TRIMCyp expression in Old World primates Macacanemestrina and Macaca fascicularisGreg Brennan†, Yury Kozyrev‡, and Shiu-Lok Hu†‡§¶

Departments of †Microbiology and ‡Pharmaceutics, and §Washington National Primate Research Center, University of Washington, Seattle, WA 98121

Edited by John M. Coffin, Tufts University School of Medicine, Boston, MA, and approved January 10, 2008 (received for review October 5, 2007)

Primates have evolved a variety of restriction factors that preventretroviral replication. One such factor, TRIM5�, mediates a posten-try restriction in many Old World primates. Among New Worldprimates, Aotus trivirgatus exerts a similar early restriction medi-ated by TRIMCyp, a TRIM5–cyclophilin A (CypA) chimera resultingfrom a CypA retrotransposition between exons 7 and 8 of theTRIM5 gene. Macaca nemestrina do not express TRIM5�; therefore,we asked whether these animals and related Old World primatesexpress TRIMCyp. RT-PCR of total RNA from M. nemestrina andMacaca fascicularis yielded three TRIMCyp amplification products,one of which is predicted to encode a TRIMCyp chimera containinga full-length CypA. Unlike A. trivirgatus, genomic sequencing of M.nemestrina and M. fascicularis identifies a CypA retrotranspositionin the 3� untranslated region of the TRIM5 locus. There is �78%homology between the predicted protein sequences of Old Worldand New World primate TRIMCyp, with most of the differencesfound in the TRIM5-derived sequence. Notably, exon 7 is absentfrom both M. nemestrina and M. fascicularis TRIMCyp. Neither M.nemestrina nor M. fascicularis TRIMCyp could restrict HIV-1 orsimian immunodeficiency virus SIVmac in an in vitro infectivityassay. The discovery of TRIMCyp in both M. nemestrina and M.fascicularis indicates that TRIMCyp expression may be more com-mon among Old World primates than previously believed. Con-vergent evolution of TRIMCyp in both Old World and New Worldprimates suggests that TRIMCyp may have provided evolutionaryadvantages.

convergent evolution � cyclophilin A � HIV � retroviruses � TRIM5�

The current HIV pandemic resulting from cross-species trans-missions of simian immunodeficiency viruses SIVcpz or

SIVsmm to humans has been well documented (1). However, themechanisms enabling such transmissions are not yet fully un-derstood. Mammals have evolved several restriction factorscapable of inhibiting the replication of certain retroviruses in aspecies-specific manner. One of the best described primatehost-restriction factors is TRIM5�, which is expressed in mostOld World primates.

Macaca mulatta TRIM5� exerts an early, postentry block toHIV-1 replication (2). TRIM5� is a member of the tripartitemotif family of proteins, characterized by the ordered N-terminal to C-terminal expression of a RING domain, B-Box,and coiled coil, also known as an RBCC domain (3). TheTRIM5� isoform also expresses a B30.2 domain at the Cterminus, which is required for recognition of the incomingretroviral capsid (4–7). Changes to the B30.2 domain have beenshown to dramatically affect the breadth and potency ofTRIM5�-mediated anti-retroviral activity. For instance, theability to restrict HIV-1 replication may be conferred to Homosapiens TRIM5� by changing a single residue to the amino acidfound in M. mulatta TRIM5�, R332P (8). Similarly, site-directedmutagenesis studies have demonstrated that mutations aroundthe cyclophilin A (CypA) binding loop of HIV-1 capsid effect thepotency of TRIM5�-mediated restriction, suggesting thatthe B30.2 domain interacts with or near the CypA binding loop(5, 9, 10).

New World primates Aotus trivirgatus exert a postentry re-striction to HIV-1 mediated by a TRIM5–CypA chimera. Se-quencing the A. trivirgatus TRIM5 gene identified a LINE-1-mediated retrotransposition of CypA into intron 7, resulting inthe expression of a fusion protein called TRIMCyp, which isunique to the Aotus genus (11–13). TRIMCyp retains theN-terminal tripartite motif of all TRIM family members, but theB30.2 domain of TRIM5� is replaced by the CypA domain.Functionally, TRIM5� and TRIMCyp are similar, preventingreverse transcription of incoming viruses at an early postentrystage. However, TRIMCyp exerts a more potent restriction toincoming retroviruses than TRIM5�, and unlike TRIM5�,TRIMCyp-mediated restriction is sensitive to cyclosporin A. A.trivirgatus has been shown to express only TRIMCyp, notTRIM5� (11–12).

We recently demonstrated that the Old World primatesMacaca nemestrina do not express TRIM5�. Instead, theytranscribe novel isoforms TRIM5� and TRIM5� (14). Theseisoforms likely arise because of a single-nucleotide polymor-phism (SNP) at the intron 6 splice acceptor. TRIM5� is trun-cated proximal to the B30.2 domain, and TRIM5� lacks 9 aaencoded by exon 7 located at the N terminus of the B30.2domain. However, neither isoform restricts HIV-1 in vitro. Wetherefore asked whether M. nemestrina or related macaquespecies express TRIMCyp to compensate for the loss ofTRIM5�-mediated retroviral restriction.

ResultsOld World Primates M. nemestrina and Macaca fascicularis bothExpress TRIMCyp. Total RNA was isolated from M. nemestrina, M.mulatta, and M. fascicularis. We performed RT-PCR by using aforward TRIM5-specific primer and a reverse CypA-specificprimer. To facilitate later functional analyses, the forwardprimer was designed to replace the start codon with a hemag-glutinin (HA) epitope tag. As expected, no amplification prod-ucts were identified in M. mulatta total RNA. However, multipleamplification products were found in both M. nemestrina and M.fascicularis (Fig. 1).

The amplification products from each species were gel-purified, cloned, and sequenced. Sequence analysis of eachcDNA species confirmed that animals from both species ex-pressed three TRIMCyp transcripts, with a CypA-coding regionjoined to exon 4, exon 5, or exon 6 of TRIM5. Because the 5�portion of the CypA gene is 2 nt out of frame with its codingregion, the TRIMCyp exon 4 and TRIMCyp exon 5 transcripts

Author contributions: G.B., Y.K., and S.-L.H. designed research; G.B. and Y.K. performedresearch; G.B., Y.K., and S.-L.H. analyzed data; and G.B. and S.-L.H. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The sequences reported in this paper have been deposited in the GenBankdatabase (accession nos. EU371628–EU371639 and EU371641).

See Commentary on page 3177.

¶To whom correspondence should be addressed at: University of Washington, 3000 West-ern Avenue, Seattle, WA 98121. E-mail: hus@bart.rprc.washington.edu.

© 2008 by The National Academy of Sciences of the USA

www.pnas.org�cgi�doi�10.1073�pnas.0709511105 PNAS � March 4, 2008 � vol. 105 � no. 9 � 3569–3574

MIC

ROBI

OLO

GY

SEE

COM

MEN

TARY

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

020

are predicted to encode fusion proteins with a truncated CypAdomain (Fig. 2A). Only the TRIMCyp exon 6 transcript ispredicted to express a TRIMCyp protein of 469 aa in M.nemestrina and M. fascicularis, comparable to the 475-aa TRIM-Cyp from A. trivirgatus (Fig. 2B). Therefore, although theseobservations are consistent with the description of multipleTRIMCyp isoforms in A. trivirgatus, the identities of the isoformsare different. The TRIMCyp isoform with the full-length CypAjoined to exon 6 will be referred to as TRIMCyp.

Old World Primate TRIMCyp Results from a Retrotransposition of CypAin the 3� Untranslated Region (UTR) of the TRIM5 Gene. To determinethe origin of the TRIMCyp transcripts observed in the OldWorld primates, we looked for a CypA retrotransposition inTRIM5. Genomic DNA was isolated from M. mulatta, M. nem-estrina, and M. fascicularis peripheral blood monocytes (PBMC).We amplified a fragment from TRIM5 exon 6 to exon 8. Theamplification product from all three species was identical in size(�650 bp, Fig. 3A), indicating the absence of additional DNAsequences inserted in the M. fascicularis and M. mulatta TRIM5intron 7. Sequence analysis of the TRIM5 locus shows that, unlikeA. trivirgatus, there is no CypA retrotransposition in Old Worldprimate TRIM5 intron 7 (Fig. 3B). We also identified a G-to-TSNP at the M. fascicularis TRIM5 intron 6 splice acceptoridentical to a SNP that we previously found in M. nemestrina (14).

PCR amplification using a forward TRIM5 exon 5-specificprimer and a reverse CypA-specific primer yielded an amplifi-cation product of �4.5 kb in M. nemestrina and M. fascicularis(Fig. 3C). Sequence analysis of this amplification product iden-tifies a CypA gene in the 3� UTR of the TRIM5 locus (Fig. 3D).Nucleotide sequences directly upstream of the CypA gene areconsistent with consensus sequences used by LINE-1-mediatedretrotransposition (15).

Alignment of Old World and New World TRIMCyp. Alignment of thepredicted amino acid sequences of A. trivirgatus, M. nemestrina,and M. fascicularis TRIMCyp identifies 97% homology betweenthe two macaque species and 78–79% homology between theOld World and New World primates. The majority of thesequence diversity between A. trivirgatus and the macaqueTRIMCyp is located in the RBCC region, with the coiled coiland the downstream linker 2 region accumulating most of theamino acid changes (Fig. 2B).

A comparison of Old World and New World primate TRIM-Cyp revealed two notable features between the linker 2 regionand the CypA domain. One is the absence in Old World primateTRIMCyp of 9 aa (amino acids 231–240, EPTEVQRYW in A.trivirgatus) encoded by TRIM5 exon 7. This sequence is pre-dicted to encode an �-helix, which is postulated to play animportant role in maintaining the restriction function ofTRIM5� (14, 16, 17). The other notable feature is the region

immediately upstream of the CypA coding sequence. Alignmentof the nucleotide sequences in this region with the CypA genefrom H. sapiens (GenBank accession no. NM�021130) or from M.mulatta chromosome 14 (GenBank accession no.NW�001100384.1, nucleotides 666274–676274) (18, 19) indicatesthat this junction sequence is most likely derived from thesequences upstream of the CypA coding regions in their respec-tive species, similar to what has been postulated for A. trivirgatusTRIMCyp (11, 12).

In contrast to the TRIM5-derived sequences, the CypA do-main is well conserved between Old World and New Worldprimate TRIMCyp. We compared these TRIMCyp CypA do-mains with the crystal structure of CypA complexed with 25-merof HIV-1 Gag polyprotein, which includes the CypA bindingloop (20). This crystal structure identifies two primary interact-ing regions on CypA equivalent to amino acids 357–375 andamino acids 404–428 in TRIMCyp. The C-terminal interactingregion of the CypA domain is conserved among the threespecies. There are four amino acid differences in the N-terminalinteracting region of the CypA domain. However, all of theresidues that are postulated to interact directly with Gag (20, 21)are conserved in all three primate species (Fig. 2B). Comparisonof CypA sequences in M. fascicularis TRIMCyp identifies dif-ferences at two amino acid residues, 356 (H/R) and 399 (I/T)(Figs. 2B and 3D), which most likely represents intraspeciespolymorphism because these sequences were derived from dif-ferent animals.

Neither M. nemestrina TRIMCyp nor M. fascicularis TRIMCyp RestrictHIV-1 in Vitro. To evaluate the ability of these Old World primateTRIMCyp proteins to restrict HIV-1, TRIMCyp from M. nem-estrina and M. fascicularis was N-terminally HA-tagged andstably transduced into CrFK cells. Expression and intracellularlocalization of each HA-tagged TRIMCyp isoform was exam-ined by immunoblotting, f low cytometry, and immunofluores-cence analyses. Immunoblotting analysis confirmed TRIMCypexpression at similar levels for each isoform tested (Fig. 4A).Flow cytometry confirmed that the majority of CrFK cellsexpress the transduced gene, and immunofluorescence analysisestablished that each TRIMCyp isoform was expressed in thecytoplasm (Fig. 4B).

We infected these TRIMCyp-expressing CrFK cells withHIV-1 pNL4–3�env-eGFP or env-minus SIVmac239 enhancedGFP (eGFP) pseudotyped with vesicular stomatitis virus(VSV)-G. GFP-positive CrFK cells were quantified 48 h postin-fection by flow cytometry analysis. As expected, A. trivirgatusTRIMCyp inhibited HIV-1 replication in vitro �100-fold. How-ever, neither M. nemestrina TRIMCyp nor M. fascicularis TRIM-Cyp inhibited the replication of HIV-1 in these cell lines (Fig.5A). As expected, none of the TRIMCyp isoforms tested re-stricted SIVmac239 in vitro (Fig. 5B).

DiscussionWe report the expression of TRIMCyp molecules in two OldWorld primates, M. nemestrina and M. fascicularis. Previously,the only published description of TRIMCyp expression was fromthe Aotus genus of New World primates. Subsequent to ourinitial finding, we became aware of M. nemestrina TRIMCypcDNA sequences deposited in GenBank by Su and colleagues(GenBank accession nos. DQ308404–DQ308406). In this com-munication, we corroborate the expression of TRIMCyp notonly in M. nemestrina but also in M. fascicularis. We also examinethe genetic structure of the TRIMCyp transcript and the func-tionality of TRIMCyp proteins from both species. Unlike NewWorld primates, in which TRIMCyp arose through CypA retro-transposition into intron 7 of the TRIM5 gene, the CypA domainin Old World primates resulted from retrotransposition into the3� UTR of the TRIM5 locus. The discovery of TRIMCyp in these

Fig. 1. RT-PCR amplification of TRIM5� and TRIMCyp from Old Worldprimate total RNA. Total RNA was isolated from Owl monkey kidney (OMK)cells, M. mulatta PBMC, M. fascicularis PBMC, and M. nemestrina PBMC.Primers were designed to amplify either TRIM5� (odd lanes) or TRIMCyp (evenlanes). Lane L, kb� ladder; lane 1, dH2O TRIM5�; lane 2, dH2O TRIMCyp; lane3, OMK TRIM5�; lane 4, OMK TRIMCyp; lane 5, M. mulatta TRIM5�; lane 6,M. mulatta TRIMCyp; lane 7, M. fascicularis TRIM5�; lane 8, M. fascicularisTRIMCyp; lane 9, M. nemestrina TRIM5; and lane 10, M. nemestrina TRIMCyp.

3570 � www.pnas.org�cgi�doi�10.1073�pnas.0709511105 Brennan et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

020

animals raises the possibility that expression of these moleculesmay be more prevalent in primate species than originallythought.

Coexpression of TRIM5 and TRIMCyp transcripts in M.nemestrina and M. fascicularis most likely results from alter-native splicing. We previously reported that the M. nemestrinaTRIM5 gene contains a G-to-T SNP altering the canonicalsplice acceptor of intron 6. The use of alternative downstreamsplice acceptors results in two novel TRIM5 isoforms (14). Inthis article, we report that the same SNP exists in the intron6 splice acceptor of M. fascicularis TRIM5. This polymorphismtherefore appears to be more common among Old Worldmonkeys than previously thought, because the same G-to-TSNP also has been observed in a minority of M. mulatta (W.Johnson, personal communication). Alternative splicing also

may occur in animals that retain the canonical intron 6 spliceacceptor. Because this acceptor sequence (AG) is followed byanother AG dinucleotide in the exon 7 of most macaqueTRIM5 genes examined, the tandem repeat of AG dinucleo-tides may cause ambiguity in splice acceptor site utilization,resulting in alternative splicing. In support of this notion, a 2-ntdeletion has been identified in TRIM5� transcripts from M.mulatta that are homozygous in their AG intron 6 sequences(T. Kodama, personal communication).

In contrast to A. trivirgatus TRIMCyp, macaque counterpartsidentified in this study failed to restrict HIV-1. Alignment of thepredicted protein sequences identifies significant sequence di-versity between Old World and New World primate TRIMCyp,with most differences found in the TRIM5-derived sequence. Itremains to be determined whether and to what extent sequence

Fig. 2. Alignment of A. trivirgatus, M. nemestrina, and M. fascicularis TRIMCyp predicted amino acid sequences. (A) Schematic diagram (not to scale) of thegenomic organization of the TRIM5 and CypA loci on chromosome 14 (top). The shaded region indicates the exon 8 3� UTR. Exons encoded by each TRIMCypsplice variant are indicated below. R, RING; B, B-Box; CC, coiled coil. (B) Alignment of TRIMCyp exon 6 predicted amino acid sequences from M. nemestrina andM. fascicularis and the published A. trivirgatus TRIMCyp sequence. The N-terminal methionine is omitted in this alignment because the start codon was replacedwith a 5� HA epitope tag during cDNA amplification. The RING, B-Box, coiled coil, and CypA domains are indicated by labeled bars over the sequence. Aminoacid differences between M. nemestrina and M. fascicularis are indicated by bold letters.

Brennan et al. PNAS � March 4, 2008 � vol. 105 � no. 9 � 3571

MIC

ROBI

OLO

GY

SEE

COM

MEN

TARY

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

020

diversity in this region contributes to the difference in the abilityof New World and Old World primate TRIMCyp to restrictHIV-1. However, TRIMCyp chimerae containing non-TRIM5RBCC domains have been shown to restrict HIV-1 despite thelow homology between the RBCC domains of the chimerae andthe A. trivirgatus TRIMCyp (22), suggesting the structure of theRBCC domain may be more important than the primary se-quence. Old World primates also differ from A. trivirgatus by thelack of exon 7 in their TRIMCyp sequence. This exon ispredicted to encode an �-helix that may play an important rolein the structure and function of TRIM5� (16, 17). M. nemestrinaTRIM5� excludes exon 7 and is incapable of restricting HIV-1in vitro; however, a chimera composed of H. sapiens exons 2–7and M. nemestrina exon 8 acquires the ability to restrict HIV-1(14, 23). Similarly, the linker 2 region, which includes exon 7, alsohas been shown to contribute to the ability of CypA constructsto restrict HIV-1 (24) Therefore, the role of exon 7 sequences inmaintaining the functionality of TRIM5 and TRIMCyp proteinswarrants further examination.

Although we have yet to confirm the in vivo expression of OldWorld primate TRIM5 and TRIMCyp at the protein level, thepossibility of coexpression of these proteins in M. nemestrina andM. fascicularis may have interesting biological implications. Aprevious study has demonstrated that coexpression of NewWorld primate TRIMCyp and human TRIM5� preventsTRIM5�-mediated restriction of ‘‘N-tropic’’ strains of the mu-rine leukemia virus (N-MLV) (25). This inhibition requiresinteractions between the two proteins, suggesting heteromultim-erization disrupts the structure necessary for recognition ofHIV-1 capsid. For this reason, it will be of interest to determinewhether both TRIM5� and TRIMCyp are indeed expressed inM. fascicularis and, if so, whether coexpression of these mole-cules may inhibit or modulate TRIM5�-mediated restriction.

The functional significance of TRIMCyp in Old World pri-mates described here is unknown. Despite their failure to restrictHIV-1 or SIV replication, it remains to be seen whethermacaque TRIMCyp will restrict other retroviruses. Alterna-tively, Old World primate TRIMCyp may represent an evolu-tionary vestige resulting from ancient retroviral challenges,similar to a function recently ascribed to H. sapiens TRIM5�

(26). If so, the different strategies used by Old World and NewWorld primates to express TRIMCyp molecules may representan example of convergent evolution that resulted in competitiveadvantage for these very divergent primate species againstretroviral challenges. Further studies will be needed to elucidatethe evolutionary and functional role of these molecules.

Materials and MethodsAnimals. Three M. nemestrina, two M. fascicularis, and two M. mulatta housedat the Washington National Primate Research Center were used in this study.Whole blood was collected from these animals in heparinized Vacutainertubes in accordance with an Institutional Animal Care and Use Committeeapproved protocol.

RNA/DNA Samples. The TRIMCyp gene from M. nemestrina and M. fasciculariswas cloned and sequenced at genomic and cDNA levels. PBMC were isolatedfrom heparinized whole blood by Lymphoprep density centrifugation. TotalRNA was isolated from 1 � 107 PBMC by using the RNeasy Mini Kit (Qiagen)following the manufacturer’s suggested protocol. DNA was isolated from 5 �106 PBMC with the Puregene DNA Purification Kit (Gentra Systems) followingthe manufacturer’s suggested protocol.

PCR and Cloning: cDNA and Genomic DNA Amplification. Amplification ofTRIMCyp cDNA from total RNA was performed by using the SuperScript IIIOne-Step RT-PCR with Platinum Taq kit (Invitrogen) using primersXhoIHA5TRIM (14) and CypA*RMCSNotI (5�-GTATATGCGGCCGCTTATTC-GAGTTGTCCACAGTCAG-3�). Primers MneT5Aex51.23 (5�-GGTGTGGATG-GCATCATTAAAAG-3�) and CypA*RMCSNotI were used to amplify a fragmentof the TRIM5 gene from PBMC genomic DNA with Platinum PCR Super MixHigh Fidelity (error rate: 1.8 � 0.4 � 10�6; Invitrogen). Approximately 100 ngof total RNA or DNA from each macaque and 0.4 �M concentrations of eachprimer were used in each reaction. The RT-PCR assay was incubated at 55°C for30 min, then 94°C for 2 min, followed by 40 cycles of 94°C for 15 s, 60°C for 30 s,and 68°C for 90 s, then a 68°C hold for 5 min and storage at 4°C. The PCR assaywas incubated at 94°C for 2 min, followed by 30 cycles of 94°C for 30 s, 60°Cfor 1 min, and 68°C for 10 min, then a 68°C hold for 10 min and storage at 4°C.One microliter of undiluted PCR product from each amplification was clonedby using the Strataclone PCR cloning kit (Stratagene) following the manufac-turer’s suggested protocol.

Sequencing and Data Analysis: cDNA and Genomic Sequencing. Nine cDNAclones from three M. nemestrina, three cDNA clones from one M. fascicularis,and one genomic clone from each species were sequenced by the DNA

Fig. 3. Genomic characterization of Old World primate TRIMCyp. (A) PCR amplification of the TRIM5 gene from M. mulatta, M. nemestrina, and M. fascicularisPBMC genomic DNA using a forward primer specific for exon 6 and a reverse primer specific for exon 8. Lane 1, M. mulatta; lane 2, M. fascicularis; lane 3, M.nemestrina; and lane 4, kb� DNA ladder. (B) Sequence of the M. fascicularis TRIM5 gene from intron 6 to exon 8. The underlined sequence indicates the locationcorresponding to the site of CypA retrotransposition in A. trivirgatus. (C) PCR amplification of TRIM5 exon 8 and CypA. M. nemestrina and M. fascicularis PBMCgenomic DNA was PCR-amplified with a forward primer specific for TRIM5 exon 5 and a reverse primer specific for CypA. Lane 1, kb� DNA ladder; lane 2, dH2O;lane 3, M. fascicularis; and lane 4, M. nemestrina. (D) Sequence of the M. fascicularis TRIM5 3� UTR and retrotransposed CypA. Lowercase letters indicate the TRIM53� UTR. Italicized and underlined letters identify the conserved retrotransposition sequence. Capital letters identify the retrotransposed CypA sequence inTRIMCyp. Bold letters highlight the start and stop codons of the CypA coding sequence.

3572 � www.pnas.org�cgi�doi�10.1073�pnas.0709511105 Brennan et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

020

Sequencing and Gene Analysis Center (Department of Pharmaceutics, Univer-sity of Washington) using the Big Dye V 3.1 reaction mix (Applied Biosystems)on an Applied Biosystems 3100 Genetic Analyzer. Primers M13F, T583.100,T5F356.375, MneT5ex4F, MneT5ex6F, T5R217.198, MneT5ex4R, MneT5ex6R(14), MneT5CyF1 (5�-GTTCTTCGACATTGCCGTCG-3�), MneT5CyR1 (5�-TTCTGT-GAAAGCAGGAACCC-3�), and MneT5CyR2 (5�-CTCTCCTGAGCTACAGAAGG-3�) were used to sequence cDNA clones, and primers T5Aex51.23,MneT5Int5F2 (5�-CCTCTCTTGATATGTCTCAG-3�), MneT5ex6F, T5F1205.1224(5�-GCTTCCAACCTGATGCAATG-3�), MneT5ex8F, MneT5CyF1, MneT5CyF2,CypA*RMCSNotI, MneT5CyR1, MneT5CyR2, MneT5ex8R1, and MneT5ex6Rwere used to sequence genomic clones (14).

Sequencing products were assembled with Sequencher 4.1 (Gene Codes).The resulting genetic and predicted amino acid sequences were analyzed byusing Vector NTI Advance 10.1.1 (Invitrogen).

Stable Transduction of CrFK Cells with TRIMCyp Genes. CrFK and 293T cells werecultured in DMEM/10% FBS. TRIMCyp cDNA with a 5� HA tag was cloned intothe pLPCX retroviral expression vector. 293T cells were used to produceretroviral vectors as described in ref. 27. These supernatants were used toinfect CrFK cells, and 24 h after infection, 4 �g/ml puromycin was added to themedium to select for cells stably transduced with the gene of interest.

Immunoblotting. HA-tagged TRIMCyp proteins expressed by stably transducedCrFK cells were detected by Western blotting of whole-cell lysates. Cells werelysed in Laemlli-SDS sample buffer (Bio-Rad). Samples were normalized bytotal protein (OD280) concentration. The SDS/PAGE was performed on a4–12% NuPAGE denaturing gel (Invitrogen), and separated bands were trans-ferred onto a nitrocellulose membrane (Bio-Rad). The primary antibody for

detection of the HA tag was HA.11 (Covance). After the binding of primaryantibody, the membrane was incubated with a goat anti-mouse alkalinephosphatase-conjugated secondary antibody (Sigma–Aldrich).

Flow Cytometry and Immunofluorescence Assay. Expression and localization ofHA-tagged TRIMCyp proteins inside the expressing cells was analyzed bystaining with anti-HA Alexa Fluor 488 conjugate (Invitrogen). Stably trans-duced CrFK cells were permeabilized by using the BD Cytofix/Cytoperm Plusintracellular staining kit (Becton Dickinson). An aliquot of 106 cells was ana-lyzed by flow cytometry with a FACScan (Becton Dickinson) to enumerate cellsexpressing the HA-tagged TRIM5 protein. Another aliquot of cells was grownon glass slides, permeabilized, and stained as described above, then analyzedand photographed on an inverted microscope (Leica Microsystems) with agreen fluorescence filter set.

Infectivity Assays. CrFK cells stably transduced with TRIMCyp or empty vectorwere seeded in 6-well culture plates in DMEM, 10% FBS, and 4 �g/ml puro-mycin at a density of 2 � 105 cells per well. After 24 h, the cells were washedin PBS and infected with 3-fold dilutions of VSV-G pseudotyped HIV-1 pNL4–3or SIVmac239, prepared as described in ref. 28, in 1 ml of DMEM, 10% FBS, and5 �g/ml polybrene. Two hours postinfection, the virus-containing supernatantwas replaced with 2 ml of DMEM and 10% FBS. At 48 h postinfection, infectedcells expressing GFP were enumerated by flow cytometric analysis using aFACScalibur (Becton Dickinson) and analyzed by FlowJo software (TreeStar, Inc.).

Fig. 4. Expression of TRIMCyp by stably transformed CrFK cells. (A) Immu-noblotting analysis of HA-tagged TRIMCyp proteins in CrFK whole-cell lysates.Lane 1, Protein Plus ladder; lane 2, M. nemestrina TRIMCyp; lane 3, M. mulattaTRIM5�; and lane 4, M. fascicularis TRIMCyp. (B) Analysis of stably transducedCrFK cells by flow cytometry (Left) and immunofluorescence assays (Right).The stably transduced TRIMCyp protein is indicated below each image. Amonoclonal anti-HA Alexa Fluor conjugate 488 antibody was used in bothassays to label the HA-tagged TRIMCyp in each cell line. The percentages ofcells expressing M. nemestrina and M. fascicularis TRIMCyp are 85% and 94%,respectively. Fig. 5. Infectivity of HIV-1 and SIVmac in CrFK cells stably transduced with

TRIMCyp. Serial dilutions of VSV-G pseudotyped HIV-1 pNL4–3-�env-eGFP (A)or env-minus SIVmac239-eGFP (B) were used to infect stably transduced CrFKcells. Forty-eight hours after infection, the percentage of GFP-positive cellswas determined by flow cytometric analysis.

Brennan et al. PNAS � March 4, 2008 � vol. 105 � no. 9 � 3573

MIC

ROBI

OLO

GY

SEE

COM

MEN

TARY

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

020

ACKNOWLEDGMENTS. We thank Michael Emerman (Fred Hutchinson CancerResearch Center, Seattle, WA) for providing the A. trivirgatus TRIMCyp CrFKcells and plasmids mGP, pL-VSV-G, and pCMV tat; Welkin Johnson and RuchiNewman for their help with the establishment of stably transduced CrFK cells;

and David Evans (New England National Primate Research Center, Southbor-ough, MA) for providing the env-minus SIVmac239 eGFP plasmid. This workwas supported by National Institutes of Health Grants P51 RR00016 (to theWashington National Primate Research Center) and K08 AI061738 (to G.B.).

1. Sharp PM, et al. (2001) The origins of acquired immune deficiency syndrome viruses:Where and when? Philos Trans R Soc London B 356:867–876.

2. Stremlau M, et al. (2004) The cytoplasmic body component TRIM5� restricts HIV-1infection in Old World monkeys. Nature 427:848–853.

3. Reddy BA, Etkin LD, Freemont PS (1992) A novel zinc finger coiled-coil domain in afamily of nuclear proteins. Trends Biochem Sci 17:344–345.

4. Li Y, Li X, Stremlau M, Lee M, Sodroski J (2006) Removal of arginine 332 allows humanTRIM5� to bind human immunodeficiency virus capsids and to restrict infection. J Virol80:6738–6744.

5. Owens CM, et al. (2004) Binding and susceptibility to postentry restriction factors inmonkey cells are specified by distinct regions of the human immunodeficiency virustype 1 capsid. J Virol 78:5423–5437.

6. Owens CM, Yang PC, Gottlinger H, Sodroski J (2003) Human and simian immunodefi-ciency virus capsid proteins are major viral determinants of early, postentry replicationblocks in simian cells. J Virol 77:726–731.

7. Stremlau M, et al. (2006) Specific recognition and accelerated uncoating of retroviralcapsids by the TRIM5� restriction factor. Proc Natl Acad Sci USA 103:5514–5519.

8. Yap MW, Nisole S, Stoye JP (2005) A single amino acid change in the SPRY domain ofhuman TRIM5� leads to HIV-1 restriction. Curr Biol 15:73–78.

9. Ylinen LM, Keckesova Z, Wilson SJ, Ranasinghe S, Towers GJ (2005) Differentialrestriction of human immunodeficiency virus type 2 and simian immunodeficiencyvirus SIVmac by TRIM5� alleles. J Virol 79:11580–11587.

10. Song H, et al. (2007) A single amino acid of the human immunodeficiency virus type 2capsid affects its replication in the presence of cynomolgus monkey and humanTRIM5�. J Virol 81:7280–7285.

11. Nisole S, Lynch C, Stoye JP, Yap MW (2004) A Trim5-cyclophilin A fusion protein foundin owl monkey kidney cells can restrict HIV-1. Proc Natl Acad Sci USA 101:13324–13328.

12. Sayah DM, Sokolskaja E, Berthoux L, Luban J (2004) Cyclophilin A retrotranspositioninto TRIM5 explains owl monkey resistance to HIV-1. Nature 430:569–573.

13. Ribeiro IP, et al. (2005) Evolution of cyclophilin A, TRIMCyp retrotransposition in NewWorld primates. J Virol 79:14998–15003.

14. Brennan G, Kodama T, Kozyrev Y, Hu SL (2007) Novel TRIM5 isoforms expressed byMacaca nemestrina. J Virol 81:12210–12217.

15. Jurka J (1997) Sequence patterns indicate an enzymatic involvement in integration ofmammalian retroposons. Proc Natl Acad Sci USA 94:1872–1877.

16. Woo JS, et al. (2006) Structural and functional insights into the B30.2/SPRY domain.EMBO J 25:1353–1363.

17. Woo JS, Suh HY, Park SY, Oh BH (2006) Structural basis for protein recognition byB30.2/SPRY domains. Mol Cell 24:967–976.

18. Gibbs RA, et al. (2007) Evolutionary and biomedical insights from the rhesus macaquegenome. Science 316:222–234.

19. Hernandez RD, et al. (2007) Demographic histories and patterns of linkage disequi-librium in Chinese and Indian rhesus macaques. Science 316:240–243.

20. Zhao Y, Chen Y, Schutkowski M, Fischer G, Ke H (1997) Cyclophilin A complexed witha fragment of HIV-1 gag protein: Insights into HIV-1 infectious activity. Structure(London) 5:139–146.

21. Braaten D, Ansari H, Luban J (1997) The hydrophobic pocket of cyclophilin is thebinding site for the human immunodeficiency virus type 1 Gag polyprotein. J Virol71:2107–2113.

22. Yap MW, Dodding MP, Stoye JP (2006) Trim-cyclophilin A fusion proteins can restricthuman immunodeficiency virus type 1 infection at two distinct phases in the viral lifecycle. J Virol 80:4061–4067.

23. Ohkura S, Yap MW, Sheldon T, Stoye JP (2006) All three variable regions of the TRIM5�

B30.2 domain can contribute to the specificity of retrovirus restriction. J Virol 80:8554–8565.

24. Javanbakht H, et al. (2007) The ability of multimerized cyclophilin A to restrictretrovirus infection. Virology 367:19–29.

25. Berthoux L, Sebastian S, Sayah DM, Luban J (2005) Disruption of human TRIM5�

antiviral activity by nonhuman primate orthologues. J Virol 79:7883–7888.26. Kaiser SM, Malik HS, Emerman M (2007) Restriction of an extinct retrovirus by the

human TRIM5� antiviral protein. Science 316:1756–1758.27. Bartz SR, Vodicka MA (1997) Production of high-titer human immunodeficiency virus

type 1 pseudotyped with vesicular stomatitis virus glycoprotein. Methods 12:337–342.28. Reed SE, Staley EM, Mayginnes JP, Pintel DJ, Tullis GE (2006) Transfection of mamma-

lian cells using linear polyethylenimine is a simple and effective means of producingrecombinant adeno-associated virus vectors. J Virol Methods 138:85–98.

3574 � www.pnas.org�cgi�doi�10.1073�pnas.0709511105 Brennan et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

020