![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article 
Journal of Virology, June 2008, p. 5368-5380, Vol. 82, No. 11
0022-538X/08/$08.00+0 doi:10.1128/JVI.02751-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Departamento de Genética del Desarrollo y Fisiología Molecular, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Cuernavaca, Morelos 62210, Mexico
Received 27 December 2007/ Accepted 24 March 2008
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Molecular chaperones and folding enzymes are important components of the ER machinery; their principal function consists of maintaining the quality of the ER (17) through avoiding the nonspecific interactions among newly synthesized proteins, maintaining them in a proper state to fold, and marking the misfolded proteins for degradation (18, 22, 28, 47, 52). Two folding systems have been proposed to contribute to this quality control (19, 33). One of these is composed of grp78 (also known as BiP), grp94 (also known as endoplasmin), and protein disulfide isomerase (PDI), although other proteins have been isolated in complexes with these chaperones (26, 35). The second system, known as the "calnexin-calreticulin cycle," is integrated by calnexin, calreticulin, and ERp57 (9, 16, 48).
The first system, known as the "grp78 system," works by recognizing hydrophobic residues in the target proteins; the binding and release of these proteins are regulated by ATP hydrolysis (24, 27, 30). The second system acts on glycosylated proteins; when a core glycan (Glc3-Man9-GlcNAc2) is added to newly synthesized proteins, two terminal glucoses are removed, and the monoglucosylated proteins are then recognized by the lectin-like protein calnexin or calreticulin, which exposes them in turn to ERp57, a thiol-disulfide oxidoreductase (19). Calnexin is a transmembrane protein, while calreticulin has a luminal disposition.
Rotaviruses, the leading cause of severe dehydrating diarrhea in infants and young children worldwide, are nonenveloped viruses formed by three concentric layers of protein that enclose a double-stranded RNA genome. The outermost layer is composed of glycoprotein VP7, which forms the smooth surface of the virus. From this surface, spike-like structures formed by VP4 project (10). In infected cells, large cytoplasmic inclusions termed viroplasms are formed, and these are thought to be the sites where double-layered particles (DLPs) assemble. The DLPs then mature by budding from the viroplasm structures into the adjacent ER membrane, modified by the viral glycoproteins VP7 and NSP4. During this process, mediated by the interaction of DLPs with NSP4, the particles acquire a transient membrane envelope that contains VP4, NSP4, and VP7 (10, 49) and that is later removed to yield the mature triple-layered particles (TLPs). The mechanism of removal of the transient lipid envelope is largely unknown, although VP7 has been reported to be important for this step (32).
VP7 is an integral ER membrane glycoprotein oriented to the luminal side of the ER. It is 326 amino acids long and is modified in most rotavirus strains by N glycosylation at asparagine residue 69 and by the establishment of several disulfide bonds that are important for the correct folding of the protein and maturation of viral particles (8, 60). NSP4, a 175-amino-acid-long nonstructural protein, is modified at its amino terminus by two high-mannose N-glycosylation chains (asparagines 8 and 18). The protein spans the membrane only once, with the glycosylated amino-terminal region oriented to the luminal side of the ER and the carboxy-terminal region (amino acids 44 to 175) oriented to the cytoplasmic side (1).
It is known that the correct assembly of infectious virus particles depends on the glycosylation of NSP4, the formation of disulfide bonds in VP7, and adequate ATP and calcium concentrations in the ER, since drugs such as brefeldin, tunicamycin, calcium ionophores, and thapsigargin block the morphogenetic process of the virus and noninfectious enveloped lipid particles accumulate in the lumen of the ER (29, 36, 38, 40, 51). On the other hand, it has been reported that the molecular chaperones grp78 and grp94 increase their levels of expression during rotavirus infection and interact with the viral structural polypeptides VP7 and VP4 (7, 65). In addition, PDI has been reported to interact with the nonglycosylated form of NSP4 and with VP7 (38) and calnexin has been reported to interact with NSP4 (37). However, the significance of these interactions to virus assembly has not been determined, and our knowledge of the cellular proteins relevant for the morphogenesis of rotaviruses is still very limited.
In this work, we evaluated the relevance to rotavirus assembly of members of the two ER folding systems through silencing their expression by RNA interference (RNAi). We found that grp78, PDI, calnexin, and calreticulin are important for the formation of infectious virus but that grp94 and ERp57 are not. The involved proteins contribute to promote the timely trimming of the carbohydrate chains of VP7 and NSP4, the correct formation of VP7 disulfide bonds, and the incorporation of properly folded VP7 into TLPs.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
siRNA transfection. The sequence and relative positions of the small interfering RNAs (siRNAs) evaluated in this work are shown in Table 1; these were purchased from Dharmacon Research (Lafayette, CO). As a negative control, a previously reported irrelevant siRNA to the green fluorescent protein (GFP) (32) was used. siRNA transfection was carried out in 80%-confluent cell monolayers cultured in 48-well plates or in 25-cm2 flasks that were transfected with 100 µl or 1,000 µl, respectively, of a mixture containing 60 µg/ml Oligofectamine (Invitrogen) and the amount of each siRNA indicated in Table 1 in MEM without serum. The transfection mixture was added to cells previously washed with MEM and incubated for 8 h at 37°C. After this time, the transfection mixture was removed and the cells were washed with MEM and kept in this medium without serum and without antibiotics for 16 h at 37°C; the cells were then washed and kept in MEM supplemented with 2% fetal bovine serum for 48 h at 37°C. Two to four different siRNAs specific for the grp78, grp94, and calnexin genes were initially designed. After their initial evaluation, the most-effective siRNAs were used throughout this work (Table 1). To silence the expression of PDI, we used a reported siRNA sequence (59), and for calreticulin and ERp57, siGenome Smartpools from Dharmacon were used.
|
Immunoblots. Cells transfected with siRNAs against the molecular chaperones were lysed at 72 h posttransfection with Laemmli sample buffer, and the proteins were separated by sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to nitrocellulose membranes (Millipore, Bedford, MA). Membranes were blocked with 5% nonfat dry milk in PBS with 0.2% Tween 20 and incubated at room temperature (RT) with the primary antibodies indicated below in PBS containing 0.1% milk and 0.2% Tween 20, followed by an incubation with secondary, species-specific, horseradish peroxidase-conjugated antibodies. The peroxidase activity was developed with Western Lightning Chemiluminescence Reagent Plus (PerkinElmer Life Sciences) by following the manufacturer's instructions. The intensity of each band was quantified by densitometry, using the Image Pro Plus 5.0 software.
Immunofluorescence. MA104 cells were grown on coverslips and infected as described above. The cells were fixed at the times indicated in Fig. 1 with 2% paraformaldehyde in PBS for 20 min at RT. After this time, the cells were washed four times with PBS and permeabilized by incubation with PBS-1% bovine serum albumin (BSA), 0.5% Triton X-100, 50 mM NH4Cl for 15 min at RT. After being washed four times for 5 min each time with 50 mM NH4Cl in PBS with gentle swirling, the cells were incubated for 1 h at RT with primary antibodies diluted in blocking buffer (50 mM NH4Cl, 1% BSA in PBS) and then rinsed four times with 50 mM NH4Cl in PBS. The coverslips were then incubated with the appropriate Alexa antibody-labeled secondary antibodies in 50 mM NH4Cl in PBS for 1 h at RT. The cells were washed four times with PBS and mounted on glass slides using Fluoprep (BioMérieux). The slides were analyzed and captured with a model E600 epifluorescence microscope coupled to a model DXM1200 digital still camera (Nikon). The images were digitally captured and prepared in Adobe Photoshop 7.0.
|
RIPA. Twenty microliters of radiolabeled lysates, or 40 µl of purified TLPs desalted and treated or not treated with 3 mM EGTA for 10 min at 37°C, was incubated with 2 µl of MAb M60 or 5 µl of the C-239 polyclonal antibody in 80 µl of radioimmunoprecipitation assay (RIPA) buffer (1% Triton X-100, 1% deoxycholic acid, 50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% SDS, 20 µg phenylmethylsulfonyl fluoride per ml, and 2 µg aprotinin and leupeptin per ml) for 1 h at RT. The samples were centrifuged at 13,000 x g for 5 min, and the supernatants were mixed with protein A-Sepharose 4B (25 µl; Zymed) and then incubated for 1 h at RT in a shaker. The protein A-Sepharose coupled to the immune complexes was pelleted at 6,000 x g for 2 min and washed three times with RIPA buffer. The immune complexes were solubilized in 20 µl of Laemmli sample buffer containing 1% β-mercaptoethanol, boiled for 3 min, and analyzed by SDS-12.5% PAGE.
Binding assay. MA104 cells grown in 48-well plates were washed twice and incubated with MEM without serum for 30 min at 37°C. After this time, the MEM was removed, and 500 µl of a solution of 1% bovine albumin in PBS was added to the cells and incubated for 1 h at 37°C. The cells were then washed with an ice-cold solution of 0.5% BSA in PBS and incubated for 1 h at 4°C with purified TLPs diluted in ice-cold MEM. After this time, the cells were washed three times with ice-cold PBS containing 0.5% BSA and lysed with 50 µl of lysis buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% Triton X-100). The cells were frozen and thawed twice, and the amount of virus present in the lysate was determined by an enzyme-linked immunosorbent assay (ELISA), as described previously (67).
Real-time reverse transcription-PCR.
MA104 cells grown in 48-well plates were transfected with siRNAs, and 72 h posttransfection, the cells were lysed with TRIzol (Invitrogen) and total RNA was purified according to the manufacturer's instructions. Total RNA was treated with RNA-free DNase (Ambion) to remove the possible contamination of DNA. The level of molecular-chaperone mRNAs was determined by one-step real-time reverse transcription-PCR, and the primers designed for the amplifications are shown in Table 2. Each reaction tube contained 120 ng of total RNA, 12.5 µl of Sybr green master mix (2x) (Applied Biosystems), 0.125 µl of reverse transcriptase (50 U/µl) (Applied Biosystems), 0.25 µl of an RNase inhibitor (20 U/µl) (Applied Biosystems), and 1 µl of each primer (2.5 pmol/µl) in a total volume of 25 µl. Amplification was carried out in an ABI Prism 7500 sequence detector system (Applied Biosystems), using the following protocol: reverse transcription at 48°C for 30 min, reverse transcription inactivation at 95°C for 10 min, 40 PCR cycles of 95°C for 15 s and 60°C for 1 min, and a dissociation phase at 60°C to 95°C for 30 min. The results were normalized to the levels of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA detected in each RNA sample. The decrease was calculated by the 2–
CT method, where CT is the threshold cycle (31).
|
|
Statistical analysis. All statistical evaluations were carried out with a one-way analysis of variance test and a Tukey posttest using the GraphPad Prism 4.0 software.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To determine if the change in distribution of the luminal chaperones correlates with the progression of the virus replication cycle, cells were infected and analyzed by immunofluorescence at different times postinfection. The distribution of the chaperones grp94, PDI, and calreticulin showed similar dynamics (shown for PDI in Fig. 1B). They showed a typical ER distribution at 4 hpi; however, as the infection proceeded, the chaperones began to accumulate next to viroplasms in some cells (see the data for 6 hpi), and this effect became evident in almost all cells by 8 hpi.
Silencing the expression of ER chaperones by RNAi. Given previous data indicating that the chaperones grp78, grp94, PDI, and calnexin interact with rotaviral proteins (37, 39, 65) and the change in localization observed in this work for some of these proteins in virus-infected cells, we explored their relevance, as well as that of calreticulin and ERp57, for the morphogenetic process of the virus. For this, the expression of chaperones was silenced by RNAi; two to four different siRNAs were tested for each gene, and the most effective siRNAs were subsequently used in this work (Table 1).
In these experiments, the cells were transfected with an siRNA, and 72 h posttransfection, the cells were harvested and the proteins detected either by immunofluorescence or by immunoblotting, as described in Materials and Methods. The viability of the cells transfected with the different siRNAs was not affected, compared to that of control mock-transfected and nontransfected cells, as judged by a cytotoxicity assay (live cell/dead cell assay; Molecular Probes) (data not shown). By immunofluorescence, the levels of expression of the chaperones were significantly decreased when the corresponding siRNAs were used; however, a complete absence of the silenced proteins was not observed for any of them, probably due to the long half-lives reported for these proteins and/or to a partial knockdown of their expression (45, 55; data not shown). In addition, based on these experiments, the transfection efficiency achieved with the siRNAs was found to be around 75%.
By immunoblot analysis, all siRNAs were found to decrease the accumulated amount of the corresponding protein by 59 to 87%, compared to that of cells transfected with an irrelevant siRNA (Fig. 2A). When the siRNA to grp78 was used, the level of this protein was reduced to 28% of that in control cells; however, under these conditions, the levels of grp94, PDI, calreticulin, and ERp57 increased 2.2- to 5.2-fold, while the amount of calnexin did not change. Similarly, in cells transfected with the siRNA to calreticulin, the expression levels of grp94 and PDI proteins increased 3.9- and 2.5-fold, respectively (Fig. 2A). The siRNAs directed to grp94, PDI, calnexin, and Erp57 altered the expression only of the proteins for which they were designed. To confirm the effectiveness of the siRNAs, the level of each targeted mRNA was quantified by real-time reverse transcription-PCR (Fig. 2B), using total cellular RNA obtained 72 h after siRNA transfection. As shown in Fig. 2B, the levels of the various mRNAs paralleled those observed for the corresponding proteins in the immunoblots, the only exceptions being the level of ERp57 mRNA, which increased twofold, and the level of PDI mRNA when the siRNA to calreticulin was used, which did not change; these levels did not correlate with the levels of the encoded proteins under these conditions (Fig. 2A).
|
|
We evaluated the possibility that knocking down the expression of the molecular chaperones had a general effect on the physiology of the ER, leading to a nonspecific effect on all proteins that transit through this organelle, regardless of whether they are specifically assisted or not by chaperones. For this, we tested the effect of silencing these proteins on the production of influenza virus, since the proper maturation of influenza virus hemagglutinin, but not of grp78, has been shown to depend on the presence of calnexin and calreticulin (42). As expected, the yield of influenza virus, as determined by a hemagglutination assay, was not affected when the expression of grp78 was silenced, but it decreased to 50% in cells infected in the presence of the siRNAs to calnexin and calreticulin (data not shown). These results strongly suggest that rotavirus specifically needs the actions of these proteins for the correct maturation of the infectious viral progeny.
Silencing gpr78 and calnexin decreases the specific infectivities of rotavirus TLPs and their ability to bind to the cell surface. We decided to further characterize the effect of silencing grp78 and calnexin on the assembly of rotavirus, since, in addition to reducing infectious viral yields when the corresponding siRNAs were used, these proteins have been reported to interact with rotavirus polypeptides (37, 65). Since the knockdown of grp94 had no effect on the yield of virus progeny, it was used as a control in these assays. To determine if the decreased production of infectious viral progeny induced by the siRNAs was due to a reduced or a defective assembly of infectious particles in the absence of the chaperones, cells transfected with the siRNA to grp78, calnexin, or grp94 or with an irrelevant siRNA were infected 72 h posttransfection, and the viral proteins were metabolically labeled with [35S]methionine and [35S]cysteine for 6 h starting at 6 hpi. The labeled virions were purified by CsCl density centrifugation, and the bands corresponding to DLPs and TLPs were collected and quantified as described in Materials and Methods. Of interest, when the expression of either grp78 or calnexin was knocked down, even though the total amount of particles assembled (DLPs plus TLPs) was similar to that assembled in control cells transfected with either the irrelevant siRNA or the siRNA to grp94, the TLP-to-DLP ratio increased (Fig. 4A), indicating that the assembly of TLPs was more efficient in cells having decreased levels of grp78 or calnexin. However, the TLPs assembled under these conditions had a specific infectivity (FFU/cpm) about 50% lower than that of TLPs obtained from control cells or cells transfected with the siRNA to grp94, suggesting that despite the fact that TLPs assembled more efficiently in grp78- and calnexin-deficient cells, the structures of these particles were faulty. These TLPs were also found to have a density slightly lower (1.35 g/cm3) than that of the particles isolated from control cells (1.36g/cm3). The viral proteins from DLPs and TLPs purified under the various conditions were found to be present in apparently the same stoichiometric ratios by SDS-PAGE (data not shown).
|
The viral particles assembled in cells with low levels of grp78 or calnexin have an altered conformation. To evaluate potential conformational changes in virions assembled in reduced levels of grp78 or calnexin, purified TLPs were incubated with increasing concentrations of EGTA for 10 min at 37°C, a treatment known to release the outer protein layer of the virus when the calcium concentration is below a threshold that varies for different rotavirus strains (8, 54). The infectivity of TLPs isolated from cells treated with siRNAs to either grp78 or calnexin was significantly more sensitive to low concentrations of EGTA (0.025 and 0.05 mM) than that of TLPs purified from cells transfected with the siRNA to grp94 or the irrelevant siRNA (Fig. 5A).
The stability of the virus infectivity was also tested by incubation of the virions at 45°C for various periods of time. The infectivity of TLPs purified from grp78-deficient cells was more sensitive under all conditions tested than that of control TLPs or TLPs isolated from cells transfected with the siRNA to grp94. On the other hand, the infectivity of TLPs from calnexin-deficient cells was significantly more stable than that of all other TLPs (Fig. 5B). All together, these results suggest that the surface proteins of TLPs purified from grp78- and calnexin-silenced cells have structural differences with respect to TLPs transfected with the irrelevant siRNA or the siRNA to grp94. These conformational differences are most probably slight, since they could not be clearly evidenced by treatment of the purified virus particles with different proteases and SDS-PAGE analysis (data not shown).
Processing of the viral glycoproteins is altered in grp78-, calnexin-, and calreticulin-silenced cells. To analyze if the posttranslational maturation of VP7 and NSP4 was altered in cells with low levels of grp78, calnexin, grp94, or calreticulin, the expression of these proteins was silenced, and the processing of the viral glycoproteins was determined by pulse-chase experiments and SDS-PAGE analysis. Cells were transfected with the corresponding siRNA and 72 h later were infected with RRV at an MOI of 3. At 8 hpi, the cells were metabolically labeled with [35S]methionine for 5 min and then chased for the periods of time indicated in Fig. 6 and 7. The viral proteins were immunoprecipitated with MAb M60 to VP7 or with polyclonal antibody C-239 to NSP4 and analyzed by SDS-PAGE under reducing conditions (Fig. 6 and 7). These proteins were also analyzed by PAGE under nonreducing conditions (see below).
|
|
In cells transfected with the siRNA to calreticulin, the mobility of VP7 was not affected; however, the amount of VP7 immunoprecipitated by MAb M60 was higher (already after 5 min of chase) than that in cells transfected with the siRNA control, suggesting that under these conditions, the disulfide bonds of VP7 form faster (Fig. 7A and B). The gel migration of NSP4 (i.e., the processing of the protein oligosaccharides) was not affected in calreticulin-deficient cells (not shown).
PDI but not ERp57 participate in the formation of VP7 disulfide bonds. Since calnexin has not been reported to have a direct role in the formation of protein disulfide bonds, we tested the involvement of PDIs in the oxidative folding of VP7. For this, cells were transfected with siRNAs to ERp57, which is known to be associated with the calnexin-calreticulin cycle (19), and to PDI, which has been shown to interact with VP7 (38) and has been proposed to participate in the cycle of grp78 (33). Seventy-two hours after siRNA transfection, the cells were infected and pulse-labeled for 5 min at 8 hpi and then chased for the periods of time indicated in Fig. 7, and the VP7 protein was analyzed by SDS-PAGE after immunoprecipitation with MAb M60. VP7 synthesized in PDI-deficient cells was consistently less efficiently precipitated than in ERp57-silenced and control, transfected cells (Fig. 7A). The smaller amount of immunoprecipitated VP7 in PDI-silenced cells was not due to degradation of the protein, since when total labeled proteins were resolved by SDS-PAGE, the accumulated levels of VP7 were very similar under all conditions tested (not shown), while immunoprecipitated VP7 was about 50% of that of control-transfected cells (Fig. 7B), indicating that PDI and not ERp57 participates in the formation of the correct disulfide bonds in VP7.
Comparative PAGE analysis under reducing and nonreducing conditions showed no significant difference in the migration of VP7 from infected control cells from that in cells with a reduced expression of chaperones (data not shown). These results indicate that neither the siRNAs to calnexin nor those to PDI prevent the formation of S-S linkages in VP7, since the migration of a VP7 protein having reduced disulfide bonds should have been easily detected as a slower-migrating entity by PAGE analysis under nonreducing conditions. It thus seems that knocking down the expression of calnexin and PDI rather than blocking the formation of S-S bonds might prevent the formation of additional or shifted disulfide bonds that have been suggested to occur posttranslationally in VP7 (39). These additional or shifted S-S linkages seem to make the M60 epitope more available to this antibody (39).
The VP7 protein assembled into virions in cells with reduced levels of chaperones are not correctly structured. Given the above-noted results, we further characterized the conformation of VP7 assembled into virions produced in cells where the expression of the chaperones grp78, calnexin, PDI, and calreticulin was knocked down. For this, purified TLPs were immunoprecipitated with MAb M60 to VP7, before or after treatment with EGTA, to release the surface viral proteins. In the case of untreated TLPs, the VP7 antibody pulled down complete viral particles with good efficiency, as judged by the presence in the immunoprecipitated material of all structural viral proteins (Fig. 8, ipp lane without EGTA of the irrelevant-siRNA panel). When the particles were treated with EGTA, MAb M60 immunoprecipitated mostly VP7 (Fig. 8, ipp lane with EGTA in the irrelevant-siRNA panel), indicating that this protein was disassembled from the particles by calcium chelation.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The potential role of the chaperones in these events was explored by RNAi. When the expression of grp78 was knocked down, the levels of the other chaperones, with the exception of calnexin, increased. This was not surprising, since grp78 is known to bind to the luminal domain of the ER transmembrane stress-sensing glycoprotein ATF6, keeping it in an inactive form (56). ATF6 is a factor involved in the activation of the transcription of genes containing a mammalian ER stress response element, which is present in the genes encoding the molecular chaperones grp78, grp94, calreticulin, ERp57, and PDI (66). In grp78-silenced cells containing no or a small amount of this protein, ATF6 is probably released from its interaction with this protein, becoming activated and promoting the transcription of the chaperone genes (56). Calnexin is probably not induced under these conditions since its gene does not have an ER stress response element box (25).
When the expression of calreticulin was knocked down, the levels of grp94 and ERp57 mRNAs increased. Since calreticulin has also been reported to bind ATF6 (21), the mechanism for the enhancement of the transcription of these chaperones could be similar to that proposed above for grp78. However, an enhancement in the transcription of the grp78 and PDI genes was not observed, suggesting that unknown regulatory mechanisms that induce the expression of folding factors to compensate for the decreased levels of calreticulin may be activated. On the other hand, the level of the PDI and ERp57 foldases did not correlate with the observed levels of their mRNAs under conditions with reduced amounts of calreticulin, suggesting a possible posttranscriptional regulation for the synthesis of these proteins.
Grp94 and grp78. Despite the fact that grp94 modifies its intracellular distribution in infected cells (this work) and associates with VP7 and VP4 (65; data not shown), knocking down its expression had no effect on viral infectivity. As suggested for its cytoplasmic homologue, hsp90 (2), this chaperone could act as a scaffold for proteins forming the complex of the grp78/grp94 system and may not be essential for their function. On the other hand, the infectious rotavirus progeny decreased when grp78 was knocked down despite increased levels of expression of most of the other molecular chaperones tested, suggesting that grp78 is important either directly or indirectly for the assembly of infectious virus and cannot be replaced by any other chaperone.
It has previously been reported that grp78 interacts with VP7 and VP4 but not with NSP4 (65). In contrast, in this work we found that in grp78-silenced cells, the electrophoretic mobility of VP7 was not affected in pulse-chase experiments but that the mobility of NSP4 was retarded until 30 min of chase, after which time it reached the same mobility as NSP4 from control cells. This is the first report, as far as we know, in which the absence of grp78 has been shown to affect the oligosaccharide trimming of a protein. The delay in processing of the NSP4 oligosaccharide chains could be the result of a deficient folding of the protein that hampers the actions of glucosidases, as has been reported for the influenza virus hemagglutinin (41). This deficient folding could be the consequence of the lack of a direct interaction (not described so far) between grp78 and the viral protein or, alternatively, of a generally altered quality control of the ER caused by the deficiency of this chaperone (12, 20, 22).
When TLPs assembled in grp78-deficient cells were compared to TLPs produced in control cells, they were found to have altered susceptibilities to EGTA and heat treatments, to bind less efficiently to MA104 cells, to have a lower specific infectivity, and to be less efficiently immunoprecipitated by MAb M60. All these observations suggest that grp78 is involved in the quality control of TLP assembly. These results also suggest that although the carbohydrate processing and formation of disulfide bonds of VP7 do not seem to be affected in the absence of grp78, this chaperone is probably required for the correct assembly of VP7 into the viral particles.
VP4 is known to be the virus attachment protein. Thus, the deficient binding of TLPs assembled in grp78-silenced cells could result from a misfolded VP4, since this protein has been reported to interact with grp78 (65), or from an altered conformation of VP4 influenced by its interaction with an incorrectly assembled VP7 (34).
Calnexin and calreticulin. In cells with reduced levels of calnexin, the carbohydrate trimming of VP7 does not seem to be affected, although its oxidative folding was impaired, similar to what was observed for influenza virus (42). A direct interaction between VP7 and calnexin has not been reported, although it could exist and might have not been detected if it were short-lived and/or weak. Alternatively, the ER quality control dependent on the calnexin-calreticulin cycle could be affected, reducing the efficiency of PDI to establish the disulfide bridges of the protein. An affected ER quality control would also be consistent with the fact that TLPs from calnexin-silenced cells have a decreased specific infectivity, have an altered stability to EGTA and heat treatments, and have a diminished reactivity with MAb M60, which was also observed for the VP7 released by EGTA from these particles.
It has been reported that calnexin interacts with the virus nonstructural protein NSP4 (37), and in this work, the trimming of NSP4 was found to be accelerated in calnexin-deficient cells. However, the fact that castanospermine (a drug that inhibits the activity of glucosidases I and II) prevents the generation of the monoglucosylated form of the NSP4 glycans and the interaction of the protein with calnexin, while having a limited effect on virus infectivity, has led others to conclude that glucose trimming and calnexin interaction with NSP4 are not critical for virus infectivity (37). The marginal effect observed for castanospermine on virus infectivity could be explained by the fact that calnexin is known to interact with some proteins in a manner independent of monoglucosylated N-linked oligosaccharides (4, 11, 50, 63, 64). Thus, untrimmed NSP4 produced in the presence of this drug could still be bound by calnexin with a weak interaction and for a short period of time, as has been shown for the hepatitis C virus glycoproteins (5).
When the level of calreticulin was decreased, the infectivity of the viral progeny also decreased compared to that of the control virus. However, in contrast to the delayed formation of disulfide bonds observed in calnexin-deficient cells, in reduced levels of calreticulin, the folding of VP7 was found to be accelerated, as has also been reported for the maturation of the influenza virus hemagglutinin and the p62/E1 proteins of Semliki Forest virus, as well as for cellular proteins (42). The accelerated folding of VP7 may result from the decreased level of quality control in the ER in the absence of calreticulin. These altered conditions seem to have an impact not on the proper folding of VP7, as judged by the efficient immunoprecipitation of this protein by MAb M60, but rather on the correct assembly of this protein into virus particles.
PDI and ERp57. The correct formation of disulfide bonds catalyzed by members of the PDI family is frequently the limiting reaction for the proper folding of proteins in the ER (19, 53). In this work, PDI was found to be responsible for the formation and/or rearrangement of the disulfide bonds in VP7 that have been proposed to occur posttranslationally (39), while ERp57 does not seem to be involved. In PDI-silenced cells, only about 50% of the synthesized VP7 molecules were folded correctly, as judged by their interaction with MAb M60 (Fig. 7), which is comparable to the level of silencing of PDI expression (Fig. 2). Thus, it seems that this folding enzyme is limiting and critical for the formation of the disulfide bonds present in VP7. The marginal 30% reduction in infectivity of TLPs assembled in PDI-deficient cells suggests that not all VP7 molecules assembled into viral particles need to have disulfide bonds for the virus to be infectious. The fact that the disulfide bridges of VP7 are correctly formed in grp78-deficient cells indicates that even though grp78 and PDI have been reported to function in the same ER folding system, the interaction of grp78 with VP7 is not necessary for the formation of its disulfide bonds by PDI.
ER molecular chaperones and rotavirus morphogenesis. It has been shown that VP7 assembles efficiently and functionally into DLPs in vitro, with no requirement for chaperones, but that the assembly of VP4 is relatively inefficient, although sufficient to produce infectious virus. The order of addition of the two proteins is critical in reconstituting infectious virus particles in that VP4 needs to be added before VP7 (61). In contrast to the low requirements for the assembly of infectious TLPs in vitro, we found in this work that multiple chaperones and a foldase enzyme are involved in the assembly of rotavirus TLPs in vivo. The complex pathway of rotavirus morphogenesis in vivo might reflect the need for an optimized and ordered assembly process of VP4 and VP7 into DLPs. If in vivo the assembly of the surface viral proteins needs to follow the same order found in vitro, the sequestration of VP7 into the ER would allow the initial incorporation of the virus spike protein VP4 into DLPs, which is thought to occur in the cytoplasm (13), followed by the assembly of VP7 in the lumen of the ER.
Members of the two folding systems proposed to function in the ER (19, 33) appear to be required for the assembly of rotavirus TLPs. These observations are in agreement with previous findings that indicate that the calnexin-calreticulin and grp78 chaperone systems may act sequentially on protein targets (3, 43, 44). This is not surprising in the case of rotavirus, given the complex steps of virus assembly that take place in the lumen of the ER, involving at least three viral proteins (NSP4, VP7, and VP4). Further experiments are needed, however, to determine the direct interactions that exist between chaperones and viral proteins.
Given the complexity of the steps of rotavirus morphogenesis that occur in the ER, characterization of the assembly of infectious rotavirus represents a challenging, but valuable, model for increasing our knowledge of the organization and operation of the complex network of chaperones and enzymes involved in maintaining the quality control of this organelle and cellular homeostasis.
|
ACKNOWLEDGMENTS |
|---|
This work was partially supported by grant 55005515 from the Howard Hughes Medical Institute. L.M.-A. is a recipient of scholarships from the National Council of Science and Technology—Mexico and DGEP/UNAM.
|
FOOTNOTES |
|---|
Published ahead of print on 2 April 2008.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CT method. Methods 25:402-408.[CrossRef][Medline]
