Our findings demonstrate a novel association of Nef with the host cell exocytic machinery that has implications for understanding mechanisms involved in intercellular transfer of Nef and other HIV-1 proteins. We identified EXOC1, EXOC2, EXOC3, EXOC4, and EXOC6 as Nef-associated proteins via mass spectrometry analysis of Nef immunocomplexes isolated from Jurkat cells, and showed this association was disrupted by mutations that abrogate the ability of Nef to associate with and activate Pak2 kinase. Furthermore, association of wild-type, but not mutant Nef, with EXOC1, EXOC2, EXOC3, and EXOC4 was verified by co-immunopreciptation assays in Jurkat cells. Importantly, shRNA-mediated depletion of EXOC2 abrogated Nef-mediated enhancement of nanotube formation in Jurkat cells. Together, these results suggest that the exocyst complex is likely to be a key effector mediating Nef’s ability to promote nanotube formation, and may mediate some of its other functions as well (e.g. microvesicle secretion). DOCK2 and ELMO1 were previously reported as Nef-interacting proteins [29
]; however, we detected DOCK2 and ELMO1 in all of the mass spectrometry samples, including empty vector control samples, and the number of peptides detected did not differ significantly in the presence or absence of Nef, suggesting these proteins are not specifically associated with Nef (Additional file 1
: Table S1 and Additional file 2
: Table S2).
Our finding that Nef associates with EXOC1-4 and EXOC6 in Jurkat cells is consistent with a recent publication from the Krogan lab, which reported that Nef associates with EXOC4 based on affinity tagging and mass spectrometry analysis using a different Nef allele (HxB2 Nef fused to a 2X Strep and 3X FLAG purification tag) expressed in Jurkat cells [17
]. In this study, MiST (mass spectrometry interaction statistics) assigned this interaction a score of .769, where a threshold of .75 indicates significance [17
]. The authors were not able to reproduce Nef-EXOC4 association by co-immunoprecipitation assays in transfected 293T cells, however, raising the possibility that a cell-type-specific factor(s) may be required for Nef-EXOC4 association [17
Nef-mediated enhancement of T-cell activation requires stimulation via TCR or PMA/PHA (Figure
and ). Our proteomic analysis was performed using unstimulated Jurkat cells. Moreover, FACS analysis of CD25 expression detected little or no expression of T-cell activation markers (Figure upper panel and ). These observations suggest that the enhanced nanotube formation and secretion of microvesicles from Nef-expressing cells reported in prior studies [13
] may be due to a direct effect of Nef on the exocytic machinery, rather than an indirect effect of Nef-mediated enhancement of T-cell activation. Microarray data from Jurkat cells expressing SIV Nef indicated that exocyst complex components were upregulated only 1.5 to 2-fold [68
], and gene expression profiles of CD4+ T-cells from HIV-infected patients compared to uninfected individuals showed only minor effects on levels of transcripts encoding exocyst complex proteins [69
]. Thus, our mass spectrometry data is likely to reflect direct or indirect association of Nef with the exocyst complex, rather than Nef-induced enhancement at the level of gene expression or enhancement of T-cell activation.
In light of our finding that Nef associates with exocyst complex subunits, protein interaction network modelling (Figure ), EXOC2 shRNA-mediated inhibition of nanotube formation in Nef-expressing cells, and review of literature, including prior reports that Nef enhances microvesicle secretion and nanotube formation [13
], we propose a potential model to unify these pathways (Figure ). Our proposed model is based on established Nef-mediated enhancement of endocytosis and exocytosis, and predicted Nef-mediated enhancement of exocyst complex assembly. Nef binding to adapter protein 2
μ-subunit (AP2M1) enhances endocytosis and trafficking of Nef to recycling endosomes [70
]. Upon Nef-mediated activation of Pak2, Pak2 activates Aurora-A [71
], which phosphorylates a major exocyst complex assembly regulator, RalA GTPase [72
]. RalA mediates assembly of exocyst complex components, enabling polarized docking with membrane-associated EXOC1 and EXOC7 [58
] on endosomes and/or recycling endosome-derived vesicles. Interaction of Nef with a protein complex that includes the exocyst complex and RalA may therefore lead to Nef-mediated formation of nanotubes. Non-filamentous assembly of actin may lead to formation of nodules filled with exocyst-tethered vesicles, which transfer to bystander cells following plasma membrane rupture. Thus, our findings, together with previous studies [13
], suggest potential involvement of RalA and the exocyst complex in Nef-mediated formation of nanotubes and enhanced release of exosome-like vesicle clusters.
Figure 6 Proposed model of Nef-mediated enhancement of nanotube formation via the exocyst complex. Based on proteomics data and protein interaction networks shown in Table and Figure , a model of Nef-mediated enhancement of nanotube (more ...)
Interactor(s) that mediate Nef association with exocyst complex subunits are not yet clear. One possibility is that Nef may interact directly with a component(s) of the exocyst complex. In this scenario, Nef might function as a viral homolog of a cellular exocyst complex regulator downstream of Pak2 activation. Among the exocyst complex components detected in our mass spectrometry analysis, the greatest number of peptides were derived from EXOC4 (Table ), the same component identified by Jager et al.
as a Nef-interacting protein in their proteomic screen [17
]. Taken together, these data suggest EXOC4 represents a potential candidate for an exocyst complex protein that may bind directly to Nef. Given predicted protein interaction networks (Figure ), Nef-AP2M1 binding is another potential means of Nef-exocyst complex association. Co-immunoprecipitation of Nef with Rab11 has been reported [77
], raising the alternative possibility that Rab11 might bridge between Nef and the exocyst complex via its interaction with EXOC6 [78
]. Our co-immunoprecipitation assays were conducted in lysis buffer containing a concentration of detergent sufficient to disrupt membranes (1% Triton X-100); therefore, it seems unlikely that Nef-exocyst complex association could be attributed to membrane bridging. Both Nef and the exocyst complex are targeted to lipid rafts [30
], and both have been linked to nanotube formation [13
]. As such, Nef and the exocyst complex are likely to co-localize at plasma membrane sites to mediate nanotube formation. Further studies are required to determine whether Nef association with the exocyst complex is direct or indirect, and whether clathrin adapter complex protein(s), Rab11, or other yet unknown protein(s) bridge between Nef and the exocyst complex. RNAi-mediated knockdown experiments targeting nodes of the Nef-exocyst interactome network (Figure ) will help to elucidate the mechanism(s) by which Nef hijacks exocyst function to enhance nanotube formation. Additionally, future studies utilizing Nef mutants to map determinants of Nef-exocyst association will be important to clarify the functional relationships between Nef-Pak2 and Nef-exocyst interactions.
Several nodes of our Nef-exocyst interactome (Figure ) are in agreement with results of prior proteomic and RNAi-based studies of HIV interactions with the host cell (Additional file 6
: Figure S3). In their proteomic study with Jurkat cells, Jager et al
. also identified Vav and the clathrin adapter proteins AP1S1A and AP3S1 as Nef-interacting proteins [17
]. Furthermore, siRNA screens indicated that AP2M1, PIP5K1C, RAB11A, RalB, and RalBP1 expression is important for HIV-1 replication [81
]. Therefore, in addition to the role of Nef-mediated enhancement of nanotube formation, it is possible that interaction of HIV-1 proteins with Nef-exocyst interactome components (Figure and Additional file 6
: Figure S3) may have broader functional implications for viral replication. Future studies to examine the effects of RNAi-mediated knockdown of EXOC2 and other Nef-exocyst interactome components will be important to understand the contribution of these interactions to HIV replication and pathogenesis.
There are striking similarities between mechanisms by which nanotubes, filopodia, virological synapses, and immunological synapses are formed, which include cytoskeletal polarization, membrane protrusion, and recruitment of proteins to discrete membrane domains that serve as foci for endocytosis and exocytosis [42
]. Furthermore, exocyst complex function is required for formation of filopodia and nanotubes [66
]. Together with our findings that Nef associates with exocyst components and EXOC2 is required for Nef-mediated enhancement of nanotube formation, these observations provide further support for a model in which formation of Nef-induced nanotubes [13
] involves Nef-mediated regulation of the exocyst complex.
The relationship of our findings to Nef-mediated enhancement of viral replication and cellular activation remains an open question. Cell-to-cell transfer is an important mode of HIV-1 dissemination within the host; for CD4+ T-cells, intercellular transfer is 2 to 3 logs more efficient at supporting viral replication than HIV-1 infection with cell-free inoculum [42
]. Nef enhances formation of the virological synapse, which is structurally similar to intercellular nanotubes [42
]. Thus, Nef-mediated regulation of the exocyst complex may also play a role in intercellular transmission of HIV-1 via the virological synapse. Exosomes or microvesicles that contain Nef [24
] or proinflammatory molecules such as cytokines [86
] contribute to cell activation [25
] and activation-induced apoptosis [24
]. As such, Nef may increase chronic immune activation in part by upregulating exocyst complex function and microvesicle secretion.