Previous Article | Next Article ![]()
Journal of Virology, May 2007, p. 4895-4899, Vol. 81, No. 9
0022-538X/07/$08.00+0 doi:10.1128/JVI.02829-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

First Department of Forensic Science, National Research Institute of Police Science, Kashiwa 277-0882, Japan,1 CREST, Japan Science and Technology Agency, Saitama 332-0012, Japan,2 Department of Biochemistry, Graduate School of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, Japan,3 International Research Center for Infectious Diseases, Institute of Medical Science, University of Tokyo, Tokyo 108-8639, Japan,4 Department of Pathological Sciences, School of Veterinary Medicine, University of WisconsinMadison, Madison, Wisconsin 53706,5 Special Pathogens Laboratory, Department of Virology 1, National Institute of Infectious Diseases, Gakuen 4-7-1, Musashimurayama, Tokyo 208-0011, Japan6
Received 21 December 2006/ Accepted 29 January 2007
|
|
|---|
|
|
|---|
The RNA genome of MARV encodes seven polypeptides, including the glycoprotein (GP), the nucleoprotein (NP), RNA-dependent RNA polymerase (L), VP35, VP30, VP40, and VP24. VP40 is the most abundant virion matrix protein and plays a key role in virus assembly and budding (15, 16, 31). GP is the only surface protein of filoviruses and is assumed to be responsible for binding to cellular receptors and for fusion of the viral envelope with the cellular membrane in the course of viral entry into the cells (2). The nucleocapsid complex, which contains NP, VP35, L, and VP30, encapsulates the viral genome.
Recent studies have indicated that viral matrix proteins play critical roles during the late stage of virus budding in many enveloped RNA viruses, including retro-, rhabdo-, filo-, arena-, and orthomyxoviruses, and when expressed alone in cells, they are released in the form of virus-like particles (VLPs). These viral proteins possess a so-called L-domain, containing PT/SAP, PPXY, and YPXL, which are motifs critical for efficient budding (3, 4, 6-13, 21, 25, 27, 30, 34-36, 39). Most of the host factors that interact with the L domain are involved in the class E vacuolar protein-sorting pathway, suggesting that budding into the lumen of multivesicular bodies (MVBs) in late endosomes and viral budding at the plasma membrane are topologically identical and share a common mechanism. MARV VP40 protein is sufficient for the release of VLPs (15, 16, 31) and contains a PPXY motif near its N terminus (Fig. 1B), but the viral L domain and the cellular factors required for its budding have yet to be determined.
![]() View larger version (23K): [in a new window] |
FIG. 1. Analysis of the putative L domain within MARV VP40. (A) MARV VLPs produced by expression of VP40. At 24 h after transfection of 293T cells with pMV-VP40 (WT), filamentous particles bud from the membrane ruffles. Black arrows, longitudinally sectioned VLP; black arrowhead, transversely sectioned VLP; white arrowheads, ruffling membranes. Electron microscopy was performed as described previously (24). Bar, 1 µm. (B) Schematic representation of the VP40 L-domain mutants used in this study, showing sequence changes. (C) COS-7 cells were transfected with pMV-VP40 (WT), pMV-VP40-APPY, pMV-VP40-PAPY, or pMV-VP40-PPPA. WB using anti-VP40 antiserum detected cell- and VLP-associated VP40. (D) The intensities of the bands for cell- and VLP-associated VP40 in panel C were quantified as described previously (32). The budding efficiency of VLPs induced by VP40-WT (VLP/cellular) was set at 1.0. The data are averages and standard deviations from three independent experiments.
|
The PPPY motif within VP40 is important for efficient VLP production. First, to confirm that the expression of MARV VP40 in cells can induce the budding of VLPs that are morphologically identical to virions, we constructed a VP40 expression vector for the wild type (WT), pMV-VP40, by insertion of the coding region of VP40, amplified by PCR using pTM-VP40 as a template, into the pCAAGS vector (16, 23) and analyzed the VP40-expressing cells by transmission electron microscopy (Fig. 1A). MARV VP40 expression induced membrane ruffling and budding of filamentous particles, as shown previously (1, 17). These structures were not seen in cells transfected with the control plasmid (data not shown). To examine the role of the PPPY sequence within VP40 in MARV budding, we constructed expression vectors for VP40 mutants with single point mutations within the PPPY motif (Fig. 1B). COS-7 cells (1 x 105) were transfected with each of these plasmids (1 µg) using TransIT-LT1 (Mirus Bio Corp., Madison, WI). At 48 h after transfection, the cell supernatants were cleared of cell debris by centrifugation (13,000 x g; 10 min), and then VLPs were pelleted through a 20% sucrose cushion by ultracentrifugation (345,000 x g; 30 min at 4°C). VLPs and cells were lysed with lysis A buffer (32, 36) and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, followed by Western blotting (WB) using anti-VP40 antibody, which was prepared from rabbits immunized with the recombinant MARV VP40 protein. As shown in Fig. 1C, the levels of VP40-induced VLPs in the case of L-domain mutants, APPY, PAPY, and PPPA, were lower than that in the WT, although similar levels of VP40 were synthesized in cells expressing either the WT or an L-domain mutant. The efficiency of VLP production was different among the three L-domain mutants. Relative levels of production of VLPs from cells expressing APPY, PAPY, and PPPA were 72%, 52%, and 30% of the WT level, respectively (Fig. 1C and D). These results indicate that the PPPY motif functions as an L domain and that the tyrosine residue is the most critical for the budding function of this L domain. The PPXY motif has been reported to interact with cellular E3 ubiquitin ligases, such as Nedd4, LDI-1, and BUL1 (7, 14, 29, 33, 37). Although we examined the effects of overexpression of dominant-negative mutants of Nedd4 and BUL1 on MARV VP40-induced VLP budding, they had no effect on the efficiency of VLP budding (data not shown). These observations suggest the involvement of an E3 protein(s) other than Nedd4 and BUL1 in MARV budding.
Contributions of NP and GP to the VP40-induced VLP budding. It has been reported that NP and GP enhance the release of Ebola VP40-induced VLPs (19). Therefore, we examined the contributions of NP and GP to MARV VP40-induced VLP budding. The expression plasmids for MARV NP and GP, pMV-NP and pMV-GP were constructed as described for pMV-VP40. COS-7 cells (1 x 105) were cotransfected with 0.5 µg of pMV-VP40 along with 0.5 µg of pMV-NP and/or 0.5 µg of pMV-GP. The levels of expression of cellular VP40 were approximately equivalent among all samples independent of coexpression of NP and/or GP, while the release of VLPs was markedly facilitated by coexpression of NP and GP (Fig. 2A and B).
![]() View larger version (23K): [in a new window] |
FIG. 2. Contributions of NP and GP to VP40-induced VLP budding. (A) COS-7 cells were transfected with pMV-VP40 (WT) in combination with pMV-NP, pMV-GP, and the control vector. Cell lysates and VLPs were collected, and WB using rabbit anti-GP, -NP, and -VP40 antibodies was performed to detect GP, NP, and VP40 (28). (B) The intensities of the bands for cell- and VLP-associated VP40 in panel A were quantified as described in the legend to Fig. 1. The budding efficiency of VLPs induced by VP40 alone (VLP/cellular) was set at 1.0. The data are averages and standard deviations from three independent experiments. (C) COS-7 cells were transfected with pMV-VP40-PPPA and the combination of pMV-NP, pMV-GP, and the control vector. Cell lysates and VLPs were collected, and WB was performed to detect GP, NP, and VP40. (D) The budding efficiency of VLPs was analyzed quantitatively as described for panel B.
|
Tsg101 is involved in MARV VP40 budding. To investigate the mechanism of MARV VP40-induced VLP budding in further detail, we examined the involvement of Tsg101a component of the ESCRT-I complex known to participate in the cellular MVB sorting pathway and budding of some virusesin MARV VLP budding using small interfering RNA (siRNA) (32). As shown in Fig. 3A, specific depletion of Tsg101 by siRNA significantly reduced the release of VP40-induced VLPs, indicating that Tsg101 is involved in VLP budding. We also examined whether exhaustion of Tsg101 has an effect on the budding of VLPs composed of VP40, GP, and NP. The VLP budding from cells coexpressing VP40, GP, and NP was also markedly decreased by specific depletion of Tsg101, although the reduction rate of VP40/GP/NP-induced VLP release by Tsg101 depletion was lower than that of VP40-induced VLPs (Fig. 3B). These results clearly show that Tsg101 is involved in MARV VLP budding as a host factor.
![]() View larger version (26K): [in a new window] |
FIG. 3. Tsg101 is involved in MARV VP40-induced VLP budding. (A and B) 293T cells were pretreated with siRNA specific for Tsg101 (siTsg101) or control RNA (siCont) 1 day before plasmid transfection (5). The following day, these cells were transfected with pMV-VP40 alone (A) or together with pMV-GP and pMV-NP (B) and siTsg101 or siCont. At 48 h after transfection, cell lysates and VLPs were collected. Endogenous Tsg101 was detected with mouse anti-Tsg101 antibody (C-2; Santa Cruz Biotechnology, Inc.) (upper panel). Viral proteins were also detected by WB. (C) To detect the incorporation of Tsg101 into VLPs, 293T cells were cotransfected with pTsg101-Myc and control vector (Cont), pMV-VP40, or pMV-VP40-PPPA. VLPs released from cells were collected, and VLPs containing equal amounts of VP40 were analyzed by WB. The presence of Tsg101 in VLPs was detected by WB using rabbit anti-Myc antibody. (D) The presence of Tsg101 in VP40/GP/NP-induced VLPs was also examined by cotransfection with pMV-GP and pMV-NP. (E) GST pulldown analysis. (Upper panel) Coomassie blue (CBB) staining of the GST-VP40 and GST-VP40 PPPY. (Lower panel) The immobilized GST fusion proteins were incubated with 293T cell lysates overexpressing Tsg101-Myc, washed extensively, and eluted, followed by detection by WB using anti-Tsg101 antibody. Input, Tsg101 in 1/12 volume of the input cell lysates.
|
Tsg101 has also been reported to be involved in the budding of human immunodeficiency virus type 1, human T-cell leukemia virus type 1, lymphocytic choriomeningitis virus (LCMV), Lassa virus, and Ebola virus. With the exception of LCMV, all these viruses possess the PT/SAP motif and interact with Tsg101 via this motif (3, 5, 20, 26, 32). Although LCMV does not have the PT/SAP motif, the STAP sequence, which is similar to PT/SAP, is present upstream of PPPY within the viral matrix Z protein. Previously, Licata et al. showed that Tsg101 was still incorporated into VLPs induced by the ATAP mutant of Ebola VP40, in which alanine is substituted for the first proline in the PTAP motif (18). In addition, it has also been reported that substitution of the first proline residue of the PTAP motif of human immunodeficiency virus p6 by alanine resulted in only moderate reduction of the binding affinity of p6 for Tsg101 (5). Therefore, in LCMV, the STAP sequence may function as an L domain instead of PT/SAP. However, there is no sequence similar to PT/SAP in MARV VP40.
We performed a glutathione S-transferase (GST) pulldown assay to examine whether Tsg101 interacts with VP40 depending on the PPPY motif. Recombinant GST-VP40 and GST-VP40
PPPY proteins were expressed in Escherichia coli BL21-Gold and purified using glutathione-Sepharose 4B beads (Amersham Biosciences, Uppsala, Sweden). The immobilized GST fusion proteins were incubated for 2 h at 4°C with 293T cell lysates overexpressing Tsg101-Myc, washed extensively, and eluted, followed by detection by WB. As shown in Fig. 3E, the interaction between VP40 and Tsg101 was confirmed, while VP40
PPPY abolished the ability to bind to Tsg101. Thus, we clearly showed that the interaction between Tsg101 and VP40 depends on the PPPY motif within VP40. This is the first report that Tsg101 interacts with viral matrix protein depending on PPPY, but not PT/SAP, and regulates viral budding. At present, it is not clear whether Tsg101 binds directly to the PPPY motif in VP40 or indirectly to VP40 via another cellular factor(s). Further studies are required to determine how Tsg101 interacts with MARV VP40 and participates in MARV budding.
It has been demonstrated that MARV VP40 interacts with the membranes of late endosomes in the course of viral infection (15). The transport of MARV VP40 involves its accumulation in MVBs followed by redistribution of VP40-enriched membrane clusters to the plasma membrane (16). Thus, VP40 is transported through the retrograde late-endosomal pathway, while GP is redistributed from the trans-Golgi network into the VP40-containing MVBs, suggesting that budding complexes assemble at late-endosomal surfaces and are then transported to the cell surface. MVBs would provide the platform for formation of membrane structures that bud MARV from the cell surface (17). Tsg101 is one of the components of ESCRT-I and is involved in the MVB sorting pathway. Therefore, it is plausible to suggest that Tsg101 recruits VP40 to MVB and participates in MARV budding there. Taken together, these results strongly suggest that MARV budding utilizes the cellular MVB sorting pathway.
The results of the present study provide important insights into the molecular aspects of MARV replication and will facilitate the development of anti-MARV therapy.
This work was supported by grants from the Japan Society for the Promotion of Science (JSPS) and the Japan Science and Technology Agency (JST).
Published ahead of print on 14 February 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»