|Home | About | Journals | Submit | Contact Us | Français|
The viral restriction factor SERINC5 inhibits HIV-1 infection via unknown mechanisms. Sood and co-workers now show that SERINC5 suppresses HIV-1 fusogenicity and increases sensitivity to neutralizing antibodies by perturbing the folding of the fusion machinery. This work advances our understanding of host-virus interactions and provides a compelling case for considering the host immune system in studies of restriction factor mechanisms.
Despite the enormous impact of the HIV-1 pandemic on human health and the intense scrutiny to which the virus has been subjected, many facets of HIV-1 biology have eluded comprehensive explanation for many years. Among these are viral accessory proteins, which have evolved to circumvent host factors that restrict retroviral infection. Perhaps the most enigmatic of these HIV-1 accessory proteins is Nef. Nef is known to interact with and regulate a myriad of host proteins and processes and has long been known to enhance viral infectivity (1). However, unlike other HIV-1 accessory proteins, none of the well-studied functions of Nef appeared to explain the enhanced viral infectivity, suggesting that an as-yet-unidentified host factor must inhibit HIV-1 infectivity and, in turn, be countered by Nef.
In 2015, back-to-back publications revealed the identities of the Nef-targeted host factors to be SERINC5 (serine incorporator 5) and SERINC3 (2, 3); the SERINC proteins are multipass membrane-spanning proteins, of which little is known beyond a role in the synthesis of serine-containing lipids. Of the five SERINC proteins, only SERINC3 and SERINC5 have activity against HIV-1, and SERINC5 is significantly more potent. These papers, and others that followed, quickly established the importance of SERINC5 as a genuine viral restriction factor. For example, where Nef is absent, retroviruses have independently evolved proteins to counter SERINC5 activity, such as glycoGag of murine leukemia virus (2, 3) and the S2 accessory protein of equine infectious anemia virus (4). Furthermore, a study of primate lentiviruses in the wild revealed a correlation between their prevalence in the susceptible population and the ability to antagonize SERINC5 (5). We have since learned that Nef counters SERINC5 by preventing its incorporation into virions. In the absence of Nef, SERINC5 packages into virions and appears to inhibit viral fusion at a post-receptor-binding step; however, the mechanism by which SERINC5 impairs fusion is not clear, and there is a significantly greater impact on viral infectivity than on fusion (2, 3). Intriguingly, it has been reported that many primary isolates of HIV-1 are resistant to this fusion-defect and that resistance is conferred by the envelope glycoprotein (Env) in the viral membrane (6). This protein binds to receptors on the target cell and mediates fusion between the virus and host membranes. Primary isolates were not entirely impervious to the effects of SERINC5, however, as they exhibited increased sensitivity to neutralizing antisera (6). Thus, the field has been left with two outstanding questions: how does SERINC5 work at a biochemical level to inhibit laboratory isolates and what differences explain the distinct behavior of primary isolates?
The work of Sood and colleagues (7) correspondingly tackles these challenges in two parts, first dealing with the impact of SERINC5 on viral fusion, and second with the sensitivity of SERINC5-containing virions to neutralizing antibodies. The authors first recapitulated earlier studies, showing a profound sensitivity of the HXB2 laboratory isolate of HIV-1 to SERINC5 and that the ability of SERINC5 to inhibit productive infection was more potent than the inhibition of fusion, which is an early step of the virus infection cycle. This observation had previously been interpreted as an indication that SERINC5 trapped viruses in a dead-end hemifusion state, where fusion is initiated, but does not allow passage of the viral core into the cytoplasm, thus blocking productive infection of the target cell. The authors tested this hypothesis by treating cells with chlorpromazine to promote completion of fusion by any hemifused viruses. Contrary to expectations, chlorpromazine had no impact on infection, suggesting that stalled fusion is not the defect induced by SERINC5. Instead, the authors provided evidence, collected from a variety of fusion assays, that the initiation of pore formation is inhibited. These data advance our understanding of the mechanism of action of SERINC5 against sensitive laboratory isolates of HIV-1 and suggest that SERINC5 may be acting on Env to impair fusion in sensitive isolates. These data do not, however, explain the comparative lack of effect on primary isolates of HIV-1.
To search for a SERINC5 activity to which primary isolates might be sensitive, the second part of the study addresses the influence of SERINC5 on Env, as detected by the host immune system. The authors saw enhanced sensitivity of SERINC5-containing virions to neutralization by antibodies bearing laboratory or primary Env proteins; the enhanced sensitivity was observed with antibodies that target cryptic epitopes that are exposed during fusion. The impact of SERINC5 on folding of the fusion machinery was further verified utilizing inhibitory peptides that bind to the pre-hairpin intermediate, blocking formation of the active fusion machinery. Again, the presence of SERINC5 in the absence of Nef caused an increased susceptibility to these peptides.
Although the precise mechanism by which Serinc5 impairs the correct folding of the fusion machinery remains unknown, the data by Sood and colleagues, as well as by other investigators (6, 7), are consistent with a perturbation of Env folding, leading to increased exposure of cryptic neutralizing epitopes. A slowing of fusion could explain the reduced infection efficiency; alternatively, the observation of increased binding of free virus by neutralizing antibodies in the absence of receptor, suggests that SERINC5 may promote Env adopting a more “open” conformation, comparable with that seen during fusion (8), thereby increasing the sensitivity to neutralizing antibodies. This could be interpreted as SERINC5 triggering a premature activation of the fusion machinery and may explain the greater resistance of primary isolates to the effects of SERINC5; Env proteins of primary isolates have been shown to resist exposure of potentially neutralizing epitopes to a greater degree than laboratory isolates (9). The closed structure of primary isolates is presumed to play a role in concealing conserved functional motifs from neutralizing antibodies, and Nef has also been linked to this protection (10). SERINC5 and Nef appear, therefore, to be significant contributors to the game of hide-and-seek played between HIV-1 and the host's antibodies.
The detailed examination by Sood and colleagues of laboratory isolates clearly indicates a defect in small-pore formation, although their data also suggest this is may not be the major factor restricting primary isolates. Rather, the ability of SERINC5 to increase virion susceptibility to neutralizing antibodies by interfering with the folding of the fusion machinery provides the most compelling explanation yet for the mechanism of this restriction factor (Fig. 1). Although the story is not yet complete and the molecular interactions between the proteins involved remain unknown, this work provides a strong argument that a thorough understanding of antiviral mechanisms often requires the context of the host-immune response.
This work was supported in part by National Institutes of Health Grants AI120860 and GM103368. The authors declare that they have no conflicts of interest with the contents of this article.