- Top of page
- Materials and Methods
- Supporting Information
Herpes simplex virus 1 (HSV1) is an enveloped virus that uses undefined transport carriers for trafficking of its glycoproteins to envelopment sites. Screening of an siRNA library against 60 Rab GTPases revealed Rab6 as the principal Rab involved in HSV1 infection, with its depletion preventing Golgi-to-plasma membrane transport of HSV1 glycoproteins in a pathway used by several integral membrane proteins but not the luminal secreted protein Gaussia luciferase. Knockdown of Rab6 reduced virus yield to 1% and inhibited capsid envelopment, revealing glycoprotein exocytosis as a prerequisite for morphogenesis. Rab6-dependent virus production did not require the effectors myosin-II, bicaudal-D, dynactin-1 or rabkinesin-6, but was facilitated by ERC1, a factor involved in linking microtubules to the cell cortex. Tubulation and exocytosis of Rab6-positive, glycoprotein-containing membranes from the Golgi was substantially augmented by infection, resulting in enhanced and targeted delivery to cell tips. This reveals HSV1 morphogenesis as one of the first biological processes shown to be dependent on the exocytic activity of Rab6.
Intracellular trafficking between organelles is regulated by the Rab group of small GTPases . Specific Rabs are associated with distinct organelles and their associated transport vesicles, enabling accurate delivery of membranes and cargo to specific compartments. There are over 60 human Rabs which coordinate the processes of vesicle formation, transport, tethering and fusion, by interacting with specific effector proteins that bind to the GTP-bound form of their respective Rabs . In recent years a number of studies have provided growing evidence for the exploitation of Rabs by enveloped viruses in their morphogenesis and egress pathways. For example, influenza A and respiratory syncytial virus have both been shown to utilize the Rab11 pathway for envelopment and budding [3-5]. Likewise, Rabs 7A and 9, both present on the membranes of late endosomes/multivesicular bodies, have been implicated in human immunodeficiency virus morphogenesis [6, 7], while the early endosome localized Rab5 together with Rab7 are involved in the formation of hepatitis C virus replication complexes .
Herpes simplex virus 1 (HSV1) is a large enveloped DNA virus with a complex structure comprising up to 40 structural proteins . The genome-containing capsid is surrounded by the virus envelope containing up to 15 glycoproteins, and the tegument where over 20 virus encoded proteins are packaged [9, 10]. HSV1 is released from the infected cell following an intricate morphogenesis pathway that has been the subject of much debate [11, 12]. After capsid assembly in the nucleus, capsids are transported to the cytoplasm by first acquiring a primary envelope at the inner nuclear membrane that is lost by fusion with the outer nuclear membrane – termed an envelopment-deenvelopment step – releasing naked capsids into the cytosol [12, 13]. These free capsids acquire their tegument proteins and are re-enveloped at a site in the cytoplasm that remains a point of contention [12, 14], with the favoured location of HSV1 envelopment commonly cited as the trans-Golgi network (TGN), where virus glycoproteins are known to localize [13, 15, 16]. Virus production is sensitive to the endoplasmic reticulum (ER)-to-Golgi inhibitor Brefeldin A (BFA) [17, 18], and is inhibited by the depletion of Rab1, a Rab known to be involved in ER-to-Golgi transport [19-22], suggesting that virus glycoproteins must access the Golgi/TGN prior to envelopment. Virus glycoproteins also localize to the plasma membrane (PM) in advance of any capsid accumulation in the cytoplasm [23-26].
Recently, we have published work providing evidence that, rather than wrapping at the TGN, HSV acquires its envelope from glycoprotein-containing endocytic membranes recently retrieved from the PM in a Rab5-dependent mechanism , in agreement with earlier studies from others . Depletion of Rab5 resulted in virus glycoproteins being trapped at the cell surface due to an inhibition of Rab5-dependent endocytosis, a subsequent failure in virus capsid wrapping and a 95% drop in virus yield. Hence, we suggest that virus envelope proteins are first transported through the Golgi/TGN to the PM ahead of capsid release from the nucleus, and are then retrieved into the recycling endocytic pathway to provide the final wrapping membranes. To identify additional Rabs involved in the HSV1 life cycle, we have now conducted an siRNA screen of 60 human Rabs to define their requirement in virus replication. The major factor that we have identified for production of infectious virus, over and above the aforementioned Rab1 or Rab5, is the Golgi-associated Rab6. In the absence of Rab6, HSV1 envelope proteins were unable to reach the cell surface and were retained in the Golgi/TGN during infection or when expressed in isolation. In infected cells, the retention of glycoproteins at the Golgi/TGN resulted in the subsequent accumulation of naked capsids which were not associated with infectivity. Notably, HSV1 infection induced the tubulation and redistribution of Rab6 positive membranes from the Golgi/TGN to peripheral membranes, delivering large amounts of envelope constituents to the PM in a pathway that requires the Rab6 effector ERC1 for optimal efficiency. Hence, we conclude that HSV1 activates a Rab6-specific exocytic pathway to transport virus glycoproteins from the Golgi/TGN to the PM, and provide a membrane population that is subsequently used for virus wrapping by endocytic retrieval. This reveals that the Rab6 specific post-Golgi pathway is fundamental to HSV1 envelopment.
- Top of page
- Materials and Methods
- Supporting Information
In this study we present the first screen of human Rab GTPases in a virus infection of human cells. After screening sixty Rabs, we found that the depletion of only four pools reduced HSV1 yield by over 10 fold – in increasing order of effect from Rab11 (12.5 fold), Rab5 (21 fold), Rab1 (53 fold) to Rab6 (130 fold). Of these four, Rabs 1, 5 and 11 have been identified previously by us and others [21, 22], but the conclusion that Rab6 is an important factor in HSV1 morphogenesis is a new finding. Rab6 is the major Rab GTPase associated with the trans cisternae of the Golgi and the TGN . It has been shown to play several roles in membrane trafficking, being identified as having a role in COP1-independent retrograde trafficking from the Golgi to the ER of certain proteins such as Golgi glycosylation enzymes and Shiga toxin . Additionally Rab6 is believed to be involved in retrograde intra-Golgi trafficking . However, in recent years there have been a number of studies invoking a role for Rab6 in vesicular trafficking from the Golgi/TGN to the PM [36, 37, 53], a specific role that would fit with our own findings.
Within the TGN, cargos are sorted into tubules away from resident Golgi proteins, and motors such as kinesin dock onto these tubules which then undergo extrusion along MTs, before fission from the Golgi/TGN occurs to form post-Golgi carriers (PGCs) . Although not absolutely essential for constitutive secretion, Rab6 was previously reported to stimulate this MT-based transport of some PGCs from the Golgi/TGN and regulate their targeting to PM sites . Moreover, recent biochemical characterisation of PGCs has revealed a specific class of carrier that is positive for Rab6 . Our data shows that Rab6 is required for the trafficking of HSV1 glycoproteins to the PM, and that glycoprotein-containing membranes on the way to the cell periphery contain Rab6. Interestingly, our studies on reporter proteins expressed by transient transfection revealed that the requirement for Rab6 in TGN-to-PM transport was variable, with transport of the tsVSV-G protein to the PM being severely attenuated in the absence of Rab6, but secretion of Gaussia luciferase being unaffected by its depletion. This is in agreement with other studies showing Rab6 is required for VSV-G transport to the PM [36, 44], but is in contrast to Rab6 positive PGCs described in a recent publication which were shown to be distinct from VSV-G transporting carriers . This combination of results may point to Rab6 being critical in the transport of specific classes of proteins to the cell surface, particularly viral transmembrane proteins, but not luminal proteins.
There are two aspects to the utilization of Rab6 by HSV1 that have a consequence for our understanding of Rab6 function – firstly that Rab6 is necessary for glycoprotein transport to the PM, and secondly that HSV1 activates the Rab6-dependent transport pathway. In considering the requirement for Rab6 in post-Golgi trafficking, there are several points in the pathway at which its activity could be essential, including tubule formation, fission, transport on MTs and fusion at the PM. The non-processive motor myosin II has been implicated in efficient fission of Rab6 positive vesicles from the Golgi/TGN, and has been shown to be a component of a class of Rab6 positive PGCs [37, 45]. Moreover, myosin II has previously been identified as a potential binding partner of an HSV1 structural protein, while inhibition of myosin II activity was shown to reduce virus yield . Although this seemed a likely target to investigate, our studies here indicate that inhibition of myosin II activity, either via the drug blebbistatin, or siRNA knockdown of NMHCIIA and NMHCIIB, did not affect glycoprotein localization to the PM and resulted in only a twofold drop in virus production (a reduction that was specific to the NMHCIIB isoform) suggesting that any involvement of myosin II in virus envelope trafficking is limited. In agreement with these results, a very recent study on a class of PGCs shown to contain Rab6 and myosin II has also shown that myosin II is not required for the biogenesis of these PGCs , whereas the fission factor protein kinase D (PKD), which was also present in the PGCs, was required for their production [45, 55, 56]. Moreover, as depletion or inhibition of PKD in HSV1 infected cells has been shown to reduce virus release , it seems likely that PKD, rather than myosin II, is involved in the fission of the Rab6 positive, HSV1 glycoprotein-containing membranes from the Golgi/TGN.
Our time-lapse analyses of infected cells indicate that virus infection may positively activate Rab6-dependent transport to the cell surface, with the number and length of Rab6 positive tubules leaving the Golgi/TGN being greatly enhanced, and long tubules frequently observed to leave the Golgi/TGN area, delivering a large amount of membrane to the cell surface. Such tubular PGCs have been described before in experiments using the tsVSV-G reporter [58, 59], but not in the abundance observed here in HSV1 infected cells. These tubules undergo fission from the Golgi/TGN, and move to the PM on what appear to be specific tracks with rates appropriate for kinesin transport on MTs. The concentration of GFP-Rab6 and glycoproteins at the tips of infected cells is indicative of the presence of specific MT delivery routes to those regions of the cell periphery. Interestingly, although KIF5B, conventional kinesin heavy chain, has been shown to enhance Rab6 vesicle movement on microtubules , KIF5B depletion here had no effect on virus yield, suggesting there may be a potential redundancy among the kinesin motors that the virus can recruit for glycoprotein vesicle transport. Moreover, virus replication is known to continue in the absence of MTs in tissue culture monolayers at least, albeit to lower levels , suggesting that although MT-directed delivery to peripheral domains is required for optimal virus production, it is not absolutely essential. This has been shown to be the case for tsVSV-G, where nocodazole treatment of cells retarded but did not inhibit its delivery to the PM, which occurred at random sites rather than targeted regions of the PM . Nonetheless, the situation may be different in neurons where MTs are considered to be particularly important for the virus to transport its components down the relatively long axon to the cell termini for transfer to epithelial cells, although it is still a point of debate as to whether the virus is transported on axonal MTs as fully enveloped virions or as separate capsid and envelope entities [61-63]. As neurons specifically express Rab6B , it would be of interest to determine if this isoform of Rab6 is important for HSV1 morphogenesis in neuronal cells.
In uninfected cells, the domains to where GFP-Rab6 positive vesicles are targeted have been shown to be enriched in the Rab6 binding partner ERC1 , a cortical protein that plays an accessory role in linking distal MT ends to cortical platforms via CLASP proteins . It has been postulated that this interaction facilitates docking and/or fusion at the PM . It is of note then that the ERC1 effector was the only effector here that when depleted resulted in a drop in virus yield, reducing it by 10 fold, and causing a detectable reduction in gD levels at the cell surface, suggesting that ERC1 may facilitate optimal glycoprotein incorporation into the PM. Interestingly, a recent publication demonstrated the presence of sites on the PM which become enriched in glycoproteins early in HSV1 infection and which require an intact MT network to form . Moreover, ERC1 has been identified in proteomic screens for Dengue virus and hepatitis C virus interacting proteins [65, 66], suggesting that ERC1 may be a common cellular factor utilized in the replication of enveloped viruses.
Rab6 is also involved in recruiting both the dynein–dynactin motor complex and the KIF20A kinesin like protein Rabkinesin-6 to the Golgi complex, but efficient depletion of the relevant Rab6 effectors – the dynactin components bicaudal-D1, D2 or dynactin 1 or KIF20A [42, 47, 48] – indicated that none of these factors are required in the HSV1 replication cycle. As Rab6 recruitment of dynactin by its interaction with bicaudal D and dynactin 1 is involved in Golgi-to-ER retrograde trafficking , these results indicate that there is likely to be no role for the retrograde trafficking of virus glycoproteins in the infected cell. Moreover, depletion of bicaudal-D2, but not -D1, caused a twofold increase in HSV1 virus yield, implying that disruption of the bicaudal-D2 interaction with Rab6 may allow more Rab6 to be recruited for the transport of HSV1-specfic PGCs. Interestingly, bicaudal-D1 and Rab6 have been implicated in the replication of the betaherpesvirus HCMV, by an interaction with its structural protein pp150 [67, 68]. However, the fact that bicaudal-D1 does not appear to be involved in HSV1 replication points to these viruses using Rab6 by different mechanisms.
The results presented here on the role of Rab6 in HSV1 infection have allowed us to extend our recently published model of HSV1 envelopment where we described the wrapping of HSV1 capsids in membranes of the recycling endocytic network, and demonstrated that Rab5 is involved in the endocytic retrieval of virus glycoproteins from the cell surface . An important feature of this pathway is that virus envelope proteins must travel to the cell surface before reaching the final site of envelopment. Our results show that Rab6 is involved in transporting envelope proteins from the Golgi/TGN to the cell surface, prior to endocytosis to wrapping membranes. Hence, although endocytosis carries on normally in Rab6 depleted cells, the endocytosed membranes do not contain virus envelope or tegument proteins, and cannot function as wrapping membranes.
Given the profound effect of Rab6 depletion on HSV1 production, the molecular components of this pathway represent useful targets to understand HSV1 exploitation of the cellular secretory pathway and investigate new methods of interfering with virus replication. Moreover, HSV1 infection is likely to prove a useful and sensitive tool for investigation of the role of Rab6 in Golgi-to-PM transport, providing valuable insight not only into the cell biology of the virus–host relationship but also the activity of Rab6 itself.