In this study, we confirmed adoption of an "open" Env to expose the CD4 binding site for efficient CD4 binding and infection of primary macrophages in two additional RPs with coreceptor switch. Moreover, changes in neutralizing antibody and T20 sensitivity around the time of X4 virus emergence are consistent with a more responsive quaternary configuration of the viral envelope spike. In agreement with reports in HIV-1-infected individuals, high virus replication is a strong predictor for R5-to-X4 conversion in SHIVSF162P3N-infected macaques. A certain level of viremia may provide the conditions optimal for coreceptor switching or, alternatively, the Env changes required for coreceptor switching may predispose to higher levels of virus replication. Adoption of an "open" Env conformation that confers structural flexibility to accommodate mutational changes, and high levels of virus replication to overcome the genetic hurdles for R5-to-X4 evolution therefore contribute to the coreceptor switch in nonhuman primates. Since an "open" Env conformation also renders the virus more susceptible to antibody neutralization, our findings help to explain the infrequent and late appearance of X4 virus in HIV-1 infection when the immune system deteriorates.
Diminished entry efficiency of the evolving R5 viruses following the time of switch, with increase in PSC-RANTES sensitivity that is suggestive of poorer CCR5 usage was seen in DG08 studied here (Figure 2) and in BR24 and CA28 reported earlier . Interestingly, these are the three R5 SHIVSF162P3N-infected RPs that harbored pure X4 virus populations. Decrease in the efficiency of CCR5 usage had also been reported in HIV-1 infected individuals with detectable CXCR4-using variants [6, 34, 35], suggestive of a decline in R5 fitness. HIV-1 and SHIV viruses with a pure X4 phenotype generally have higher replication kinetics in vitro [60–64] and in vivo [65–68]. Coupled with the broader cellular host range of X4 viruses in vivo [69, 70], they are likely to out-compete R5 viruses once established. The decrease in the selective pressure for R5 viruses to maintain optimal replicative capacity and fitness with X4 emergence therefore, could be a reason for the decrease in entry efficiency and CCR5 utilization of R5 viruses in SHIV-infected macaques with pure X4 viruses. Nevertheless, R5 variants persist following the appearance of X4 viruses in these animals, perhaps because of their distinct target cell range [69, 70].
Transmitted/founder viruses in acute and early HIV-1 infection have been reported to be sCD4 resistant and to replicate poorly in monocyte-derived macrophages that express low levels of the CD4 receptor [3, 71, 72]. Neutralization resistance in vitro is often accompanied by an increase in HIV-1 dependence on CD4 for entry [73–75]. Indeed, founder viruses were shown to require high levels of CD4 for entry , which may explain their deficiency in infecting CD4low cells. Consistent with these findings in humans, viruses that are sCD4 resistant and infected primary macrophages less efficiently predominated during acute infection in R5 SHIVSF162P3N-infected RP macaques DE86, BT78, CC39 and DN57, and in the two RPs we previously studied , regardless of coreceptor switching. The exception was DG08, where the acute (w2) viruses were observed to be sCD4 sensitive but did not infect macrophages efficiently. This however, was rapidly replaced two weeks later by variants that are more resistant to sCD4, reinforcing the view that these viruses are selected for during primary infection. As this was observed despite intravenous inoculation (BR24, CA28, DE86, BT78, CC39) which circumvented mucosal barriers, and in macaques that failed to mount a strong humoral or cellular immune response, these factors are unlikely to be playing a role in the early expansion of sCD4-resistant viruses. We speculate that the abundance of target cells with high levels of cell surface CD4 expression in the new hosts maybe an underlying determinant for the expansion and propagation of viruses that are more dependent on CD4 for entry during acute SHIVSF162P3N infection in macaques and HIV-1 infection in humans. Further research on the importance of CD4 receptor density in HIV infection as well as the kinetics of virus replication in infected cells may help to understand the selective forces that govern the biological phenotype of HIV-1 transmitted founder viruses.
Enhanced infection of primary macrophages was seen in the RP macaques at end-stage disease irrespective of coreceptor switch, indicative of a strong selective pressure to replicate in cells with lower CD4 cell-surface expression levels over the infection course. Adaptation to macrophages has also be observed in SIV-infected RP macaques , and isolates from late in disease of HIV-1 infected individuals have been reported to be more macrophage-tropic compared to those from early in infection [78–80], suggestive of similar selection pressure in humans and rhesus monkeys. Presumably, as discussed above, ongoing virus replication in both hosts leads to loss in target cells (e.g., memory T cells) with high receptor expression levels, putting pressure on the virus to change so that it can use CD4 more efficiently to infect alternate CCR5+ cells such as macrophages that express low amounts of receptor. Infection-associated immune activation may also decrease receptor density, since CD4 is down-regulated with T-cell activation [81–83]. In RPs with coreceptor switch (BR24, CA28, DE86, DG08), infection of CD4low cells was achieved by adoption of a less constrained and more “open” Env conformation that exposes the CD4 binding site for better CD4 binding, allowing flexibility in accommodating the structural remodeling needed for the change in coreceptor preference. The V3 loop mutations that confer CXCR4 usage in R5 SHIVSF162P3N-infected RP monkeys and in HIV-1 infected individuals also render the virus sCD4 sensitive, consistent with the notion that improved CD4 binding is a necessary prerequisite for acquisition of CXCR4 use [23, 84, 85]. In RP macaques that did not switch, R5 viruses appeared to have evolved to compensate for inadequate CD4 binding by improving CCR5 usage and/or the requisite post-binding conformational changes for virus entry in response to non-optimal CD4. It is possible that given time, these latter macaques (e.g. BT78, CC39) could also adopt an “open” Env conformation to facilitate CD4 binding and infection of primary macrophages, setting the stage to tropism switch. Alternatively, the presence of host immune response may be a hindering factor. This is best illustrated in DN57, which, contrary to the other RP monkeys, mounted a persistent albeit weak anti-SHIV antibody response. R5 viruses evolved in this macaque to be highly sCD4 sensitive, with IC50 values comparable to those seen in DE86, suggestive of a similar strong selective pressure for better CD4 usage. The evolving R5 viruses in DN57 also infected macrophages efficiently, but this was not accompanied by a corresponding increase in CD4 binding that is indicative of exposure of the receptor binding site. Rather, the Envs in DN57 evolved to adapt to less-than-optimal CD4 density by changes in receptor and/or coreceptor contact, or by acquiring a higher propensity for proceeding to fusion. Thus, although the pressure to infect cell targets that express low levels of CD4 is present in SHIVSF162P3N-infected RP macaques regardless of coreceptor switch, the mechanisms employed to overcome this selective pressure vary depending on other conditions in the host.