To further develop a small animal model that is highly permissive for de novo infection by HIV-1, we sought to expand on our earlier results in hCD4/hCCR5-tg rats [3, 7, 12, 13]. Here, we generated rats that transgenically express hCycT1 in a cell type-specific manner to explore their suitability for enhancing the susceptibility of this important cell type in the multi-tg rat model.
In infected rat T-cells, hCycT1 transgene expression boosted early HIV gene expression ~3- to 7-fold compared to n-tg littermates. Importantly, cultures from ~1/4 of hCycT1-tg rats reached levels of HIV gene expression within the average range of infected human T-cells. However, while hCycT1 expression was clearly required, no correlation could be established between steady-state expression levels of hCycT1 and the ability of infected cultures to support early HIV gene expression. This finding is consistent with a scenario in which endogenous factors in concert with hCycT1 play a decisive role in regulating HIV LTR-driven gene expression in rat T-cells. Alternatively, only a subset of CD4 T-cells may be capable of translating hCycT1 expression into an enhanced gene expression and in these cells transgene expression levels may still correlate with the functional impact.
In principle, this could involve the recruitment of transcription factors (e.g., NF-κB, SP1, NFAT) to the 5'-LTR as well as factors regulating the activity of P-TEFb. Interestingly, no cellular gene is as sensitive to the availability of P-TEFb as the genes of HIV-1 (for review ). So far, several positive regulators of P-TEFb and HIV-1 LTR transactivation have been reported, including bromodomain protein Brd4 [29, 30], NF-κB , and the DNA-dependent ATPase subunit Brm of the SWI/SNF chromatin-remodeling complex . Negative regulators have also been found, including the noncoding 7SK small nuclear RNA [33, 34], HEXIM1 , the DRB sensitivity-inducing factor (DSIF) , and negative elongation factor (NELF) . Furthermore, transcription factor recruitment at contiguous LTR regions is partly dependent on histone acetylation as well as the viral Tat protein . Conceivably, the interaction of such endogenous factors with the Tat/hCycT1-containing P-TEFb complex may be different in rats than in humans. Similarly, dominant-negative activities of CycT1 or CycT2 of rat origin have to be considered in a hCycT1-tg context.
Recently, Sun et al. reported that hCycT1 transgenesis in mice, employing the identical transgene vector, also resulted in a cell type-specific expression in CD4 T-cells, macrophages as well as microglia [25, 26]. Importantly, crossing HIV-1JR-CSF-tg mice, which carry two to four proviruses, with these hCycT1-tg mice revealed a marked increase in the production of infectious HIV-1 in both T-cells and macrophages . HIV-1 p24 concentrations in supernatants were still lower than the levels typically reached in dynamically infected human cultures. In a more refined characterization of T-cells from these hCycT1-tg mice, Zhang et al. demonstrated HIV-1 RNA expression per infected T-cell was only 10% of that of human references , indicating that the transcriptional deficit also in this rodent had only partially been overcome by hCycT1 transgenesis. Of note, this analysis required a secondary T-cell receptor stimulation to circumvent a peri-integrational block in primary mouse T-cells, which is absent in the rat species . Interestingly, infected rat macrophages pose an exception to the otherwise species-specific impairment at the level of early HIV-1 gene expression . As shown herein, tg expression of hCycT1 does not translate into an enhanced viral gene expression in this primary cell type, and already macrophages from n-tg rats are at a level comparable to human MDM. This may, in part, relate to the ability of HIV-1 to exploit a distinct set of nuclear transcription factors and alternative mechanisms of transcriptional regulation in macrophages compared to other cell types, including T-cells (for review [39–41]). In addition, hCycT1-dependence and species-specific differences of HIV LTR driven gene expression appear to be less pronounced or even absent when levels of gene expression per cell are low. This was observed both for macrophages (Fig. 5C) and T-cells infected with the HIV-1 strain R7/3 with apparently lower intrinsic LTR activity (Fig. 8C, compare left and right panel). Notably, macrophages from hCD4/hCCR5-tg rats are the only primary non-human cells reported to allow a productive HIV-1 infection, albeit at lower levels than in human MDM [4, 12, 42]. The general ability of macrophages to support HIV LTR-driven transcription and, related to that, their responsiveness to transgenic hCycT1 expression, appears to be remarkably different in rats and mice .
Zhang et al. also characterized the late phase in primary T-cells from hCycT1-tg mice ex vivo and report post-transcriptional defects at the levels of Gag expression, Gag processing, Gag release and virus infectivity . They estimated that the post-integration defects alone add up to a 300- to 500-fold reduction in the yield of infectious virus after a single cycle of HIV-1 replication. It is currently unclear why T-cells from provirus-tg mice appear to be much less restricted in supporting these steps of the HIV-1 replication cycle [24–26]. Limitations for several distinct late-phase steps in rodent cells have been proposed, including the Rev-dependent export of HIV-1 RNA [1, 9, 43–45] Gag trafficking and virion assembly [1, 9, 10, 45], both inhibitory and supportive functions of TRIM family members , as well as the Vif-resistance of mouse APOBEC3G . These findings are of high relevance for future analyses of the late-phase defect in T-cells from rats and possible additional genetic modifications of the host.
We identify HIV-1R7/3 YU-2 Env GFP as a virus, the behavior of which in tg rat T-cells ex vivo displays key characteristics of a spreading, primarily cell-to-cell-mediated infection: first, GFP expression from the nef locus, a surrogate for early viral gene expression in infected cells, increased continuously over periods of 2 weeks with peak levels comparable to human reference cultures. This was not a general property of replication-competent HIV-1 GFP reporter viruses since another R5 strain failed to recapitulate this phenotype. Second, this increase was sensitive to efavirenz addition 18 h p.i. and also not seen for an env -deficient single-round virus. This demonstrates the requirement for multiple rounds of replication involving reverse transcription. Third, the lack of significant p24 levels in supernatants from infected rat T-cell cultures and the successful transfer of infection to naïve cultures only through cells, but not supernatant from infected cultures, indicates a primarily cell-to-cell-mediated spread. Levels of p24 antigen per infected rat T-cell as well as the overall percentage of infected T-cells in culture may have been too low to detect a spread by quantification of cell-associated p24 levels. This challenging observation warrants further investigations. It will be interesting to investigate the genetic determinants in HIV-1R7/3 YU-2 Env GFP that underlie its ability to propagate in tg rat T-cells. Through such a genetic approach and forced adaptation of this or other HIV-1 strains for replication in the improved transcriptional context of triple-tg rat T-cells, the evolution of a highly rat-adapted HIV-1 strain may be feasible. Moreover, the molecular characterization of such an adapted strain could greatly facilitate the identification of host determinants that are critical regulators of late phase-steps of HIV replication.