An HIV-1 capsid binding protein TRIM11 accelerates viral uncoating

TRIM11 is reported to interfere with both early and late stages of viral replication [17, 18]. We previously reported that TRIM11 inhibits HIV-1 reverse transcription during early steps of the viral replication cycle [18]. However, the mechanism by which TRIM11 decreases viral DNA is still unclear. In this study, we demonstrated that TRIM11 both purified from E. coli. and expressed in cells interacts with in vitro assembled HIV-1 capsid. Importantly, the PLA assay showed that TRIM11 associates with HIV-1 capsid during viral infection. We then revealed that TRIM11 accelerates HIV-1 uncoating during viral infection in both HEK293 cells and THP-1-derived macrophages. The effect of TRIM11 on uncoating is dependent on microtubule dynamics, whereas independent of proteasomal or lysosomal pathway. All of these findings support the restriction of HIV-1 transduction by TRIM11, while MLV transduction is not influenced by TRIM11. The HIV-1 CA mutant G89V is insensitive to the effect of TRIM11, which strengthened the possibility that the capsid of HIV-1 could be the determinant for restriction by TRIM11. Thus, our work presents the first human TRIM family member that recognizes HIV-1 capsid and accelerates its uncoating.

Multiple HIV-1 capsid binding proteins have been identified [15, 16, 25, 27–30], most of which, like CSPF6, NUP153, NUP358, TNPO3 and CypA, were shown as dependency factors for HIV-1 replication [25, 28–30]. These factors aid properly uncoating of HIV-1 capsid or escorting viral DNA into nucleus. In the other side, some capsid-binding proteins could perturb uncoating and impede HIV-1 replication. MxB was identified as an IFN-? inducible restriction factor that could block viral DNA nuclear entry by interacting with HIV-1 capsid to increase its stability [16]. In this study, we introduced TRIM11 as a new HIV-1 capsid binding factor that will tilt the balance of uncoating process in favor of accelerating when it is overexpressed in cells, which will result in reduced viral reverse transcription levels.

TRIM11 is not the only human TRIM family member that could associate with HIV-1 capsid. TRIM6 and TRIM34, which are closely related to TRIM5? in sequence similarity, have been reported to bind in vitro assembled HIV-1 capsid, but they do not have the ability to restrict HIV-1 replication [20]. By analysis of extensive TRIM5?_rh mutants, Yang et al., demonstrated that binding to HIV-1 capsid is necessary, but not sufficient, for HIV-1 restriction [31]. Cytoplasmic body formation and E3 ubiquitin ligase activity were also implicated in TRIM5?_rh-mediated capsid disruption and restriction of reverse transcription [32–34]. The restriction activity of TRIM5? was recently reported to correlate with its ability to induce TAK-1 dependent innate immune signaling [35]. Thus, in addition to binding with capsid, some domains of TRIM5?_rh must function in a cryptic mechanism to accelerate viral capsid uncoating. In accordance to these findings, the chimeric proteins TRIM-CypA and Trim-NUP153(C), which use different domain for binding to capsid, sustain the restriction ability of TRIM5?_rh [36–38]. Then other than binding to HIV-1 capsid, how TRIM11 exerts the function of accelerating uncoating is still unknown.

We previously reported that TRIM11 decreased HIV-1 reverse transcription in a proteasome independent manner [18]. In accordance with this result, here we found that the effect of TRIM11 on viral uncoating is independent of proteasome. Our findings suggested that the mechanism by which TRIM11 accelerates HIV-1 uncoating is not exactly the same as that of TRIM5?_rh which can be blocked by MG132 [4, 5]. Furthermore, premature disassembly of HIV-1 imposed by TRIM11 is also independent of the lysosomal pathway. In contrast, microtubule dynamics is essential for TRIM11-mediated uncoating and reduction of reverse transcription. Since microtubule dynamics are involved in uncoating [22], we speculate that this cellular apparatus, that originally supports early stages of HIV-1 replication, may be utilized by antiviral proteins like TRIM5?_rh and human TRIM11 to accelerate uncoating.

We compared the effects of TRIM11 and TRIM5?_rh on viral uncoating, reverse transcription and transduction, and found that although they have comparable effects on viral reverse transcription, TRIM11 restricts HIV-1 transduction less potently than TRIM5?_rh does. We attribute this to the fact that TRIM11 lacks ability to impede the accumulation of nuclear viral DNA. In addition, HIV-1 reverse transcription levels were significantly increased and the viral capsid was more stable in TRIM11 knockdown THP-1-derived macrophages, suggesting that physiological levels of TRIM11 in HIV-1 relevant cells impede viral reverse transcription by accelerating uncoating process. Although TRIM11 expression was shown negatively correlated with HIV-1 replication in monocytes [39], it seems like TRIM11 does not act as a restriction factor like TRIM5?_rh that can be a barrier for retrovirus cross species transmission. Unlike Rhesus macaques, human is the natural host of HIV-1, which means the virus has evolved well enough to conquer potential obstacles in the host. We previously reported that HIV-1 Vpr regulated TRIM11 expression levels bidirectionally [18]. However, whether the regulation of TRIM11 by Vpr is a strategy used by HIV-1 to overcome the restriction effect of TRIM11 remained unclear. We speculate that this regulation by Vpr is more relevant to the regulation of innate immune signaling, as Vpr has not been reported to be involved in HIV-1 uncoating. Another scenario is that maybe the restriction of TRIM11 on HIV-1 replication is covered by other cellular factors. Thus, changing TRIM11 expression levels by genetic means trumps the effect of other cellular factors, and evidences its influence on HIV-1 uncoating process. Further investigation into the mechanism by which HIV-1 escapes TRIM11-mediated restriction of early replication is needed.