To study the potential role of golgin-97 in VV morphogenesis and replication we employed an RNAi approach to suppress golgin-97 expression inside the infected cell. To induce RNAi we tested two double-stranded (ds) oligoribonucleotides, dsRNA-1 and dsRNA-2 with non-overlapping target sites within golgin-97 mRNA (). In order to analyze RNAi dose response and optimize the conditions, each dsRNA was titrated. As a negative control for potential dsRNA-induced side effects we used double-stranded oligoribonucletides with minimized sequence homology to vertebrate transcripts. demonstrates that golgin-97 expression was significantly repressed by both dsRNAs at relatively low concentrations. Decreasing protein levels corresponded well with increasing amounts of each dsRNA and reached a minimum at 30 pmol (10%) and 20 pmol (18%) for dsRNA-1 and dsRNA-2, respectively (). Anti-actin antibodies used as a control for non-specific RNAi effect on expression of unrelated proteins showed that inhibition of golgin-97 expression did not interfere with actin accumulation (). Confirming the immunoblotting results, immunofluorescence experiments (, supplementary Fig S1
) showed striking difference in levels of golgin-97 accumulation and redistribution between the negative RNAi control and each dsRNA.
Fig 1 (A) dsRNA-1 and dsRNA-2 target sites within golgin-97 mRNA. (B) RNAi dose response curve. Hela cells were transfected with indicated amounts of dsRNA-1, dsRNA-2, or the negative RNAi control. The relative levels of golgin-97 expression, “dsRNA-1” (more ...)
Fig 2 VV replication in golgin-97 depleted cells by confocal laser scanning microscopy. HeLa cells transfected with dsRNA-1, dsRNA-2 or negative RNAi control (“NC”) were infected with VV WR at a MOI of 5.0 pfu/cell and fixed at 24 hpi. The cells (more ...)
To test whether golgin-97 is required for VV replication, HeLa cells were transfected with either dsRNA-1 or dsRNA-2 and infected with VV Western Reserve (WR) strain at a multiplicity of infection (MOI) of 5.0 plaque forming units (pfu) per cell. Surprisingly, VV replication was significantly affected by both dsRNAs. Reduction in virus titers estimated as 85% (dsRNA-1) and 73% (dsRNA-2) correlated with the loss of golgin-97 expression () implying that the protein is essential for virus replication.
Table 1 Relative reduction in golgin-97 expression and virus replication efficiency in the presence of dsRNA-1 and dsRNA-2. Hela cells transfected with dsRNA-1, dsRNA-2 or negative RNAi control were infected with VV WR at an MOI of 5.0 pfu/cell. At 48 hpi the (more ...)
To decipher the inhibitory effects of golgin-97 depletion on VV replication the infected cells were analyzed by immunofluorescence. Repression of the protein expression did not appear to interfere with either virus entry into the cell or its ability to establish replication factories (, supplementary Fig S1
). Golgin-97 accumulated normally inside the factories in the presence of the negative RNAi control and as expected was absent in dsRNA-1 and dsRNA-2 treated samples (, supplementary Fig S1
). These results suggested that golgin-97 may act at the late steps of virus infection such as virion assembly/maturation or virion translocation within the infected cell. This interpretation would be consistent with the fact that golgin-97 is a membrane protein (Barr, 1999
; Gleeson et al., 2004
; Munro and Nichols, 1999
) and that it was shown to play a role in exocytosis, the pathway used by many enveloped viruses to exit the cell (Lu, Tai, and Hong, 2004
). To test this hypothesis we examined virion morphology at 24 hpi by electron microscopy. In the presence of the negative RNAi control, virion assembly and maturation were not affected (). In contrast, depletion of golgin-97 resulted in the loss of MV particles and dramatic increase in the number of IVs (). To rule out the possibility of a delay in virus morphogenesis, we compared the data of virus replication assays at 24hpi and 48hpi. There was no significant difference or noticeable increase in relative efficiency of virus replication for 48hpi-samples (data not shown). Interestingly, some of the examined cells also contained variable amounts of IV with nucleoid (IVn), an intermediate form in IV maturation stage that normally can be seen at early hours in infection cycle () and abnormal looking particles that are likely to be an intermediate between IVn and MV (; IVn′). This observation could be explained by incomplete suppression of golgin-97 expression and indicate that the protein may act at IVn formation, although we can not exclude that it may be important for the next stage, IVn-MV as well.
Fig 3 VV morphogenesis in golgin-97 depleted cells by electron microscopy of ultrathin sections. HeLa cells transfected with negative RNAi control (A and D), dsRNA-1 (B, E, G–I) or dsRNA-2 (C and F) were infected with VV WR at an MOI of 5.0 pfu/cell. (more ...)
Based on the data described here and the results of our previous experiments (Alzhanova and Hruby, 2006
) we propose a model of the MV biogenesis () and suggest a host cell protein, golgin-97 as a key player in this process. Although the mechanisms and the functions of the protein are yet to be determined, one of the possible scenarios may be that when attached to the TGN membrane at the C-terminus, golgin-97 delivers its fragments along with associated protein factors required for maturation of the virus core and, potentially, virus membrane to the virion assembly sites upon its relocation inside the factories. Supporting this hypothesis, a processed form of transiently expressed C-terminal but not N-terminal FLAG: golgin-97 fusion protein was detected inside the virions (Alzhanova and Hruby, 2006
). In this process, golgin-97 is likely to associate with the proteins comprising insoluble core fraction and incorporate into IVn particles while they are being assembled. Alternatively, golgin-97 may be attached to the nascent viral membranes via a viral or cellular protein co-factors in a manner similar to its association with the TGN membrane (Lu and Hong, 2003
; Panic et al., 2003
; Setty et al., 2003
; Wu et al., 2004
) and become passively incorporated during IV/IVn assembly. In this case, the protein may simply play a role of a structural protein: its packaging may contribute to the formation and/or stabilization of the condensed core structure. In addition to this, the surface exposed part of the protein may be involved in the transportation of the assembled MV particles to the TGN compartment where they acquire the secondary membrane and mature into WV. The latter would support the existing hypothesis about the TGN as the origin of the secondary membrane. Further investigation of golgin-97 functions in virus morphogenesis and the experimental verification of the proposed model will undoubtedly provide new insights on the mechanisms of host-pathogen interactions, revealing yet another example of how viruses exploit cellular proteins and pre-existing pathways to serve their needs. As to continuous efforts in developing smallpox vaccines and discovery of anti-poxvirus drugs, these data have identified new potential targets, virus proteins that facilitate golgin-97 delivery inside the virus factories and its incorporation into the virions, as well as golgin-97 itself.
Fig. 4 Working model. During poxvirus infection, a host cell protein, golgin-97 translocates into the viral factories. Being anchored to the TGN membrane, the protein may transport the membrane fragments and associated protein factors important for virus maturation (more ...)