Analysis of genetic factors related to preferential selection of the NSP1 gene segment observed in mixed infection and multiple passage of rotaviruses

Summary. Reassortment is one of the major evolutionary mechanisms of the rotavirus genome. Preferential selection (assortment) of the NSP1 gene segment from either of the parental viruses after coinfection of these viruses has been reported as a notable finding in reassortment. To analyze genetic factors which are associated with preferential selection of the rotavirus NSP1 gene segment into progeny viruses, mixed infection and multiple passages were performed using two panels of rotaviruses, i.e., bovine rotavirus A5 clones, and simian rotavirus SA11 and five strains of SA11-based single NSP1 gene-substitution reassortants. In the first experiment, three A5 clones (A5-10, A5-13, and A5-16) that had genetically distinct NSP1 genes in the same genetic background were used. In coinfection of these A5 clones, it was noted that the A5-10 NSP1 gene, which encodes an incomplete protein product due to presence of a nonsense codon at an unusual position, was selected more preferentially than the A5-13 NSP1 gene with intact length and structure. The A5-16 NSP1 gene, with a deletion of 500 bp, was least efficiently selected. In the second experiment, we prepared two reassortants, SOF and SRF, which have NSP1 genes from rotavirus strains OSU and RRV, respectively, in the genetic background of SA11, which were used together with previously prepared reassortants SKF, SDF, and SNF, which had NSP1 genes from strains KU, DS1, and K9, respectively. Among the 6 NSP1 genes analyzed, the NSP1 gene from SKF was most preferentially selected, followed by SNF, SOF, SDF, SA11, and SRF, in that order. Although SOF exhibited less growth efficacy than SA11, the growth rates of other reassortants were similar to that of SA11. These findings suggest that for the occurrence of preferential selection of the NSP1 gene, production of the intact NSP1 protein may not be involved, but the presence of intact length of the NSP1 gene may be required. Furthermore, it was also found that genetic similarity based on primary structure of this gene is not related to the selectivity of the NSP1 gene.     

Analysis on reassortment of rotavirus NSP1 genes lacking coding region for cysteine-rich zinc finger motif

Summary.  Rotavirus clones A5–10 and A5–16 isolated from a bovine rotavirus strain A5 possess NSP1 gene which has a point mutation generating a nonsense codon and a 500 base-deletion, respectively. As a result, the two A5 clones encode truncated NSP1 product which lacks cysteine-rich region forming zinc finger motif. In order to analyze reassortment of these mutated NSP1 gene with RNA segments from heterologous strains, we investigated a number of reassortant clones derived from coinfection with either A5–10, A5–16 or a reference strain A5–13 (possessing intact NSP1 gene) and either simian rotavirus SA11 or human rotavirus KU. In coinfection with SA11 and A5–13, selection rates of A5–13 segments in reassortants ranged approximately from 20 to 70% (46% for NSP1 gene). In contrast, in the reassortment between SA11 and A5–10 or between SA11 and A5–16, selection rates of NSP1 gene from A5–10 and A5–16 were only 1% (one clone) and 0%, respectively. In reassortants from crosses KU × A5-clones, selection rate of A5–13 NSP1 gene decreased to 15%, while 11 reassortants with A5–10 NSP1 gene (31%) and one reassortant with A5–16 NSP1 gene (2%) were isolated. Reassortants with A5–10 NSP1 possessed a single gene (segment 9 or 11) from KU in the genetic background of A5–10. One reassortant clone (cl-55) with A5–16 NSP1 gene possessed KU gene segments 3, 4, and 8–11. When single-step growth curves were compared, the reassortant cl-55 showed almost identical growth curve to that of KU, while KU showed a better replication than A5–16. These results indicated that although A5–10 or A5–16 NSP1 gene encoding the truncated NSP1 is selected into reassortants much less efficiently than normal NSP1 gene, the reassortants with the mutated NSP1 gene and RNA segments from heterologous strains normally replicated in cultured cells. Thus, cysteine-rich region of NSP1 was not considered essential for genome segment reassortment with heterologous virus. Accepted August 29, 1998 Received July 6, 1998     

G (VP7) serotype-dependent preferential VP7 gene selection detected in the genetic background of simian rotavirus SA11

Summary We previously found the preferential selection of VP7 gene from a parent rotavirus strain SA11 with G serotype 3 (G3) in the sequential passages after mixed infection of simian rotavirus SA11 and SA11-human rotavirus single-VP7 gene-substitution reassortants with G1, G2, or G4 specificity. However, it has not been known whether or not VP7 genes derived from other strains with G3 specificity (G3-VP7 gene) are preferentially selected in the genetic background of SA11. To address this question, mixed infections followed by multiple passages were performed with a reassortant SA11-L2/KU-R1 (SKR1) (which possesses VP7 gene derived from G1 human rotavirus KU and other 10 genes of SA11 origin) and one of the five G3-rotaviruses, RRV, K9, YO, AK35, and S3. After the 10th passage, selection rates of SA11-L2/KU-R1 gene 9 (G1-VP7 gene) and gene 5 (NSP1 gene) reduced considerably (0 to 20.4%) in the clones obtained from all the coinfection experiments, while all or some of other segments were preferentially selected from SKR1 depending on the pairs of coinfection. When viral growth kinetics was examined, SKR1 exhibited better growth and reached a higher titer than any G3 viruses. Although the generated reassortants with VP7 gene and NSP1 gene derived from G3 viruses showed almost similar growth kinetics to that of SKR1 during the first 20 h of replication, the titers of these reassortants were higher than that of SKR1 after 36h postinfection. The results obtained in this study suggested that G3-VP7 gene is functionally more adapted to the genetic background of SA11.     

NSP5 phosphorylation regulates the fate of viral mRNA in rotavirus infected cells

Summary.  Elucidation of the function of the non-structural rotavirus proteins during infection is difficult in the absence of a reverse genetic system. To study the role of NSP5, nonstructural phosphoprotein NSP5, we constructed a reassortant strain (SACC11) in the SA11 background that harbours a heterologous segment 11 encoding a variant protein (h-NSP5). Cells infected by SACC11 produced viral polypeptides at earlier times than SA11 infected cells while showing less accumulation of genomic dsRNA. These changes suggested that NSP5 might direct viral messenger RNA to protein synthesis or genome replication. Distinct patterns of proteins were shown to form complexes with NSP5 in co-immunoprecipitation studies with SA11 and SACC11 infected cells. Recombinant h-NSP5 from either bacteria or eucaryotic cells migrated faster in PAGE suggesting that it was hypophosphorylated. Indeed, the kinase inhibitor H-7 enhanced translation of viral proteins in SA11 but not SACC11 infected cells suggesting that NSP5 function in the regulation of the fate of viral positive strand RNA is mediated by phosphorylation. Present address: LaboRetro, INSERM U412, Ecole Normale Supérieure de Lyon, 46 allée d’Italie 46, F-69364, Lyon, France. Received October 15, 2001; accepted May 20, 2002     

Isolation and molecular characterization of a naturally occurring non-structural protein 5 (NSP5) gene reassortant of group A rotavirus of serotype G2P[4] with a long RNA pattern

We report here two unusual strains of group A rotavirus, AU85 and AU102, isolated from children with diarrhea. These strains showed an unusual combination of serotype G2 and a long RNA pattern. RNA-RNA hybridization assays showed that these strains are reassortants in which a single genome segment 11 (the NSP5 gene) was derived from a Wa genogroup strain, while other 10 genome segments from a DS-1 genogroup strain. Phylogenetic analysis showed that the NSP5 gene of strain AU85 did not form cluster with Wa strain, while it belonged to the cluster of YM and other porcine strains. Phylogenetic analysis also showed that NSP5 and VP7 genes of AU85 were derived from the rotavirus circulating in the area. Both co-electrophoresis and RNA-RNA hybridization showed that AU85 and AU102 are identical strains. Moreover, the nucleotide sequence comparison between these two strains revealed that they had 100% identical NSP4, NSP5, and VP7 genes. These results suggest that AU85 was a reassortant formed relatively recently between rotaviruses belonging to the Wa and the DS-1 genogroup.     

Identical rearrangement of NSP3 genes found in three independently isolated virus clones derived from mixed infection and multiple passages of Rotaviruses

Three rotavirus variants with a rearranged RNA segment derived from the NSP3 gene were isolated in three independent experiments of coinfection and multiple passages of simian rotavirus strain SA11 and single-VP7-gene- or NSP1-gene-substitution reassortants having genetic background of SA11. Sequence analysis indicated that the three rearranged NSP3 genes had almost identical sequences and genomic structures organized by partial duplication of the open reading frame in a head-to-tail orientation following the termination codon. The junction site of the original NSP3 gene (first copy) and the duplicated portion (second copy) was identical among the three rearranged genes, while a direct repeat, i.e., a homologous sequence between the first copy and second template for duplication, typically located at the junction site, was not detected. However, short similar sequences were present at the end of the first copy and beginning of the second copy. These findings suggest that rearrangement of the NSP3 gene may occur at a certain preferential site which is related to sequence similarity between 3′-untranslated region and a region near the 5′-end of ORF.     

Serotypic characterization of outer capsid spike protein VP4 of vervet monkey rotavirus SA11 strain

Summary.  The vervet monkey rotavirus SA11, a prototype strain of group A rotaviruses, has been shown to possess VP7 serotype 3 specificity but its neutralization specificity with regard to the other outer capsid protein VP4 has not been elucidated. We thus determined its VP4 specificity by two-way cross-neutralization with guinea pig antiserum prepared with a single gene substitution reassortant that had only the VP4-encoding gene from the simian rotavirus SA11 strain and remaining ten genes from human rotavirus DS-1 strain (G serotype 2). The SA11 VP4 was related antigenically in a one-way fashion to rhesus monkey rotavirus MMU18006 VP4 (a P5B strain) and marginally to human and canine rotavirus VP4s with P serotype 5A specificity. In addition, the SA11 VP4 was shown to be distinct antigenically from those of other known P serotypes (1–4, and 6–11) as well as those of uncharacterized equine, lapine, and avian rotavirus strains. The SA11 VP4 is thus proposed for classification as a P5B serotype. Received September 2, 1997 Accepted January 8, 1998     

Assignment of simian rotavirus SA11 temperature-sensitive mutant groups B and E to genome segments

Recombinant (reassortant) viruses were selected from crosses between temperature-sensitive (ts) mutants of simian rotavirus SA11 and wild-type human rotavirus Wa. The double-stranded genome RNAs of the reassortants were examined by electrophoresis in Tris-glycine-buffered polyacrylamide gels and by dot hybridization with a cloned DNA probe for genome segment 2. Analysis of replacements of genome segments in the reassortants allowed construction of a map correlating genome segments providing functions interchangeable between SA11 and Wa. The reassortants revealed a functional correspondence in order of increasing electrophoretic mobility of genome segments. Analysis of the parental origin of genome segments in ts+ SA11/Wa reassortants derived from the crosses SA11 tsB(339) X Wa and SA11 tsE(1400) X Wa revealed that the group B lesion of tsB(339) was located on genome segment 3 and the group E lesion of tsE(1400) was on segment 8.     

Determinants of rotavirus stability and density during CsCl purification.

The stability of rotavirus infectivity during CsCl gradient purification and subsequent storage was examined using our standard SA11 wild type (SA11-Cl3), the SA11 4F variant (SA11-4F), bovine rotavirus B223, and a panel of bi- and triparental reassortants derived from these parental viruses. Viral stability was determined by the recovery of infectivity at each step during a standard CsCl purification protocol. SA11-4F was the most stable parent (91-93% recovery), SA11-Cl3 had intermediate stability (10-21% recovery), and B223 was least stable (0.5-7% recovery). Among the reassortants, the recovery varied from 0.5 to 88.6% of the initial infectivity and was determined primarily by the parental origin of genome segment 4. The greatest loss of infectivity occurred during Freon extraction, with smaller losses during the CsCl gradient, and the smallest loss during the virus pelleting step. Comparison of the stability of viruses grown in the presence or absence of exogenous trypsin revealed that, in general, viruses grown in the absence of trypsin were more stable during purification. During 4-5 months storage at 4 degrees, the differences in stability of parental and reassortant viruses were not as dramatic as during purification and were not significantly affected by the presence or absence of trypsin during growth. However, survival during storage was as low as 4% and as high as 100% and was also primarily dependent on the parental origin of genome segment 4. It was noted that bovine rotavirus B223 had higher density in CsCl than either SA11-Cl3 or SA11-4F. The observation of heterogeneity in density was investigated using reassortants. These results indicated that all reassortants had intermediate density and suggested that physical interactions among the structural proteins were responsible for the heterogeneity in density. The possible roles of viral structural proteins in rotavirus stability and the relationship between the stability and the density are discussed.     

Preferential selection of VP7 gene from a parent rotavirus strain (SA11) in sequential passages after mixed infection with SA11 and SA11-human rotavirus single-VP7 gene-substitution reassortants

Summary We studied the competitive growth among SA11-L2(G3) and its single-human VP7 gene-substitution reassortants SA11-L2/KU-R1(G1) and SA11-L2/DS1-R1(G2), which have the genetic background of SA11-L2, during sequential passages after mixed infection. When the same infectious units (m.o.i. of 5 p.f.u./cell) of SA11-L2 and a reassortant SA11-L2/KU-R1 were inoculated onto and passaged in MA104 cells, 88% of the virus clones isolated from the culture fluid at the 3rd passage belonged to G3, and all the clones from the 10th passage had G3 specificity. Even when SA11-L2/KU-R1 with titer 10 times higher than that of SA11-L2 was used in the coinfection, the predominance of clones with G3-VP7 was observed. Although G2 clones slightly surpassed G1 clones in number in the mixed culture of SA11-L2/KU-R1 and SA11-L2/DS1-R1, G3 clones predominated in the virus progeny from a mixed culture infected with the same titers of SA11-L2, SA11-L2/KU-R1, and SA11-L2/DS1-R1. However, no significant difference in viral growth was detected among SA11-L2 and the two reassortants.     

Preparation and characterization of a neutralizing monoclonal antibody directed to VP4 of rotavirus strain K8 which has unique VP4 neutralization epitopes

Summary For selecting the neutralizing monoclonal antibodies (N-MAbs) directed to VP4 of rotavirus strain K8, which has unique VP4 neutralization epitopes, we prepared several reassortant viruses by mixed infection of two different strains K8 (serotype 1) and P (serotype 3) in vitro: three reassortant clones having VP4 of K8 and VP7 of P and four clones having VP4 of P and VP7 of K8. By using these reassortants in screening hybridomas, a N-MAb (K8-2C12) directed to strain K8-specific VP4 was obtained. The MAb K8-2C12 neutralized only K8 when tested against numerous strains of different serotypes, while in enzyme-linked immunosorbent assay this MAb reacted also with simian rotavirus SA11 (serotype 3), bovine rotavirus NCDV (serotype 6), and human rotavirus (HRV) strain 69M (serotype 8). Neutralization-resistant mutants of K8 were selected by the K8-2C12 antibody and VP4 amino acid sequences of the mutants were determined. Single amino acid substitution was detected in the three mutant clones at position 394, which is included in the major cross-reactive neutralization region identified in other rotaviruses.     

Genome heterogeneity of SA11 rotavirus due to reassortment with "O" agent

Derivatives of the rotavirus SA11-H96 strain, isolated in 1958 from an overtly healthy vervet monkey, have been used extensively to probe the viral life cycle. To gain insight into the phenotypic and genotypic differences among SA11 isolates, we sequenced the segmented double-stranded RNA genomes of SA11-H96 (P5B[2]:G3), two SA11-4F-like viruses (P6[1]:G3), two SA11-4F-like viruses with gene 5 rearrangements, and relevant segments of SA11 temperature-sensitive mutants and the "O" (Offal) agent (P6[1]:G8), a rotavirus isolated in 1965 from abattoir waste. This analysis indicates that the only complete genomic sequence previously reported for SA11 (Both) is instead that of a reassortant, originating like the SA11-4F-like viruses, from the introduction of an "O" agent gene into the SA11 genetic background. These results, combined with identification of mutations that correlate with altered growth properties and ts phenotype, emphasize the importance of considering segment origin and sequence variation in interpreting experimental outcomes with SA11 strains.     

Culture adaptation and characterization of group A rotaviruses causing diarrheal illnesses in Bangladesh from 1985 to 1986.

Group A rotaviruses collected between 1985 and 1986 during comprehensive surveillance of treated diarrheal episodes occurring in a rural Bangladesh population were culture adapted and characterized by electropherotype, serotype, and subgroup. Of 454 episodes of rotavirus-associated diarrhea, rotaviruses were culture adapted from 381 (84%), and 335 contained 11 electrophoretically identical segments in unpassaged and cultured preparations. These 335 comprised 69 different electropherotypes with between 1 (32 isolates) and 79 representatives. The persistence of specific rotavirus strains within the study population, as defined by the detection of viruses with particular electropherotypes, was generally limited to a period of only a few months. All 335 isolates were serotyped by neutralization with hyperimmune antisera to prototype rotavirus strains representative of serotypes 1 to 4, i.e., Wa, DS-1, P, and ST-3. It was found that 80, 48, 119, and 88 isolates belonged to serotypes 1 to 4, respectively. The concentrations of hyperimmune antisera required to neutralize these isolates, however, were at least threefold greater than those needed to neutralize the homologous strains. Therefore, the isolates appeared to have altered neutralization epitopes from their prototype strains. Furthermore, the serotype 4 isolates were consistently shown to be much more closely related to the serotype 4B VA70 strain than the serotype 4A ST-3 strain. All but two isolates identified as serotypes 1, 3, or 4 had long electropherotypes and were subgroup II, and all but one serotype 2 isolate were subgroup I and had short electropherotypes. The three disparate strains appeared to be genetic reassortants. Evidence is presented that dual infections required for reassortant formation were not uncommon. Thus, formation of multiple reassortants may have been a cause for the observed rapid shift in viral strains within the study population.     

Full genomic analysis of a porcine–bovine reassortant G4P[6] rotavirus strain R479 isolated from an infant in China

During the 2004 surveillance of rotaviruses in Wuhan, China, a G4P[6] rotavirus strain R479 was isolated from a stool specimen collected from a 2‐year‐old child with diarrhea. The strain R479 had an uncommon subgroup specificity I + II, and analysis of the VP6 gene suggested that it was related to porcine rotaviruses. In the present study, full‐length nucleotide sequences of all the RNA segments of R479 were determined and analyzed phylogenetically to identify the origin of individual RNA segments. According to the rotavirus genotyping system based on 11 RNA segments, the genotype of R479 was expressed as G4‐P[6]‐I5‐R1‐C1‐M1‐A1‐N1‐T7‐E1‐H1. This genotype includes the porcine‐like VP6 genotype (I5) and bovine‐like NSP3 genotype (T7). Phylogenetic analysis revealed that R479 genes encoding VP1, VP2, VP3, VP6, VP7, VP8*, NSP1, NSP4, and NSP5 were more closely related to those of porcine rotaviruses than human or other animal rotaviruses. In contrast, it was remarkable that the NSP3 gene of R479 was genetically closely related to only a bovine rotavirus strain UK. The NSP2 gene of R479 was also unique and clustered with only the G5P[8] human strain IAL28 and G3P[24] simian strain TUCH. These results suggested that R479 may be a reassortant virus having the NSP3 gene from a bovine rotavirus in the genetic background of a porcine rotavirus, with an NSP2 gene related to the porcine‐human reassortant strain IAL28. To our knowledge, R479 is the first porcine–bovine reassortant rotavirus isolated from a human. J. Med. Virol. 82:1094–1102, 2010. © 2010 Wiley‐Liss, Inc.     

Mutations selected in rotavirus enterotoxin NSP4 depend on the context of its expression

The rotavirus NSP4 protein is cytotoxic when transiently expressed in cells and is capable of inducing secretory diarrhea in neonatal mice. NSP4 consists of 175 amino acids, and sequences important for its toxic effects have been mapped to the carboxy-terminal half of the protein. In this report, we compared NSP4-encoding nucleotide sequences recovered from cell lines engineered to express NSP4 from human rotavirus strain Wa with NSP4 sequences recovered from cells persistently infected with either Wa or simian rotavirus strain SA11. In cells stably transfected with Wa NSP4, we found that proline(138) was changed to either serine or threonine. However, in cells persistently infected with SA11, we found that phenylalanine(33) was changed to leucine, and in cells persistently infected with Wa, no changes were observed in NSP4. Expression of Wa NSP4 in Caco-2 cells resulted in increased cell-doubling times and decreased cell viability in comparison to cells expressing NSP4-serine(138) or NSP4-threonine(138). This result suggests that sequence polymorphism at residue 138 in Wa NSP4 influences the cytotoxicity of the protein. Therefore, mutations in the carboxy-terminal half of NSP4 are selected when NSP4 is expressed in cells in the absence of other viral proteins, but not in the context of viral replication. These findings suggest that cytotoxic functions of NSP4 are not operant during natural rotavirus infection.     

Whole-genome consensus sequence analysis of a South African rotavirus SA11 sample reveals a mixed infection with two close derivatives of the SA11-H96 strain

Whole-genome, sequence-independent amplification and 454^® pyrosequencing of a rotavirus SA11 cell culture sample with an unknown passage history yielded consensus sequences of twelve complete genome segments. Two distinct sequences for genome segment 8 (encoding NSP2) were present, indicating a mixed infection with two rotavirus SA11 strains. The genotypes of the viruses were G3-P[2]-I2-R2-C5-M5-A5-Nx-T5-E2-H5, where x was either 5 or 2. The strains were named RVA/Simian-tc/ZAF/SA11-N5/1958/G3P[2] and RVA/Simian-tc/ZAF/SA11-N2/1958/G3P[2]. The genotype (N2) and sequence of genome segment 8 of RVA/Simian-tc/ZAF/SA11-N2/1958/G3P[2] were identical to that of the bovine rotavirus O agent. Five novel amino acids were detected in minor population variants of three genome segments. Genome segment 1 (VP1) has a high nucleotide substitution rate, but the substitutions are synonymous. Distance matrices and Bayesian molecular clock phylogenetics showed that SA11-N2 is a reassortant containing genome segment 8 from the O agent, whereas SA11-N5 is a very close derivative of the prototype SA11-H96.     

Characterization of a rotavirus rearranged gene 11 by gene reassortment

Summary.  The effect of replacement of gene 11 of rotavirus SA-11 by a gene carrying a head to tail duplication obtained from a swine rotavirus strain was studied. The swine rotavirus strain with a duplicated gene (CC86) exhibits both a phenotype that allows to overgrow other viral strains when coinfected and an increased plaque size when plated in both CV-1 and MA-104 monkey kidney cells. Using reassortment methods the duplicated gene of the swine rotavirus was introduced into the SA-11 virus, replacing the regular gene 11. The reassorted strain was characterized to find out the origin of each of the other viral gene segments. Based on electrophoretic mobilities segments 1, 2, 3, 5, 7, 8 and 10 were identified as of SA-11. The SA-11 origin of the segments 4, 6 and 9 was confirmed by neutralization with polyclonal and monoclonal antibodies and by ELISA. The results suggest that the new reassortant virus was a monoreassortant carrying SA-11 genes except the duplicated gene originated from the swine virus CC86. The ability to in vivo replicate and to synthesize viral proteins was identical in the reassorted virus and the parental strains. Sequence analysis indicates that the new phenotype does not originate in the duplication of gene 11 but possibly from mutations in the coding region of NSP5 gene that may result in different phosphorylation patterns of the protein. Received January 30, 1998 Accepted April 6, 1998     

Humoral immune response to rotavirus NSP4 enterotoxin in Spanish children

The rotavirus non-structural protein 4 (NSP4) has been shown to play a crucial role in rotavirus-induced diarrhea, acting as a viral enterotoxin. It has also been demonstrated that antibody to NSP4 can reduce the severity of rotavirus-induced diarrhea in newborn mice. Two recombinant baculoviruses, expressing the NSP4 protein from the SA11 and Wa rotavirus strains, genotypes A and B, respectively, were used to produce and purify these glycoproteins, which were applied as antigen in an enzyme-linked immunosorbent assay (ELISA) to test the specific antibody response to NSP4 in human sera. Serum samples from 30 children convalescing from a rotavirus infection, from 54 healthy children under 5-years-old, and from 49 adults were tested to determine the presence of antibodies to the viral enterotoxin and to rotavirus structural proteins. Seventy percent of the sera from rotavirus-infected children showed an IgG antibody response to either one or both NSP4 proteins used in this study, although the response was weak. However, IgG antibodies towards either one or both NSP4 proteins were only detected in 26% of the non-convalescent healthy children and in only 18% of the adults. No serum IgA antibodies towards NSP4 were found in this study. IgG antibody recognition of the NSP4 protein from the SA11 and Wa rotavirus strains was not always heterotypic.