| |
New Protein That Stops HIV From Replicating
|
| |
| |
--Clue Found to AIDS Transmission to Humans
--Could lead to Gene Therapy
This past week there were several articles on the internet related to a protein identified in monkeys called Trim 5 may have altered genetically over years and that may be the reason why rhesus monkeys could not be HIV-infected. But after the gene changed humans could be infected.
This report contains the published reports in Current Biology and articles in February Nature discussing these findings. At the very end of my report is the original published research article from Current Biology by Jonathan Stoye.
"Scientists Find Clue to AIDS Origins, New Therapy"
Reuters Jan 10, 2005
Patricia Reaney A single difference between a human gene and the one in rhesus monkeys that blocks HIV infection in the animals could offer both insight into the origins of the AIDS epidemic and a potential gene therapy, said authors of a study published Monday. If humans had the same gene, "we probably would never have had AIDS. I believe it is a key change,' said Dr. Jonathan Stoye, head of virology at London's National Institute of Medical Research and an author of the study. In laboratory experiments, it is much more difficult to infect monkey cells with HIV than to infect human cells. Something differed in the animal cells to block infection. A gene called Trim 5 alpha was found to be the reason. In monkeys, but not in humans, it stops HIV from replicating. Stoye and colleagues found one specific change in a protein that was key to blocking HIV. In substituting a human protein with a monkey protein, the team found they could make human cells HIV-resistant. Stoye's team believes introducing the gene carrying that one change back into human cells would make those cells resistant to HIV. "What we are suggesting one might try doing is to purify HIV negative cells out of a patient who has been infected, introduce the gene with this one modification and then put them back into people," said Stoye. More laboratory research, followed by animal tests and human trials is necessary, he said. The full report, "A Single Amino Acid Change in the SPRY Domain of Human Trim5α Leads to HIV-1 Restriction," was published in Current Biology (2005;15:73-78). Published report in Current Biology is dense so I put it at the end of this report which contains more easily understood reports in Nature journal.
"...Although we do not know the sequence of the ancestral Trim5, we note that rhesus and African green monkey sequences contain a proline at this position, and these species are resistant to HIV-1. This implies that during evolution, changes at this position affecting sensitivity to HIV-1 infection may have occurred. Without such changes it seems unlikely that the cross-species infection(s) leading to the current AIDS epidemic [30] would have occurred..."
Nature Medicine
Restrictions on HIV
The identity of a protein that is key to understanding why humans, but not monkeys, are susceptible to HIV-1 is revealed by Matthew Stemlau et al. in the 26 February Nature. This protein, TRIM-5alpha, counteracts HIV-1 infection in macaques but allows HIV-1 infection in humans.
TRIM-5alpha is encoded by the gene Lv1, which was shown several years ago to shield monkey cells from infection. Until now, no one knew what Lv1 actually encoded. To find out, Stemlau et al. took a very straightforward approach: they transformed cells with a monkey cDNA library and selected for cells that were resistant to HIV-1 based vectors. Out came TRIM-5alpha.
Not much is known about TRIM proteins, but there are at least 36 of them in humans, many of which localize to discrete cellular compartments. TRIM-5alpha, for instance, forms speckles in the cytoplasm—consistent with the view that it could hijack HIV-1 before it enters the nucleus to undergo reverse transcription. Many other mechanisms could come into play: for instance, one well-known TRIM family member, PML, adds a ubiquitin-like moiety, SUMO, to target proteins. However it works, TRIM-5alpha appears to target the capsid protein of HIV-1.
ORIGINAL ARTICLE
HIV: Replication trimmed back
Nature 427, 791 - 793 (26 February 2004)
STEPHEN P. GOFF
Stephen P. Goff is at the Howard Hughes Medical Institute, Department of Biochemistry and Molecular Biophysics, Columbia University College of Physicians and Surgeons, New York 10032, USA.
A long-standing point of intrigue has been how certain non-human primates are resistant to HIV-1. The discovery in macaque monkeys of a protein that resides in mysterious cytoplasmic bodies holds the key.
Mammals have lived with retroviruses throughout their history. To avoid the detrimental effects of these RNA viruses, mammals have evolved a large number of genes to inhibit their replication. Certain non-human primates strongly 'restrict' HIV-1, blocking infection soon after this retrovirus enters the cell and before it starts making DNA from its RNA template during reverse transcription. Sodroski and colleagues (page 848 of this issue1) have now isolated a gene, known as TRIM5alpha, that is responsible for this restriction in rhesus macaques. The work uncovers key players in the early steps of virus replication, and may lead to new approaches towards inhibiting HIV-1.
The prototypical gene blocking early events in the retroviral life cycle, the Fv1 locus, was first characterized in the 1970s. Fv1 blocks particular mouse leukaemia viruses (MuLVs) soon after reverse transcription. Whether or not a virus falls victim to Fv1 depends on its coat protein; in particular, a single residue at position 110 of the MuLV capsid coat determines this sensitivity. An important feature of the block is that it can be overcome if the virus is present in extremely high numbers. Fv1 was found to correspond to the capsid gene of a virus resident in the mouse genome, suggesting that it might compete with the incoming virus for critical host targets2. How Fv1 acts remains unclear, but there must be a specific interaction with the capsid protein of the incoming virus. Subsequently, many human cell lines have also been found to be resistant to particular MuLVs, an activity conferred by a gene named Ref1 (refs 3, 4). Remarkably, the capsid is also targeted, and virus sensitivity is again controlled by residue 110. The Ref1 gene of humans has not yet been isolated.
Excitement in the field increased with the realization that restriction was not limited to mouse viruses, but that many primates, including rhesus macaques and owl monkeys, restrict HIV-1 (refs 3, 5--9). The gene responsible, dubbed Lv1 (ref. 7), in some species restricted HIV-1 but not the related monkey virus SIVMAC. From a comparison of viruses containing portions of both HIV-1 and SIV genomes, the target of Lv1 was pinpointed to be the HIV-1 capsid protein10, 11. Sodroski and colleagues1 have now isolated the gene responsible from a rhesus monkey complementary DNA library by selecting for cells that manifest resistance to HIV-based vectors. The success of this approach is testimony to the power of direct genetic selections. The active cDNA encodes TRIM5alpha.
TRIM5alpha, a protein resulting from differential processing of TRIM5 transcripts, passes all the tests as the main HIV-1 resistance factor present in primates1. First, expression of the cDNA in human cells renders them resistant to HIV-1 infection, and the induced block occurs at the right stage of the viral life cycle. The rhesus TRIM5alpha cDNA has potent antiviral activity, but its human counterpart does not, despite both proteins being expressed well. And whereas the rhesus gene product is active against HIV-1, it is only very weakly active against SIVMAC, and not at all active against MuLVs. Furthermore, TRIM5alpha blocks versions of the viruses that contain both macaque and human sequences — but only if the HIV-1 capsid region is present. Thus TRIM5alpha targets the HIV-1 capsid protein, as expected. Finally, 'knock-down' of TRIM5alpha, using small interfering RNA molecules that specifically target TRIM5alpha transcripts for destruction, relieved the block against HIV-1 in rhesus cells, but had no effect on MuLV infection.
What is known about TRIM proteins? Not much. They are defined by a cluster of three different protein motifs: a RING motif, which is rich in cysteines and binds zinc; one or two so-called B boxes, which also bind zinc; and a coiled-coil domain that is probably involved in the formation of protein complexes. All individual TRIM proteins can probably aggregate with their peers. Some might also form alliances with other TRIM proteins.
A look through the human genome reveals at least 37 TRIM family members12. The most famous member is PML, which can fuse with the retinoic acid receptor to cause some forms of leukaemia. PML also shows antiviral activity against many viruses. It resides in mysterious nuclear bodies of uncertain function, to which it may localize other proteins and selected messenger RNAs13. Each of the various TRIM proteins seems to localize to particular compartments within cells, forming discrete structures to which they entice other proteins. TRIM5 forms cytoplasmic speckles12, consistent with the fact that it blocks HIV-1 infection before reverse transcription.
How might TRIM5alpha block an incoming virus in the rhesus macaques? The work by Sodroski and colleagues1 offers a few possible explanations. Infection of human cells by HIV-1 induces PML to move into the cytoplasm and bind the virus14; the ability of TRIM proteins to aggregate into large structures and to attract other proteins into them suggests that TRIM5alpha might be binding and trapping the incoming virus in the rhesus macaques. Another possibility is that TRIM5alpha could influence whether, and how, the capsid is modified, and thus its localization. One such modification, reported for PML, is the addition of a 'SUMO' group. There are hints that the capsid proteins of MuLV and HIV-1 are indeed SUMOylated, and that this may be required for infection. Alternatively, TRIM5alpha might interfere with the normal uncoating of the viral RNA core that precedes reverse transcription. This could involve interactions with cyclophilin A, a protein thought to affect the stability of the HIV-1 core. Indeed, there are strong hints that HIV-1 restriction is modulated by cyclophilin A (refs 10, 15).
A final possibility is that TRIM5alpha initiates the degradation of the incoming HIV-1 particle. Other TRIM5 isoforms, including TRIM5delta, behave as ubiquitin ligases16 — they add ubiquitin to other proteins, which labels them for destruction by the proteasome, the cell's disposal system. Perhaps TRIM5alpha could transfer ubiquitin directly to the capsid. In support of this idea, treating human cells with proteasome inhibitors increased their susceptibility to HIV-1 infection17. Time will tell which of these ideas, if any, pan out.
Several straightforward questions arise. Does TRIM5alpha bind directly to the HIV-1 capsid protein? Does it cause its relocalization, modification or degradation? Does TRIM5alpha comprise a subunit of a ubiquitin ligase — or of a similarly acting SUMO transferase? Is reverse transcription itself blocked by TRIM5alpha, or is the resultant DNA degraded after synthesis? To what else might TRIM5alpha bind? Are TRIM proteins somehow involved in the mechanism of action of the Fv1, or the human Ref1, resistance genes?
All we can say for certain is that right now many laboratories are working furiously to be the first to find the answers to these questions. When the mechanism of action of TRIM5alpha is uncovered, the next goal will be to recreate its effects in a therapeutic treatment.
References
1. Stremlau, M. et al. Nature 427, 848--853 (2004).
2. Best, S., Le Tissier, P., Towers, G. & Stoye, J. Nature 382, 826--829 (1996).
3. Towers, G. et al. Proc. Natl Acad. Sci. USA 97, 12295--12299 (2000).
4. Besnier, C. et al. J. Virol. 77, 13403--13406 (2003).
5. Hofmann, W. et al. J. Virol. 73, 10020--10028 (1999).
6. Munk, C., Brandt, S. M., Lucero, G. & Landau, N. R. Proc. Natl Acad. Sci. USA 99, 13843--13848 (2002).
7. Cowan, S. et al. Proc. Natl Acad. Sci. USA 99, 11914--11919 (2002).
8. Besnier, C., Takeuchi, Y. & Towers, G. Proc. Natl Acad. Sci. USA 99, 11920--11925 (2002).
9. Hatziioannou, T., Cowan, S., Goff, S. P., Bieniasz, P. D. & Towers, G. J. EMBO J. 22, 385--394 (2003).
10. Dorfman, T. & Gottlinger, H. G. J. Virol. 70, 5751--5757 (1996).
11. Owens, C. M. et al. J. Virol. 77, 726--731 (2003).
12. Reymond, A. et al. EMBO J. 20, 2140--2151 (2001).
13. Borden, K. L. Mol. Cell. Biol. 22, 5259--5269 (2002).
14. Turelli, P. et al. Mol. Cell 7, 1245--1254 (2001).
15. Towers, G. J. et al. Nature Med. 9, 1138--1143 (2003).
16. Xu, L. et al. Exp. Cell Res. 288, 84--93 (2003).
17. Schwartz, O. et al. J. Virol. 72, 3845--3850 (1998).
The cytoplasmic body component TRIM5alpha restricts HIV-1 infection in Old World monkeys
Letters to Nature
Nature 427, 848 - 853 (26 February 2004)
MATTHEW STREMLAU1, CHRISTOPHER M. OWENS1, MICHEL J. PERRON1, MICHAEL KIESSLING1, PATRICK AUTISSIER2 & JOSEPH SODROSKI1,3
1 Department of Cancer Immunology and AIDS, Dana-Farber Cancer Institute, Department of Pathology, Division of AIDS, Harvard Medical School, Boston, Massachusetts 02115, USA
2 Division of Viral Pathogenesis, Beth Israel Deaconess Medical Center, Department of Medicine, Division of AIDS, Harvard Medical School, Boston, Massachusetts 02115, USA
3 Department of Immunology and Infectious Diseases, Harvard School of Public Health, Boston, Massachusetts 02115, USA
Host cell barriers to the early phase of immunodeficiency virus replication explain the current distribution of these viruses among human and non-human primate species1-4. Human immunodeficiency virus type 1 (HIV-1), the cause of acquired immunodeficiency syndrome (AIDS) in humans, efficiently enters the cells of Old World monkeys but encounters a block before reverse transcription2-4. This species-specific restriction acts on the incoming HIV-1 capsid5-7 and is mediated by a dominant repressive factor7-9. Here we identify TRIM5alpha, a component of cytoplasmic bodies, as the blocking factor. HIV-1 infection is restricted more efficiently by rhesus monkey TRIM5alpha than by human TRIM5alpha. The simian immunodeficiency virus, which naturally infects Old World monkeys10, is less susceptible to the TRIM5alpha-mediated block than is HIV-1, and this difference in susceptibility is due to the viral capsid. The early block to HIV-1 infection in monkey cells is relieved by interference with TRIM5alpha expression. Our studies identify TRIM5alpha as a species-specific mediator of innate cellular resistance to HIV-1 and reveal host cell components that modulate the uncoating of a retroviral capsid.
Recombinant HIV-1 expressing green fluorescent protein and pseudotyped with the vesicular stomatitis virus (VSV) G glycoprotein (denoted HIV-1--GFP) can efficiently infect the cells of many mammalian species including humans, but not those of Old World monkeys4-9. Here we used a murine leukaemia virus vector to transduce human HeLa cells, which are susceptible to HIV-1--GFP infection, with a complementary DNA library prepared from primary rhesus monkey lung fibroblasts (PRL cells). Two independent HeLa clones resistant to HIV-1--GFP infection, but susceptible to infection with recombinant simian immunodeficiency virus (SIV--GFP) or murine leukaemia virus (MLV--GFP), were identified in a screen (Methods).
The only monkey cDNA insert common to both HIV-1--GFP-resistant clones was predicted to encode TRIM5alpha, a member of the tripartite motif (TRIM) family of proteins containing RING domains, B-boxes and coiled coils11-13. TRIM5alpha also contains a carboxy-terminal B30.2 (SPRY) domain not found in the other TRIM5 isoforms (ref. 13 and Fig. 1a). The natural functions of TRIM5alpha, or of the cytoplasmic bodies in which the TRIM5 proteins localize13, 14, are unknown. One TRIM5 isoform has been shown to have ubiquitin ligase activity typical of RING-containing proteins14. TRIM5 proteins are expressed constitutively in many tissues13, consistent with the pattern of expression expected for the HIV-1-blocking factor in monkeys4.
HeLa cells stably expressing rhesus monkey TRIM5alpha (TRIM5alpharh) and control HeLa cells containing empty vector were incubated with different amounts of recombinant HIV-1--GFP, SIV--GFP and MLV--GFP. Expression of TRIM5alpharh resulted in a marked inhibition of infection by HIV-1--GFP, whereas MLV--GFP infected control and TRIM5alpharh-expressing HeLa cells equivalently (Fig. 1b, c). TRIM5alpharh inhibited infection by SIV--GFP less efficiently than that by HIV-1--GFP (Fig. 1c). Stable TRIM5alpharh expression also inhibited the replication of infectious HIV-1 in HeLa-CD4 cells, which express the receptors for HIV-1 (ref. 15 and Fig. 1d). The replication of a simian--human immunodeficiency virus (SHIV) chimaera, which contains core proteins (including the capsid protein) of SIVmac (ref. 16), was not inhibited in these TRIM5alpharh-expressing cells. When the infections were done with eightfold more HIV-1 and SHIV, similar results were obtained (Supplementary Information). We conclude that expression of TRIM5alpharh specifically and efficiently blocks infection by HIV-1, and exerts a slight inhibitory effect on infection by SIVmac.
To investigate the viral target of the TRIM5alpharh-mediated restriction, HeLa cells expressing TRIM5alpharh or control HeLa cells were incubated with recombinant HIV-1--GFP, SIV--GFP, SIV(HCA-p2)--GFP or HIV(SCA)--GFP. SIV(HCA-p2)--GFP is identical to SIV--GFP, except that the SIV capsid and adjacent p2 sequences have been replaced by those of HIV-1 (ref. 17), and SIV(HCA-p2)--GFP has been shown to be susceptible to the block in Old World monkey cells5, 7. HIV(SCA)--GFP is identical to HIV-1--GFP, except that most of the capsid protein has been replaced by that of SIV5, and HIV(SCA)--GFP has been shown to be less susceptible than HIV-1 to the block in Old World monkey cells5. We found that HIV-1--GFP and SIV(HCA-p2)--GFP infections were restricted to the same extent in TRIM5alpharh-expressing HeLa cells, whereas infections by SIV--GFP and HIV(SCA)--GFP were less restricted in these cells (Fig. 1e). We conclude that capsid sequences influence viral susceptibility to the TRIM5alpharh-mediated restriction.
To determine the level at which HIV-1 infection is blocked by TRIM5alpharh, we used a real-time polymerase chain reaction (PCR) assay to detect viral cDNA at various times after incubating VSV G-pseudotyped HIV-1 with either HeLa cells expressing TRIM5alpharh or control HeLa cells (Fig. 2). After 2 h, the levels of early reverse transcripts were comparable in TRIM5alpharh-expressing and control HeLa cells. At later time points, both early and late reverse transcripts were barely detectable in TRIM5alpharh-expressing cells; by contrast, both early and late viral cDNAs were abundant in control cells. As the VSV G glycoproteins support entry into these cells equivalently (see MLV--GFP infection in Fig. 1b, c), these data indicate that early events after HIV-1 entry are impaired in cells expressing TRIM5alpharh. Directly or indirectly, TRIM5alpharh disrupts viral cDNA synthesis or accelerates its decay.
The predicted sequences (Fig. 1a) of TRIM5alpharh and its human orthologue, TRIM5alphahu, show differences that might account for the species-specific nature of the early block to HIV-1 infection. To test this hypothesis, TRIM5alpharh--HA and TRIM5alphahu--HA proteins, containing C-terminal epitope tags from influenza haemagglutinin, were expressed stably in HeLa cells. Similar levels of TRIM5alpharh--HA and TRIM5alphahu--HA were expressed in HeLa cells (Fig. 3a). Neither TRIM5alpharh--HA nor TRIM5alphahu--HA affected the efficiency with which HeLa cells were infected by MLV--GFP (Fig. 3b). TRIM5alphahu--HA inhibited HIV-1--GFP infection less efficiently than did TRIM5alpharh--HA. Parallel experiments with TRIM5alpha proteins lacking the epitope tags verified that fusion of HA to the C terminus of TRIM5alpharh and TRIM5alphahu did not affect the efficiency with which these proteins suppressed HIV-1 infection (Supplementary Information). By contrast, the modest inhibitory effect of TRIM5alpharh on SIV--GFP infection was not seen for the epitope-tagged TRIM5alpharh--HA protein (Fig. 3b). Neither TRIM5alphahu nor TRIM5alphahu--HA exerted significant effects on SIV--GFP infection. Thus, differences in the TRIM5alpha proteins of humans and monkeys probably contribute to the species-specific differences in susceptibility to HIV-1 infection.
We examined the ability of rhesus monkey TRIM5 variants to inhibit HIV-1 infection. Differential splicing of TRIM5 transcripts results in the production of several isoforms that lack the TRIM5alpha C terminus13. For example, the first 300 amino acid residues of the TRIM5gammarh isoform are identical to those of TRIM5alpharh, but the TRIM5gammarh sequence diverges and stops thereafter. To examine the contribution of the C-terminal B30.2 (SPRY) domain of TRIM5alpharh to the inhibition of HIV-1 infection, we tested an epitope-tagged version of the TRIM5gammarh isoform. We also examined whether an intact RING domain is required for TRIM5alpharh-mediated inhibition of HIV-1 infection. Studies of proteins containing RING domains, including TRIM5delta, indicate that alteration of one or both of the two most amino-terminal cysteines inactivates ubiquitin ligase activity14, 18. Thus, we created and tested two HA-tagged TRIM5alpharh mutants, TRIM5alpharh-C15A and TRIM5alpharh-C15A/C18A. Expression of TRIM5gammarh and the RING domain cysteine mutants was comparable to that of wild-type TRIM5alpharh (Fig. 3c). TRIM5gammarh did not inhibit HIV-1--GFP infection and, compared with wild-type TRIM5alpharh, TRIM5alpharh-C15A and TRIM5alpharh-C15A/C18A were significantly less effective in inhibiting HIV-1--GFP infection (Fig. 3d). We conclude that both the RING domain at the N terminus and the B30.2 (SPRY) domain at the C terminus of TRIM5alpharh contribute to the HIV-1-inhibitory activity of this protein.
The TRIM5gammarh, TRIM5alpharh-C15A and TRIM5alpharh-C15A/C18A variants were also examined for dominant-negative activity in relieving the block to HIV-1--GFP infection in PRL cells. We found that HIV-1--GFP infected PRL cells expressing TRIM5gammarh more efficiently than it infected control PRL cells transduced with empty vector or PRL cells expressing the RING domain cysteine mutants of TRIM5alpharh (Fig. 3e). Thus, TRIM5gammarh acts in a dominant-negative manner to suppress the restriction to HIV-1 in PRL cells.
To determine whether TRIM5alpharh is necessary for the restriction to HIV-1 infection in Old World monkey cells, short interfering RNAs (siRNAs) directed against TRIM5alpharh were used to downregulate its expression. Several different siRNAs directed against TRIM5alpharh, including siRNA 1 and siRNA 3 (Methods), reduced TRIM5alpharh--HA expression in HeLa cells (Fig. 4a and data not shown). As expected, siRNA 3 did not inhibit the expression of TRIM5alpharh-escape, a wild-type HA-tagged protein encoded by an expression vector with six silent mutations in the siRNA 3 target sequence (Fig. 4a). Transfection of PRL cells with the TRIM5alpharh-directed siRNAs, but not with a control siRNA, resulted in a marked increase in the efficiency of HIV-1--GFP infection (Fig. 4b--d). The siRNAs targeting TRIM5alpharh did not affect MLV--GFP infection (Fig. 4c, d, and data not shown). The increase observed after transfection of siRNA 3 was diminished when the PRL cells were transduced with a vector expressing TRIM5alpharh-escape (Fig. 4c, d). Although the restriction to HIV-1 infection is not as strong in LLC-MK2 cells as it is in PRL cells, siRNA directed against TRIM5alpharh also increased the efficiency of HIV-1--GFP infection in LLC-MK2 cells (Supplementary Information). We conclude that TRIM5alpharh is an essential factor for the early block to HIV-1 in Old World monkey cells.
The maintenance of a strong block to HIV-1 in Old World monkeys implies a selective advantage, presumably imposed by the presence of HIV-1-like viruses during the evolution of this primate lineage. Here, TRIM5alpharh has been identified as a key mediator of the monkey cell restriction to HIV-1. Expression of TRIM5alpharh was necessary for the early block to HIV-1 in monkey cells and sufficient for establishing such a block in human cells. Differences in the expression level, or polymorphisms in TRIM5alpharh or its associated cofactors, may account for the different degrees of HIV-1 restriction observed in several established Old World monkey cell lines4.
The mechanism of the TRIM5alpharh-mediated block to HIV-1 infection requires further investigation, but the importance of the TRIM5alpharh RING domain and the capsid specificity of the restriction indicate that TRIM5alpharh may directly bind and ubiquitinate the HIV-1 capsid. HIV-1 capsid mutants19 that bind the monkey cell restriction factor efficiently, but still infect monkey cells, might be resistant to the downstream effects of TRIM5alpha binding, including ubiquitination. Capsid modification by ubiquitin could adversely affect uncoating, which occurs soon after HIV-1 entry20. Studies of HIV-1 Gag mutants suggest that capsid disassembly shows precise requirements; that is, both increases and decreases in capsid stability are detrimental to HIV-1 replication21.
Although human and Old World monkey cells are susceptible to infection by HIV-1 and SIVmac, respectively, the TRIM5alpha proteins from these cells showed some ability to repress the infecting virus. Variations in expression or polymorphisms in TRIM5alpha may thus influence the course of natural infection by these viruses. The treatment of human target cells with proteasome inhibitors results in an increase in the early phase of HIV-1 infection22, which suggests that HIV-1-suppressive processes involving ubiquitination may be operative in human cells. TRIM5alphahu was less effective in suppressing HIV-1 and SIVmac infection than was TRIM5alpharh. Notably, it has been suggested that HIV-1 capsids bind the Old World monkey restriction factor more efficiently than do SIVmac capsids7-9. Apparently, each virus has evolved in its natural host to achieve an acceptably low level of TRIM5alpha interaction. Vigorous, detrimental capsid disassembly may result when efficiently binding capsids, like those of HIV-1, encounter more effective TRIM5alpha proteins, like those expressed in simian cells.
The TRIM proteins constitute a large family whose members exhibit diverse functions and cellular locations13. A nuclear TRIM protein, PML, has been reported to inhibit the replication or expression of various viruses, including HIV-1 (refs 23--26). The function of cytoplasmic bodies, which contain a subset of TRIM family members13, is unknown; our results suggest that cytoplasmic bodies may contribute to innate cellular resistance to viruses. The induction of expression of some TRIM proteins26-28 by interferon and the expansion of this gene family in parallel with the metazoan lineage are consistent with this idea.
Understanding the early species-specific restrictions to HIV-1 replication, in conjunction with the late block owing to APOBEC3G29, may suggest approaches to the development of animal models of HIV-1 infection. Furthermore, insight into the uncoating process, hitherto a poorly understood aspect of the retroviral life cycle, should facilitate intervention.
Methods
Screen for HIV-1-resistant cells A PRL4 cDNA library (3.2 times 106 independent clones) was inserted into the pLIB vector (Clontech) and used to transduce 3 times 106 HeLa cells. After 3 d, 6 times 106 transduced cells were reseeded in batches of 5 times 105 cells in 10-cm dishes and incubated with sufficient HIV-1--GFP to infect at least 99% of the cells. About 0.5% of the cells were selected for absence of fluorescence by a FACS Vantage SE cell sorter (Becton Dickinson). Collected cells were allowed to grow into 500-cell colonies and subjected to a second round of HIV-1--GFP infection at a high multiplicity of infection. GFP-negative colonies were identified by fluorescence microscopy, cloned and expanded.
We selected 313 HeLa clones from seven sequential screens and tested them for susceptibility to HIV-1--GFP and SIV--GFP. Two clones with a selective block to HIV-1--GFP were identified. Eleven cDNA inserts from these two clones were recovered by PCR amplification of genomic DNA samples with oligonucleotide primers specific for the pLIB vector and the following conditions: 1 cycle at 95 °C for 1 min, 40 cycles of 95 °C for 1 min, 68 °C for 1 min and 72 °C for 5 min, and then 1 cycle at 72 °C for 10 min. Each cDNA was subcloned into the EcoRI and ClaI restriction sites of the pLPCX vector (Clontech) and sequenced. The cDNA encoding TRIM5alpharh was the only monkey cDNA present in both HIV-1-resistant HeLa clones and was the only cDNA subsequently confirmed to inhibit HIV-1 infection.
Creation of cells stably expressing TRIM5alpha variants We co-transfected pLPCX vectors containing TRIM5 cDNAs or control empty pLPCX vectors into 293T cells by using pVPack-GP and pVPack-VSV-G (both from Stratagene) packaging plasmids. The resulting virus particles were used to transduce 5 times 105 HeLa cells in the presence of 5 µg ml-1 polybrene, and the cells were subjected to selection in 1 µg ml-1 puromycin (Sigma). Cells were transduced with either a pLPCX--TRIM5alpharh(fl) vector containing full-length TRIM5alpharh cDNA, which includes the 5'- and 3'-untranslated regions (UTRs), or a pLPCX--TRIM5alpharh(cds) vector containing only the amino acid coding sequence of the TRIM5alpharh cDNA.
Infection with viruses expressing GFP We prepared HIV-1--GFP, SIV--GFP, HIV(SCA)--GFP and SIV(HCA-p2)--GFP viruses as described4, 5, 17, and MLV--GFP by co-transfecting 293T cells with 15 µg of pFB-hrGFP, 15 µg of pVPack-GP and 4 µg of pVPack-VSV-G (all from Stratagene). HIV and SIV viral stocks were quantified by measuring reverse transcriptase (RT) activity as described16. MLV RT activity was determined by the same procedure except that 20 mM MnCl2 was used instead of MgCl2. For infections, 3 times 104 HeLa cells or 2 times 104 PRL cells seeded in 24-well plates were incubated in the presence of virus for 24 h. Cells were washed and returned to culture for 48 h, and then subjected to FACS analysis with a FACScan (Becton Dickinson).
Infection with replication-competent virus Stocks of replication-competent HIV-1HXBc2 and SHIVHXBc2 (ref. 16) were prepared from supernatants of 293T cells transfected with the respective proviral clones. Spreading infections were initiated with stocks normalized according to RT activity, and replication was monitored over time by analysing culture supernatants for RT activity.
Quantitative real-time PCR Virus stocks derived from the transfection of 293T cells were treated with 50 U ml-1 Turbo DNase (Ambion) for 60 min at 37 °C. Cells (2 times 105) were infected with 2 times 104 RT units of VSV G-pseudotyped HIV-1--GFP or control HIV-1--GFP lacking envelope glycoproteins, and genomic DNA was isolated at various time points (0--48 h). We quantified early HIV-1 reverse transcripts with primers ert2f and ert2r and the ERT2 probe8, and late HIV-1 reverse transcripts with primers MH531 and MH532 and the probe LRT-P30 as described. Reaction mixtures contained Taqman universal master mix (PE Biosystems), 300 nM primers, 100 nM probe and 500 ng of genomic DNA. The PCR conditions were 2 min at 50 °C and 10 min at 95 °C, followed by 40 cycles of 15 s at 95 °C and 1 min at 60 °C, on an ABI Prism 7700 (Applied Biosystems).
Cloning of TRIM5 isoforms and TRIM5alpharh mutagenesis The human TRIM5alpha open reading frame (ORF) was amplified from a kidney cDNA library (Clontech), using primers derived from National Center for Biotechnology Information (NCBI) RefSeq NM_033034 and inserted into pLPCX (Clontech) to create pTRIM5alphahu. The predicted sequence of this TRIM5alphahu differs in three amino acid residues from NCBI RefSeq NM_033034, which is derived from the TRIM5alpha of the retinoic-acid-induced NT2 neuronal precursor line. We amplified the rhesus monkey TRIM5gamma ORF from the PRL cDNA library with primers derived from NCBI RefSeq NM_033092 and inserted it into pLPCX to create pTRIM5gammarh. In the pTRIM5alpharh--HA and pTRIM5alphahu--HA plasmids, an in-frame sequence encoding the influenza virus HA epitope tag was included at the 3' end of the TRIM5alpharh and TRIM5alphahu sequences, respectively. The pTRIM5gammarh--HA plasmid contains a sequence encoding the initiator methionine and the HA epitope tag at the 5' end of the TRIM5gammarh sequence. Cysteine to alanine changes (C15A and C15A/C18A) were introduced into the TRIM5alpharh RING domain by using the Quick-change mutagenesis kit (Stratagene).
Immunoblotting HA-tagged proteins were expressed in HeLa cells by transfection with Lipofectamine 2000 (Invitrogen) or by transduction with pLPCX vectors as described above. The HA-tagged proteins were detected in whole-cell lysates (100 mM (NH4)2SO4, 20 mM Tris-HCl (pH 7.5), 10% glycerol, 1% Nonidet P40) by western blotting with horseradish peroxidase (HRP)-conjugated 3F10 antibody (Roche). beta-Actin was detected with A5441 antibody (Sigma).
RNA interference Eight siRNAs directed against TRIM5alpharh were designed using the siRNA Selection Program (Whitehead Institute for Biomedical Research, 2003) and purchased from Dharmacon RNA Technologies: siRNA 1, 5'-GCUCAGGGAGGUCAAGUUGdTdT-3'; siRNA 2, 5'-GAGAAAGCUUCCUGGAAGAdTdT-3'; siRNA 3, 5'-GCCUUACGAAGUCUGAAACdTdT-3'; siRNA 4, 5'-GGAGAGUGUUUCGAGCUCCdTdT-3'; siRNA 5, 5'-CCUUCUUACACACUCAGCCdTdT-3'; siRNA 6, 5'-CGUCCUGCACUCAUCAGUGdTdT-3'; siRNA 7, 5'-CAGCCUUUCUAUAUCAUCGdTdT-3'; siRNA 8, 5'-CUCCUGUCUCUCCAUGUACdTdT-3'.
Four of the siRNAs (1--4) were directed against TRIM5 coding sequences common to the messenger RNAs of all of the TRIM5 isoforms. Four of the siRNAs (5--8) were directed against the 3' UTR specific to the mRNAs encoding TRIM5alpha and TRIM5alt epsilon. Two of the siRNAs (2 and 8) showed some toxicity to HeLa cells and were not studied further. After transfection into PRL cells, the remaining six siRNAs all showed ability to relieve the block to HIV-1--GFP infection (Fig. 4 and data not shown). A control siRNA (Nonspecific control duplex 1, 5'-AUGAACGUGAAUUGCUCAAUU-3'; Dharmacon RNA Technologies) was included in the experiments. HeLa cells (1 times 105) or PRL cells (5 times 104) were seeded in six-well plates and transfected with 120 nM siRNA by using 10 µl of Oligofectamine (Invitrogen). After 48 h, cells were reseeded for HIV-1--GFP infection. In some experiments, HeLa and PRL cells were transduced with a pLPCX vector encoding TRIM5alpharh-escape or an empty pLPCX vector as a control. After 48 h, the cells were transfected with siRNA; and after 2 d, the cells were replated and used for infection. In the vector encoding TRIM5alpha-escape, silent mutations at the siRNA 3 recognition site changed the wild-type sequence from 5'-GCCTTACGAAGTCTGAAAC-3' to 5'-GGTTAACGAAGAGCGAAAC-3'.
References
1. LaBonte, J., Babcock, G., Patel, T. & Sodroski, J. Blockade of human immunodeficiency virus (HIV-1) of New World monkey cells occurs primarily at the stage of virus entry. J. Exp. Med. 196, 431--435 (2002)
2. Shibata, R., Sakai, H., Kawamura, M., Tokunaga, K. & Adachi, A. Early replication block of human immunodeficiency virus type 1 in monkey cells. J. Gen. Virol. 76, 2723--2730 (1995)
3. Himathongkham, S. & Luciw, P. A. Restriction of HIV-1 (subtype B) replication at the entry step in rhesus macaque cells. Virology 219, 485--488 (1996)
4. Hofmann, W. et al. Species-specific, postentry barriers to primate immunodeficiency virus infection. J. Virol. 73, 10020--10028 (1999)
5. Owens, C. M., Yang, P. C., Go 2acutettlinger, H. & Sodroski, J. Human and simian immunodeficiency virus capsid proteins are major viral determinants of early, post-entry replication blocks in simian cells. J. Virol. 77, 726--731 (2003)
6. Kootstra, N. A., Munk, C., Tonnu, N., Landau, N. R. & Verma, I. M. Abrogation of postentry restriction of HIV-1 based lentiviral vector transduction in simian cells. Proc. Natl Acad. Sci. USA 100, 1298--1303 (2003)
7. Cowan, S. et al. Cellular inhibitors with Fv-1-like activity restrict human and simian immunodeficiency virus tropism. Proc. Natl Acad. Sci. USA 99, 11914--11919 (2002)
8. Besnier, C., Takeuchi, Y. & Towers, G. Restriction of lentivirus in monkeys. Proc. Natl Acad. Sci. USA 99, 11920--11925 (2002)
9. Munk, C., Brandt, S. M., Lucero, G. & Landau, N. R. A dominant block to HIV-1 replication at reverse transcription in simian cells. Proc. Natl Acad. Sci. USA 99, 13843--13848 (2002)
10. Kanki, P. J., Alroy, J. & Essex, M. Isolation of T-lymphotropic retrovirus related to HTLV-III/LAV from wild-caught African green monkeys. Science 230, 951--954 (1985)
11. Reddy, B. A., Etkin, L. D. & Freemont, P. S. A novel zinc finger coiled-coil domain in a family of nuclear proteins. Trends Biochem. Sci. 17, 344--345 (1992)
12. Borden, K. L. RING fingers and B-boxes: zinc-binding protein--protein interaction domains. Biochem. Cell Biol. 76, 351--358 (1998)
13. Reymond, A. et al. The tripartite motif family identifies cell compartments. EMBO J. 20, 2140--2151 (2001)
14. Xu, L. et al. BTBD1 and BTBD2 colocalize to cytoplasmic bodies with the RBCC/tripartite motif protein, TRIM5delta. Exp. Cell Res. 288, 84--93 (2003)
15. Feng, Y., Broder, C. C., Kennedy, P. E. & Berger, E. A. HIV-1 entry co-factor: functional cDNA cloning of a seven-transmembrane, G protein-coupled receptor. Science 272, 872--877 (1996)
16. Li, J., Lord, C. I., Haseltine, W., Letvin, N. L. & Sodroski, J. Infection of cynomolgus monkeys with a chimeric HIV-1/SIVmac virus that expresses the HIV-1 envelope glycoproteins. J. AIDS 5, 639--646 (1992)
17. Dorfman, T. & Go 2acutettlinger, H. G. The human immunodeficiency virus type 1 capsid p2 domain confers sensitivity to the cyclophilin-binding drug SDZ NIM 811. J. Virol. 70, 5751--5757 (1996)
18. Waterman, H., Levkowitz, G., Alroy, I. & Yarden, Y. The RING finger of c-CBL mediates desensitization of the epidermal growth factor receptor. J. Biol. Chem. 274, 22151--22154 (1999) 19. Owens, C. M. et al. Binding and susceptibility to post-entry restriction factors in monkey cells are specified by distinct regions of the human immunodeficiency virus type 1 capsid. J. Virol. (in the press)
20. Fassati, A. & Goff, S. P. Characterization of intracellular reverse transcription complexes of human immunodeficiency virus type 1. J. Virol. 75, 3626--3635 (2001)
21. Forshey, B. M., von Schwedler, U., Sundquist, W. I. & Aiken, C. Formation of a human immunodeficiency virus type 1 core of optimal stability is crucial for viral replication. J. Virol. 76, 5667--5677 (2002) 22. Schwartz, O., Maréchal, V., Friguet, B., Arenzana-Seisdedos, F. & Heard, J.-M. Antiviral activity of the proteasome on incoming human immunodeficiency virus type 1. J. Virol. 72, 3845--3850 (1998)
23. Turelli, P. et al. Cytoplasmic recruitment of INI1 and PML on incoming HIV preintegration complexes: interference with early steps of viral replication. Mol. Cell 7, 1245--1254 (2001)
24. Marcello, A. et al. Recruitment of human cyclin T1 to nuclear bodies through direct interaction with the PML protein. EMBO J. 22, 2156--2166 (2003)
25. Bonilla, W. V. et al. Effects of promyelocytic leukemia protein on virus-host balance. J. Virol. 76, 3810--3818 (2002)
26. Chee, A. V., Lopez, P., Pandolfi, P. P. & Roizman, B. Promyelocytic leukemia protein mediates interferon-based anti-herpes simplex virus 1 effects. J. Virol. 77, 7101--7105 (2003)
27. Rhodes, D. A. et al. The 52,000 MW Ro/SS-A autoantigen in Sjogren's syndrome/systemic lupus erythematosus (Ro52) is an interferon-gamma inducible tripartite motif protein associated with membrane proximal structures. Immunology 106, 246--256 (2002)
28. Toniato, E. et al. TRIM8/GERP RING finger protein interacts with SOCS-1. J. Biol. Chem. 277, 37315--37322 (2002)
29. Mariani, R. et al. Species-specific exclusion of APOBEC3G from HIV-1 virions by Vif. Cell 114, 21--31 (2003)
30. Butler, S., Hansen, M. S. T. & Bushman, F. A quantitative assay for HIV DNA integration in vivo. Nature Med. 7, 631--634 (2001)
Report
A Single Amino Acid Change in the SPRY Domain of Human Trim5Éø Leads to HIV-1 Restriction
Current Biology, Vol 15, 73-78, 11 January 2005
Melvyn W. Yap, Sébastien Nisole,1 and Jonathan P. Stoye
Division of Virology, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, United Kingdom
SUMMARY
Retroviral restriction factors are cellular proteins that interfere with retrovirus replication at a postpenetration, preintegration stage in the viral life cycle [1--3]. The first restriction activity described was the mouse Fv1 gene [4]. Three different alleles of Fv1, capable of restricting different murine leukaemia viruses (MLV), have been characterized at the molecular level [5, 6]. Two further activities, Ref1, which acts on MLV, and Lv1, which acts on lentiviruses, have been identified in other mammalian species [7--9]. Recently, it has become clear that Ref1 and Lv1 are encoded by the same gene, Trim5Éø, which inhibits retrovirus replication in a species-specific manner [10--14]. A series of chimeras between the human and rhesus monkey Trim5 genes were created to map and identify these specificity determinants. The Trim5Éø SPRY domain was found to be responsible for targeting HIV-1 restriction. By contrast, N-MLV restriction appears dependent on both the coiled-coil domain and the SPRY domain. A single amino acid substitution (R332P) in the human Trim5Éø can confer the ability to restrict HIV-1, suggesting that small changes during evolution may have profound effects on our susceptibility to cross-species infection.
Human Trim5Éø restricts N-tropic MLV, but not B-tropic MLV or HIV-1, whereas rhesus monkey Trim5Éø restricts N-tropic MLV and HIV-1, but not B-tropic MLV [11--14]. The human and rhesus forms of Trim5Éø can also be distinguished by using the L117H mutation of N-MLV, which is restricted by human, but not rhesus monkey Trim5Éø [13]. A series of chimeras between the human and rhesus gene was constructed with overlapping PCR (Figure 1) to examine which domain of Trim5Éø was involved in determining the specificity of retroviral restriction. The structures of these chimeras (and all subsequent constructs) were confirmed by sequencing.
Trim5Éø has a modular design with a series of distinct motifs. These include the RING motif, a B-Box, and a coiled-coil domain, which together form the tripartite RBCC domain characterizing Trim proteins, as well as a SPRY domain [15]. These different domains were exchanged between the two proteins. The chimeras HR3, HR2, and HR1 contain the human Trim5Éø RING motif with progressively longer substitutions of rhesus C-terminal residues. A reciprocal series was also made (RH3, RH2, and RH1) where parts of the rhesus Trim5Éø are substituted by human sequences. Restriction was measured by a previously described two-color FACS assay [13, 16]. Briefly, nonrestricting cells were transduced with an MLV-based vector carrying the Trim gene and EYFP. Two days posttransduction, the cells were challenged with tester viruses incorporating various capsids and encoding the EGFP marker. The percentage of Trim/EYFP positive cells that were transduced with the EGFP marker was then scored against that of Trim/EYFP negative cells. Restriction was expressed as a ratio of EGFP+, EYFP+ cells to EGFP+, EYFP- cells. A ratio of less than or equal to 0.3 was taken as restriction, whereas a ratio between 0.7 and 1 indicated the absence of restriction. A typical result is shown in Figure 1A. As observed previously, human Trim5Éø restricted N-MLV and the L117H mutant of N-MLV, but not B-MLV or HIV-1 [13]. Rhesus Trim5Éø restricted HIV-1 as well as N-MLV, but not B-MLV or N L117H. These restricting phenotypes were used as standards for comparison with the restriction profile of the chimeric molecules.
All chimeras could restrict N-MLV, but not B-MLV, indicating that the proteins were still functional after exchanging domains (Figure 1B). Substitution of the SPRY domain from the human Trim5Éø with the rhesus SPRY domain in HR3 is sufficient to confer restriction to HIV-1. The reciprocal chimera, where the SPRY of rhesus Trim5Éø was replaced with the human form (RH3), lost the ability to restrict HIV-1. This implied that the SPRY domain of Trim5Éø was sufficient and necessary for recognizing and restricting HIV-1. The determinants for N-MLV restriction seemed to be more complex. Replacing the SPRY domain of rhesus Trim5Éø with the human form (RH3) enabled it to restrict N L117H in addition to N-MLV, implying that the SPRY domain was also involved in targeting the restriction of N-MLV. However, substitution of the human SPRY domain with the corresponding rhesus sequences (HR3) did not abolish its ability to restrict N L117H. It was only when the coiled-coil domain as well as the SPRY of human Trim5Éø was replaced with that from rhesus (HR2) that restriction of N L117H was lost. These results suggested that both the coiled-coil and the SPRY domains play roles in determining the specificity of MLV restriction.
Two more sets of chimeras were constructed (Figure 2) to further delineate the regions of the SPRY domain involved in restriction. HR4 to HR7 consisted of human Trim5Éø with decreasing lengths of rhesus sequences substituted within the SPRY domain, whereas RH4 to RH7 formed a reciprocal series where the rhesus Trim5Éø was substituted with human SPRY sequences. As expected, all the chimeras could restrict N-MLV, but not B-MLV, indicating that they were functional with respect to restriction. HR4 was able to restrict HIV-1, but the restriction was absent when the rhesus sequences after residue 374 were replaced with human ones from HR5 to HR7. This suggested that the region determining specificity for HIV-1 restriction was between amino acids 322 to 374 of the rhesus Trim5Éø. RH4, which does not restrict HIV-1, lacked the rhesus sequence in this region. When the rhesus sequence was included in the chimera RH5, restriction of HIV-1 was restored, confirming the requirement of this region in HIV-1 recognition and restriction. The requirements for MLV restriction were found to be somewhat different from that of HIV-1. HR4 to HR7 could restrict N L117H. This was expected as both human Trim5Éø and HR3 (Figure 1B) could restrict this mutant. The chimeras RH4 to RH7 (Figure 2) contain decreasing amount of human SPRY sequences compared to RH3 (Figure 1B). RH4 and RH5 could restrict N L117H, but the restriction was abolished when human residues before amino acid 401 were excluded, in RH6 and RH7. This suggested that the SPRY region in human Trim5Éø influencing N-MLV restriction involved residues up to amino acid 400. Taken with the previous data, these results indicated that the specificity determinants of MLV restriction encompassed a region somewhat larger than that which governed HIV-1 restriction.
We next sought to identify individual residues between amino acids 322 and 374 in the SPRY domain that affected restriction. Between the human and rhesus sequence, there was a difference of seven amino acids, two contiguous, the others scattered throughout the region. In addition, a stretch of six residues starting from amino acid 335 in the human sequence was replaced by 8 different residues in the rhesus sequence. The residues at these positions in the human Trim5Éø were replaced with the corresponding residues from rhesus monkey by site-directed mutagenesis. The mutants were then assayed for restriction of HIV-1 and MLV (Figure 3) . Only two positions in this region affected restriction to HIV-1. The arginine to proline change at residue 332 and the exchange of the six human residues at position 335 for the eight amino acid rhesus sequence conferred the ability to restrict HIV-1 to the human Trim5Éø. Conversely, changing each of these positions in the rhesus Trim5Éø to the corresponding residues in the human protein did not abolish restriction of HIV-1. Indeed, when both positions of the human protein were substituted by the rhesus residues, the resulting double mutant had a restricting phenotype that was similar to the two single mutants. This suggested that both of these positions contributed to the restriction and that they might be degenerate. Hence, the loss of one of these positions could be compensated by the presence of the other.
The corresponding change from proline to arginine in position 334 in the rhesus protein did not confer restriction to N L117H, suggesting that unlike HIV-1 restriction, this position was not involved in MLV restriction. Exchanging the eight residues at position 337 of the rhesus protein with the six found in humans resulted in the restriction of N L117H, suggesting that the position was involved in MLV restriction. However, corresponding changes in the human sequence to the rhesus residues did not abolish restriction of N L117H. This suggests that other regions in the human Trim5Éø might be contributing degenerately to restricting N L117H. Indeed, this could involve sequences found in the coiled-coil region (Figure 1B).
We have previously shown that Trim1 could also restrict N-MLV, but not HIV-1 [13]. Like Trim5, it consists of an SPRY domain after the tripartite RBCC motif. Trim1 was first identified as a homolog of the Opitz syndrome gene Trim18 (Mid1) to which it showed 83% similarity at the amino acid level [17, 18]. However Trim18, unlike Trim1, does not restrict N-MLV (Figure 4) . Formally, this might be due to a lack of protein expression. Alternatively, it provides a further opportunity to examine the role of the SPRY domain in retroviral restriction. The C-terminal portion of African green monkey Trim1, cloned from Vero cells, was used to replace the corresponding region in Trim18 beginning from residue 314 (Figure 4). This chimera (Vero Trim18/Trim1) was active against N-MLV as well as N-MLV containing the N82D mutation, but not B-MLV, a restriction phenotype typical of Trim1. The reciprocal chimera Vero Trim1/Trim18 did not restrict N-MLV and N N82D. Taken together, these results provided further evidence for the involvement of the SPRY domain as the recognition domain in the process of retroviral restriction.
In owl monkeys, a psuedogene encoding the cyclophilin A (CypA) protein is inserted within the seventh intron of the Trim5Éø gene [19, 20]. It is thought that the cyclosporine A (CsA) sensitive Lv1 activity of OMK cells [21] can be explained by the presence of a fusion protein in which most of the Trim5Éø SPRY domain is replaced by CypA [19, 20]. This implies that the function provided by SPRY can be replaced by CypA; because CypA is known to bind HIV-1 CA in a CsA sensitive manner [22], this suggests that SPRY is providing binding as well as specificity functions. To examine this idea further, we replaced the SPRY domain of human and rhesus Trim5Éøs with the CypA insertion of the OMK Trim5Éø gene (Figure 5) . Both novel chimeras restricted wild-type HIV-1 (Figure 5), but not a mutant form that is not bound by CypA, CA G89V. In addition, the substitutions abolished activity against N-MLV, which does not bind CypA [22]. Replacing the SPRY domain of Trim5Éø from African green monkey yielded identical results (data not shown). This confirmed that CypA was performing the function of CA recognition and binding in OMK Trim5-CypA.
We have previously found that an artificially created OMK Trim5Éø (V6), which contained the owl monkey SPRY domain in place of CypA, was inactive against both HIV-1 and MLV [20]. To further probe the role of SPRY in CA recognition and binding, we asked if the SPRY domain from rhesus Trim5Éø could rescue the restriction activity of OMK Trim5 V6. Replacing the SPRY of OMK Trim5 V6 in OMKRhSPRY with that from rhesus restored its activity against HIV-1 (Figure 5), suggesting that this domain alone could account for the targeting of HIV-1 CA. However, OMKRhSPRY was not active against N-MLV. Intriguingly, the sequence of the coiled-coil domain of OMK Trim5 is very different from that of the rhesus protein. In the light of the earlier result (Figure 1) suggesting the involvement of the coiled-coil domain in MLV restriction, the failure of OMKRhSPRY to restrict might be due to the lack of MLV CA recognition sequences in the coiled-coil domain of OMK Trim5.
Our results indicate that the SPRY domain is the primary determinant of restriction specificity and suggest that this specificity might be mediated through SPRY binding of capsid. The SPRY domain was originally characterized in proteins that regulated intracellular signaling [23]. It can occur as a subdomain within the B30.2 domain [24], which is also found in Trim1 and Trim18 [17, 18]. Although often associated with Trim proteins, SPRY domains are also found in unrelated proteins such as venom toxins and proteins with immunoglobulin-like folds [24]. The functions of B30.2 and SPRY domains are still not known but might include specific protein-protein interactions. Our studies suggest that in Trim proteins, they play a crucial role in the interaction with infectious agents such as retroviruses to abolish infection. The SPRY domain seems to function independently from the tripartite motif, because it can be exchanged between different primate Trim5s and can even be replaced with CypA, with associated changes in restriction specificity. It is tempting to speculate that these proteins, together with the other members of the Trim family, might be involved in innate immunity against different infectious agents. Interestingly, mutations in the B30.2 domain in Trim18 (Mid1) [25, 26] and Trim20 (Pyrin) [27, 28] are associated with genetic defects. It has been suggested [25--28] that these mutations are preserved during evolution, because they offer enhanced protection against disease-causing organisms in a manner analogous to the survival of the sickle cell alleles of haemoglobin [29].
The region imparting restriction specificity appears to lie between amino acids 332--340 of human Trim5Éø. This is the region showing greatest diversity between the sequences of human, rhesus, and African green monkey [11--13], and we are currently carrying out a detailed phylogenetic analysis of this region to correlate sequence variation with changes in restriction function. Our most provocative initial finding is the observation that a single amino acid substitution (R332P) in the human Trim5Éø protein can confer the ability to restrict HIV-1 replication. Although we do not know the sequence of the ancestral Trim5, we note that rhesus and African green monkey sequences contain a proline at this position, and these species are resistant to HIV-1. This implies that during evolution, changes at this position affecting sensitivity to HIV-1 infection may have occurred. Without such changes it seems unlikely that the cross-species infection(s) leading to the current AIDS epidemic [30] would have occurred. In this context, it is tempting to speculate that the R332P and/or RYQTFV335LFTFPSLT derivatives of human Trim5Éø might be suitable candidate(s) for gene therapy of HIV-1 infection, perhaps introduced in combination with agents that target other steps in the viral life cycle such as the Vif-insensitive, K128D derivative of human APOBEC3G [31--33] or HIV-1 siRNA [34].
|
|
| |
| |
|
|
|
