Biological Differences between Vesicular Stomatiti
http://www.100md.com
病菌学杂志 2005年第6期
Department of Microbiology, University of Alabama School of Medicine, Birmingham, Alabama
ABSTRACT
We previously generated recombinant vesicular stomatitis viruses (VSV) based on the Indiana serotype genome which contained either the homologous glycoprotein gene from the Indiana serotype (VSIV-GI) or the heterologous glycoprotein gene from the New Jersey serotype (VSIV-GNJ). The virus expressing the GNJ gene was more pathogenic than the parental VSIV-GI virus in swine, a natural host (26). For the present study, we investigated the biological differences between the GI and GNJ proteins that may be related to the differences in pathogenesis between VSIV-GI and VSIV-GNJ. We show that the capacities of viruses with either the GNJ or GI glycoprotein to infect cultured cells differ depending on the pH. VSIV-GNJ could infect cells at acidic pHs, while the infectivity of VSIV-GI was severely reduced. VSIV-GNJ infection was also more sensitive to inhibition by ammonium chloride, indicating that the GNJ protein had a lower pH threshold for membrane fusion. We applied selective pressure to VSIV-GI by growing it at successively lower pH values and isolated variant viruses in which we identified amino acid changes that conferred low-pH-resistant infectivity. Repeated passage in cell culture at pH 6.8 resulted in the selection of a VSIV-GI variant (VSIV-6.8) that was similar to VSIV-GNJ regarding its pH- and ammonium chloride-dependent infectivity. Sequence analysis of VSIV-6.8 revealed that it had a single amino acid substitution in the amino-terminal region of the glycoprotein (F18L). This alteration was shown to be responsible for the observed phenotype by site-directed mutagenesis of a VSIV-GI full-length cDNA and analysis of the recovered engineered virus. A further adaptation of VSIV-6.8 to pHs 6.6 and 6.4 resulted in additional amino acid substitutions in areas of the glycoprotein that were not previously implicated in attachment or fusion.
INTRODUCTION
Isolates of Vesicular stomatitis virus (VSV) are enveloped, nonsegmented, negative-strand RNA viruses that belong to the genus Vesiculovirus in the Rhabdoviridae family. VSV infects domestic animals such as cattle, swine, and horses, causing vesicular lesions in the tongue, oral tissues, udders, and hooves (23, 38). Important economic losses in milk and meat production are associated with VSV outbreaks, not only due to the virus infection itself but also due to quarantines, trade barriers, and other governmental measures aimed to control virus spread (30). These measures are particularly pertinent because the lesions induced by VSV are similar to those of foot-and-mouth disease, one of the most contagious animal diseases.
The VSV genome is delivered to the cytoplasm of host cells, where replication occurs, via receptor-mediated endocytosis of viral particles and subsequent pH-induced fusion of the viral envelope with the endosomal membrane (29). The VSV G protein, the sole viral surface glycoprotein, is required for attachment and fusion. Attachment seems to occur via electrostatic interactions between positive charges on the glycoprotein surface and negatively charged phospholipids in the plasma membrane (3, 27, 40), and fusion involves structural changes in the protein driven by the low pH of the endosomal compartment (4, 29, 46). The VSV G protein is distinct in several ways from other viral fusion proteins. For example, conformational changes in the VSV G protein induced by the pH are reversible, and VSV G can exist in different conformational forms (8, 9, 32, 34). Furthermore, VSV G does not contain an obvious hydrophobic sequence which would be indicative of a fusion peptide. Biochemical and mutational studies have indicated that VSV G has an internal fusion peptide between residues 118 and 139, but mutations at residues 395 to 418 also affect membrane fusion activity (13, 15, 16, 24, 42, 43, 48). Additionally, His-148 and His-149 have been suggested to directly participate in VSV fusion (5). Recently, the membrane-proximal domain encompassing residues 449 to 462 has been shown to contribute to G protein-mediated membrane fusion (19, 20).
There are two major serotypes of VSV, Indiana (VSIV) and New Jersey, which are distinguished by neutralizing antibodies against the G protein (6, 22). In addition to their antigenic structures (21), the Indiana (GI) and New Jersey (GNJ) glycoproteins also differ in the number (511 and 517, respectively) and composition of amino acids (only 50% identity) (17, 39), in posttranslational modifications (7), and in folding (25, 28). Correspondingly, Indiana and New Jersey strains are not equally important regarding VSV pathogenesis. Outbreaks caused by New Jersey strains are more frequent and more severe than those caused by Indiana strains (2, 36, 44). We recently reported that this can be related, at least in part, to the G glycoprotein, as the replacement of the GI glycoprotein with the GNJ protein in the background of a recombinant VSIV substantially increased the pathogenicity of the virus in swine (26). However, the basis for these differences is not known. For this study, we investigated the biological differences between the GI and GNJ proteins that may help to explain their role in VSV pathogenesis. The VSV GI glycoprotein structure has been shown to be very sensitive to even mildly acidic pHs, at which the exposure of hydrophobic domains has been reported to occur (4). However, the relevance of this observation for VSV biology, if any, remains to be addressed. Another question that has not been clarified is whether or not the GI and GNJ glycoproteins are equally sensitive to pH changes. For this study, we examined the effect of low pHs on the infectivity of VSIV-GI and VSIV-GNJ in cell culture. We demonstrated that VSIV-GNJ was able to replicate at acidic pH values at which VSIV-GI replication was almost completely inhibited. In addition, by using the selective pressure of growth at increasingly acidic pHs, we identified amino acid changes in the GI glycoprotein that increased the ability of VSIV-GI to infect cells at low pHs.
MATERIALS AND METHODS
Virus and cells. VSIV-GI and VSIV-GNJ were recovered from cDNA clones and have been described previously (26, 45). The San Juan isolate of the Indiana serotype of VSV provided the template for all genes except for the G protein gene, which was derived from either the Indiana strain Orsay (VSIV-GI) or the New Jersey strain 95COB (VSIV-GNJ) (37). Baby hamster kidney (BHK-21) cells were used to recover viruses from cDNAs and for the selection of VSIV-GI variant viruses by passage at acidic pHs. Vero-76 cells were used for growth experiments involving acidic pHs or ammonium chloride and for plaque assays.
Virus replication at acidic pHs. Vero-76 cells growing in six-well plates were washed with Hanks' balanced salt solution (HBSS) and infected at a multiplicity of infection (MOI) of 0.01 PFU per cell in minimum essential medium without sodium bicarbonate containing 2% fetal calf serum (MEM2) adjusted to pH 7.2, 7.0, 6.8, 6.6, or 6.4 with 50 mM HEPES. Viruses were adsorbed for 1 h at 37°C at atmospheric concentrations of CO2 to avoid acidification of the medium. After the adsorption period, the inoculum was removed and the monolayers were washed with HBSS. Fresh MEM2 at the desired pH was added, and the cells were incubated for an additional 17 to 22 h. Supernatant fluids were harvested and clarified by low-speed centrifugation, and viral yields were determined by plaque assays on Vero-76 cells. In some cases (indicated in the figure legends), viruses were preincubated for up to 60 min at 37°C in the corresponding medium prior to the inoculation of cells.
Selection of VSIV-GI mutant viruses by passage at low pHs. (i) Virus selection at pH 6.8 (VSIV-6.8). BHK-21 cells growing in 60-mm-diameter dishes were infected at an MOI of 0.01 in MEM2, pH 6.8. Viruses were adsorbed for 1 h at 37°C at the atmospheric CO2 concentration. After the adsorption period, fresh MEM2, pH 6.8, was added, and the cells were incubated for an additional 48 h. The supernatant fluid was harvested and clarified by low-speed centrifugation. The viral yield was determined by a plaque assay on Vero-76 cells, and 0.01 PFU/cell was used to infect new cells. After five passages, the virus was amplified by the infection of BHK-21 cells at a low MOI in MEM2, pH 6.8, pelleted by centrifugation at 21,000 rpm for 90 min in a Ty30 rotor, and purified by centrifugation in a 15 to 45% (wt/vol) sucrose gradient. Finally, the virus was concentrated by centrifugation and resuspended in Dulbecco's modified Eagle medium containing 2% fetal calf serum (DMEM2).
(ii) Virus selection at pH 6.6 (VSIV-6.6). The supernatant from VSIV-6.8 passage three was used to infect BHK-21 cells at an MOI of 0.01 in MEM2, pH 6.6, and the virus was passaged, amplified, and purified as described above for VSIV-6.8, except that the pH during the passages and amplification was maintained at 6.6 instead of 6.8.
(iii) Virus selection at pH 6.4 (VSIV-6.4). The supernatant from VSIV-6.6 passage three was used to infect BHK-21 cells at an MOI of 0.01 in MEM2, pH 6.4, and the virus was passaged, amplified, and purified as described above for VSIV-6.8, except that the pH during the passages and amplification was maintained at 6.4 instead of 6.8. RNAs purified from BHK-21 cells infected with the variant viruses were amplified by reverse transcription-PCR (RT-PCR) with primers specific for the matrix (M) and polymerase (L) genes by use of a OneStep RT-PCR kit (QIAGEN), and the nucleotide sequences of the G genes were determined.
Recovery of recombinant viruses from full-length cDNA clones containing introduced mutations in the glycoprotein gene to confirm phenotype. The full-length cDNA clone from which VSIV-GI was recovered (see above) was digested with the restriction enzyme BstBI, and the fragment containing the GI protein gene was subcloned into the plasmid pBluescript-GUS (31) by conventional cloning techniques. The GI gene was mutagenized by the use of a QuikChange site-directed mutagenesis kit (Stratagene) and specific mutagenic primers. The mutagenized BstBI fragments were sequenced to confirm that they contained the desired nucleotide changes and then used to replace the original fragment in the VSIV-GI full-length cDNA clone. Viruses were recovered by transfection of BHK-21 cells as previously described (1, 45), and the RNAs from BHK-21 cells infected with the recovered viruses were amplified and the G gene was sequenced to confirm that the introduced nucleotide changes were maintained during the recovery process.
Virus replication in the presence of ammonium chloride. Confluent Vero-76 cells in six-well plates were incubated in MEM2, pH 7.2, containing 0, 1, 2, 4, 8, or 16 mM ammonium chloride for 30 min at 37°C. The medium was removed, and the cells were infected at an MOI of 0.01 in the same medium. After 1 h of adsorption at 37°C, fresh medium containing ammonium chloride was added, and the cells were incubated for an additional 15 h. Supernatant fluids were collected and clarified by low-speed centrifugation, and viral yields were determined by plaque assays on Vero-76 cells.
RESULTS
VSIV-GI and VSIV-GNJ differ in their capacities to grow at acidic pHs. To test whether the serotype of the glycoprotein expressed by VSV influences virus infectivity at mildly acidic pHs, we examined the capacities of two recombinant viruses that differed only in their glycoprotein genes, VSIV-GI and VSIV-GNJ, to replicate in cell culture at acidic pH values. Vero-76 cells were infected at an MOI of 0.01 with VSIV-GI or VSIVGNJ in HEPES-buffered MEM2 at pH values that decreased in steps of 0.2 pH units from 7.2 to 6.4 and were then incubated at 37°C. At 17 h postadsorption, supernatant fluids were collected and clarified by low-speed centrifugation, and the virus yields were titrated by plaque assays on Vero-76 cells. Both virus infections yielded similar amounts of infectious particles at pH 7.2, but the titer of VSIV-GI started to decline when replication was carried out at pH 7.0 and dropped almost 30-fold when replication was done at pH 6.8, whereas at the same pH values VSIV-GNJ replication levels remained essentially unchanged (Fig. 1). At pHs 6.6 and 6.4, the amounts of infectious particles of both viruses in supernatants were severely reduced (Fig. 1). The virus titers in the supernatants correlated with the cytopathic effect observed (data not shown). At pHs 7.2 and 7.0, the cell monolayers infected with VSIV-GI were completely destroyed, while at pH 6.8 and below, cytopathic effects were not obvious, except for a few patches of infection at pH 6.8. However, the cell monolayer was completely destroyed when the cells were infected with VSIV-GNJ at pH 6.8.
pH-induced reversible changes in virions, but not in cells, are responsible for the decreased ability of VSIV-GI to infect at acidic pHs. To investigate how the VSIV-GI and VSIV-GNJ virus infectivities were affected by incubation at a low pH (6.8), we diluted the viruses in medium at pH 7.2 or pH 6.8 and incubated them for 0 to 60 min at 37°C prior to infection. Vero-76 cells were then infected at an MOI of 0.01 with these viruses at the same pH used for incubation. At 22 h postadsorption, supernatant samples were harvested and clarified by centrifugation at low speed, and virus titers were determined by plaque assays on Vero-76 cells (Fig. 2). At pH 7.2, VSIV-GI and VSIV-GNJ virus yields were similar, regardless of the preincubation time at 37°C (Fig. 2A). However, at pH 6.8, VSIV-GNJ virus yields were unaffected by the previous incubation at 37°C, whereas VSIV-GI virus yields decreased rapidly with increased times of incubation and dropped more than 100-fold when the virus was preincubated at pH 6.8 for 1 h at 37°C (Fig. 2B). These results show that at pH 6.8, VSIV-GI infectivity was decreased by changes affecting the virions. However, it does not exclude the possibility that changes in the cells could also influence virus infectivity. To test this, we incubated the cells for 1 h at pH 7.2 or 6.8 prior to infection with VSIV-GI and also preincubated the virus for 0 or 60 min in MEM2, pH 6.8, at 37°C. After 1 h of adsorption at 37°C, fresh MEM2, pH 6.8, was added, and the cells were incubated at 37°C for an additional 22 h. Supernatant fluids were taken and assayed for virus yield by a plaque assay. Infectivity was diminished when the virus, but not the cells, was preincubated at pH 6.8 (Fig. 2C). This experiment showed that the exposure of cells to pH 6.8 did not measurably influence VSIV-GI infectivity.
Using a soluble form of VSIV glycoprotein, Crimmins et al. (9) showed that the conformational change detected when the protein was exposed to a pH close to 5 was reversible. We analyzed whether the loss of VSIV-GI infectivity after incubation at pH 6.8 was recovered by a subsequent incubation at pH 7.2. Vero-76 cells were infected at an MOI of 0.1 with VSIV-GI that had been previously incubated in MEM2 at 37°C as follows: (i) 2 h at pH 7.2, (ii) 1 h at pH 7.2 and 1 h at pH 6.8, (iii) 1 h at pH 6.8 and 1 h at pH 7.2, or (iv) 2 h at pH 6.8. After 1 h of adsorption at 37°C, the inoculum was removed, the cells were washed two times with HBSS, and fresh MEM2 at pH 6.8 was added to prevent additional rounds of infection. At 21 h postadsorption, supernatants were collected and clarified by low-speed centrifugation, and the virus yield was determined by a standard plaque assay on Vero-76 cells. The VSIV-GI infectivity dropped >100-fold when the virus was incubated for 2 h at pH 6.8 or incubated for 1 h at pH 7.2 and 1 h at pH 6.8, but the infectivity was recovered when the virus that had previously been incubated at pH 6.8 was subsequently incubated at pH 7.2 (Fig. 2D). These and the above results demonstrate that the changes in the VSIV-GI virions responsible for the loss of infectivity at pH 6.8 were reversible.
The ability of VSIV-GI to infect at acidic pHs can be increased by a single amino acid substitution in the amino-terminal region of the glycoprotein. Since VSIV-GI and VSIV-GNJ have the same VSIV genomic backbone, which differs only in the glycoprotein expressed, the differences observed between the two viruses regarding pH-dependent infectivity should reside in this protein. To identify the amino acid(s) or region(s) of the glycoprotein that modulates the ability of VSV to infect at acidic pHs, we utilized selective pressure by growing the virus at successively decreasing pH values. VSIV-GI was passaged five times at a low MOI (0.01) in BHK cells at pH 6.8 as described in Materials and Methods. The virus from passage five was amplified, purified through a sucrose gradient, and resuspended in DMEM2. All viral stocks were made in the same cell line and in the same medium. After five successive passages at pH 6.8, the recovered virus (VSIV-6.8) showed a pH-dependent infectivity similar to that of VSIV-GNJ (Fig. 3A). Sequence analysis of RNAs from VSIV-6.8 revealed a single nucleotide substitution in the glycoprotein gene that resulted in an amino acid change at position 18 from phenylalanine to leucine (F18L). To confirm that this change was responsible for the VSIV-6.8 phenotype, we introduced this single nucleotide mutation by site-directed mutagenesis into the full-length cDNA clone from which VSIV-GI was derived and recovered the recombinant virus (rVSIV-6.8) by transfection of BHK-21 cells. While the VSIV-GI virus yield was decreased >1,000-fold at pH 6.8 compared with the virus yield at pH 7.2, the VSIV-6.8 and rVSIV-6.8 virus yields were essentially the same and remained high at pH 6.8 (Fig. 3B). This result demonstrated that the amino acid substitution F18L increased the ability of VSIV-GI to grow at mildly acidic pHs.
Further selection of VSIV-GI variant viruses at pHs 6.6 and 6.4 leads to additional amino acid substitutions in different areas of the glycoprotein. To explore whether further adaptation to pHs 6.6 and 6.4 would select additional mutations in the glycoprotein, we used the VSIV-6.8 virus from passage three to infect BHK-21 cells at pH 6.6. After five passages at pH 6.6, the virus was amplified, purified by sucrose gradient centrifugation, and resuspended in DMEM2. Passage three of this virus (VSIV-6.6) was used to select a virus variant at pH 6.4 (VSIV-6.4). We used passage three in both cases to generate the following pH-adapted viruses because at this passage there was an evident increase in the in vitro cytopathic effect of the virus, indicating that a variant virus was already present in the viral population. This stepwise selection facilitated the process of selection, as direct selection at pH 6.6 or 6.4 from the original stock yielded little or no virus. Sequence analysis of the VSIV-6.6 viral RNA by RT-PCR showed that in addition to the F18L mutation found in the VSIV-6.8 virus, VSIV-6.6 had two nucleotide substitutions that led to the amino acid changes Q301R and K462R in the G protein (Fig. 4C). The VSIV-6.4 glycoprotein had the three amino acid substitutions found in VSIV-6.6 plus an additional nucleotide change that led to the amino acid replacement H65R (Fig. 4C). Viruses VSIV-6.6 and VSIV-6.4 both grew well at pH 6.6, and although the viral titers decreased at pH 6.4, they were >10 times higher than those of VSIV-6.8 (Fig. 4A). The three mutations found in VSIV-6.6 were introduced into the VSIV-GI full-length cDNA, and the recombinant virus was recovered by transfection of BHK cells (rVSIV-6.6). To analyze the individual contributions of the Q301R and K462R mutations to the ability of VSIV-6.6 to grow at pH 6.6, we introduced these mutations separately along with the F18L mutation into the VSIV-GI cDNA clone by QuikChange mutagenesis. The corresponding viruses were recovered after transfection of the cDNAs into BHK-21 cells. The G sequences of all recombinant viruses were confirmed by nucleotide sequencing of RT-PCR products of viral RNAs from infected cells and were found to contain only the introduced mutations. rVSIV-6.6 and the virus containing the Q301R mutation grew as well as VSIV-6.6 at pH 6.6, but in contrast, the titer of the virus containing the K462R mutation was almost 100-fold lower at that pH (Fig. 4B). These results demonstrate that the Q301R substitution is a major contributor to the ability of VSIV-6.6 to grow at low pHs. However, the K462R change also contributed, since the virus containing this change yielded higher titers than VSIV-6.8 at pH 6.6 (Fig. 4B).
Effect of ammonium chloride on virus infectivity. VSV enters the host cell via receptor-mediated endocytosis, and the nucleocapsid is liberated in the cytoplasm by fusion of the viral envelope and the endosomal membrane. This process is mediated by conformational changes in the G glycoprotein at acidic pHs and can be blocked with compounds, such as ammonium chloride, that prevent endosomal acidification (29). We analyzed whether GI and GNJ behaved distinctly in an assay of virus fusion inhibition by ammonium chloride. The infectivities of VSIV-GI and VSIV-GNJ were analyzed in the presence of various amounts of ammonium chloride ranging from 1 to 16 mM. Samples from the supernatant fluids were taken at 15 h postadsorption, and viral titers were determined by plaque assays. A progressive decrease in the virus titer was observed with increasing concentrations of ammonium chloride for both viruses (Fig. 5). However, VSIV-GNJ infectivity declined faster than that of VSIV-GI, and at 8 mM ammonium chloride, the difference was almost 100-fold (Fig. 5). These data suggest that VSIV-GNJ requires a more acidic pH to fuse with the endosomal membrane than does VSIV-GI, and these findings appear to correlate with the results from the experiments showing that VSIV-GNJ remained infectious at lower pHs than did VSIV-GI. There was also a correlation between the pHs at which VSIV-GI variant viruses were selected and their sensitivities to ammonium chloride, i.e., the lower the pH at which the virus was selected, the lower the concentration of chemical that was needed to inhibit viral infection (Fig. 5). In conclusion, the experiments with ammonium chloride are in agreement with the experiments involving VSV infectivity at various pHs and emphasize the different sensitivities of the GI and GNJ glycoproteins to acidic pHs, a phenomenon that may have important consequences for VSV biology in vivo.
DISCUSSION
We previously reported that the replacement of the glycoprotein gene of a recombinant vesicular stomatitis Indiana virus (VSIV-GI) with the glycoprotein gene of the New Jersey serotype renders the virus (VSIV-GNJ) more pathogenic to swine, a natural host (26). This correlates with field observations showing that New Jersey strains are more pathogenic than Indiana strains (2, 36, 44). In the present study, we have initiated a characterization of the biological differences between the GI and GNJ glycoproteins that may help us to understand the above results. In particular, we have analyzed the effect of mildly acidic pHs on the in vitro infectivities of VSIV-GI and VSIV-GNJ. We found that VSIV-GI infectivity decreased rapidly when the infection of cells was carried out at pH values below 7.2, while VSIV-GNJ still produced high virus titers at pH 6.8. VSIV-GI, but not VSIV-GNJ, quickly lost its infectivity after incubation at pH 6.8 prior to cell infection. However, VSIV-GI infectivity could be recovered by reincubation at pH 7.2. These results indicate that a reversible conformational change took place in the GI glycoprotein that made the VSIV-GI virus less able to infect cells at pH 6.8. Thus, this work shows that the GI and GNJ glycoproteins are not equally sensitive to mildly acidic pH values. These findings raise the possibility that these differences may be relevant to the epidemiology and/or pathogenesis of the virus. For example, our results suggest that the Indiana strains are more readily inactivated by environmental pH variations than the New Jersey strains during natural VSV infection and/or transmission.
There are several steps at which exposure to low pHs could block VSV replication, including attachment, internalization, fusion, budding, and release from infected tissues. Differences in pH sensitivity also may be translated into differences in oligomerization (12), maturation (25, 28), and/or transport to the cell surface (18). An attractive explanation for the different sensitivities of GI and GNJ to low pHs can be discussed in the context of the previously published "three-state model" that relates pH-dependent conformational transitions of the GI protein to mechanisms of viral fusion (8, 32, 33). At a neutral pH, the GI protein is in a native (tense) state. At slightly acidic pHs, the GI protein undergoes a proton-driven shift either to an activated (relaxed) state, which is fusion active, or to an inactivated (desensitized) state, which is fusion inactive. The active state is achieved when the conformational changes result in the movement of the fusion active region of the protein into the target membrane. The desensitized state occurs when this region fails to insert into the membrane. These changes are reversible (9), so there is a pH-dependent equilibrium between the different conformational states of the glycoprotein (8, 32). We can interpret our results in light of this model as follows: at mildly acidic pH values, the GI conformation shifts to the desensitized state and the protein is thus not able to initiate fusion and to infect, while GNJ is still in the native state and is fusion competent. In this case, GNJ would have a more stable structure, such that more acidic conditions are required to trigger the conformational changes leading to desensitization. The ability of VSIV-GNJ to grow at low pH values correlated with its increased sensitivity to ammonium chloride, which also indicated a requirement for more acidic conditions for fusion and endosomal release for this virus (14). According to this observation, the pH values governing not only desensitization, but also the other conformational states of GNJ, would be shifted to more acidic values compared to those for GI.
To examine whether GI could be modified to respond to mildly acidic pHs in a similar fashion as GNJ, we exerted selective pressure on VSIV-GI viruses by passage at increasingly lower pH values, from 6.8 to 6.4. The recovered, passaged viruses were able to infect cells at progressively lower pHs and were also increasingly more sensitive to ammonium chloride. Although a shift of the pH optimum of fusion has been used as a criterion to characterize the fusion peptides of several viral proteins, including GI (16, 48), the mutations observed in the low-pH-adapted VSIV-GI viruses described here did not map to any region of the GI glycoprotein reported to be directly involved in fusion (see the introduction). The only amino acid substitutions that were in regions that are known to be functionally important for VSV biology were F18L and K462R. The K462R mutation is localized in a protein segment involved in budding and fusion potentiation (20, 35), but we have shown that this amino acid change is less important than Q301R for the ability of VSIV-6.4 to infect at low pHs. Peptides corresponding to the amino-terminal region of GI, where the F18L mutation was mapped, have been shown to be pH-dependent hemolysins and hemagglutinins (41), although mutations in this region that abolished the hemolytic activity of the peptide did not affect the pH-dependent fusion activity of the glycoprotein (47).
Except for F18L, the three other amino acid changes found in the low-pH-adapted viruses were changes to arginine, suggesting that positive charges are important for modulating pH-induced conformational changes in GI. This is consistent with the observation that the K462R mutation had less impact on the phenotype of selected viruses than did H65R or Q301R. We do not know the mechanism(s) by which all of the observed amino acid changes influence GI sensitivity to pH, but the results described here resemble those reported previously for variant influenza virus hemagglutinins selected for their ability to grow in cells treated with amantadine hydrochloride, which raises the endosomal pH (10). Mutant viruses that fuse at elevated pH values contain several changes, most of them involving charged amino acids localized in different regions of the hemagglutinin primary sequence that appear to alter intra- and intersubunit contacts, destabilizing part of the protein structure. This would lower the energy required to trigger the conformational change (10, 11). Our case would be the opposite, as the amino acid substitutions in GI would be predicted to lead to a more stable structure so that a lower pH is required to destabilize it. It is worth mentioning that arginine has a side chain with three nitrogen atoms, which confer a high potential to form hydrogen bonds that would increase the stability of the protein.
Mutant GI proteins showing a reduced pH threshold for fusion have been described previously (15, 43, 48), and recombinant VSIV viruses encoding some of these mutated glycoproteins have been recovered that are more sensitive to chloroquine than the wild type (14). All of the reported mutations are located in or close to the putative fusion domain (residues 118 to 139). However, these viruses were not tested for stability and for the ability to replicate at acidic pH values. Under normal pH conditions, these viruses were attenuated in cell culture (14). VSIV-6.6 and, especially, VSIV-6.4 were also attenuated in cell culture compared to VSIV-GI in an experiment using normal medium (pH above 7.4), but VSIV-6.8 and VSIV-GNJ were not, despite also being more sensitive to inhibition by lysosomotropic agents. The case of VSIV-GNJ is particularly interesting, since this virus is more pathogenic than VSIV-GI in a natural host, swine (26).
In summary, we found that the ability of VSV to grow at mildly acidic pHs was a function of the glycoprotein, as demonstrated by the replacement of the GI gene with the GNJ gene. In addition, using selective pressure for growth at successively lower pHs, we identified amino acid substitutions in VSIV variant viruses which enhanced the ability of the viruses to replicate at acidic pHs. The importance of these substitutions for the observed phenotypes was confirmed by an analysis of the effects of specific changes engineered into the wild-type VSIV genome. The results presented here may be relevant to understanding the biology of VSV in nature. We are currently exploring whether the more severe pathogenicity of VSIV-GNJ than that of VSIV-GI in the natural host may be related to the differential sensitivities of the GI and GNJ proteins to pH variations.
ACKNOWLEDGMENTS
We thank the members of the G. W. Wertz and L. A. Ball laboratories for helpful discussions during preparation of the manuscript.
This work was supported by NIH grant R37AI12464 to G.W.W. I. Martínez was a recipient of a fellowship from the "Ministerio de Sanidad y Consumo" (Spain) (BAE 01/5006).
REFERENCES
Ball, L. A., C. R. Pringle, B. Flanagan, V. P. Perepelitsa, and G. W. Wertz. 1999. Phenotypic consequences of rearranging the P, M, and G genes of vesicular stomatitis virus. J. Virol. 73:4705-4712.
Bridges, V. E., B. J. McCluskey, M. D. Salman, H. S. Hurd, and J. Dick. 1997. Review of the 1995 vesicular stomatitis outbreak in the western United States. J. Am. Vet. Med. Assoc. 211:556-560.
Carneiro, F. A., M. L. Bianconi, G. Weissmuller, F. Stauffer, and A. T. Da Poian. 2002. Membrane recognition by vesicular stomatitis virus involves enthalpy-driven protein-lipid interactions. J. Virol. 76:3756-3764.
Carneiro, F. A., A. S. Ferradosa, and A. T. Da Poian. 2001. Low pH-induced conformational changes in vesicular stomatitis virus glycoprotein involve dramatic structure reorganization. J. Biol. Chem. 276:62-67.
Carneiro, F. A., F. Stauffer, C. S. Lima, M. A. Juliano, L. Juliano, and A. T. Da Poian. 2003. Membrane fusion induced by vesicular stomatitis virus depends on histidine protonation. J. Biol. Chem. 278:13789-13794.
Cartwright, B., and F. Brown. 1972. Serological relationships between different strains of vesicular stomatitis virus. J. Gen. Virol. 16:391-398.
Chen, S. S. 1991. Myristylation of the envelope glycoprotein of vesicular stomatitis virus. Intervirology 32:193-197.
Clague, M. J., C. Schoch, L. Zech, and R. Blumenthal. 1990. Gating kinetics of pH-activated membrane fusion of vesicular stomatitis virus with cells: stopped-flow measurements by dequenching of octadecylrhodamine fluorescence. Biochemistry 29:1303-1308.
Crimmins, D. L., W. B. Mehard, and S. Schlesinger. 1983. Physical properties of a soluble form of the glycoprotein of vesicular stomatitis virus at neutral and acidic pH. Biochemistry 22:5790-5796.
Daniels, R. S., J. C. Downie, A. J. Hay, M. Knossow, J. J. Skehel, M. L. Wang, and D. C. Wiley. 1985. Fusion mutants of the influenza virus hemagglutinin glycoprotein. Cell 40:431-439.
Doms, R. W., M. J. Gething, J. Henneberry, J. White, and A. Helenius. 1986. Variant influenza virus hemagglutinin that induces fusion at elevated pH. J. Virol. 57:603-613.
Doms, R. W., D. S. Keller, A. Helenius, and W. E. Balch. 1987. Role for adenosine triphosphate in regulating the assembly and transport of vesicular stomatitis virus G protein trimers. J. Cell. Biol. 105:1957-1969.
Durrer, P., Y. Gaudin, R. W. Ruigrok, R. Graf, and J. Brunner. 1995. Photolabeling identifies a putative fusion domain in the envelope glycoprotein of rabies and vesicular stomatitis viruses. J. Biol. Chem. 270:17575-17581.
Fredericksen, B. L., and M. A. Whitt. 1998. Attenuation of recombinant vesicular stomatitis viruses encoding mutant glycoproteins demonstrate a critical role for maintaining a high pH threshold for membrane fusion in viral fitness. Virology 240:349-358.
Fredericksen, B. L., and M. A. Whitt. 1996. Mutations at two conserved acidic amino acids in the glycoprotein of vesicular stomatitis virus affect pH-dependent conformational changes and reduce the pH threshold for membrane fusion. Virology 217:49-57.
Fredericksen, B. L., and M. A. Whitt. 1995. Vesicular stomatitis virus glycoprotein mutations that affect membrane fusion activity and abolish virus infectivity. J. Virol. 69:1435-1443.
Gallione, C., and J. K. Rose. 1983. Nucleotide sequence of a cDNA clone encoding the entire glycoprotein from the New Jersey serotype of vesicular stomatitis virus. J. Virol. 46:162-169.
Gaudin, Y., C. Tuffereau, P. Durrer, A. Flamand, and R. W. Ruigrok. 1995. Biological function of the low-pH, fusion-inactive conformation of rabies virus glycoprotein (G): G is transported in a fusion-inactive state-like conformation. J. Virol. 69:5528-5534.
Jeetendra, E., K. Ghosh, D. Odell, J. Li, H. P. Ghosh, and M. A. Whitt. 2003. The membrane-proximal region of vesicular stomatitis virus glycoprotein G ectodomain is critical for fusion and virus infectivity. J. Virol. 77:12807-12818.
Jeetendra, E., C. S. Robison, L. M. Albritton, and M. A. Whitt. 2002. The membrane-proximal domain of vesicular stomatitis virus G protein functions as a membrane fusion potentiator and can induce hemifusion. J. Virol. 76:12300-12311.
Keil, W., and R. R. Wagner. 1989. Epitope mapping by deletion mutants and chimeras of two vesicular stomatitis virus glycoprotein genes expressed by a vaccinia virus vector. Virology 170:392-407.
Kelley, J. M., S. U. Emerson, and R. R. Wagner. 1972. The glycoprotein of vesicular stomatitis virus is the antigen that gives rise to and reacts with neutralizing antibody. J. Virol. 10:1231-1235.
Letchworth, G. J., L. L. Rodriguez, and J. Del Barrera. 1999. Vesicular stomatitis. Vet. J. 157:239-260.
Li, Y., C. Drone, E. Sat, and H. P. Ghosh. 1993. Mutational analysis of the vesicular stomatitis virus glycoprotein G for membrane fusion domains. J. Virol. 67:4070-4077.
Machamer, C. E., R. W. Doms, D. G. Bole, A. Helenius, and J. K. Rose. 1990. Heavy chain binding protein recognizes incompletely disulfide-bonded forms of vesicular stomatitis virus G protein. J. Biol. Chem. 265:6879-6883.
Martinez, I., L. L. Rodriguez, C. Jimenez, S. J. Pauszek, and G. W. Wertz. 2003. Vesicular stomatitis virus glycoprotein is a determinant of pathogenesis in swine, a natural host. J. Virol. 77:8039-8047.
Mastromarino, P., C. Conti, P. Goldoni, B. Hauttecoeur, and N. Orsi. 1987. Characterization of membrane components of the erythrocyte involved in vesicular stomatitis virus attachment and fusion at acidic pH. J. Gen. Virol. 68:2359-2369.
Mathieu, M. E., P. R. Grigera, A. Helenius, and R. R. Wagner. 1996. Folding, unfolding, and refolding of the vesicular stomatitis virus glycoprotein. Biochemistry 35:4084-4093.
Matlin, K. S., H. Reggio, A. Helenius, and K. Simons. 1982. Pathway of vesicular stomatitis virus entry leading to infection. J. Mol. Biol. 156:609-631.
McCluskey, B. J., H. S. Hurd, and E. L. Mumford. 1999. Review of the 1997 outbreak of vesicular stomatitis in the western United States. J. Am. Vet. Med. Assoc. 215:1259-1262.
Oomens, A. G., A. G. Megaw, and G. W. Wertz. 2003. Infectivity of a human respiratory syncytial virus lacking the SH, G, and F proteins is efficiently mediated by the vesicular stomatitis virus G protein. J. Virol. 77:3785-3798.
Pak, C. C., A. Puri, and R. Blumenthal. 1997. Conformational changes and fusion activity of vesicular stomatitis virus glycoprotein: [125I]iodonaphthyl azide photolabeling studies in biological membranes. Biochemistry 36:8890-8896.
Puri, A., S. Grimaldi, and R. Blumenthal. 1992. Role of viral envelope sialic acid in membrane fusion mediated by the vesicular stomatitis virus envelope glycoprotein. Biochemistry 31:10108-10113.
Puri, A., J. Winick, R. J. Lowy, D. Covell, O. Eidelman, A. Walter, and R. Blumenthal. 1988. Activation of vesicular stomatitis virus fusion with cells by pretreatment at low pH. J. Biol. Chem. 263:4749-4753.
Robison, C. S., and M. A. Whitt. 2000. The membrane-proximal stem region of vesicular stomatitis virus G protein confers efficient virus assembly. J. Virol. 74:2239-2246.
Rodriguez, L. L. 2002. Emergence and re-emergence of vesicular stomatitis in the United States. Virus Res. 85:211-219.
Rodriguez, L. L., T. A. Bunch, M. Fraire, and Z. N. Llewellyn. 2000. Re-emergence of vesicular stomatitis in the western United States is associated with distinct viral genetic lineages. Virology 271:171-181.
Rodriguez, L. L., and S. T. Nichol. 1999. Vesicular stomatitis virus, p. 1910-1919. In R. G. Webster and A. Granoff (ed.), Encyclopedia of virology. Academic Press, London, United Kingdom.
Rose, J. K., and C. J. Gallione. 1981. Nucleotide sequences of the mRNAs encoding the vesicular stomatitis virus G and M proteins determined from cDNA clones containing the complete coding regions. J. Virol. 39:519-528.
Schlegel, R., T. S. Tralka, M. C. Willingham, and I. Pastan. 1983. Inhibition of VSV binding and infectivity by phosphatidylserine: is phosphatidylserine a VSV-binding site? Cell 32:639-646.
Schlegel, R., and M. Wade. 1985. Biologically active peptides of the vesicular stomatitis virus glycoprotein. J. Virol. 53:319-323.
Shokralla, S., R. Chernish, and H. P. Ghosh. 1999. Effects of double-site mutations of vesicular stomatitis virus glycoprotein G on membrane fusion activity. Virology 256:119-129.
Shokralla, S., Y. He, E. Wanas, and H. P. Ghosh. 1998. Mutations in a carboxy-terminal region of vesicular stomatitis virus glycoprotein G that affect membrane fusion activity. Virology 242:39-50.
Stallknecht, D. E., D. E. Perzak, L. D. Bauer, M. D. Murphy, and E. W. Howerth. 2001. Contact transmission of vesicular stomatitis virus New Jersey in pigs. Am. J. Vet. Res. 62:516-520.
Whelan, S. P., L. A. Ball, J. N. Barr, and G. T. Wertz. 1995. Efficient recovery of infectious vesicular stomatitis virus entirely from cDNA clones. Proc. Natl. Acad. Sci. USA 92:8388-8392.
White, J., K. Matlin, and A. Helenius. 1981. Cell fusion by Semliki Forest, influenza, and vesicular stomatitis viruses. J. Cell Biol. 89:674-679.
Woodgett, C., and J. K. Rose. 1986. Amino-terminal mutation of the vesicular stomatitis virus glycoprotein does not affect its fusion activity. J. Virol. 59:486-489.
Zhang, L., and H. P. Ghosh. 1994. Characterization of the putative fusogenic domain in vesicular stomatitis virus glycoprotein G. J. Virol. 68:2186-2193.(Isidoro Martinez and Gail)
ABSTRACT
We previously generated recombinant vesicular stomatitis viruses (VSV) based on the Indiana serotype genome which contained either the homologous glycoprotein gene from the Indiana serotype (VSIV-GI) or the heterologous glycoprotein gene from the New Jersey serotype (VSIV-GNJ). The virus expressing the GNJ gene was more pathogenic than the parental VSIV-GI virus in swine, a natural host (26). For the present study, we investigated the biological differences between the GI and GNJ proteins that may be related to the differences in pathogenesis between VSIV-GI and VSIV-GNJ. We show that the capacities of viruses with either the GNJ or GI glycoprotein to infect cultured cells differ depending on the pH. VSIV-GNJ could infect cells at acidic pHs, while the infectivity of VSIV-GI was severely reduced. VSIV-GNJ infection was also more sensitive to inhibition by ammonium chloride, indicating that the GNJ protein had a lower pH threshold for membrane fusion. We applied selective pressure to VSIV-GI by growing it at successively lower pH values and isolated variant viruses in which we identified amino acid changes that conferred low-pH-resistant infectivity. Repeated passage in cell culture at pH 6.8 resulted in the selection of a VSIV-GI variant (VSIV-6.8) that was similar to VSIV-GNJ regarding its pH- and ammonium chloride-dependent infectivity. Sequence analysis of VSIV-6.8 revealed that it had a single amino acid substitution in the amino-terminal region of the glycoprotein (F18L). This alteration was shown to be responsible for the observed phenotype by site-directed mutagenesis of a VSIV-GI full-length cDNA and analysis of the recovered engineered virus. A further adaptation of VSIV-6.8 to pHs 6.6 and 6.4 resulted in additional amino acid substitutions in areas of the glycoprotein that were not previously implicated in attachment or fusion.
INTRODUCTION
Isolates of Vesicular stomatitis virus (VSV) are enveloped, nonsegmented, negative-strand RNA viruses that belong to the genus Vesiculovirus in the Rhabdoviridae family. VSV infects domestic animals such as cattle, swine, and horses, causing vesicular lesions in the tongue, oral tissues, udders, and hooves (23, 38). Important economic losses in milk and meat production are associated with VSV outbreaks, not only due to the virus infection itself but also due to quarantines, trade barriers, and other governmental measures aimed to control virus spread (30). These measures are particularly pertinent because the lesions induced by VSV are similar to those of foot-and-mouth disease, one of the most contagious animal diseases.
The VSV genome is delivered to the cytoplasm of host cells, where replication occurs, via receptor-mediated endocytosis of viral particles and subsequent pH-induced fusion of the viral envelope with the endosomal membrane (29). The VSV G protein, the sole viral surface glycoprotein, is required for attachment and fusion. Attachment seems to occur via electrostatic interactions between positive charges on the glycoprotein surface and negatively charged phospholipids in the plasma membrane (3, 27, 40), and fusion involves structural changes in the protein driven by the low pH of the endosomal compartment (4, 29, 46). The VSV G protein is distinct in several ways from other viral fusion proteins. For example, conformational changes in the VSV G protein induced by the pH are reversible, and VSV G can exist in different conformational forms (8, 9, 32, 34). Furthermore, VSV G does not contain an obvious hydrophobic sequence which would be indicative of a fusion peptide. Biochemical and mutational studies have indicated that VSV G has an internal fusion peptide between residues 118 and 139, but mutations at residues 395 to 418 also affect membrane fusion activity (13, 15, 16, 24, 42, 43, 48). Additionally, His-148 and His-149 have been suggested to directly participate in VSV fusion (5). Recently, the membrane-proximal domain encompassing residues 449 to 462 has been shown to contribute to G protein-mediated membrane fusion (19, 20).
There are two major serotypes of VSV, Indiana (VSIV) and New Jersey, which are distinguished by neutralizing antibodies against the G protein (6, 22). In addition to their antigenic structures (21), the Indiana (GI) and New Jersey (GNJ) glycoproteins also differ in the number (511 and 517, respectively) and composition of amino acids (only 50% identity) (17, 39), in posttranslational modifications (7), and in folding (25, 28). Correspondingly, Indiana and New Jersey strains are not equally important regarding VSV pathogenesis. Outbreaks caused by New Jersey strains are more frequent and more severe than those caused by Indiana strains (2, 36, 44). We recently reported that this can be related, at least in part, to the G glycoprotein, as the replacement of the GI glycoprotein with the GNJ protein in the background of a recombinant VSIV substantially increased the pathogenicity of the virus in swine (26). However, the basis for these differences is not known. For this study, we investigated the biological differences between the GI and GNJ proteins that may help to explain their role in VSV pathogenesis. The VSV GI glycoprotein structure has been shown to be very sensitive to even mildly acidic pHs, at which the exposure of hydrophobic domains has been reported to occur (4). However, the relevance of this observation for VSV biology, if any, remains to be addressed. Another question that has not been clarified is whether or not the GI and GNJ glycoproteins are equally sensitive to pH changes. For this study, we examined the effect of low pHs on the infectivity of VSIV-GI and VSIV-GNJ in cell culture. We demonstrated that VSIV-GNJ was able to replicate at acidic pH values at which VSIV-GI replication was almost completely inhibited. In addition, by using the selective pressure of growth at increasingly acidic pHs, we identified amino acid changes in the GI glycoprotein that increased the ability of VSIV-GI to infect cells at low pHs.
MATERIALS AND METHODS
Virus and cells. VSIV-GI and VSIV-GNJ were recovered from cDNA clones and have been described previously (26, 45). The San Juan isolate of the Indiana serotype of VSV provided the template for all genes except for the G protein gene, which was derived from either the Indiana strain Orsay (VSIV-GI) or the New Jersey strain 95COB (VSIV-GNJ) (37). Baby hamster kidney (BHK-21) cells were used to recover viruses from cDNAs and for the selection of VSIV-GI variant viruses by passage at acidic pHs. Vero-76 cells were used for growth experiments involving acidic pHs or ammonium chloride and for plaque assays.
Virus replication at acidic pHs. Vero-76 cells growing in six-well plates were washed with Hanks' balanced salt solution (HBSS) and infected at a multiplicity of infection (MOI) of 0.01 PFU per cell in minimum essential medium without sodium bicarbonate containing 2% fetal calf serum (MEM2) adjusted to pH 7.2, 7.0, 6.8, 6.6, or 6.4 with 50 mM HEPES. Viruses were adsorbed for 1 h at 37°C at atmospheric concentrations of CO2 to avoid acidification of the medium. After the adsorption period, the inoculum was removed and the monolayers were washed with HBSS. Fresh MEM2 at the desired pH was added, and the cells were incubated for an additional 17 to 22 h. Supernatant fluids were harvested and clarified by low-speed centrifugation, and viral yields were determined by plaque assays on Vero-76 cells. In some cases (indicated in the figure legends), viruses were preincubated for up to 60 min at 37°C in the corresponding medium prior to the inoculation of cells.
Selection of VSIV-GI mutant viruses by passage at low pHs. (i) Virus selection at pH 6.8 (VSIV-6.8). BHK-21 cells growing in 60-mm-diameter dishes were infected at an MOI of 0.01 in MEM2, pH 6.8. Viruses were adsorbed for 1 h at 37°C at the atmospheric CO2 concentration. After the adsorption period, fresh MEM2, pH 6.8, was added, and the cells were incubated for an additional 48 h. The supernatant fluid was harvested and clarified by low-speed centrifugation. The viral yield was determined by a plaque assay on Vero-76 cells, and 0.01 PFU/cell was used to infect new cells. After five passages, the virus was amplified by the infection of BHK-21 cells at a low MOI in MEM2, pH 6.8, pelleted by centrifugation at 21,000 rpm for 90 min in a Ty30 rotor, and purified by centrifugation in a 15 to 45% (wt/vol) sucrose gradient. Finally, the virus was concentrated by centrifugation and resuspended in Dulbecco's modified Eagle medium containing 2% fetal calf serum (DMEM2).
(ii) Virus selection at pH 6.6 (VSIV-6.6). The supernatant from VSIV-6.8 passage three was used to infect BHK-21 cells at an MOI of 0.01 in MEM2, pH 6.6, and the virus was passaged, amplified, and purified as described above for VSIV-6.8, except that the pH during the passages and amplification was maintained at 6.6 instead of 6.8.
(iii) Virus selection at pH 6.4 (VSIV-6.4). The supernatant from VSIV-6.6 passage three was used to infect BHK-21 cells at an MOI of 0.01 in MEM2, pH 6.4, and the virus was passaged, amplified, and purified as described above for VSIV-6.8, except that the pH during the passages and amplification was maintained at 6.4 instead of 6.8. RNAs purified from BHK-21 cells infected with the variant viruses were amplified by reverse transcription-PCR (RT-PCR) with primers specific for the matrix (M) and polymerase (L) genes by use of a OneStep RT-PCR kit (QIAGEN), and the nucleotide sequences of the G genes were determined.
Recovery of recombinant viruses from full-length cDNA clones containing introduced mutations in the glycoprotein gene to confirm phenotype. The full-length cDNA clone from which VSIV-GI was recovered (see above) was digested with the restriction enzyme BstBI, and the fragment containing the GI protein gene was subcloned into the plasmid pBluescript-GUS (31) by conventional cloning techniques. The GI gene was mutagenized by the use of a QuikChange site-directed mutagenesis kit (Stratagene) and specific mutagenic primers. The mutagenized BstBI fragments were sequenced to confirm that they contained the desired nucleotide changes and then used to replace the original fragment in the VSIV-GI full-length cDNA clone. Viruses were recovered by transfection of BHK-21 cells as previously described (1, 45), and the RNAs from BHK-21 cells infected with the recovered viruses were amplified and the G gene was sequenced to confirm that the introduced nucleotide changes were maintained during the recovery process.
Virus replication in the presence of ammonium chloride. Confluent Vero-76 cells in six-well plates were incubated in MEM2, pH 7.2, containing 0, 1, 2, 4, 8, or 16 mM ammonium chloride for 30 min at 37°C. The medium was removed, and the cells were infected at an MOI of 0.01 in the same medium. After 1 h of adsorption at 37°C, fresh medium containing ammonium chloride was added, and the cells were incubated for an additional 15 h. Supernatant fluids were collected and clarified by low-speed centrifugation, and viral yields were determined by plaque assays on Vero-76 cells.
RESULTS
VSIV-GI and VSIV-GNJ differ in their capacities to grow at acidic pHs. To test whether the serotype of the glycoprotein expressed by VSV influences virus infectivity at mildly acidic pHs, we examined the capacities of two recombinant viruses that differed only in their glycoprotein genes, VSIV-GI and VSIV-GNJ, to replicate in cell culture at acidic pH values. Vero-76 cells were infected at an MOI of 0.01 with VSIV-GI or VSIVGNJ in HEPES-buffered MEM2 at pH values that decreased in steps of 0.2 pH units from 7.2 to 6.4 and were then incubated at 37°C. At 17 h postadsorption, supernatant fluids were collected and clarified by low-speed centrifugation, and the virus yields were titrated by plaque assays on Vero-76 cells. Both virus infections yielded similar amounts of infectious particles at pH 7.2, but the titer of VSIV-GI started to decline when replication was carried out at pH 7.0 and dropped almost 30-fold when replication was done at pH 6.8, whereas at the same pH values VSIV-GNJ replication levels remained essentially unchanged (Fig. 1). At pHs 6.6 and 6.4, the amounts of infectious particles of both viruses in supernatants were severely reduced (Fig. 1). The virus titers in the supernatants correlated with the cytopathic effect observed (data not shown). At pHs 7.2 and 7.0, the cell monolayers infected with VSIV-GI were completely destroyed, while at pH 6.8 and below, cytopathic effects were not obvious, except for a few patches of infection at pH 6.8. However, the cell monolayer was completely destroyed when the cells were infected with VSIV-GNJ at pH 6.8.
pH-induced reversible changes in virions, but not in cells, are responsible for the decreased ability of VSIV-GI to infect at acidic pHs. To investigate how the VSIV-GI and VSIV-GNJ virus infectivities were affected by incubation at a low pH (6.8), we diluted the viruses in medium at pH 7.2 or pH 6.8 and incubated them for 0 to 60 min at 37°C prior to infection. Vero-76 cells were then infected at an MOI of 0.01 with these viruses at the same pH used for incubation. At 22 h postadsorption, supernatant samples were harvested and clarified by centrifugation at low speed, and virus titers were determined by plaque assays on Vero-76 cells (Fig. 2). At pH 7.2, VSIV-GI and VSIV-GNJ virus yields were similar, regardless of the preincubation time at 37°C (Fig. 2A). However, at pH 6.8, VSIV-GNJ virus yields were unaffected by the previous incubation at 37°C, whereas VSIV-GI virus yields decreased rapidly with increased times of incubation and dropped more than 100-fold when the virus was preincubated at pH 6.8 for 1 h at 37°C (Fig. 2B). These results show that at pH 6.8, VSIV-GI infectivity was decreased by changes affecting the virions. However, it does not exclude the possibility that changes in the cells could also influence virus infectivity. To test this, we incubated the cells for 1 h at pH 7.2 or 6.8 prior to infection with VSIV-GI and also preincubated the virus for 0 or 60 min in MEM2, pH 6.8, at 37°C. After 1 h of adsorption at 37°C, fresh MEM2, pH 6.8, was added, and the cells were incubated at 37°C for an additional 22 h. Supernatant fluids were taken and assayed for virus yield by a plaque assay. Infectivity was diminished when the virus, but not the cells, was preincubated at pH 6.8 (Fig. 2C). This experiment showed that the exposure of cells to pH 6.8 did not measurably influence VSIV-GI infectivity.
Using a soluble form of VSIV glycoprotein, Crimmins et al. (9) showed that the conformational change detected when the protein was exposed to a pH close to 5 was reversible. We analyzed whether the loss of VSIV-GI infectivity after incubation at pH 6.8 was recovered by a subsequent incubation at pH 7.2. Vero-76 cells were infected at an MOI of 0.1 with VSIV-GI that had been previously incubated in MEM2 at 37°C as follows: (i) 2 h at pH 7.2, (ii) 1 h at pH 7.2 and 1 h at pH 6.8, (iii) 1 h at pH 6.8 and 1 h at pH 7.2, or (iv) 2 h at pH 6.8. After 1 h of adsorption at 37°C, the inoculum was removed, the cells were washed two times with HBSS, and fresh MEM2 at pH 6.8 was added to prevent additional rounds of infection. At 21 h postadsorption, supernatants were collected and clarified by low-speed centrifugation, and the virus yield was determined by a standard plaque assay on Vero-76 cells. The VSIV-GI infectivity dropped >100-fold when the virus was incubated for 2 h at pH 6.8 or incubated for 1 h at pH 7.2 and 1 h at pH 6.8, but the infectivity was recovered when the virus that had previously been incubated at pH 6.8 was subsequently incubated at pH 7.2 (Fig. 2D). These and the above results demonstrate that the changes in the VSIV-GI virions responsible for the loss of infectivity at pH 6.8 were reversible.
The ability of VSIV-GI to infect at acidic pHs can be increased by a single amino acid substitution in the amino-terminal region of the glycoprotein. Since VSIV-GI and VSIV-GNJ have the same VSIV genomic backbone, which differs only in the glycoprotein expressed, the differences observed between the two viruses regarding pH-dependent infectivity should reside in this protein. To identify the amino acid(s) or region(s) of the glycoprotein that modulates the ability of VSV to infect at acidic pHs, we utilized selective pressure by growing the virus at successively decreasing pH values. VSIV-GI was passaged five times at a low MOI (0.01) in BHK cells at pH 6.8 as described in Materials and Methods. The virus from passage five was amplified, purified through a sucrose gradient, and resuspended in DMEM2. All viral stocks were made in the same cell line and in the same medium. After five successive passages at pH 6.8, the recovered virus (VSIV-6.8) showed a pH-dependent infectivity similar to that of VSIV-GNJ (Fig. 3A). Sequence analysis of RNAs from VSIV-6.8 revealed a single nucleotide substitution in the glycoprotein gene that resulted in an amino acid change at position 18 from phenylalanine to leucine (F18L). To confirm that this change was responsible for the VSIV-6.8 phenotype, we introduced this single nucleotide mutation by site-directed mutagenesis into the full-length cDNA clone from which VSIV-GI was derived and recovered the recombinant virus (rVSIV-6.8) by transfection of BHK-21 cells. While the VSIV-GI virus yield was decreased >1,000-fold at pH 6.8 compared with the virus yield at pH 7.2, the VSIV-6.8 and rVSIV-6.8 virus yields were essentially the same and remained high at pH 6.8 (Fig. 3B). This result demonstrated that the amino acid substitution F18L increased the ability of VSIV-GI to grow at mildly acidic pHs.
Further selection of VSIV-GI variant viruses at pHs 6.6 and 6.4 leads to additional amino acid substitutions in different areas of the glycoprotein. To explore whether further adaptation to pHs 6.6 and 6.4 would select additional mutations in the glycoprotein, we used the VSIV-6.8 virus from passage three to infect BHK-21 cells at pH 6.6. After five passages at pH 6.6, the virus was amplified, purified by sucrose gradient centrifugation, and resuspended in DMEM2. Passage three of this virus (VSIV-6.6) was used to select a virus variant at pH 6.4 (VSIV-6.4). We used passage three in both cases to generate the following pH-adapted viruses because at this passage there was an evident increase in the in vitro cytopathic effect of the virus, indicating that a variant virus was already present in the viral population. This stepwise selection facilitated the process of selection, as direct selection at pH 6.6 or 6.4 from the original stock yielded little or no virus. Sequence analysis of the VSIV-6.6 viral RNA by RT-PCR showed that in addition to the F18L mutation found in the VSIV-6.8 virus, VSIV-6.6 had two nucleotide substitutions that led to the amino acid changes Q301R and K462R in the G protein (Fig. 4C). The VSIV-6.4 glycoprotein had the three amino acid substitutions found in VSIV-6.6 plus an additional nucleotide change that led to the amino acid replacement H65R (Fig. 4C). Viruses VSIV-6.6 and VSIV-6.4 both grew well at pH 6.6, and although the viral titers decreased at pH 6.4, they were >10 times higher than those of VSIV-6.8 (Fig. 4A). The three mutations found in VSIV-6.6 were introduced into the VSIV-GI full-length cDNA, and the recombinant virus was recovered by transfection of BHK cells (rVSIV-6.6). To analyze the individual contributions of the Q301R and K462R mutations to the ability of VSIV-6.6 to grow at pH 6.6, we introduced these mutations separately along with the F18L mutation into the VSIV-GI cDNA clone by QuikChange mutagenesis. The corresponding viruses were recovered after transfection of the cDNAs into BHK-21 cells. The G sequences of all recombinant viruses were confirmed by nucleotide sequencing of RT-PCR products of viral RNAs from infected cells and were found to contain only the introduced mutations. rVSIV-6.6 and the virus containing the Q301R mutation grew as well as VSIV-6.6 at pH 6.6, but in contrast, the titer of the virus containing the K462R mutation was almost 100-fold lower at that pH (Fig. 4B). These results demonstrate that the Q301R substitution is a major contributor to the ability of VSIV-6.6 to grow at low pHs. However, the K462R change also contributed, since the virus containing this change yielded higher titers than VSIV-6.8 at pH 6.6 (Fig. 4B).
Effect of ammonium chloride on virus infectivity. VSV enters the host cell via receptor-mediated endocytosis, and the nucleocapsid is liberated in the cytoplasm by fusion of the viral envelope and the endosomal membrane. This process is mediated by conformational changes in the G glycoprotein at acidic pHs and can be blocked with compounds, such as ammonium chloride, that prevent endosomal acidification (29). We analyzed whether GI and GNJ behaved distinctly in an assay of virus fusion inhibition by ammonium chloride. The infectivities of VSIV-GI and VSIV-GNJ were analyzed in the presence of various amounts of ammonium chloride ranging from 1 to 16 mM. Samples from the supernatant fluids were taken at 15 h postadsorption, and viral titers were determined by plaque assays. A progressive decrease in the virus titer was observed with increasing concentrations of ammonium chloride for both viruses (Fig. 5). However, VSIV-GNJ infectivity declined faster than that of VSIV-GI, and at 8 mM ammonium chloride, the difference was almost 100-fold (Fig. 5). These data suggest that VSIV-GNJ requires a more acidic pH to fuse with the endosomal membrane than does VSIV-GI, and these findings appear to correlate with the results from the experiments showing that VSIV-GNJ remained infectious at lower pHs than did VSIV-GI. There was also a correlation between the pHs at which VSIV-GI variant viruses were selected and their sensitivities to ammonium chloride, i.e., the lower the pH at which the virus was selected, the lower the concentration of chemical that was needed to inhibit viral infection (Fig. 5). In conclusion, the experiments with ammonium chloride are in agreement with the experiments involving VSV infectivity at various pHs and emphasize the different sensitivities of the GI and GNJ glycoproteins to acidic pHs, a phenomenon that may have important consequences for VSV biology in vivo.
DISCUSSION
We previously reported that the replacement of the glycoprotein gene of a recombinant vesicular stomatitis Indiana virus (VSIV-GI) with the glycoprotein gene of the New Jersey serotype renders the virus (VSIV-GNJ) more pathogenic to swine, a natural host (26). This correlates with field observations showing that New Jersey strains are more pathogenic than Indiana strains (2, 36, 44). In the present study, we have initiated a characterization of the biological differences between the GI and GNJ glycoproteins that may help us to understand the above results. In particular, we have analyzed the effect of mildly acidic pHs on the in vitro infectivities of VSIV-GI and VSIV-GNJ. We found that VSIV-GI infectivity decreased rapidly when the infection of cells was carried out at pH values below 7.2, while VSIV-GNJ still produced high virus titers at pH 6.8. VSIV-GI, but not VSIV-GNJ, quickly lost its infectivity after incubation at pH 6.8 prior to cell infection. However, VSIV-GI infectivity could be recovered by reincubation at pH 7.2. These results indicate that a reversible conformational change took place in the GI glycoprotein that made the VSIV-GI virus less able to infect cells at pH 6.8. Thus, this work shows that the GI and GNJ glycoproteins are not equally sensitive to mildly acidic pH values. These findings raise the possibility that these differences may be relevant to the epidemiology and/or pathogenesis of the virus. For example, our results suggest that the Indiana strains are more readily inactivated by environmental pH variations than the New Jersey strains during natural VSV infection and/or transmission.
There are several steps at which exposure to low pHs could block VSV replication, including attachment, internalization, fusion, budding, and release from infected tissues. Differences in pH sensitivity also may be translated into differences in oligomerization (12), maturation (25, 28), and/or transport to the cell surface (18). An attractive explanation for the different sensitivities of GI and GNJ to low pHs can be discussed in the context of the previously published "three-state model" that relates pH-dependent conformational transitions of the GI protein to mechanisms of viral fusion (8, 32, 33). At a neutral pH, the GI protein is in a native (tense) state. At slightly acidic pHs, the GI protein undergoes a proton-driven shift either to an activated (relaxed) state, which is fusion active, or to an inactivated (desensitized) state, which is fusion inactive. The active state is achieved when the conformational changes result in the movement of the fusion active region of the protein into the target membrane. The desensitized state occurs when this region fails to insert into the membrane. These changes are reversible (9), so there is a pH-dependent equilibrium between the different conformational states of the glycoprotein (8, 32). We can interpret our results in light of this model as follows: at mildly acidic pH values, the GI conformation shifts to the desensitized state and the protein is thus not able to initiate fusion and to infect, while GNJ is still in the native state and is fusion competent. In this case, GNJ would have a more stable structure, such that more acidic conditions are required to trigger the conformational changes leading to desensitization. The ability of VSIV-GNJ to grow at low pH values correlated with its increased sensitivity to ammonium chloride, which also indicated a requirement for more acidic conditions for fusion and endosomal release for this virus (14). According to this observation, the pH values governing not only desensitization, but also the other conformational states of GNJ, would be shifted to more acidic values compared to those for GI.
To examine whether GI could be modified to respond to mildly acidic pHs in a similar fashion as GNJ, we exerted selective pressure on VSIV-GI viruses by passage at increasingly lower pH values, from 6.8 to 6.4. The recovered, passaged viruses were able to infect cells at progressively lower pHs and were also increasingly more sensitive to ammonium chloride. Although a shift of the pH optimum of fusion has been used as a criterion to characterize the fusion peptides of several viral proteins, including GI (16, 48), the mutations observed in the low-pH-adapted VSIV-GI viruses described here did not map to any region of the GI glycoprotein reported to be directly involved in fusion (see the introduction). The only amino acid substitutions that were in regions that are known to be functionally important for VSV biology were F18L and K462R. The K462R mutation is localized in a protein segment involved in budding and fusion potentiation (20, 35), but we have shown that this amino acid change is less important than Q301R for the ability of VSIV-6.4 to infect at low pHs. Peptides corresponding to the amino-terminal region of GI, where the F18L mutation was mapped, have been shown to be pH-dependent hemolysins and hemagglutinins (41), although mutations in this region that abolished the hemolytic activity of the peptide did not affect the pH-dependent fusion activity of the glycoprotein (47).
Except for F18L, the three other amino acid changes found in the low-pH-adapted viruses were changes to arginine, suggesting that positive charges are important for modulating pH-induced conformational changes in GI. This is consistent with the observation that the K462R mutation had less impact on the phenotype of selected viruses than did H65R or Q301R. We do not know the mechanism(s) by which all of the observed amino acid changes influence GI sensitivity to pH, but the results described here resemble those reported previously for variant influenza virus hemagglutinins selected for their ability to grow in cells treated with amantadine hydrochloride, which raises the endosomal pH (10). Mutant viruses that fuse at elevated pH values contain several changes, most of them involving charged amino acids localized in different regions of the hemagglutinin primary sequence that appear to alter intra- and intersubunit contacts, destabilizing part of the protein structure. This would lower the energy required to trigger the conformational change (10, 11). Our case would be the opposite, as the amino acid substitutions in GI would be predicted to lead to a more stable structure so that a lower pH is required to destabilize it. It is worth mentioning that arginine has a side chain with three nitrogen atoms, which confer a high potential to form hydrogen bonds that would increase the stability of the protein.
Mutant GI proteins showing a reduced pH threshold for fusion have been described previously (15, 43, 48), and recombinant VSIV viruses encoding some of these mutated glycoproteins have been recovered that are more sensitive to chloroquine than the wild type (14). All of the reported mutations are located in or close to the putative fusion domain (residues 118 to 139). However, these viruses were not tested for stability and for the ability to replicate at acidic pH values. Under normal pH conditions, these viruses were attenuated in cell culture (14). VSIV-6.6 and, especially, VSIV-6.4 were also attenuated in cell culture compared to VSIV-GI in an experiment using normal medium (pH above 7.4), but VSIV-6.8 and VSIV-GNJ were not, despite also being more sensitive to inhibition by lysosomotropic agents. The case of VSIV-GNJ is particularly interesting, since this virus is more pathogenic than VSIV-GI in a natural host, swine (26).
In summary, we found that the ability of VSV to grow at mildly acidic pHs was a function of the glycoprotein, as demonstrated by the replacement of the GI gene with the GNJ gene. In addition, using selective pressure for growth at successively lower pHs, we identified amino acid substitutions in VSIV variant viruses which enhanced the ability of the viruses to replicate at acidic pHs. The importance of these substitutions for the observed phenotypes was confirmed by an analysis of the effects of specific changes engineered into the wild-type VSIV genome. The results presented here may be relevant to understanding the biology of VSV in nature. We are currently exploring whether the more severe pathogenicity of VSIV-GNJ than that of VSIV-GI in the natural host may be related to the differential sensitivities of the GI and GNJ proteins to pH variations.
ACKNOWLEDGMENTS
We thank the members of the G. W. Wertz and L. A. Ball laboratories for helpful discussions during preparation of the manuscript.
This work was supported by NIH grant R37AI12464 to G.W.W. I. Martínez was a recipient of a fellowship from the "Ministerio de Sanidad y Consumo" (Spain) (BAE 01/5006).
REFERENCES
Ball, L. A., C. R. Pringle, B. Flanagan, V. P. Perepelitsa, and G. W. Wertz. 1999. Phenotypic consequences of rearranging the P, M, and G genes of vesicular stomatitis virus. J. Virol. 73:4705-4712.
Bridges, V. E., B. J. McCluskey, M. D. Salman, H. S. Hurd, and J. Dick. 1997. Review of the 1995 vesicular stomatitis outbreak in the western United States. J. Am. Vet. Med. Assoc. 211:556-560.
Carneiro, F. A., M. L. Bianconi, G. Weissmuller, F. Stauffer, and A. T. Da Poian. 2002. Membrane recognition by vesicular stomatitis virus involves enthalpy-driven protein-lipid interactions. J. Virol. 76:3756-3764.
Carneiro, F. A., A. S. Ferradosa, and A. T. Da Poian. 2001. Low pH-induced conformational changes in vesicular stomatitis virus glycoprotein involve dramatic structure reorganization. J. Biol. Chem. 276:62-67.
Carneiro, F. A., F. Stauffer, C. S. Lima, M. A. Juliano, L. Juliano, and A. T. Da Poian. 2003. Membrane fusion induced by vesicular stomatitis virus depends on histidine protonation. J. Biol. Chem. 278:13789-13794.
Cartwright, B., and F. Brown. 1972. Serological relationships between different strains of vesicular stomatitis virus. J. Gen. Virol. 16:391-398.
Chen, S. S. 1991. Myristylation of the envelope glycoprotein of vesicular stomatitis virus. Intervirology 32:193-197.
Clague, M. J., C. Schoch, L. Zech, and R. Blumenthal. 1990. Gating kinetics of pH-activated membrane fusion of vesicular stomatitis virus with cells: stopped-flow measurements by dequenching of octadecylrhodamine fluorescence. Biochemistry 29:1303-1308.
Crimmins, D. L., W. B. Mehard, and S. Schlesinger. 1983. Physical properties of a soluble form of the glycoprotein of vesicular stomatitis virus at neutral and acidic pH. Biochemistry 22:5790-5796.
Daniels, R. S., J. C. Downie, A. J. Hay, M. Knossow, J. J. Skehel, M. L. Wang, and D. C. Wiley. 1985. Fusion mutants of the influenza virus hemagglutinin glycoprotein. Cell 40:431-439.
Doms, R. W., M. J. Gething, J. Henneberry, J. White, and A. Helenius. 1986. Variant influenza virus hemagglutinin that induces fusion at elevated pH. J. Virol. 57:603-613.
Doms, R. W., D. S. Keller, A. Helenius, and W. E. Balch. 1987. Role for adenosine triphosphate in regulating the assembly and transport of vesicular stomatitis virus G protein trimers. J. Cell. Biol. 105:1957-1969.
Durrer, P., Y. Gaudin, R. W. Ruigrok, R. Graf, and J. Brunner. 1995. Photolabeling identifies a putative fusion domain in the envelope glycoprotein of rabies and vesicular stomatitis viruses. J. Biol. Chem. 270:17575-17581.
Fredericksen, B. L., and M. A. Whitt. 1998. Attenuation of recombinant vesicular stomatitis viruses encoding mutant glycoproteins demonstrate a critical role for maintaining a high pH threshold for membrane fusion in viral fitness. Virology 240:349-358.
Fredericksen, B. L., and M. A. Whitt. 1996. Mutations at two conserved acidic amino acids in the glycoprotein of vesicular stomatitis virus affect pH-dependent conformational changes and reduce the pH threshold for membrane fusion. Virology 217:49-57.
Fredericksen, B. L., and M. A. Whitt. 1995. Vesicular stomatitis virus glycoprotein mutations that affect membrane fusion activity and abolish virus infectivity. J. Virol. 69:1435-1443.
Gallione, C., and J. K. Rose. 1983. Nucleotide sequence of a cDNA clone encoding the entire glycoprotein from the New Jersey serotype of vesicular stomatitis virus. J. Virol. 46:162-169.
Gaudin, Y., C. Tuffereau, P. Durrer, A. Flamand, and R. W. Ruigrok. 1995. Biological function of the low-pH, fusion-inactive conformation of rabies virus glycoprotein (G): G is transported in a fusion-inactive state-like conformation. J. Virol. 69:5528-5534.
Jeetendra, E., K. Ghosh, D. Odell, J. Li, H. P. Ghosh, and M. A. Whitt. 2003. The membrane-proximal region of vesicular stomatitis virus glycoprotein G ectodomain is critical for fusion and virus infectivity. J. Virol. 77:12807-12818.
Jeetendra, E., C. S. Robison, L. M. Albritton, and M. A. Whitt. 2002. The membrane-proximal domain of vesicular stomatitis virus G protein functions as a membrane fusion potentiator and can induce hemifusion. J. Virol. 76:12300-12311.
Keil, W., and R. R. Wagner. 1989. Epitope mapping by deletion mutants and chimeras of two vesicular stomatitis virus glycoprotein genes expressed by a vaccinia virus vector. Virology 170:392-407.
Kelley, J. M., S. U. Emerson, and R. R. Wagner. 1972. The glycoprotein of vesicular stomatitis virus is the antigen that gives rise to and reacts with neutralizing antibody. J. Virol. 10:1231-1235.
Letchworth, G. J., L. L. Rodriguez, and J. Del Barrera. 1999. Vesicular stomatitis. Vet. J. 157:239-260.
Li, Y., C. Drone, E. Sat, and H. P. Ghosh. 1993. Mutational analysis of the vesicular stomatitis virus glycoprotein G for membrane fusion domains. J. Virol. 67:4070-4077.
Machamer, C. E., R. W. Doms, D. G. Bole, A. Helenius, and J. K. Rose. 1990. Heavy chain binding protein recognizes incompletely disulfide-bonded forms of vesicular stomatitis virus G protein. J. Biol. Chem. 265:6879-6883.
Martinez, I., L. L. Rodriguez, C. Jimenez, S. J. Pauszek, and G. W. Wertz. 2003. Vesicular stomatitis virus glycoprotein is a determinant of pathogenesis in swine, a natural host. J. Virol. 77:8039-8047.
Mastromarino, P., C. Conti, P. Goldoni, B. Hauttecoeur, and N. Orsi. 1987. Characterization of membrane components of the erythrocyte involved in vesicular stomatitis virus attachment and fusion at acidic pH. J. Gen. Virol. 68:2359-2369.
Mathieu, M. E., P. R. Grigera, A. Helenius, and R. R. Wagner. 1996. Folding, unfolding, and refolding of the vesicular stomatitis virus glycoprotein. Biochemistry 35:4084-4093.
Matlin, K. S., H. Reggio, A. Helenius, and K. Simons. 1982. Pathway of vesicular stomatitis virus entry leading to infection. J. Mol. Biol. 156:609-631.
McCluskey, B. J., H. S. Hurd, and E. L. Mumford. 1999. Review of the 1997 outbreak of vesicular stomatitis in the western United States. J. Am. Vet. Med. Assoc. 215:1259-1262.
Oomens, A. G., A. G. Megaw, and G. W. Wertz. 2003. Infectivity of a human respiratory syncytial virus lacking the SH, G, and F proteins is efficiently mediated by the vesicular stomatitis virus G protein. J. Virol. 77:3785-3798.
Pak, C. C., A. Puri, and R. Blumenthal. 1997. Conformational changes and fusion activity of vesicular stomatitis virus glycoprotein: [125I]iodonaphthyl azide photolabeling studies in biological membranes. Biochemistry 36:8890-8896.
Puri, A., S. Grimaldi, and R. Blumenthal. 1992. Role of viral envelope sialic acid in membrane fusion mediated by the vesicular stomatitis virus envelope glycoprotein. Biochemistry 31:10108-10113.
Puri, A., J. Winick, R. J. Lowy, D. Covell, O. Eidelman, A. Walter, and R. Blumenthal. 1988. Activation of vesicular stomatitis virus fusion with cells by pretreatment at low pH. J. Biol. Chem. 263:4749-4753.
Robison, C. S., and M. A. Whitt. 2000. The membrane-proximal stem region of vesicular stomatitis virus G protein confers efficient virus assembly. J. Virol. 74:2239-2246.
Rodriguez, L. L. 2002. Emergence and re-emergence of vesicular stomatitis in the United States. Virus Res. 85:211-219.
Rodriguez, L. L., T. A. Bunch, M. Fraire, and Z. N. Llewellyn. 2000. Re-emergence of vesicular stomatitis in the western United States is associated with distinct viral genetic lineages. Virology 271:171-181.
Rodriguez, L. L., and S. T. Nichol. 1999. Vesicular stomatitis virus, p. 1910-1919. In R. G. Webster and A. Granoff (ed.), Encyclopedia of virology. Academic Press, London, United Kingdom.
Rose, J. K., and C. J. Gallione. 1981. Nucleotide sequences of the mRNAs encoding the vesicular stomatitis virus G and M proteins determined from cDNA clones containing the complete coding regions. J. Virol. 39:519-528.
Schlegel, R., T. S. Tralka, M. C. Willingham, and I. Pastan. 1983. Inhibition of VSV binding and infectivity by phosphatidylserine: is phosphatidylserine a VSV-binding site? Cell 32:639-646.
Schlegel, R., and M. Wade. 1985. Biologically active peptides of the vesicular stomatitis virus glycoprotein. J. Virol. 53:319-323.
Shokralla, S., R. Chernish, and H. P. Ghosh. 1999. Effects of double-site mutations of vesicular stomatitis virus glycoprotein G on membrane fusion activity. Virology 256:119-129.
Shokralla, S., Y. He, E. Wanas, and H. P. Ghosh. 1998. Mutations in a carboxy-terminal region of vesicular stomatitis virus glycoprotein G that affect membrane fusion activity. Virology 242:39-50.
Stallknecht, D. E., D. E. Perzak, L. D. Bauer, M. D. Murphy, and E. W. Howerth. 2001. Contact transmission of vesicular stomatitis virus New Jersey in pigs. Am. J. Vet. Res. 62:516-520.
Whelan, S. P., L. A. Ball, J. N. Barr, and G. T. Wertz. 1995. Efficient recovery of infectious vesicular stomatitis virus entirely from cDNA clones. Proc. Natl. Acad. Sci. USA 92:8388-8392.
White, J., K. Matlin, and A. Helenius. 1981. Cell fusion by Semliki Forest, influenza, and vesicular stomatitis viruses. J. Cell Biol. 89:674-679.
Woodgett, C., and J. K. Rose. 1986. Amino-terminal mutation of the vesicular stomatitis virus glycoprotein does not affect its fusion activity. J. Virol. 59:486-489.
Zhang, L., and H. P. Ghosh. 1994. Characterization of the putative fusogenic domain in vesicular stomatitis virus glycoprotein G. J. Virol. 68:2186-2193.(Isidoro Martinez and Gail)