Correspondence Hiroaki Takeuchi and Tetsuro Matano, International Research Center for Infectious Diseases, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan. Tel: +81 3 6409 2078; Fax: +81 3 6409 2076; email: email@example.com and firstname.lastname@example.org
Viral replication requires the help of host cell factors, whose species specificity may affect viral tropism. On the other hand, there exist host factors that restrict viral replication. The anti-viral system mediated by some of these restriction factors, which is termed intrinsic immunity and is distinguished from conventional innate and adaptive immunity, has been described as playing an important role in making species-specific barriers against viral infection. Here, we describe the current progress in understanding of such restriction factors against retroviral replication, focusing on TRIM5α and APOBEC, whose anti-retroviral effects have recently been recognized. Additionally, we mention cyclophilin A, which is essential for HIV-1 replication in human cells and may affect viral tropism. Understanding of these host factors would contribute to identification of the determinants for viral tropism.
apolipoprotein B mRNA-editing enzyme-catalytic subunit
human immunodeficiency virus type 1
lentivirus susceptibility factor 1
murine leukemia virus
old world monkey
restriction factor 1
African green monkey simian immunodeficiency virus
simian immunodeficiency virus mac
tripartite interaction motif 5
tripartite interaction motif 5 α
virus infectivity factor
Among host factors exerting inhibitory effects on viral replication, the Fv-1 and the Fv-4 loci have been known to confer resistance to Friend virus infection in mice and their restriction mechanisms have been intensively investigated (1–5). The latter, Fv-4, corresponds to a defective endogenous provirus with an ecotropic MLV-like env gene. Entry of ecotropic MLV into cells expressing Fv-4 is inhibited by binding of the Fv-4 env gene product, Fv-4 Env, to the ecotropic MLV receptor, and this receptor interference has been reported to play a central role in resistance to Friend virus infection (6). Additionally, it has recently been suggested that the Fv-4 Env-mediated dominant negative effect on MLV Env function contributes to efficient resistance in Fv-4-expressing mice (7). On the other hand, MLV restriction mediated by the Fv-1 gene product, Fv-1, occurs after viral entry into the cells but before the integration step during the viral replication cycle (8). The two main alleles of Fv-1, Fv-1n and Fv-1b, confer resistance to replication of B-tropic and N-tropic MLV respectively (9). Although the precise mechanism of Fv-1-mediated restriction remains unclear, comparison of viral genome sequences between B-tropic and N-tropic MLV has indicated the 110th amino acid in Gag CA as the viral determinant for the tropism, suggesting that MLV Gag CA is the target for the host factor involved in this restriction (10, 11).
Ref-1, which shows a similar pattern of inhibition of retroviral replication with Fv-1, is known as a restriction factor in mammalian cells apart from murine cells. The viral determinant for Ref-1-mediated restriction has been reported to be at the Gag CA residue, corresponding to the 110th in MLV Gag CA involved in Fv-1-mediated restriction (12). Ref-1 and Fv-1 both show viral restriction post-viral entry in the early phase of the retroviral replication cycle, but the exact point of restriction has been indicated to be different: the former, Ref-1-mediated restriction, occurs at the step prior to reverse transcription, while the latter, Fv-1 mediated restriction, is considered to occur post-reverse transcription (8). Further, restriction of HIV-1 replication post-viral entry has been reported in OWM cells that support efficient replication of SIVmac, and the existence of a restriction factor, Lv-1, responsible for this resistance of non-human primate cells to HIV-1 replication has been suggested (13–15).
TRIM5α: A HOST FACTOR RESTRICTING HIV-1 REPLICATION POST-VIRAL ENTRY IN PRIMATE CELLS
Recently, two independent groups have identified the α-isoform of TRIM5, TRIM5α, as a restriction factor responsible for resistance of monkey cells to HIV-1 infection and shown that restriction of HIV-1 replication by TRIM5α derived from rhesus and owl monkeys but not efficiently from humans (16, 17) (Fig. 1). Subsequent studies have revealed virus-specific restriction activities by TRIM5α and its homologues derived from humans and non-human primates (18–24). For instance, restriction by rhesus monkey TRIM5α is efficient against HIV-1 but inefficient against SIVmac and undetectable against MLV (Fig. 1).
TRIM5α is a trimeric cytoplasmic protein (25, 26) consisting of RING finger, B-box, coiled-coil, and SPRY (B30.2) domains (27, 28). The coiled-coil domain is indispensable for TRIM5α multimerization, and both the coiled-coil and the SPRY domains are required for its binding to the virion core (28, 29).
TRIM5α-midiated restriction of HIV-1 replication is considered to occur after viral entry in the early phase of the viral replication cycle, but its precise mechanism remains unclear and several possibilities have been proposed. First, it has been suggested that binding of TRIM5α to the virion CA after viral entry may accelerate or abrogate the process of its uncoating and disruption, resulting in inhibition of HIV-1 replication (29). Second, involvement of ubiquitin in TRIM5α-mediated restriction has been suggested by recent reports showing that a mutation in its RING finger domain decreases the restriction ability of TRIM5α (16, 30–32) and that recovery from the restriction occurs in the presence of proteosome inhibitors (32, 33), although this is controversial (29, 31, 32). Furthermore, some reports have shown TRIM5α-mediated inhibition of viral cDNA nuclear import as well as viral cDNA synthesis (34, 35). In addition to restriction at the early phase of retroviral replication cycle, TRIM5α has recently been shown to inhibit virus production by accelerating degradation of viral Gag protein (36).
APOBEC: AN ENZYMATIC HOST FACTOR RESTRICTING RETROVIRAL REPLICATION
HIV-1 replication in primary CD4+ T lymphocytes and monocytes is dependent on the presence of an HIV-1 accessory protein, Vif, which has been reported to work in a host cell-specific manner (37, 38). Vif is required for infectious HIV-1 production from some immortalized human T cell lines such as CEM (termed non-permissive) but not in others such as CEM-SS (termed permissive), and the existence of a restriction factor whose anti-retroviral activity can be abrogated by Vif has been suggested in the case of the non-permissive cells (39–45). Comparison of the non-permissive and permissive cells has revealed Apo3G, a member of the APOBEC family of cytidine deaminases, to be the restriction factor responsible for inhibition of vif-deleted HIV-1 replication in human non-permissive cells (46). Unlike TRIM5α and Fv-1, the target of Apo3G-mediated restriction is not viral CA, but viral single-stranded cDNA synthesized during reverse transcription. It is packaged into virus particles produced from Apo3G-expressing cells and inhibits viral replication after viral entry into the cells (Fig. 2). HIV-1 Vif can inhibit the uptake of Apo3G into the virion by inducing polyubiquitylation and proteosomal degradation of cellular Apo3G, resulting in abrogation of Apo3G-mediated restriction (47) (Fig. 2).
Several mechanisms for Apo3G-mediated restriction against HIV-1 infection have been reported. First, it has been reported that the cytidine deaminase activity of Apo3G can induce hypermutation (a large number of G-to-A substitutions) in proviral DNA during reverse transcription, resulting in failure of infectious HIV-1 production (48–53) (Fig. 2). Second, the possibility of Apo3G-mediated inhibition of tRNA annealing or processing during reverse transcription has been shown (54–56). Additional mechanisms, including inhibition at the step of viral plus-strand cDNA transfer, have also been suggested (56–58).
Restriction of retroviral infection by Apo3G derived from non-human species has also been reported (51, 59–62), and the Vif-Apo3G interaction is considered to be species-specific (51, 63). Indeed, it has been indicated that hApo3G is insensitive to SIVagm Vif while agmApo3G is insensitive to HIV-1 Vif, and that the determinant for this specificity is at the 128th residue in Apo3G (51, 64–67). However, a recent report has shown that SIVagm Vif can support SIVagm replication in an hApo3G-positive human T cell line (A3.01): vif-deleted SIVagm replication was severely restricted with accumulation of G-to-A mutations in the viral genome, suggesting ambiguity of species specificity (68).
Restriction of HIV-1 and SIV replication by other members of the APOBEC family has been reported, although it might not be as efficient as Apo3G. Thus, APOBEC proteins are now considered to be a new class of host restriction factors against retroviral replication (61, 69). For instance, human APOBEC3F can inhibit HIV-1 replication in the absence of Vif (59, 60, 70, 71) whereas human APOBEC3B does so even in the presence of Vif (59, 72, 73). Association of deaminase activity with Apo3G-mediated restriction of HIV-1 replication has been strongly suspected, but the possibility of involvement of deaminase activity-independent mechanisms in this restriction has also been suggested (57, 58). Indeed, several groups have reported Apo3G and APOBEC3F variants lacking in deaminase activity without loss of restriction activity, as well as variants lacking in restriction activity without loss of deaminase activity (74–77). Additionally, a recent report has suggested a deaminase-independent inhibitory effect of Apo3G on viral DNA synthesis following reverse transcription (78). However, this possibility is controversial (79, 80), and the precise mechanism for Apo3G-mediated restriction of HIV-1 infection remains unclear.
CYCLOPHILIN A: A HOST FACTOR INVOLVED IN RETROVIRAL REPLICATION
CypA, a ubiquitous protein, was first identified as the target of CsA, an immunosuppressive reagent (81). CypA has proline-isomerase activity that catalyzes cis-trans isomerization of the Pro residue (82, 83). CsA binding to CypA inhibits this isomerase activity (83). CypA binding to HIV-1 Gag CA has been shown by analysis using the yeast two-hybrid system (84). The Ala-Gly-Pro-Ile residues from the 88th to the 91st in CA are the key portion for its binding to the active site of CypA (85–87). Interestingly, the peptidyl-prolyl bond between the 89th Gly (Gly89) and the 90th Pro (Pro90) exhibits the trans conformation, (in contrast to the cis conformation usually observed in other known CypA targets) (87, 88), and this Pro90 residue but not other Pro is considered to be critical for the binding of CA to CypA. Thus, it has been suggested that CypA acts as a molecular chaperone without exerting cis-trans isomerase activity on HIV-1 CA (87). However, this is still controversial and the possibility of CypA-mediated cis-trans isomerization of the Gly89-Pro90 peptidyl-prolyl bond has also been suggested (88).
It has been well established that CypA promotes HIV-1 replication after viral entry in the early phase in human cells (85, 89–95) (Fig. 3). CypA is efficiently incorporated into the virion produced from HIV-1-infected cells through interaction with CA in the context of Gag polyprotein (Fig. 3). Disruption of CypA incorporation into the virion by CsA administration or by Gag mutations resulted in reduction in infectivity of the produced viruses (85, 89, 91, 95–98). Several reports have shown that both CA dimerization and CypA multimerization are required for the efficient CA-CypA binding which is critical for HIV-1 infectivity (99, 100). Recently, promotion of HIV-1 replication by post-entry interaction of CA with CypA in target cells has been shown, suggesting the importance of CypA for efficient HIV-1 replication (94, 101, 102).
Only retroviruses with CA capable of binding to CypA exhibit CypA-dependent viral replication (84, 85, 90, 92, 95). This suggests involvement of CA-CypA interaction in the determination of retroviral tropism (14, 16, 17, 34, 98, 101–111). The effect of CypA on SIV replication in human cells has not been clearly determined but, for the first time, a recent study has shown that human CypA exerts an inhibitory effect on vif-deleted SIV replication, which may be recovered by SIV Vif excluding the CypA from the virion (112). This Vif function can be distinguished from the anti-hApo3G function of Vif described above.
PERSPECTIVES: HOST FACTORS INVOLVED IN THE POST-ENTRY STEP OF THE EARLY PHASE OF RETROVIRAL REPLICATION
Restriction of HIV-1 replication in non-human primate cells after viral entry occurs at the step prior to reverse transcription, and TRIM5α plays a crucial role in this restriction (13–15, 101, 106, 113–115). Interestingly, CypA-dependency is considered to occur at the same step in the retroviral replication cycle (116).
Several groups have reported modest restriction of HIV-1 replication by human TRIM5α, a restriction which is not altered by disruption of the CA-CypA interaction or by elimination of endogenous CypA (117–119). In contrast, restriction of HIV-1 replication by the TRIM5α derived from OWM (such as rhesus macaques) is abrogated by CsA-mediated or small interfering RNA-mediated inhibition of OWM-derived CypA function, indicating involvement of CypA in TRIM5α-mediated restriction of HIV-1 replication in OWM cells (117, 118, 120). In the owl monkey (a new world monkey), a CypA-TRIM5α-fusion protein has been found and involvement of CypA in the restriction has been suggested (17). Thus, CypA may exert restriction activity against HIV-1 replication in association with TRIM5α in non-human primates but not in humans. Elucidation of the key factors involved in this difference in CypA function between non-human primates and humans may contribute to understanding of the species-specific restriction mechanism against retroviral replication.