- Email: [email protected]
Identification of the simian rotavirus SA11 genome segment 3 product
VIROLOGY
163, 26-32 (1988)
Identification
of the Simian Rotavirus
SAl 1 Genome Segment
3 Product
MING LIU,* PAUL A. OFFIT,t AND MARY K. ESTES*,+,’ Departments of *Virology and Epidemiology and of +Medicine, Baylor College of Medicine, Houston, Texas 77030, and tDivision Infectious Diseases, The Children’s Hospital of Philadelphia, University of Pennsylvania, Philadelphia, Pennsylvania 79 104 Received August
13, 1987; accepted November
of
4. 1987
Previous studies on rotavirus gene-coding assignments failed to clearly identify the simian rotavirus SAl 1 genome segment 3 protein product. This question was reexamined by using new conditions of electrophoresis with improved resolution of proteins in the high-molecular-weight range. Our results showed that the SAl 1 genome segment 3 codes for a protein with an apparent mol wt of 88,000. This protein normally comigrates with the protein product of genome segment 4. The gene 3 protein was located in viral core particles by comparing the electrophoretic patterns of [36S]methionine-Iabeled viral polypeptides from infected cells and from purified double-shelled, single-shelled, and core particles. To confirm the identity of the gene 3 product, we also studied two reassortant viruses in which genome segment 3 was reassorted from each of two parental virus strains (SAl 1 and NCDV). The gene 3 and gene 4 products of these viruses were identified by (i) their separation by two different polyacrylamide gel systems, (ii) their location in distinct particle types, (iii) their differential sensitivity to trypsin digestion, and (iv) their distinctive protease peptide maps. We propose that the genome segment 3 product be called VP3 and that the gene 4 product be named VP4 from now on. 0 1999 Academic Press, Inc. also were established that allowed better resolution of the high-molecular-weight proteins. After identification of the protein product of the SAl 1 genome segment 3, the authenticity of this gene product was further examined by several biochemical parameters. The function of the gene 3 product was investigated by determining whether it was a structural protein. Finally, with a complete gene assignment for SAl 1, the molar ratios of the structural polypeptides in double-shelled particles were determined.
INTRODUCTION Rotaviruses are now widely recognized as the major etiologic viral agent of gastroenteritis in infants and young children throughout the world (Estes el al., 1983; Cukor and Blacklow, 1984). The rotavirus genome contains 11 segments of double-stranded (ds) RNA that range in molecular weight from 0.2 X 1O6 to 2.2 x 106. Identification of the genome segment 3 (gene 3) protein product has remained unclear, although the protein-coding assignments for the 11 rotavirus genome segments have been reported by several groups of researchers. For example, previous studies have concluded that the gene 3 product was a structural protein (Smith et al., 1980), a nonstructural protein (Arias ef a/., 1982), or a structural protein that was translated poorly in vitro and was synthesized and processed rapidly (McCrae and McCorquodale, 1982). These studies all approached this problem by comparing the proteins synthesized in infected cells to those translated in cell-free systems. We were not able to unequivocally identify a gene 3 protein using similar methods (Mason et al., 1983). In this study new approaches were used to identify the gene 3 protein product. The electrophoretic patterns of the proteins of two parental and two reassortant viruses that each contained a distinct genome segment 3 were compared, and new conditions for electrophoresis of the proteins on polyacrylamide gels ’
To whom
0042-6822/88
requests
for
reprints
$3.00
Copyright Q 1998 by Academic Press, Inc. All ughts of reproduction I” any form reserved.
MATERIALS
AND METHODS
Viruses and cells Two parental (the simian rotavirus SAl 1 and the bovine rotavirus NCDV) and two reassortant [N3 and S3,10, derived and characterized as described by Offit et a/. (1986)] viruses were used for this study. The reassortant virus N3 is a single-gene reassortant that contains all of the SAl 1 genome segments except segment 3, which is derived from NCDV. The S3,lO reassortant virus contains all of the NCDV gene segments except segments 3 and 10, which are derived from SAl 1. All viruses were propagated at low multiplicity (-0.1 PFU/cell) in fetal rhesus monkey (MA1 04) cells in the presence of trypsin and were assayed as previously described (Estes et al., 1979). MA1 04 cells were grown in medium 199 supplemented with 5% fetal bovine serum, 0.03% glutamine, 0.075% sodium bicarbonate, 100 U of penicillin/ml, and 100 pg of
should be addressed. 26
IDENTIFICATION
streptomycin/ml 1986). Analysis
as previously
described
OF SAll
(Chan et al.,
of viral genotypes
Radiolabeling of intracellular viral polypeptides cell-free translation products
In some experiments rotavirus mRNA was prepared and translated in rabbit reticulocyte lysates as described by Mason et al. (1980). Isotopic
The genotype of each virus was determined by analyzing the electrophoretic migration patterns of 32P-isbeled RNA on polyacrylamide gels essentially as described previously (Gombold and Ramig, 1986). Briefly, MA1 04 ceil monolayelrs in 35-mm dishes were inoculated with trypsin-activated virus, and RNA was labeled in 1 .O ml of medium -I99 that lacked serum but contained 7.5 pg of actinomycin D and 75 @Ci of 32P-or-thophosphoric acid. After a 2-day incubation at 37” the cells were pelleted and lysed in 0.5% Nonidet-P40, and RNA was extracted with phenol and precipitated with ethanol. The 32P-Yabeled dsRNAs were subjected to electrophoresis on 8% polyacrylamide slab gels (45 cm long and 0.75 mm thick) and were run for 24-26 hr at 350 V. After electrophoresis, the gels were dried and exposed to Kodak X-Omat AR X-ray film. The origin of each genome segment of both reassortants was determined by comparison with the migration of segments from the parental viruses run on the same gel. and
The methods used to radiolabel rotavirus proteins were described by Ericson et al. (1982). Briefly, MA104 cells in 35-mm dishes (3 X 1O5 cells) were infected with trypsin-activated virus at a multiplicity of 20-30 PFU/cell or were mock-infected. After a 1-hr adsorption period at 37” the inoculum was washed off, and minimum essential medium lacking serum but containing 5 pg of actinomycin D/ml was added. Cells were starved in the same medium without methionine for 30 min before the pulse, and [35S]methionine (20 &i/ml) was added. After a 30-min labeling period the cells were harvested in 200 ~1 of RIPA buffer (150 mM NaCI, 1% sodium deoxycholate, 1% Triton X-100, 0.1% sodrum dodecyl sulfate, 10 mM Tris-hydrochloride, pH 7.2 and 1% Trasylol). In some cases, lysates were made in the presence of 2 pg of tunicamycin (Sigma Chemical Co., St. Louis, MO) per milliliter, which was added to the media after virus adsorption. For radiolabeling of viral proteins at early stages of infection, MA1 04 cells were pretreated with 5 pg of actinomycin D/ml for 1 hr before infection and were infected as above except that 5 pg of actinomycin D/ml was included in the inoculum. The cells were labeled by pulsing with 40 &i of [35S]methionine/ml for 10 min immediately after washing off the inoculum or they were pulsed after different time periods within the first 2 hr after infection.
27
GENE 3 PRODUCT
labeling and purification
of virus
Radiolabeled viruses were grown in MA1 04 cells in 850-cm2 roller bottles as previously described (Estes et a/., 1981). Radiolabeled virus was prepared by adding [35S]methionine (sp act 10 &i/ml) or 3H-L-amino acid hydrolysate (sp act 20 &i/ml) (ICN Pharmaceuticals, Irvine, CA) per milliliter of medium. After a 1-hr virus adsorption period, the monolayers were washed to remove residual trypsin, and the incubation was continued for 18 hr in medium lacking serum, trypsin, and methionine or amino acids. Virus was purified from infected cells by centrifugation in CsCl gradients. To produce single-shelled particles, purified doubleshelled virus was treated with 5 mM EDTA to remove the outer capsid as described by Cohen er al. (1979) and the resulting single-shelled particles were spun on CsCl gradients. Purified core particles were obtained by treating single-shelled particles with 1.5 M CaCI, followed by pelleting through a 45% sucrose cushion as previously described (Bican et al., 1982). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis Electrophoresis of polypeptides was performed on polyacrylamide gels using two different ratios of acrylamide-bisacrylamide cross-linkage to form the gels. Gels with 1% cross-linkage and the discontinuous buffer system of Laemmli (1970) but containing 0.5 M urea were used as previously described (Mason et al., 1980). Gels with 4% cross-linkage were used to obtain optimal resolution of the gene 3 product. Protein samples were suspended in electrophoresis buffer and boiled for 2 min immediately before electrophoresis. The resultant gels were prepared for fluorography (Bonner and Laskey, 1974) and exposed to X-ray film (Kodak X-Omat AR) at -70”. Molecular weight determinations were estimated by comparison of the relative mobilities of the viral polypeptides with those of 14C-labeled markers from New England Nuclear (Boston, MA). The molecular weight markers included phosphorylase b (94,000), bovine serum albumin (68,000), ovalbumin (43,000), carbonic anhydrase (30,000), and cytochrome c (12,500). Quantitation of radioactivity polyacrylamide gels
from
Exposed X-ray films were developed with a Kodak X-Omat M20 processor and read with a Kontes fiber optics scanner. Data were integrated and plotted by a
LIU. OFFIT, AND ESTES
28
Hewlett Packard 3390 integrator connected to the scanner. Different exposures of films and different amounts of virus were analyzed to obtain data within the linear exposure range of the film. In some cases, radioactivity in polypeptide bands was determined by placing excised gel slices in glass scintillation vials and solubilizing them in 0.2 ml of 70% perchloric acid and 0.4 ml of 30% hydrogen peroxide with shaking at 60” (Mahin and Lofberg, 1966). The radioactivity in the solubilized gel slices was determined by counting in Liquiscint scintillation cocktail (National Diagnostics Laboratories, Somerville, NJ) in a scintillation spectrometer.
S3,lO NCDV
N3 SAll
NCDV
SAIl
Peptide mapping The relationships between SAl 1 high-molecularweight polypeptides were analyzed by comparing staphylococcal V8 protease or chymotrypsin peptide maps as described by Cleveland et a/. (1977). Individual polypeptide bands in viral particles that were initially resolved on 6% polyacrylamide gels with 4% cross-linkage were excised from the gels and digested with different concentrations of protease. The partial peptide patterns were compared on fluorograms after exposure for 2-3 months at -70”. RESULTS Characterization of the genotypes reassortant viruses
of parental
and
The genomic patterns of the parental and reassortant viruses were confirmed before proceeding with this study. Figure 1 shows the 32P-labeled dsRNAs from cells infected with these viruses. Compared with the parental SA11 and NCDV RNA patterns, the single-gene reassortant N3 had RNA segments that comigrated with all of the RNA segments from SAl 1 except gene 3. The gene 3 of the N3 reassor-tant comigrated with the NCDV gene 3. Similarly, the reassortant S3,lO had the same RNA segments as NCDV, except for genes 3 and 10 which were from SAll. Resolution
of the gene 3 protein
After confirmation of the RNA patterns, the protein patterns of these viruses were analyzed. Figure 2 shows the analysis of [35S]methionine-labeled viral proteins seen when cells were infected at a high multiplicity with trypsin-activated viruses. Both a schematic of the patterns observed and the actual protein profiles are shown. Optimal resolution between the high-molecular-weight proteins of the parental and reassortant viruses required electrophoresis of the proteins on 6%
FIG. 1. RNA patterns of parental and reassortant viruses. The electrophoretic mobilities of individual segments of 3’P-labeled dsRNAs from the parental (NCDV or SAll) or reassortant (N3 or S3,lO) viruses were analyzed by electrophoresis on 8% polyacrylamide gels. The origin of each RNA segment of the reassortants was determined by comparison with the parental RNAs and confirmed those reported originally by Offit et a/. (1986). The double arrowheads indicate the genome 3 segments.
polyacrylamide gels with 1% bisacrylamide cross-linkage. The 6% gel system allowed us to separate the gene 3 and gene 4 protein products of NCDV, and the migration order of the four high-molecular-weight NCDV proteins in this gel system was gene 1 product-gene 2 product-gene 4 product-gene 3 product. The gene 3 and 4 protein products of SAll comigrated in this system. N3, the reassortant that contained only gene 3 from NCDV, showed protein migration patterns identical with those of the SAl 1 proteins except for one band. This band, which presumably represented the gene 3 product, comigrated with an NCDV band. Similarly, for the reassortant S3,lO three of the high-molecular-weight protein bands migrated parallel to the NCDV proteins. The one band that co-
IDENTIFICATION NCDV
SAl 1
S3,lO
OF SAll
N3
NCDV
-1-l
-1
29
GENE 3 PRODUCT NCDV
S3,lO
SAll
N3
NCDV
1.41
1
FIG. 2. Resolution of high-molecular-weight proteins of parental and reassortant viruses by electrophoresis on 6% polyacrylamide gels with 1% bisacrylamide cross-linkage. Lysates were prepared from [35S]methionine-labeled. rotavirus-infected cells as described under Materials and Methods. Samples were electrophoresed. and the resultant gels were prepared for fluorography and exposed to X-ray film at -70”. The left-hand panel shows a schematic of the protein patterns, whereas the actual protein bands are seen in the right-hand panel. The number beside each radioactive band represents the genome segment that codes for that protein band. Open numbers, genome segments from SAl 1. Closed numbers, genome segments from NCDV. The NCDV protein bands 2 and 4 in the left NCDV lane show an apparent slower migration than their counterparts in the S3,lO reassortant because of slight “gel smiling.”
migrated with the SAl 1 band was postulated to be the gene 3 product of SAl 1. Figure 3 shows the results obtained when these viral proteins were analyzed on a different gel system, using a 6% polyacrylamide concentration with a 4% cross-linkage. The gene 3 and 4 products of SAll could be separated on this gel system, but the gene 3 and 4 products of NCDV comigrated. Because the gene 3 product of SAl 1 migrated at the same level as the gene 3 #and4 products of NCDV, the high-molecular-weight protein of N3 displayed a pattern similar to that of the SAl 1 protein, and the S3,lO protein pattern was similar to the patterns of NCDV proteins. The gene 3 product
is a unique structural
protein
In an effc.r-t to understand the function of the gene 3 product, we determined whether it was a structural protein located in the virion. To do this SAl 1 doubleshelled, single-shelled, and core particles were purified. These particles were then analyzed on 6% poly-
NCDV
-1
SAl 1
s3.10
NCDV
N3
acrylamide gels with 4% cross-linkage, which (as shown previously) were capable of resolving the SAl 1 gene 3 and 4 products. Figure 4A shows that the gene 3 product was seen in double-shelled particles as well as in the single-shelled and core particles. The preparations of core particles analyzed in these experiments contained >95% cores when visualized after negative staining and electron microscopy. However, these particles were always aggregated. Analysis of the protein patterns in such preparations showed VPI, VP2 the gene 3 product, and VP6 (Fig. 4A, lane 3). When the counts in each of these bands in the different forms of viral particles were determined (data not shown) and normalized to VP1 for comparison, the results showed that EDTA removed the gene 4 and gene 9 products from double-shelled particles, and CaCI, treatment stripped off the majority of VP6 from single-shelled particles. In contrast, the relative amounts of VPl, VP2, and the gene 3 product remained similar in all three types of particles, confirming
NCDV
S3,lO
SAll
N3
NCDV
-1
-1 -
‘I
-
.I
FIG. 3. Resolution of high-molecular-weight proteins of parental and reassortant viruses by electrophoresis on 6% polyacrylamide gels with 4% bisacrylamide cross-linkage. Proteins of the parental and reassortant viruses were analyzed by electrophoresis as described in Fig. 2, and a schematic of the protein patterns and the actual protein bands is shown. Open numbers, genome segments from SAl 1. Closed numbers, genome segments from NCDV.
30
LIU. OFFIT, AND ESTES
B
FIG. 4. Identification of the gene 3 product in vrral partrcles, infected cells, and rabbit reticulocyte lysate translation mixtures. (A) The proteins in purified viral particles were resolved on 6% polyacrylamide gels with 4% cross-linkage. The proteins in purified doubleshelled particles (lane l), single-shelled particles generated from treating double-shelled particles with 5 mM EDTA followed by purification in CsCl gradients (lane 2) or core particles derived from treating single-shelled particles with 1.5 M CaCI, and pelleting the cores through a 40% sucrose cushion (lane 3) are shown. (B) The gene 3 product was detected in infected cells (lane l), but not in mock-infected cells (lane 2) by analysis of proteins labeled with 20 PCi of [35S]methionine per milliliter for 10 mm after a 1-hr vrrus adsorption period. Proteins were resolved by electrophoresis on 6% polyacrylamide gels with 4% cross-linkage. (C) The gene 3 product was detected by in vitro translation of purified SAl 1 total mRNAs. The 35S-labeled proteins in purified single-shelled particles (lane 1) and in translation products from rabbit reticulocyte lysates programmed with rotavirus total mRNAs (lane 2) or with no RNA (lane 3) are shown. The double arrowheads highlight the gene 3 product.
that these three proteins must make up the core particles. The patterns of proteins synthesized in virus-infected cells after pulse-labeling with [35S]methionine were examined to determine when the gene 3 product was produced. The gene 3 product was seen in virusinfected (but not in mock-infected) cells after a lo-min pulse, as early as 60-70 min postinfection (Fig. 4B). This provided additional evidence that this protein is a primary translation product. The protein profiles in lysates treated with the glycosylation inhibitor tunicamycin showed the expected shift in the molecular weights of the known glycoproteins VP7 and NS28, but the migration mobility of the gene 3 product was not affected (data not shown). These experiments also revealed that the gene 3 product was present in infected cells in amounts similar to those of the other three high-molecular-weight proteins. This contrasted with the small amount of this
protein seen in viral particles as well as in products of in vitro translation reactions, which were programmed by purified SAI 1 total mRNAs and conducted in rabbit reticulocyte lysates (Fig. 4C). Further experiments are in progress to determine the significance of this observation. The protein patterns of double-shelled viruses treated with different concentrations of ttypsin also were examined (Fig. 5) to confirm that the band identified in viral particles as the gene 3 product was not the trypsin-sensitive gene 4 product (VP3). Trypsin treatment of double-shelled particles resulted in the cleavage of VP3, but had no effect on the gene 3 product. Finally, one-dimensional peptide maps were done to compare the putative gene 1, 2, 3, and 4 products. Comparison of the peptide maps for the gene 1, 2, 3, and 4 proteins digested by staphylococcal V8 protease (Fig. 6) and chymotrypsin (data not shown) indicated that these proteins were distinct. Molecular
ratio of polypeptides
in viral particles
The percentage of each of the six polypeptides in different types of viral particles was determined by analysis of 3H-labeled viruses by the optical scanning of autoradiograms and by assuming that the radioactivity in each peak reflected the amount of each polypeptide (Table 1). The approximate molecular ratios of the polypeptides also were calculated from the aver-
FIG. 5. Distinction of the gene 3 product from the gene 4 product by trypsin treatment. The gene 3 protein was resolved by electrophoresis of proteins in single-shelled particles (lane 1) or in purified double-shelled particles (lanes 2-5) analyzed directly (lane 2) or after treatment of double-shelled particles with 10, 1, or 0.1 pg of TPCKtrypsin/ml (lanes 3, 4, and 5, respectively). Trypsin treatment was performed for 30 min at 37”. Electrophoresis was in 6% polyacrylamide gels with 4% cross-linkage. The double arrowhead indicates the gene 3 product.
IDENTIFICATION
OF SAll
GENE 3 PRODUCT
31
FIG. 6. One-dimensional V8 protease mapping of the four high-molecular-weight proteins of rotavirus SAl 1. V8 protease digestion (0, 1, or 10 pglml) was used for one-dimensional peptide mapping to compare the relationship between the products of genome segments 1, 2, 3, and 4. The maps shown were from two parallel gels of the same composition that were run under the same electrophoretic conditions at the same time. The autoradiograms for the gene 1 and gene 2 products were from one gel exposed for different time periods; the autoradiograms for the gene 3 and gene 4 products were from a second gel. The numbers above the concentrations of the enzymes indicate the genome segments that code for the high-molecular-werght proteins.
age of two determinations from 3H-labeled viruses (Table 1). Data obtained by scintillation counting of gel slices or (optical scanning of autoradiograms gave consistent results. DISCUSSION Utilizing different systems of polyacrylamide gels, the gene 3 product was resolved from comigration with other proteins, and evidence was found to indicate that this product is a primary structural protein located in the core rshell of rotavirus particles. The TABLE 1 RELATIVE PROPORTIONS OFTHE POLYPEPTIDES IN SAll DOUBLE-SHELLED PARTICLES
Virion polypeptide” VP1 VP2 VP3 VP4 VP6 VP7
Molecular weight 125,000 94,000 88,OC’O 88,OC~O 41 ,OClO 34,000
Percentage of virion proteir?
Approximate molar ratioC
2 15 0.5 1.5 51 30
1 10 0.35 1 78 55
“The virion polypeptides are designated as proposed in this manuscript. VP3 is the protein product of gene segment 3, and VP4 is the prote n product of gene segment 4. b Determined by densitometer scanning and integration of the polypeptide peaks resolved following polyacrylamide gel electrophoresis of particles labeled vvith ‘H-labeled amino acrds. ’ Calculated by dividing the average percentage of the virion protein by the molecular weight of the corresponding polypeptide. The value obtained for VP1 was taken as 1.
failure of previous investigators to unequivocally identify the gene 3 product probably was due to the special migration properties of this protein on polyacrylamide gels. In this study, the gene 3 product of NCDV could be resolved on 6% polyacrylamide gels with 1% cross-linkage, and the gene 3 protein of SAl 1 could be separated from the gene 4 protein on 6% polyacrylamide gels with 4% cross-linkage. To resolve the structural proteins of rotavirus SAl 1, a 6% polyacrylamide gel with 4% cross-linkage gel system was optimal. Our results support the studies of Smith et al. (1980), who concluded that the gene 3 product is a structural protein located in single-shelled viral partcles. Our results also extend their studies by localizing this protein to the viral inner core particles. McCrae and McCorquodale (1982) also suggested that the gene 3 product was an inner shell structural protein that was initially synthesized as a larger precursor and was immediately processed. We did not identify any precursor in our studies. The following evidence supports our identification of the gene 3 product: (i) distinction of the gene 3 protein patterns in both parental viruses and reassortant viruses that contain reasserted gene 3 segments, (ii) in vitro translation of total mRNA in rabbit reticulocyte lysates, (iii) viral protein labeling at an early stage of infection, (iv) one-dimensional peptide mapping of high-molecular-weight proteins, and (v) the unique characteristics of being a core protein and being resistant to trypsin treatment of double-shelled particles easily differentiate the gene 3 product from the gene 4 product (VP3) with which it often comigrates. Utilizing the same gel systems reported in this study, we also
32
LIU, OFFIT, AND ESTES
resolved and identified the gene 3 products in rhesus rotavirus RRV and in single-gene 3 reassortants of RRV and SAl 1 (provided by Dr. R. F. Ramig; data not shown). From our results to date we conclude that the gene 3 product is a core protein. Ramig (1983) isolated 10 different groups of temperature-sensitive mutants of SA11, and subsequently mapped mutant 339 of group B to genome segment 3 (Gombold et a/., 1985). It is of interest that this mutant has shown reduced amounts of single-stranded RNA and no detectable dsRNA synthesized at the nonpermissive temperature (Ramig, 1983). This observation, taken together with the location of the gene 3 product, suggests that this protein might function as par-t of the virion-associated RNA polymerase. Since the product of genome segment 3 represents only a small molar ratio of the virion and mRNA of genome segment 3 is not translated well in the in vitro translation system programmed with total viral transcripts, gene cloning and subsequent in vitro expression should be helpful approaches to study the function of this protein. Although the term VP3 has been widely used for the gene 4 product, we propose that the genome segment 3 product be called VP3 and that the gene 4 product be named VP4 from now on. ACKNOWLEDGMENTS We thank Mr. Edward Calomenl and Dr. Tomoyuki Tanaka for their electron microscopic work to confirm the types of purified viral particles, Dr. Wai-Kit Chan for critical comments, and Dr. Robert F. Ramig for providing the SAl l/RRV reassortants. This work was supported by Grant DK-30144 from the National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases and by a grant from the Health Care Dtvision of Richardson-Vicks (Shelton, CT).
REFERENCES ARIAS, C. F., LOPEZ, S., and ESPEJO,R. T. (1982). Gene protein products of SAl 1 simian rotavirus genome. J. Viral. 41, 42-50. BICAN. P., COHEN, J., CHARPILIENNE,A., and SCHERRER, R. (1982). Purification and characterization of bovine rotavirus cores. /. Viral. 43,1113-1117. BONNER,W. M., and LASKEY, R. A. (1974). A film detection method for tritium-labelled proteins and nucleic actds in polyacrylamide gels. Eur. J. Biochem. 46, 83-88. CHAN. W.-K., PENARANDA, M. E., CRAWFORD,S. E., and ESTES, M. K.
(1986). Two glycoproteins are produced from the rotavlrus neutralizatlon gene. Virology 151, 243-252. CLEVELAND, D. W., FISCHER, S. G., KIRSCHNER,M. W., and LAEMMLI, U. K. (1977). Peptlde mapplng by limited proteolysls in sodtum dodecyl sulfate and analysts by gel electrophoresis. /. B/o/. Chem. 252,1102-1106. COHEN, J., LAPORTE,J., CHARPILIENNE,A., and SCHERRER,R. (1979). Activation of rotavirus RNA polymerase by calcium chelation. Arch. Virol. 60, 177-l 86. CUKOR, G.. and BLACKLOW,N. R. (1984). Human viral gastroenteritis. Microbial. Rev. 48, 157-l 79. ERICSON, B. L., GRAHAM, D. Y., MASON, B. B., and ESTES. M. K. (1982). Identification, synthesis, and modifications of simian rotavirus SAl 1 polypeptides in infected cells. J. Viral. 42, 825-839. ESTES. M. K., GRAHAM, D. Y., GER~A, C. P., and SMITH, E. M. (1979). Simian rotavlrus SAl 1 replication in cell cultures. J. Viral. 31, 810-815. ESTES, M. K., GRAHAM, D. Y., and MASON, B. B. (1981). Proteolytic enhancement of rotavirus infectivity: Molecular mechanisms. /. Viral. 39, 879-888. ESTES.M. K., PALMER, E. L., and OBIJESKI,J. F. (1983). Rotaviruses: A review. Cur. Top. Microbial. lmmunol. 105, 123-l 84. GOMBOLD, J. L., ESTES,M. K., and RAMIG, R. F. (1985). Assignment of simian rotavirus SAl 1 temperature-sensitive mutant groups B and E to genome segments. Virology 143, 309-320. GOMBOLD, J. L., and RAMIG, R. F. (1986). Analysis of reassortment of genome segments in mice mixedly infected with rotaviruses SAl 1 and RRV. J. Viroi. 57, 110-l 16. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227, 680-685. MAHIN, D. I., and LOFBERG,R. I. (1966). A simplified method of sample preparation for determination of tritium, carbon-l 4. or sulfur35 in blood or tissue by liquid scintillation counting. Anal. Bio&em. 16, 500-509. MASON, B. B., GRAHAM, D. Y., and ESTES, M. K. (1980). In vitro transcription and translation of simian rotavirus SAl 1 gene products. /. VifOl. 33, 11 11-l 121. MASON, B. B., GRAHAM, D. Y., and ESTES, M. K. (1983). Biochemical mapping of the simian rotavirus SAl 1 genome. /. Viral. 46, 413-423. MCCRAE, M. A., and MC~ORQLJODALE,J. G. (1982). The molecular biology of rotaviruses. II. Identification of the protein-coding assignments of calf rotavirus genome RNA species. Virology 117, 435-443. OFFIT, P. A., BLAVAT, G., GREENBERG,H. B., and CLARK, H. F. (1986). Molecular basis of rotavirus virulence: Role of gene segment 4. /. Virol. 57, 46-49. RAMIG, R. F. (1983). Genetic studies with simian rotavirus SAl 1. ln “Double-Stranded RNA Viruses” (R. W. Compans and D. H. L. Bishop, Eds.), pp. 321-327. Elsevier, New York. SMITH, M. L., LAZDINS, I., and HOLMES, I. H. (1980). Coding asslgnments of double-stranded RNA segments of SA 11 rotavirus established by in vitro translation. /. Viral. 33, 976-982.
163, 26-32 (1988)
Identification
of the Simian Rotavirus
SAl 1 Genome Segment
3 Product
MING LIU,* PAUL A. OFFIT,t AND MARY K. ESTES*,+,’ Departments of *Virology and Epidemiology and of +Medicine, Baylor College of Medicine, Houston, Texas 77030, and tDivision Infectious Diseases, The Children’s Hospital of Philadelphia, University of Pennsylvania, Philadelphia, Pennsylvania 79 104 Received August
13, 1987; accepted November
of
4. 1987
Previous studies on rotavirus gene-coding assignments failed to clearly identify the simian rotavirus SAl 1 genome segment 3 protein product. This question was reexamined by using new conditions of electrophoresis with improved resolution of proteins in the high-molecular-weight range. Our results showed that the SAl 1 genome segment 3 codes for a protein with an apparent mol wt of 88,000. This protein normally comigrates with the protein product of genome segment 4. The gene 3 protein was located in viral core particles by comparing the electrophoretic patterns of [36S]methionine-Iabeled viral polypeptides from infected cells and from purified double-shelled, single-shelled, and core particles. To confirm the identity of the gene 3 product, we also studied two reassortant viruses in which genome segment 3 was reassorted from each of two parental virus strains (SAl 1 and NCDV). The gene 3 and gene 4 products of these viruses were identified by (i) their separation by two different polyacrylamide gel systems, (ii) their location in distinct particle types, (iii) their differential sensitivity to trypsin digestion, and (iv) their distinctive protease peptide maps. We propose that the genome segment 3 product be called VP3 and that the gene 4 product be named VP4 from now on. 0 1999 Academic Press, Inc. also were established that allowed better resolution of the high-molecular-weight proteins. After identification of the protein product of the SAl 1 genome segment 3, the authenticity of this gene product was further examined by several biochemical parameters. The function of the gene 3 product was investigated by determining whether it was a structural protein. Finally, with a complete gene assignment for SAl 1, the molar ratios of the structural polypeptides in double-shelled particles were determined.
INTRODUCTION Rotaviruses are now widely recognized as the major etiologic viral agent of gastroenteritis in infants and young children throughout the world (Estes el al., 1983; Cukor and Blacklow, 1984). The rotavirus genome contains 11 segments of double-stranded (ds) RNA that range in molecular weight from 0.2 X 1O6 to 2.2 x 106. Identification of the genome segment 3 (gene 3) protein product has remained unclear, although the protein-coding assignments for the 11 rotavirus genome segments have been reported by several groups of researchers. For example, previous studies have concluded that the gene 3 product was a structural protein (Smith et al., 1980), a nonstructural protein (Arias ef a/., 1982), or a structural protein that was translated poorly in vitro and was synthesized and processed rapidly (McCrae and McCorquodale, 1982). These studies all approached this problem by comparing the proteins synthesized in infected cells to those translated in cell-free systems. We were not able to unequivocally identify a gene 3 protein using similar methods (Mason et al., 1983). In this study new approaches were used to identify the gene 3 protein product. The electrophoretic patterns of the proteins of two parental and two reassortant viruses that each contained a distinct genome segment 3 were compared, and new conditions for electrophoresis of the proteins on polyacrylamide gels ’
To whom
0042-6822/88
requests
for
reprints
$3.00
Copyright Q 1998 by Academic Press, Inc. All ughts of reproduction I” any form reserved.
MATERIALS
AND METHODS
Viruses and cells Two parental (the simian rotavirus SAl 1 and the bovine rotavirus NCDV) and two reassortant [N3 and S3,10, derived and characterized as described by Offit et a/. (1986)] viruses were used for this study. The reassortant virus N3 is a single-gene reassortant that contains all of the SAl 1 genome segments except segment 3, which is derived from NCDV. The S3,lO reassortant virus contains all of the NCDV gene segments except segments 3 and 10, which are derived from SAl 1. All viruses were propagated at low multiplicity (-0.1 PFU/cell) in fetal rhesus monkey (MA1 04) cells in the presence of trypsin and were assayed as previously described (Estes et al., 1979). MA1 04 cells were grown in medium 199 supplemented with 5% fetal bovine serum, 0.03% glutamine, 0.075% sodium bicarbonate, 100 U of penicillin/ml, and 100 pg of
should be addressed. 26
IDENTIFICATION
streptomycin/ml 1986). Analysis
as previously
described
OF SAll
(Chan et al.,
of viral genotypes
Radiolabeling of intracellular viral polypeptides cell-free translation products
In some experiments rotavirus mRNA was prepared and translated in rabbit reticulocyte lysates as described by Mason et al. (1980). Isotopic
The genotype of each virus was determined by analyzing the electrophoretic migration patterns of 32P-isbeled RNA on polyacrylamide gels essentially as described previously (Gombold and Ramig, 1986). Briefly, MA1 04 ceil monolayelrs in 35-mm dishes were inoculated with trypsin-activated virus, and RNA was labeled in 1 .O ml of medium -I99 that lacked serum but contained 7.5 pg of actinomycin D and 75 @Ci of 32P-or-thophosphoric acid. After a 2-day incubation at 37” the cells were pelleted and lysed in 0.5% Nonidet-P40, and RNA was extracted with phenol and precipitated with ethanol. The 32P-Yabeled dsRNAs were subjected to electrophoresis on 8% polyacrylamide slab gels (45 cm long and 0.75 mm thick) and were run for 24-26 hr at 350 V. After electrophoresis, the gels were dried and exposed to Kodak X-Omat AR X-ray film. The origin of each genome segment of both reassortants was determined by comparison with the migration of segments from the parental viruses run on the same gel. and
The methods used to radiolabel rotavirus proteins were described by Ericson et al. (1982). Briefly, MA104 cells in 35-mm dishes (3 X 1O5 cells) were infected with trypsin-activated virus at a multiplicity of 20-30 PFU/cell or were mock-infected. After a 1-hr adsorption period at 37” the inoculum was washed off, and minimum essential medium lacking serum but containing 5 pg of actinomycin D/ml was added. Cells were starved in the same medium without methionine for 30 min before the pulse, and [35S]methionine (20 &i/ml) was added. After a 30-min labeling period the cells were harvested in 200 ~1 of RIPA buffer (150 mM NaCI, 1% sodium deoxycholate, 1% Triton X-100, 0.1% sodrum dodecyl sulfate, 10 mM Tris-hydrochloride, pH 7.2 and 1% Trasylol). In some cases, lysates were made in the presence of 2 pg of tunicamycin (Sigma Chemical Co., St. Louis, MO) per milliliter, which was added to the media after virus adsorption. For radiolabeling of viral proteins at early stages of infection, MA1 04 cells were pretreated with 5 pg of actinomycin D/ml for 1 hr before infection and were infected as above except that 5 pg of actinomycin D/ml was included in the inoculum. The cells were labeled by pulsing with 40 &i of [35S]methionine/ml for 10 min immediately after washing off the inoculum or they were pulsed after different time periods within the first 2 hr after infection.
27
GENE 3 PRODUCT
labeling and purification
of virus
Radiolabeled viruses were grown in MA1 04 cells in 850-cm2 roller bottles as previously described (Estes et a/., 1981). Radiolabeled virus was prepared by adding [35S]methionine (sp act 10 &i/ml) or 3H-L-amino acid hydrolysate (sp act 20 &i/ml) (ICN Pharmaceuticals, Irvine, CA) per milliliter of medium. After a 1-hr virus adsorption period, the monolayers were washed to remove residual trypsin, and the incubation was continued for 18 hr in medium lacking serum, trypsin, and methionine or amino acids. Virus was purified from infected cells by centrifugation in CsCl gradients. To produce single-shelled particles, purified doubleshelled virus was treated with 5 mM EDTA to remove the outer capsid as described by Cohen er al. (1979) and the resulting single-shelled particles were spun on CsCl gradients. Purified core particles were obtained by treating single-shelled particles with 1.5 M CaCI, followed by pelleting through a 45% sucrose cushion as previously described (Bican et al., 1982). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis Electrophoresis of polypeptides was performed on polyacrylamide gels using two different ratios of acrylamide-bisacrylamide cross-linkage to form the gels. Gels with 1% cross-linkage and the discontinuous buffer system of Laemmli (1970) but containing 0.5 M urea were used as previously described (Mason et al., 1980). Gels with 4% cross-linkage were used to obtain optimal resolution of the gene 3 product. Protein samples were suspended in electrophoresis buffer and boiled for 2 min immediately before electrophoresis. The resultant gels were prepared for fluorography (Bonner and Laskey, 1974) and exposed to X-ray film (Kodak X-Omat AR) at -70”. Molecular weight determinations were estimated by comparison of the relative mobilities of the viral polypeptides with those of 14C-labeled markers from New England Nuclear (Boston, MA). The molecular weight markers included phosphorylase b (94,000), bovine serum albumin (68,000), ovalbumin (43,000), carbonic anhydrase (30,000), and cytochrome c (12,500). Quantitation of radioactivity polyacrylamide gels
from
Exposed X-ray films were developed with a Kodak X-Omat M20 processor and read with a Kontes fiber optics scanner. Data were integrated and plotted by a
LIU. OFFIT, AND ESTES
28
Hewlett Packard 3390 integrator connected to the scanner. Different exposures of films and different amounts of virus were analyzed to obtain data within the linear exposure range of the film. In some cases, radioactivity in polypeptide bands was determined by placing excised gel slices in glass scintillation vials and solubilizing them in 0.2 ml of 70% perchloric acid and 0.4 ml of 30% hydrogen peroxide with shaking at 60” (Mahin and Lofberg, 1966). The radioactivity in the solubilized gel slices was determined by counting in Liquiscint scintillation cocktail (National Diagnostics Laboratories, Somerville, NJ) in a scintillation spectrometer.
S3,lO NCDV
N3 SAll
NCDV
SAIl
Peptide mapping The relationships between SAl 1 high-molecularweight polypeptides were analyzed by comparing staphylococcal V8 protease or chymotrypsin peptide maps as described by Cleveland et a/. (1977). Individual polypeptide bands in viral particles that were initially resolved on 6% polyacrylamide gels with 4% cross-linkage were excised from the gels and digested with different concentrations of protease. The partial peptide patterns were compared on fluorograms after exposure for 2-3 months at -70”. RESULTS Characterization of the genotypes reassortant viruses
of parental
and
The genomic patterns of the parental and reassortant viruses were confirmed before proceeding with this study. Figure 1 shows the 32P-labeled dsRNAs from cells infected with these viruses. Compared with the parental SA11 and NCDV RNA patterns, the single-gene reassortant N3 had RNA segments that comigrated with all of the RNA segments from SAl 1 except gene 3. The gene 3 of the N3 reassor-tant comigrated with the NCDV gene 3. Similarly, the reassortant S3,lO had the same RNA segments as NCDV, except for genes 3 and 10 which were from SAll. Resolution
of the gene 3 protein
After confirmation of the RNA patterns, the protein patterns of these viruses were analyzed. Figure 2 shows the analysis of [35S]methionine-labeled viral proteins seen when cells were infected at a high multiplicity with trypsin-activated viruses. Both a schematic of the patterns observed and the actual protein profiles are shown. Optimal resolution between the high-molecular-weight proteins of the parental and reassortant viruses required electrophoresis of the proteins on 6%
FIG. 1. RNA patterns of parental and reassortant viruses. The electrophoretic mobilities of individual segments of 3’P-labeled dsRNAs from the parental (NCDV or SAll) or reassortant (N3 or S3,lO) viruses were analyzed by electrophoresis on 8% polyacrylamide gels. The origin of each RNA segment of the reassortants was determined by comparison with the parental RNAs and confirmed those reported originally by Offit et a/. (1986). The double arrowheads indicate the genome 3 segments.
polyacrylamide gels with 1% bisacrylamide cross-linkage. The 6% gel system allowed us to separate the gene 3 and gene 4 protein products of NCDV, and the migration order of the four high-molecular-weight NCDV proteins in this gel system was gene 1 product-gene 2 product-gene 4 product-gene 3 product. The gene 3 and 4 protein products of SAll comigrated in this system. N3, the reassortant that contained only gene 3 from NCDV, showed protein migration patterns identical with those of the SAl 1 proteins except for one band. This band, which presumably represented the gene 3 product, comigrated with an NCDV band. Similarly, for the reassortant S3,lO three of the high-molecular-weight protein bands migrated parallel to the NCDV proteins. The one band that co-
IDENTIFICATION NCDV
SAl 1
S3,lO
OF SAll
N3
NCDV
-1-l
-1
29
GENE 3 PRODUCT NCDV
S3,lO
SAll
N3
NCDV
1.41
1
FIG. 2. Resolution of high-molecular-weight proteins of parental and reassortant viruses by electrophoresis on 6% polyacrylamide gels with 1% bisacrylamide cross-linkage. Lysates were prepared from [35S]methionine-labeled. rotavirus-infected cells as described under Materials and Methods. Samples were electrophoresed. and the resultant gels were prepared for fluorography and exposed to X-ray film at -70”. The left-hand panel shows a schematic of the protein patterns, whereas the actual protein bands are seen in the right-hand panel. The number beside each radioactive band represents the genome segment that codes for that protein band. Open numbers, genome segments from SAl 1. Closed numbers, genome segments from NCDV. The NCDV protein bands 2 and 4 in the left NCDV lane show an apparent slower migration than their counterparts in the S3,lO reassortant because of slight “gel smiling.”
migrated with the SAl 1 band was postulated to be the gene 3 product of SAl 1. Figure 3 shows the results obtained when these viral proteins were analyzed on a different gel system, using a 6% polyacrylamide concentration with a 4% cross-linkage. The gene 3 and 4 products of SAll could be separated on this gel system, but the gene 3 and 4 products of NCDV comigrated. Because the gene 3 product of SAl 1 migrated at the same level as the gene 3 #and4 products of NCDV, the high-molecular-weight protein of N3 displayed a pattern similar to that of the SAl 1 protein, and the S3,lO protein pattern was similar to the patterns of NCDV proteins. The gene 3 product
is a unique structural
protein
In an effc.r-t to understand the function of the gene 3 product, we determined whether it was a structural protein located in the virion. To do this SAl 1 doubleshelled, single-shelled, and core particles were purified. These particles were then analyzed on 6% poly-
NCDV
-1
SAl 1
s3.10
NCDV
N3
acrylamide gels with 4% cross-linkage, which (as shown previously) were capable of resolving the SAl 1 gene 3 and 4 products. Figure 4A shows that the gene 3 product was seen in double-shelled particles as well as in the single-shelled and core particles. The preparations of core particles analyzed in these experiments contained >95% cores when visualized after negative staining and electron microscopy. However, these particles were always aggregated. Analysis of the protein patterns in such preparations showed VPI, VP2 the gene 3 product, and VP6 (Fig. 4A, lane 3). When the counts in each of these bands in the different forms of viral particles were determined (data not shown) and normalized to VP1 for comparison, the results showed that EDTA removed the gene 4 and gene 9 products from double-shelled particles, and CaCI, treatment stripped off the majority of VP6 from single-shelled particles. In contrast, the relative amounts of VPl, VP2, and the gene 3 product remained similar in all three types of particles, confirming
NCDV
S3,lO
SAll
N3
NCDV
-1
-1 -
‘I
-
.I
FIG. 3. Resolution of high-molecular-weight proteins of parental and reassortant viruses by electrophoresis on 6% polyacrylamide gels with 4% bisacrylamide cross-linkage. Proteins of the parental and reassortant viruses were analyzed by electrophoresis as described in Fig. 2, and a schematic of the protein patterns and the actual protein bands is shown. Open numbers, genome segments from SAl 1. Closed numbers, genome segments from NCDV.
30
LIU. OFFIT, AND ESTES
B
FIG. 4. Identification of the gene 3 product in vrral partrcles, infected cells, and rabbit reticulocyte lysate translation mixtures. (A) The proteins in purified viral particles were resolved on 6% polyacrylamide gels with 4% cross-linkage. The proteins in purified doubleshelled particles (lane l), single-shelled particles generated from treating double-shelled particles with 5 mM EDTA followed by purification in CsCl gradients (lane 2) or core particles derived from treating single-shelled particles with 1.5 M CaCI, and pelleting the cores through a 40% sucrose cushion (lane 3) are shown. (B) The gene 3 product was detected in infected cells (lane l), but not in mock-infected cells (lane 2) by analysis of proteins labeled with 20 PCi of [35S]methionine per milliliter for 10 mm after a 1-hr vrrus adsorption period. Proteins were resolved by electrophoresis on 6% polyacrylamide gels with 4% cross-linkage. (C) The gene 3 product was detected by in vitro translation of purified SAl 1 total mRNAs. The 35S-labeled proteins in purified single-shelled particles (lane 1) and in translation products from rabbit reticulocyte lysates programmed with rotavirus total mRNAs (lane 2) or with no RNA (lane 3) are shown. The double arrowheads highlight the gene 3 product.
that these three proteins must make up the core particles. The patterns of proteins synthesized in virus-infected cells after pulse-labeling with [35S]methionine were examined to determine when the gene 3 product was produced. The gene 3 product was seen in virusinfected (but not in mock-infected) cells after a lo-min pulse, as early as 60-70 min postinfection (Fig. 4B). This provided additional evidence that this protein is a primary translation product. The protein profiles in lysates treated with the glycosylation inhibitor tunicamycin showed the expected shift in the molecular weights of the known glycoproteins VP7 and NS28, but the migration mobility of the gene 3 product was not affected (data not shown). These experiments also revealed that the gene 3 product was present in infected cells in amounts similar to those of the other three high-molecular-weight proteins. This contrasted with the small amount of this
protein seen in viral particles as well as in products of in vitro translation reactions, which were programmed by purified SAI 1 total mRNAs and conducted in rabbit reticulocyte lysates (Fig. 4C). Further experiments are in progress to determine the significance of this observation. The protein patterns of double-shelled viruses treated with different concentrations of ttypsin also were examined (Fig. 5) to confirm that the band identified in viral particles as the gene 3 product was not the trypsin-sensitive gene 4 product (VP3). Trypsin treatment of double-shelled particles resulted in the cleavage of VP3, but had no effect on the gene 3 product. Finally, one-dimensional peptide maps were done to compare the putative gene 1, 2, 3, and 4 products. Comparison of the peptide maps for the gene 1, 2, 3, and 4 proteins digested by staphylococcal V8 protease (Fig. 6) and chymotrypsin (data not shown) indicated that these proteins were distinct. Molecular
ratio of polypeptides
in viral particles
The percentage of each of the six polypeptides in different types of viral particles was determined by analysis of 3H-labeled viruses by the optical scanning of autoradiograms and by assuming that the radioactivity in each peak reflected the amount of each polypeptide (Table 1). The approximate molecular ratios of the polypeptides also were calculated from the aver-
FIG. 5. Distinction of the gene 3 product from the gene 4 product by trypsin treatment. The gene 3 protein was resolved by electrophoresis of proteins in single-shelled particles (lane 1) or in purified double-shelled particles (lanes 2-5) analyzed directly (lane 2) or after treatment of double-shelled particles with 10, 1, or 0.1 pg of TPCKtrypsin/ml (lanes 3, 4, and 5, respectively). Trypsin treatment was performed for 30 min at 37”. Electrophoresis was in 6% polyacrylamide gels with 4% cross-linkage. The double arrowhead indicates the gene 3 product.
IDENTIFICATION
OF SAll
GENE 3 PRODUCT
31
FIG. 6. One-dimensional V8 protease mapping of the four high-molecular-weight proteins of rotavirus SAl 1. V8 protease digestion (0, 1, or 10 pglml) was used for one-dimensional peptide mapping to compare the relationship between the products of genome segments 1, 2, 3, and 4. The maps shown were from two parallel gels of the same composition that were run under the same electrophoretic conditions at the same time. The autoradiograms for the gene 1 and gene 2 products were from one gel exposed for different time periods; the autoradiograms for the gene 3 and gene 4 products were from a second gel. The numbers above the concentrations of the enzymes indicate the genome segments that code for the high-molecular-werght proteins.
age of two determinations from 3H-labeled viruses (Table 1). Data obtained by scintillation counting of gel slices or (optical scanning of autoradiograms gave consistent results. DISCUSSION Utilizing different systems of polyacrylamide gels, the gene 3 product was resolved from comigration with other proteins, and evidence was found to indicate that this product is a primary structural protein located in the core rshell of rotavirus particles. The TABLE 1 RELATIVE PROPORTIONS OFTHE POLYPEPTIDES IN SAll DOUBLE-SHELLED PARTICLES
Virion polypeptide” VP1 VP2 VP3 VP4 VP6 VP7
Molecular weight 125,000 94,000 88,OC’O 88,OC~O 41 ,OClO 34,000
Percentage of virion proteir?
Approximate molar ratioC
2 15 0.5 1.5 51 30
1 10 0.35 1 78 55
“The virion polypeptides are designated as proposed in this manuscript. VP3 is the protein product of gene segment 3, and VP4 is the prote n product of gene segment 4. b Determined by densitometer scanning and integration of the polypeptide peaks resolved following polyacrylamide gel electrophoresis of particles labeled vvith ‘H-labeled amino acrds. ’ Calculated by dividing the average percentage of the virion protein by the molecular weight of the corresponding polypeptide. The value obtained for VP1 was taken as 1.
failure of previous investigators to unequivocally identify the gene 3 product probably was due to the special migration properties of this protein on polyacrylamide gels. In this study, the gene 3 product of NCDV could be resolved on 6% polyacrylamide gels with 1% cross-linkage, and the gene 3 protein of SAl 1 could be separated from the gene 4 protein on 6% polyacrylamide gels with 4% cross-linkage. To resolve the structural proteins of rotavirus SAl 1, a 6% polyacrylamide gel with 4% cross-linkage gel system was optimal. Our results support the studies of Smith et al. (1980), who concluded that the gene 3 product is a structural protein located in single-shelled viral partcles. Our results also extend their studies by localizing this protein to the viral inner core particles. McCrae and McCorquodale (1982) also suggested that the gene 3 product was an inner shell structural protein that was initially synthesized as a larger precursor and was immediately processed. We did not identify any precursor in our studies. The following evidence supports our identification of the gene 3 product: (i) distinction of the gene 3 protein patterns in both parental viruses and reassortant viruses that contain reasserted gene 3 segments, (ii) in vitro translation of total mRNA in rabbit reticulocyte lysates, (iii) viral protein labeling at an early stage of infection, (iv) one-dimensional peptide mapping of high-molecular-weight proteins, and (v) the unique characteristics of being a core protein and being resistant to trypsin treatment of double-shelled particles easily differentiate the gene 3 product from the gene 4 product (VP3) with which it often comigrates. Utilizing the same gel systems reported in this study, we also
32
LIU, OFFIT, AND ESTES
resolved and identified the gene 3 products in rhesus rotavirus RRV and in single-gene 3 reassortants of RRV and SAl 1 (provided by Dr. R. F. Ramig; data not shown). From our results to date we conclude that the gene 3 product is a core protein. Ramig (1983) isolated 10 different groups of temperature-sensitive mutants of SA11, and subsequently mapped mutant 339 of group B to genome segment 3 (Gombold et a/., 1985). It is of interest that this mutant has shown reduced amounts of single-stranded RNA and no detectable dsRNA synthesized at the nonpermissive temperature (Ramig, 1983). This observation, taken together with the location of the gene 3 product, suggests that this protein might function as par-t of the virion-associated RNA polymerase. Since the product of genome segment 3 represents only a small molar ratio of the virion and mRNA of genome segment 3 is not translated well in the in vitro translation system programmed with total viral transcripts, gene cloning and subsequent in vitro expression should be helpful approaches to study the function of this protein. Although the term VP3 has been widely used for the gene 4 product, we propose that the genome segment 3 product be called VP3 and that the gene 4 product be named VP4 from now on. ACKNOWLEDGMENTS We thank Mr. Edward Calomenl and Dr. Tomoyuki Tanaka for their electron microscopic work to confirm the types of purified viral particles, Dr. Wai-Kit Chan for critical comments, and Dr. Robert F. Ramig for providing the SAl l/RRV reassortants. This work was supported by Grant DK-30144 from the National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases and by a grant from the Health Care Dtvision of Richardson-Vicks (Shelton, CT).
REFERENCES ARIAS, C. F., LOPEZ, S., and ESPEJO,R. T. (1982). Gene protein products of SAl 1 simian rotavirus genome. J. Viral. 41, 42-50. BICAN. P., COHEN, J., CHARPILIENNE,A., and SCHERRER, R. (1982). Purification and characterization of bovine rotavirus cores. /. Viral. 43,1113-1117. BONNER,W. M., and LASKEY, R. A. (1974). A film detection method for tritium-labelled proteins and nucleic actds in polyacrylamide gels. Eur. J. Biochem. 46, 83-88. CHAN. W.-K., PENARANDA, M. E., CRAWFORD,S. E., and ESTES, M. K.
(1986). Two glycoproteins are produced from the rotavlrus neutralizatlon gene. Virology 151, 243-252. CLEVELAND, D. W., FISCHER, S. G., KIRSCHNER,M. W., and LAEMMLI, U. K. (1977). Peptlde mapplng by limited proteolysls in sodtum dodecyl sulfate and analysts by gel electrophoresis. /. B/o/. Chem. 252,1102-1106. COHEN, J., LAPORTE,J., CHARPILIENNE,A., and SCHERRER,R. (1979). Activation of rotavirus RNA polymerase by calcium chelation. Arch. Virol. 60, 177-l 86. CUKOR, G.. and BLACKLOW,N. R. (1984). Human viral gastroenteritis. Microbial. Rev. 48, 157-l 79. ERICSON, B. L., GRAHAM, D. Y., MASON, B. B., and ESTES. M. K. (1982). Identification, synthesis, and modifications of simian rotavirus SAl 1 polypeptides in infected cells. J. Viral. 42, 825-839. ESTES. M. K., GRAHAM, D. Y., GER~A, C. P., and SMITH, E. M. (1979). Simian rotavlrus SAl 1 replication in cell cultures. J. Viral. 31, 810-815. ESTES, M. K., GRAHAM, D. Y., and MASON, B. B. (1981). Proteolytic enhancement of rotavirus infectivity: Molecular mechanisms. /. Viral. 39, 879-888. ESTES.M. K., PALMER, E. L., and OBIJESKI,J. F. (1983). Rotaviruses: A review. Cur. Top. Microbial. lmmunol. 105, 123-l 84. GOMBOLD, J. L., ESTES,M. K., and RAMIG, R. F. (1985). Assignment of simian rotavirus SAl 1 temperature-sensitive mutant groups B and E to genome segments. Virology 143, 309-320. GOMBOLD, J. L., and RAMIG, R. F. (1986). Analysis of reassortment of genome segments in mice mixedly infected with rotaviruses SAl 1 and RRV. J. Viroi. 57, 110-l 16. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227, 680-685. MAHIN, D. I., and LOFBERG,R. I. (1966). A simplified method of sample preparation for determination of tritium, carbon-l 4. or sulfur35 in blood or tissue by liquid scintillation counting. Anal. Bio&em. 16, 500-509. MASON, B. B., GRAHAM, D. Y., and ESTES, M. K. (1980). In vitro transcription and translation of simian rotavirus SAl 1 gene products. /. VifOl. 33, 11 11-l 121. MASON, B. B., GRAHAM, D. Y., and ESTES, M. K. (1983). Biochemical mapping of the simian rotavirus SAl 1 genome. /. Viral. 46, 413-423. MCCRAE, M. A., and MC~ORQLJODALE,J. G. (1982). The molecular biology of rotaviruses. II. Identification of the protein-coding assignments of calf rotavirus genome RNA species. Virology 117, 435-443. OFFIT, P. A., BLAVAT, G., GREENBERG,H. B., and CLARK, H. F. (1986). Molecular basis of rotavirus virulence: Role of gene segment 4. /. Virol. 57, 46-49. RAMIG, R. F. (1983). Genetic studies with simian rotavirus SAl 1. ln “Double-Stranded RNA Viruses” (R. W. Compans and D. H. L. Bishop, Eds.), pp. 321-327. Elsevier, New York. SMITH, M. L., LAZDINS, I., and HOLMES, I. H. (1980). Coding asslgnments of double-stranded RNA segments of SA 11 rotavirus established by in vitro translation. /. Viral. 33, 976-982.