engineered to evade owl monkey TRIM-cyp, and primary CD4-positive cells from one owl monkey species, A. nancymaae, which do not support replication of this virus.
We initially explored whether differences between human and owl monkey CD4 or CXCR4 could account for this absence of replication in primary cells. Marmoset and squirrel monkey CD4s do present a significant hurdle to HIV-1 replication in these species (13), but we observed no such hurdle with A. nancymaae receptors. We did, however, observe additional sources of restriction that were likely present but not fully appreciated in previous studies of HIV-1 replication in New World primate cells (13). First, we observed that OMK cells expressed substantially less APOBEC3G than primary A. nancymaae cells and showed that when A. nancymaae APOBEC3G was introduced into OMK cells, viral replication was slowed, although not completely inhibited. HIV-1 vif could partially downregulate expression of A. nancymaae APOBEC3G, although much less efficiently than it downregulated human APOBEC3G. The partial susceptibility of the owl monkey APOBEC3G is due to the presence of an aspartic acid at residue 128, a residue shared with human APOBEC3G. In contrast, African green monkey, rhesus macaque, and, as we show here, marmoset APOBEC3G lack this aspartic acid, and as a consequence, these orthologs are not downregulated by vif (2, 19).
Differences between tetherin molecules found in OMK and primary A. nancymaae cells further explain the differences in HIV-1 replication in these cells. We observed that the OMK tetherin did not restrict HIV-1, whereas the A. nancymaae tetherin restricted vpu-positive virus efficiently. This difference mapped to residue 181 present in a region thought to be cleaved during the addition of a glycosylphosphatidylinositol link to the tetherin C terminus (10, 11, 20). We showed here that the presence of a threonine at residue 181 prevented efficient glycosylation of the protein in the endoplasmic reticulum, consistent with a conformation or topology distinct from that of functional tetherins. Surprisingly, this substitution appears to be fixed in one species of owl monkey, A. lemurinus griseimembra, but could not be observed in any of three A. vociferens or four A. nancymaae monkeys. Combined with similar analysis of the cycophilin domains of TRIM-cyp (17), our data clearly indicate that OMK cells derive from A. lemurinus griseimembra and suggest that HIV-1 would not be restricted by tetherin in this species. Unfortunately, despite its use in malaria research, this particular owl monkey species is endangered, precluding at this point in vivo adaptation of HIV-1. However, it is not clear that the I181T mutation emerged after speciation of A. lemurinus griseimembra, and at least 5 of the 10 known owl monkey species, in addition to A. vociferans, are more closely related to A. lemurinus griseimembra than is A. nancymaae. It is therefore possible that a closely related owl monkey species also expresses the I18IT tetherin variant. Given that owl monkey entry receptors promote efficient entry of HIV-1 and that owl monkey APOBEC3G is sufficiently susceptible to HIV-1 vif to permit further adaptation in vivo, identification of such an owl monkey may be of use in the development of an animal model for HIV-1 or other tetherin-restricted viruses.
Published ahead of print on 24 June 2009.
†Supplemental material for this article may be found at http://jvi.asm.org/.