Although several studies have suggested a specific role for HIV-1 Vpr in facilitating the nuclear localization of the viral DNA during infection of non-dividing cells, such as macrophages, there is still no evident correlation reported in the literature between its real contribution to this process and the known functions of Vpr in vitro, including its ability to accumulate at the NE. Given that Vpr is dynamically associated with the NE [25,28], this subcellular distribution may be a pre-requirement for one or more of its known functions. Based on the findings that Vpr is able to interact directly with some proteins of the NPC, including the nucleoporin hCG1 , we have now identified two Vpr mutants, Vpr-L23F and -K27M, that are specifically deficient for hCG1-binding. Both mutations similarly abrogate Vpr concentration at the NE both in HeLa cells and in primary human monocyte-derived macrophages, supporting the hypothesis that this nucleoporin participates in the docking of the protein to the NPC. To our knowledge, it is the first report confirming that HIV-1 Vpr efficiently accumulates at the NE in primary macrophages. However, direct evidences regarding the specific role of hCG1 in the NE concentration of Vpr are still missing, since we failed so far to significantly deplete the endogenous hCG1 protein by using the RNA interference technology.
While substitutions of the Leu23 residue have been described previously [37,38], the mutation at position 27 (K27M) was not yet identified and is particularly interesting since the K27 is well-conserved among HIV-1 isolates and constitutes the only lysine residue along the whole Vpr sequence. This residue may potentially constitute a site for post-translation modifications, such as methylation, acetylation, hydroxylation, sumoylation or ubiquitination . None of these modifications have been described previously for Vpr and our western blot analysis did not reveal any change in the level of expression and/or stability of the Vpr-K27M mutant compared to the wt protein. Interestingly, both Leu23 and Lys27 residues are located in the first N-terminal α-helix H1 (residues 17–33) of Vpr which has amphipathic properties (see on Fig. 2). The structural analysis confirms that the 3D structure and the stability of the three α-helices of Vpr are not significantly affected by the L23F and K27M substitutions, as indicated by the overall conservation of the hydrogen-bonding network of the protein. This analysis also reveals that the Leu23 and Lys27 residues are located in a close proximity at the end of the first α-helix, in a pocket that is easily accessible for protein-protein interaction with cellular partners, such as nucleoporins. We can notice that the two other mutations, A30L and F34I, impairing the NE accumulation of Vpr involve amino acids located on the same face of the first α-helix of the protein (see on Fig. 2D). However, Ala30 and Phe34 are not accessible and are rather involved in the stability of the structure by establishing hydrophobic interactions with residues of the third α-helix (55–77) of Vpr (see on Fig. 2A). Proximities between Ala30 and Leu64/Leu68 and between Phe34 and Leu64/Leu67/Leu68 have been identified from NMR experiments, indicating that Ala30 and Phe34 are directly involved in the interaction between the first and the third α-helix. Any mutation of the residues found at this interface will likely perturb the structure of the Vpr protein.
Our functional analysis raises intriguing questions concerning the functional link between Vpr accumulation at the NE and the in vitro properties of the protein, namely G2-arrest and cell death induction. Like other Vpr mutants that fail to localize at the NE, such as Vpr-A30L and -F34I, both L23F and K27M substitutions affect the Vpr-induced G2-arrest and cell death. These observations highlight two important points: i) the functional link between Vpr docking at the NE and its ability to cause a G2-arrest, and ii) the link between the G2-arrest and the pro-apoptotic activities of Vpr. First, it is possible that the accumulation at the NE could constitute a prerequisite for the Vpr-induced G2-arrest . It was reported that Vpr provoked herniations and transient ruptures of the NE, resulting in a mixing of cytoplasmic and nuclear components that could contribute to the cell cycle arrest. Nonetheless, the molecular mechanism underlying this process is still unknown, and the local bursting caused by Vpr at the NE in this report has not been confirmed to date, even in imaging experiments performed on living cells . Alternatively, the concentration of Vpr at the NE, or in a close vicinity of the nuclear pore complexes, could be required for the establishment of local interactions with some cellular partners involved in the regulation of the cell cycle. Because no mutant of HIV-1 Vpr has been described so far as disrupting the NE accumulation but keeping intact G2-arrest activity, our results confirm these observations and suggest that Vpr accumulation at the NE may be required. However, targeting to the NE is not sufficient for the G2-arrest activity, since several Vpr point mutants with substitutions in the C-terminal basic region of the protein, such as Vpr-R90K and Vpr-R80A, have been reported as defective for the G2-arrest activity (present study and Refs. [31,40,41]), while we show that both can still accumulate at the NE. Recent studies have shown that these residues may be involved in the direct recruitment of an unknown cellular factor that is required for G2-to-M transition [42-46].
Second, our results show that the pro-apoptotic activity of Vpr totally parallels the Vpr-induced G2-arrest, confirming previous reports suggesting that apoptosis is a direct consequence of the prolonged cell cycle arrest induced in Vpr-expressing cells [47-51]. Whereas other authors have suggested that these two Vpr properties were separated [22,41,52-54], it was recently demonstrated that the Vpr-induced apoptosis is directly dependent on the cell cycle arrest . Both Vpr activities are directly related to the activation of G2 checkpoint ATR-initiated DNA damage-signaling pathway.
Though it is admitted that Vpr displays evident affinity for the NE [21-23,25,28], the biological significance of this property for efficient virus replication in non-dividing cells has been explored only in a few reports [25,29]. While a previous study showed that virus expressing single-point mutants (Vpr-F34I and -H71R), which failed to concentrate at the NE, displayed decreased infectivity in macrophages , it was recently reported that a Vpr mutant with multiple substitutions, in which the 4 Leu residues, including Leu23, found within the first α-helix of the protein were replaced by Ala, failed to localize in the nucleus and rendered the virus unable to replicate in MDMs from two donors . Here, we show that the substitution of the Leu23 residue is sufficient to impair both nuclear accumulation of the Vpr protein and virus replication in primary macrophages. In addition, mutation of the Lys27 residue found in a close proximity within the first helix similarly impaired HIV-1 replication in MDMs from most donors. While the Vpr-L23F and -K27M mutant viruses displayed a wild-type level of replication in only one out of four donors, we cannot formally exclude that the replication defect observed in the three other donors may be related to the slight difference in the level of virion incorporation of the two Vpr mutants evidenced in Fig. 6A and 6B. As reported previously, the first α-helix of HIV-1 Vpr contains the main determinants required for efficient incorporation into virions [12,20,36,37].
Primary cells and especially MDMs are heterogeneous cell populations that greatly vary in their susceptibility to HIV infection [5,56]. Moreover, it is most likely that the contribution of Vpr for virus replication in macrophages relies on more than one of its activities. In addition to its role in nuclear import of the viral DNA, Vpr displays several other activities that could be important to maintain efficient virus replication in macrophages. Since it was reported that the manipulation of the cell cycle by Vpr increased viral expression in both dividing and non-dividing cells [18,57], the replication defect observed here with the mutant viruses could be, at least in part, related to the impairment of the Vpr-mediated G2-arrest. Additionally, the Vpr effect on the fidelity of the reverse-transcription process was also directly correlated with the ability of HIV-1 to efficiently replicate in primary macrophages . These findings confirm that the absence of Vpr expression is deleterious for HIV-1 replication in primary macrophages, but additional analyses are required to re-evaluate the relationship between the multiple functions of Vpr characterized in vitro and its requirement in vivo for efficient replication in the non-dividing target cells of HIV-1.