- Top of page
- Materials and Methods
Dendritic cells (DC) are crucial components of the early events of HIV infection. Dendritic cells capture and internalize HIV at mucosal surfaces and efficiently transfer the virus to CD4+ T cells in trans through infectious synapses (trans-infection pathway). Alternatively, HIV-1 replicates in DC (R5-HIV-1) (cis-infection pathway). Here, we analyzed HIV trafficking in DC during the trans-infection pathway as well as the cis-infection pathway. Confocal immunofluorescence microscopy demonstrated that after capture by DC, R5-HIV-1 and HIV-1 pseudotyped with vesicular stomatitis virus protein G colocalized in a viral compartment enriched in tetraspanins including CD81, CD82 and CD9, although at different levels, indicating a role of the viral envelope in targeting to the tetraspanin-rich compartment. Replication of R5-HIV-1 in DC (cis-infection pathway) also led to the accumulation, in an envelope-independent manner, of mature viral particles in a tetraspanin-rich compartment. A fraction of the HIV-1-containing compartments appeared directly accessible from the cell surface. In sharp contrast with the trans-infection pathway, the δ-subunit of the adaptor protein 3 (AP-3) complex was enriched on the HIV-1-containing compartment during R5-HIV-1 replication in DC (cis-infection pathway). Downregulation of AP-3 δ-adaptin reduced significantly viral particle release from HIV-1-infected DC. Together, these studies demonstrate a role for AP-3 in HIV replication in a tetraspanin-rich compartment in DC and contribute to the elucidation of the trafficking pathways required for DC–T cell transfer of HIV-1 infection, a critical step during the early events of HIV infection.
Model systems have demonstrated that dendritic cells (DC) are crucial mediators of the early events of HIV-1 infection (reviewed in 1–4). As professional antigen-presenting cells (5), DC reside in the skin and mucosae in an immature state and migrate from the periphery to the secondary lymphoid organs upon pathogen encounter.
In the absence of viral replication, DC/Langerhans cells are able to capture and internalize intact HIV-1 at mucosal surfaces and successfully transmit infection in trans to CD4+ T cells through an infectious synapse (6–12). Although the in vivo contribution of this pathway (trans-infection pathway) remains to be determined, the rapid transfer of captured infectious HIV-1 to mucosal CD4+ T cells (prior to replication) could very well establish HIV-1 replication locally, before a subsequent systemic dissemination [(13) and reviewed in 4,14].
HIV-1 infection of DC is difficult to detect both in vitro and in vivo, with HIV-1 replication in DC occurring at much lower levels than in CD4+ T cells (reviewed in 1,2). Nevertheless, DC can be preferentially infected with chemokine receptor 5 (CCR5)-using HIV-1 strains (R5-HIV-1) (15–18) and have been shown to subsequently transfer newly synthesized HIV-1 infectious particles to CD4+ T cells (7,19). This replication pathway (cis-infection pathway) is believed to have a central role in HIV-1 infection of CD4+ T cells. As DC migration to lymph nodes is a prerequisite for HIV-1 to gain access to the replication-competent site, replication in DC, even if minor, would allow the retention of an infectious HIV-1 population within DC prior to its transfer to CD4+ T cells [(13) and reviewed in (4,14)]. The preferential replication of R5-HIV-1 over chemokine receptor 4 (CXCR4)-using HIV-1 (X4-HIV-1) in DC appears to result from differential levels of coreceptors expression on the surface of immature DC (iDC) and is also partly due to a block during viral fusion with target iDC (16,17). However, HIV-1 infection of DC is still insignificant, regardless of the viral tropism, when compared with activated CD4+ T cells. The cellular HIV-1 restriction factors APOBEC3G/3F were recently shown to be responsible for the defective HIV-1 infection of DC (20). The potent host antiretroviral factors function at a post-entry level, in the same way as for resting CD4+ T cells (21).
While HIV-1 is usually considered to assemble at the plasma membrane of infected cells (reviewed in 22), several observations have hinted at the possibility that intracellular organelles could play a part in this process (23,24). Studies in primary macrophages have shown accumulation of HIV-1 components at the membranes of multivesicular bodies (MVB) as well as budding of viral particles into such organelles (23,24). Although a matter of debate (25–27), these data underline a potential cell-specific role for a cell-type-specific specialized compartment in HIV-1 assembly and infection and raise the question as to whether this could also be the situation in DC. HIV-1 budding requires the recruitment of many cellular proteins to the site of viral particle formation. Mammalian homologues of the yeast vacuolar protein sorting (VPS) proteins, such as Tsg101, have been described as essential for HIV-1 assembly and budding (28–31). Recently, however, the role of the adaptor protein complex family (AP-1 to AP-4) (reviewed in 32) on HIV-1 components trafficking has been investigated (33–35). Among those adaptors, the δ-adaptin subunit of the AP-3 complex has been shown to target HIV-1 Gag to late endosome/MVB and exerts an important role in HIV-1 assembly and production (36).
We showed previously that non-replicating X4-HIV-1 is internalized into a non-conventional, non-lysosomal, tetraspanin-rich compartment in DC prior to its subsequent presentation to T cells (trans-infection pathway) (37). Here, we describe a role for the tetraspanin-rich compartment in replication (cis-infection pathway). We show the differential recruitment of the δ-adaptin subunit of AP-3 to this compartment after HIV-1 replication in DC when compared with the trans-infection pathway. We also demonstrate a role for AP-3 δ-adaptin in viral particle release from HIV-1-infected DC.
- Top of page
- Materials and Methods
In this study, we characterize the tetraspanin-rich compartment in which HIV-1 accumulates during the trans- and cis-infection pathways, prior to the subsequent transmission of infectious viral particles to CD4+ T cells across infectious synapses (6–8,10,11,37,43,44).
We described previously that X4-HIV-1 captured by DC was internalized into a novel tetraspanin-rich compartment containing MVB (37). Here, we show using confocal immunofluorescence microscopy that R5-tropic HIV-1 is also internalized in a compartment characterized by the presence of the tetraspanin CD81, CD9 and CD82. Although this compartment contained some CD63 and HLA-DR, it was distinct from HLA-DM and LAMP-1-containing lysosomes, as previously described for X4-HIV-1 (37).
Although X4-tropic and R5-tropic HIV-1 accumulated in a CD81-rich viral compartment during the trans-infection pathway, we observed that the percentage of HIV-1 particles colocalizing with CD81 labeling was significantly different (∼90 and ∼65%, respectively). In order to evaluate the role of the viral envelope in the targeting to the compartment, we used VSVG-pseudotyped particles. Unlike HIV-1 (45), VSV requires the acidic pH-dependent triggering of its envelope protein for fusion with endosomal membranes during entry into the cells (46,47). We could demonstrate that HIV-1 targeting to the tetraspanin-rich compartment is envelope dependent during the trans-infection pathway, with a decrease in HIV-1/CD81 colocalization when we compared X4-HIV-1 with R5-HIV-1 or VSVG-HIV-1Δenv. It is possible that differences in the extent of colocalization with the tetraspanin CD81 might result from differential fusion properties of VSVG-, X4- or R5-HIV envelopes with iDC (16,17).
Infection of DC, although poor in comparison with CD4+ T cells, is likely to play a part in the systemic propagation of HIV infection in patients by transferring HIV-1 infection to CD4+ T cells after migration to lymph nodes (cis-infection pathway) (7,48). Using confocal immunofluorescence microscopy, we demonstrated that, during the cis-infection pathway, replicating R5-HIV-1 accumulated in DC in a clustered compartment containing the tetraspanins CD81, CD82 and CD9. We also observed that VSVG-HIV-1Δenv-infected DC accumulated virus in a CD81-rich compartment 6 days post-infection, indicating that targeting to the tetraspanin-rich viral compartment in DC appears envelope independent during the cis-infection pathway.
Upon or after budding, HIV-1 is known to go through a maturation stage during which p55gag is proteolytically cleaved and the capsid (CA) and nucleocapsid (NC) proteins reassembled into a conical core structure in the center of the viral particle (reviewed in 49,50). Using a monoclonal anti-p17MA antibody that specifically recognizes p17MA only, following cleavage from p55gag, we observed that HIV-1 virions present in the CD81-rich viral compartment during the cis-infection pathway are mature viral particles. Although it is difficult to completely rule out the possible recapture of newly assembled HIV-1 virions budding from the surface, the strong presence of mature HIV-1 particles in the tetraspanin-rich compartment suggests that this compartment is a major site of viral assembly in iDC.
HIV-1 Gag was shown to accumulate at membranes of MVB in human primary macrophages, directing HIV-1 subsequent assembly and budding steps to an intracellular compartment (23,51–54). This model of cell-type-dependent budding of HIV-1 in a MVB-containing compartment argues in favor of the DC tetraspanin-rich compartment as a HIV assembly site. However, this model is still a matter of vivid debate. Recent findings argue against a significant production of HIV-1 in MVB-containing late endosomes in macrophages, favoring instead HIV-1 assembly and budding from the plasma membrane, regardless of the cell type (25–27). In fact, these sites of viral budding containing MVB appear to be close or connected to the cell surface of macrophages, underlying an unexpected complexity between the plasma membrane and MVB-containing structures (26,27). In DC, our results demonstrate that only a fraction of HIV-1-containing compartments (17.5%) is accessible from the cell surface. Of note, our assay is performed at 4°C and may underestimate the number of compartments directly accessible from the extracellular milieu over a more prolonged period of time.
We showed that the δ-adaptin subunit of the AP-3 complex, a component of HIV-1 assembly machinery, is specifically enriched in the tetraspanin-rich/virus-containing compartment during HIV-1 replication in DC (cis-infection pathway) but not after HIV-1 capture by DC (trans-infection pathway). The direct interaction between HIV-1 Gag and AP-3 δ-adaptin was shown to target HIV-1 Gag to MVB and to play a functional part in HIV-1 assembly in cell lines (36), making AP-3 δ-adaptin a potential cellular factor required for HIV-1 assembly. RNA interference allowed us to demonstrate that AP-3 δ-adaptin plays a significant role in the production of HIV-1 from DC as its downregulation decreased viral particle release from HIV-1-infected cells (cis-infection pathway) (Figure 7). Concomitantly, AP-3 δ-adaptin downregulation increased cell-associated HIV-1 p24gag amounts in infected DC. Morphology of the virus-containing compartment did not show major alterations in the absence of AP-3 δ-adaptin, but we cannot rule out that the residual levels of AP-3 δ-adaptin present in DC after RNA interference are sufficient for an appropriate targeting of HIV-1 p24gag to the tetraspanin-rich HIV-1-containing compartment.
Finally, we demonstrate that HIV particles present within the tetraspanin-rich compartment in DC during the trans- and cis-infection pathways relocate to infectious synapses upon contact with autologous CD4+ T cells. As we showed previously in DC–Jurkat CD4+ T cell clusters (37), CD81, as opposed to CD63, is enriched at the DC–T cell infectious synapse, in accordance with CD81 role in the immunologic synapse (55–59).
In conclusion, our studies characterize a tetraspanin-rich compartment targeted by HIV-1 during the trans- and cis-infection pathways. Our data reveal that the DC trans-infection pathway and the DC cis-infection pathway intersect at the level of this tetraspanin compartment and share components of their respective intracellular trafficking machinery. Our study highlights the presence of HIV-1 during replication in iDC in a tetraspanin-rich compartment sharing some similarities with the MVB-containing compartment observed in macrophages during HIV-1 replication (24,26,60). Furthermore, we describe for the first time the selective recruitment of AP-3 to this tetraspanin-rich compartment in iDC during the cis-infection pathway but not during the trans-infection pathway. We also show a direct role for AP-3 δ-adaptin in viral particle release from infected DC, indicating that the tetraspanin-rich compartment is a likely site of production of viral particles. We do not rule out a contribution of direct budding from the cell surface potentially involving tetraspanins as observed in CD4+ T cells (40–42), but our data show that a significant fraction of HIV production in iDC is likely to result from production in a compartment enriched in tetraspanins as well as AP-3 δ-adaptin. Together, these studies highlight the cellular machinery required for DC–T cell transfer of HIV infection, an essential step in the early dissemination of HIV infection.