Sexual transmission of HIV-1 is suspected to be increased when HSV-2 is shed in the genital tract of dually HIV-1- and HSV-2-infected individuals [19, 20]. HSV-2 may increase the infectiousness of HIV-infected subjects, by increasing genital HIV load during an HSV-2 genital recurrence [5–8] through the transactivation of HIV-1 LTR by interaction with HSV proteins (ICPO, ICP4) or the production of pro-inflammatory chemokines known to enhance HIV-1 replication [9, 10]. The recruitment of activated CD4+ cells that markedly upregulate HIV replication in HSV-infected lesions  may also account for the high titer of HIV in genital HSV lesions. Other mechanisms increasing the infectiousness of HIV by HSV-2 may be hypothesized. HIV-1 has been reported to undergo pseudotype formation in vitro when cells are coinfected with HIV-1 and other enveloped viruses by acquiring in the HIV-1 envelope proteins of the co-infecting virus, including human T lymphotropic virus type 1 [21, 22], HIV-2 , endogenous murine retroviruses [24–26], vesicular stomatitis virus , and HSV-1 [14, 15]. Such pseudotype formation results in altered cell tropism for the next round of infection [21, 22]. Pseudotype formation between HIV-1 and HSV-2 could have an impact on sexual transmission by facilitating HIV-1 epithelial cell infection within the genital mucosa.
In the present study, we investigated whether HIV-1 and HSV-2 can interact at the cellular level by forming HIV-1 hybrid virions pseudotyped with HSV-2 envelope glycoproteins, as was previously reported for HSV-1 [14, 15]. We report that HIV-1 produced from HIV-1 and HSV-2 co-infected mononuclear cells and epithelial cells releasing high titers of both HIV-1 and HSV-2 does not acquire an increased tropism to naturally non-permissive cells. We also find no evidence of incorporation into the HIV-1 envelope of gB or gD, which are essential for the binding and entry of HSV-2 [27, 28]. In order to avoid the deleterious cytopathic effects of HSV-2 infection, we used plasmids coding for gB and gD. Despite microscopic observations of transfected cells expressing both gp160 and gB or gD and a co-localization of HIV-1 envelope proteins and HSV-2 glycoproteins, HIV-1 virions did not acquire either gB or gD. We cannot completely exclude the formation of HIV-1 pseudotypes by the incorporation of other HSV-2 proteins in the HIV-1 viral membrane. However, our observations demonstrate that no significant level of surface viral proteins involved in HSV-2 cell entry was incorporated, since no HSV-2 tropism was detected in HIV-1 particles. Nor can we exclude pseudotype formation by the integration of HIV-1 genes into the HSV-2 genome or HSV-2 genes into the HIV-1 genome, as was described in chickens with the Marek disease herpesvirus and the reticuloendotheliosis lentivirus, which can both induce T lymphomas in chickens and often coexist in the same animal [29, 30]. However such hybrid virions do not acquire a modified cell tropism [29, 30]. We observed partial colocalization of gp160 with gB and gD at the plasma membrane level of co-infected cells and HIV-1-infected P4P cells transfected with a plasmid coding for gB or gD. Such slight co-localization may be explained by the major intracellular expression of gB in early endosomes  and that the final HSV-2 envelope is thought to be acquired in the trans-Golgi network [28, 32] while gp120 is preferentially expressed at the cell surface [33, 34]. In addition, the long cytoplasmic domain of gB with 112 amino acids could be sterically inhibitory to be incorporated in a retroviral particle [35–38]. These features are consistent with the lack of gB and gD incorporation into the HIV-1 envelope.
Other studies analyzing interactions between HIV and herpesviruses did not provide evidence for the formation of HIV-1 pseudotypes . Similarly, Van Kuyk and colleagues have reported that HIV-1 obtained from EBV productively infected lymphobastoid cell lines did not acquire the EBV tropism . Toth and colleagues were unable to detect HIV-1/HCMV pseudotypes in supernatant fluids from dually infected syncytiotrophoblasts .
Our observations contrast with previous reports on HIV-1/HSV-1 pseudotype formation [14, 15]. Thus, cell-free culture supernatant from dually infected lymphocyte cell lines was able to induce HIV productive infection in cells normally resistant to HIV infection, but there was no direct evidence for incorporation of HSV-1 proteins into the HIV-1 envelope [14, 15]. Zhu et al reported the co-infection of HIV-producing H9 cells with HSV-1 tsJ12, a temperature sensitive mutant of HSV-1 . Inoculation of HeLa cells with supernatant from the coinfected H9 cells resulted in HIV production, which was not seen with supernatant from H9 cells infected only with HIV-1. However, the level of p24 antigen from H9 cells, which could be higher in co-infected H9 cells as a consequence of HIV transactivation by HSV, was not determined before inoculation of Hela cells. Calistri et al have used HIV-1IIIB produced in H9 cells and HSV-1 to co-infect CEM cells and CD4- Vero cells with the aim of evaluating pseudotype formation . High production of HIV was observed when Vero cells were infected with the CEMHSV/HIVcell-free supernatant. However, the levels of HIV replication in CEMHSV/HIV and in control CEMHIV cell-free supernatants before Vero infection, and those in Vero supernatant infected with the CEMHIV cell-free supernatants were not mentioned. In addition, HIV-1 proviral DNA was detected in Vero cells infected by the CEMHSV/HIVcell-free supernatant, but the negative control consisted of Vero cells infected with supernatant of H9 cells chronically infected by HIV-1, instead of Vero cells infected with CEMHIV cell-free supernatants. Prior HSV-1 infection of CEM cells has been shown to result in a delay of HIV-1 production, due to cell growth inhibition and apoptosis . Thus we cannot exclude a significant release of HIV-1 proviral DNA in CEMHSV/HIVsupernantant, leading to the false positive detection of HIV-1 proviral DNA in Vero cells.
In conclusion, we find no evidence of in vitro HIV-1/HSV-2 pseudotype formation in mononuclear cells or in epithelial cells. These findings suggest that the existence of HIV-1/HSV-2 pseudotype formation in the genital tract of HIV-1- and HSV-2- co-infected individuals is unlikely, and that genital HIV-1 particles highly produced under the influence of HSV-2 genital replication, as observed in HIV-1- and HSV-2- coinfected women, do not acquire properties of an epithelial cell-tropic virus, such as HSV. Finally, the synergistic interactions between HIV-1 and HSV-2 at the genital level more likely results in increased genital inoculum of both viruses, rather than in phenotype changes of genital HIV-1.