Non-structural enterotoxin (NsP4) Gene based Molecular Characterization of Caprine and Ovine Rotavirus A, India

University, Ludhiana 141004, Punjab, India. Abstract Rotavirus A (RVA) causes viral gastroenteritis in humans and animals, including calves, piglets, and foals. The current study reports the genetic characterization of the full-length enterotoxin gene, NSP4, from caprine and ovine species. Upon characterizing eight full-length NSP4 genes by sequencing, it was found that the four caprine and three ovine RVAs NSP4 genes are of E2 genotype and the sole ovine RVA isolate was found to be of E1 genotype. In the sequence and phyloanalysis of the NSP4 gene the seven E2 genotypes clustered with bovine, human, and caprine isolates from India and Bangladesh, respectively. The E1 genotype of ovine RVA was closer to human RVA isolate from India. The nucleotide per cent identity analysis revealed that all E2 genotype strains of caprine and ovine species ranged from 88.4% to 90.4% and it was found common to both the reference human RVA isolates DS-1 and AU-1. Whereas, the E1 genotype ovine strain clustered with human RVA isolates with 93.1% nucleotide per cent identity. The RVA strains circulating in caprine and ovine populations may share a common origin which is usually found in artiodactyl species because humans share a common dwelling with animals. Future studies are needed to confirm these findings of their relationship


INTRODUCTION
Rotavirus A (RVA) is among the leading causes of virus mediated gastroenteritis in humans and animals worldwide 1 . In livestock, RVA causes severe enteric infection in calves, piglets, and foals 2 . RVAs belong to the family Reoviridae which possesses 11 dsRNA segmented genome which encodes for six structural (VP1-4 and VP6-7) and six non-structural proteins (NSP1-6). On the basis of group-specific capsid protein VP6 gene, RVA is classified into nine species, namely A-I (RVA, RVB…. RVI). Of them, three species (RVA, RVB, & RVC) have been reported in small ruminants 3 . RVAs have been studied extensively for domesticated species like bovine, porcine and humans, but caprine and ovine species have remained neglected 4 . Apart from RVs other enteric viruses are being discovered and studied which are also associated with diarrhoeal problems in small ruminants [5][6][7] .
Among the different proteins of rotaviruses, NSP4 plays the role of viral enterotoxin, which has been described to induce diarrhoea in suckling mice 8 . This discovery led to an increase in scientific interest to understand the role of the NSP4 enterotoxin gene in inducing diarrhoea in different animals and human neonates. NSP4, consisting of 175 amino acids, is an endoplasmic reticulum (ER) specific glycoprotein that has a role in viral pathogenesis and morphogenesis 9 . It has three hydrophobic domains, viz -H1 domain (7-21 aa), H2 domain (29-47 aa), and H3 domain (67-85 aa), and a coiled alpha helix domain stretches from 95-137 aa 1 . The amino-terminus region is located from 1-44 aa inside the lumen of ER, whereas 45-175 aa stretch towards carboxyl-terminus contains the cytoplasmic tail (CT) known for key biological functions related to the protein, which includes a variation of Ca 2+ homeostasis by discharging Ca 2+ from the ER 10-12 , Ca 2+ and VP4 binding 1 , permeabilization into the plasma membrane 13 , recruiting the double-layered particle to the lumen of ER to be matured into triple-layered particle 14,15 , and induction of diarrhoea in neonatal mice 8,16,17 .
So far, 27 E genotypes (E1-E27) have been identified in different animal and human species established on a 85% identity cut-off value for the NSP4 gene 18,19 . Genotypes E1, E2, and E3 have been reported more commonly 18 . Usually, all the NSP4 genotypes tend to segregate according to the RVA host species 20 suggesting the dominance of specific genotypes for a particular species. To date, the RVA NSP4 and other genes of RVs have been analysed in more detail for human [21][22][23][24] , bovine 25 , porcine 26 , and avian species 27 . However, small ruminants remain the least studied species worldwide. RVA infection has been reported from all over the world, but reports from India are scanty, especially from caprine and ovine. In this report, we characterize caprine and ovine RVAs based on full-length enterotoxin gene NSP4 (n=8) to identify their genotype difference and sequence variability.

MATERIAlS AND METhODS Sampling, processing, and viral RNA extraction
The collection of samples was done from goat kids and lambs that were below six months of age, irrespective of signs of diarrhoea. Specimens were processed by preparing a 10% suspension in 1X PBS and stored further at -20°C for long time storage. The viral RNA was isolated using QIAzol Lysis Reagent (QIAGEN, USA) was suspended in 20 µl of nuclease-free water (NFW) (QIAGEN, Hilden, Germany), and kept at -20°C until further use.

cDNA preparation by Reverse Transcription
Initially, 15 µl (100-200 ng) of viral RNA was mixed with 1.0 µl (100 ng/µl) of random hexamer (QIAGEN, Hilden Germany) and 1 µl of DMSO (MB grade, Merck, Darmstadt, Germany) followed by incubation at 95°C for 5 min and 10 min snap chilling on ice. Simultaneously, 5X RT buffer, 1µl of 10mM dNTPs (QIAGEN, Hilden Germany), 1µl of (40U) RNase Inhibitor Ribolock™ (Fermentas, Vilnius, Lithuania), and 1µl of M-MLV RT enzyme (Promega, Madison, USA) were mixed accordingly and added to the snap chilled viral RNA mix making the final volume of the reaction to 25 µl and kept at 37°C for 60 min. The RT enzyme was inactivated by incubating the reaction at 80°C for 5 min. Thereafter, the cDNA prepared was employed in diagnosing RVA using custom-designed primers.

Molecular detection of rotavirus and NSP4 amplification
The RVA presence in the fecal samples was detected employing a degenerate primer pair based on the VP6 gene by the RT-PCR, as described earlier 28 . Consecutively, the positive samples were subjected to PCR amplification for the full-length NSP4 enterotoxin gene. We used published primers to amplify the full-length gene (742 bp) as described earlier 25 . The primer name, sequence and the amplicon size of the VP6 and NSP4 gene used in the diagnosis and characterization of RVA isolates are given in Table  1. Following the amplification of desired products, the amplicons thus generated were cloned and sequenced in pDrive TA cloning vector according to a previous protocol 29 . The quality of each sequence thus generated was analyzed using the Finch TV (Geospiza, Inc. UK) software version 1.4, and the sequence identity was verified using the BLAST software in GenBank (http://blast.ncbi.nlm.nih. gov/Blast.cgi).
Eight isolates showing positive results in PCR were cloned and sequenced which was outsourced to M/s Eurofins Genomics, Bangalore. Out of the 08 isolates, four samples each from caprine and ovine origin were characterized. The sequence database generated on NSP4 genes was submitted to NCBI GenBank nucleotide repository under accession numbers MT998256 to MT998263. The strain name, place of collection, species, length, and accession numbers is given in (Table 2).

Phylogenetic analysis
Phylogenetic analysis was performed for RVA NSP4 gene with different isolates from India and worldwide. A total of 08 strains of the current study were subjected to phylogenetic analyses with 57 sequences from different species all over the world retrieved from the NCBI database (Fig.  1). The analysis was done using the MEGA 7.0 software 30 . In both cases, the evolutionary history was obtained using the Maximum Likelihood (ML) method based on the Tamura-3 parameter model for computing the distances with the support of 1000 bootstrap replicates. A separate gamma distribution was used to model evolutionary frequency differences between the sites.

Per cent identity calculation and comparison of Inter-Species Variable Domain (ISVD) region
For per cent identity calculation, we analyzed current study sequences in MegAlign software packaged in DNASTAR where sequences were aligned by ClustalW, and individual distances among them were calculated. Furthermore, the rotavirus genotyping web-based tool hosted by ViPRBRC for RVA classification (https://www. viprbrc.org) was used to infer the E genotypes of all NSP4 sequences. The ORFs of all individual genes  were translated theoretically using the EditSeq tool of DNASTAR. A comparison of the pre-defined interspecies variable domain (ISVD) region was made with the assistance of the Protean tool the DNASTAR software package.

Phylogenetic analysis
The phylogenetic analysis of 8 NSP4 genes of the current study sequence revealed 7 (4 caprine and 3 ovines) of E2 genotype and 1 (ovine) E1 genotype (Fig. 1). This observation was also confirmed by the web-based rotavirus genotyping tool of ViPRBC. As per the previous literature E1 genotype for ovine RVA strain has never been reported in India earlier. In phylogenetic analysis, all the E2 type strains of the current study clustered inside a major clade comprising different E2 genotype isolates from worldwide. Out of the four caprine strains of the current study, threebranched alongside bovine RVA isolates from India whereas one appeared alongside a Bangladeshi

Per cent identity analysis
The per cent identity analysis was done with the selected reference NSP4 genotypes which have been reported in humans and domesticated artiodactyl type species like bovine and porcine. The nucleotide per cent identity analysis revealed that all the E2 genotype strains of caprine and ovine species of the current study showed the highest (88.4% to 90.4%) and equal per cent identity with two reference human RVA isolates DS-1 (AF174305) and AU-1 (D89873) ( Table 3). Ovine strain L-RAJ54 showed the highest nucleotide similarity percentage (91.3%) with reference human RVA E1 genotype isolate Wa (K02032). The current study caprine and ovine strains shared 87.4 to 93.8% and 77.7 to 95.1% of nucleotide similarity among themselves, respectively. This showed that ovine samples were more divergent among them with respect to caprine samples.

Amino acid sequence comparison of RVA NSP4 ISVD
The NSP4 strains were translated theoretically and aligned using the MegAlign software to observe any crucial changes in the ISVD region of NSP4 C-terminal (Fig. 2). The observations revealed that E2 genotype strains of current study strains contain more variation in comparison to E1 type when aligned along with reference isolates. E2 type strains showed some number of varied amino acid residues in the ISVD region (131-141 aa) whereas the single E1 type ovine strain L-RAJ54 showed variance at residue position 131, 134, 137, and 141 only.

DISCUSSION
RVAs are known to cause diarrhoea in human neonates 31 , calves, and piglets 32 , but their aetiology has not been studied well in small ruminants. The current study reports the presence of RVAs in the caprine and ovine population of India along with the characterization of its major enterotoxin gene for the first time in India. However, few reports of prevalence emerged from India in the past for small ruminants 33-37 but these reports failed to provide data on sequencebased characterization of RVA in small ruminants. Therefore, this study was designed to decipher the circulating potential of RVA in the small ruminant population of India as well as to characterize them based on sequence and phylogeny.
The phylogenetic analysis revealed a diverse population of ovine and caprine RVA based on the full-length NSP4 gene where in seven E2 type strains of the present study branched into different sub-clusters within the major E2 clade. One ovine RVA sample showed E1 genotype specificity which was found closer to a human RVA strain ChD148 (GenBank record only) of India. These observations point towards the diverse origin of RVAs in the Indian small ruminant population. Closeness to human and bovine type RVA isolates also suggest the ongoing interspecies transmission occurring between these ungulates and other mammals including humans. We aligned and observed the NSP4 amino acid sequences of the present study and found that strains possess a significant number of amino acid residue changes in the interspecies variable domain of the NSP4 C-terminus. This observation also supports the regarding the sequence and phylogenetic variation among small ruminant RVAs. Nevertheless, it will be interesting to observe whether these amino acid changes will lead to any alteration in the infectivity of these viral strains.
Overall the data presented in the study hypothesize that RVA strains circulating in these small ruminant populations may share a common origin which is usually found in artiodactyl species, but future studies are needed to confirm these findings. This hypothesis is reasonable because humans, large and small ruminants share a common dwelling place. The data from the prevalence also warrants strict surveillance measures to be taken in the future to know about the circulating genotypes of small ruminant RVAs.

ACKNOWlEDgMENTS
Authors acknowledge and express thanks to their respective Institutes and Universities. Help of Shanmuganathan in sampling is acknowledged. YSM acknowledges the Education Division, ICAR, Govt. of India for National Fellowship.