The first complete genome sequence and genetic evolution analysis of bovine norovirus in Xinjiang, China

A total of 974 calf faecal samples were collected in Xinjiang, China, from July 2020 to July 2023. The seasons and regions of sample collection were clearly recorded. The calves from which samples were taken were under three months of age. The faecal samples were diluted with 1× phosphate-buffered saline solution at a ratio of 1:3. After being shaken on a vortex oscillator, the samples were subjected to freezing at −20°C and thawing three times. All animal procedures were conducted following the guidelines in document 2020025 approved by the Animal Ethics Committee of Xinjiang Agricultural University.

Viral RNA was extracted from the samples using the TRIzol/chloroform stratification method. The reverse transcription was carried out with a Hifair III 1st Strand cDNA Synthesis Kit (gDNA digester plus) (Yeasen, Wuhan, China). The primer pairs were designed according to an epidemic strain reference sequence in the GenBank database with accession number MN122335.1. The primer sequences used for detection are shown in the Supplementary Table S1.

The complete genome of BNoV was amplified using RT-PCR. The primer sequences used for the amplification of the BNoV complete genome designed by Primer Premier 5 (Premier Biosoft International, San Francisco, CA, USA) are shown in Supplementary Table S1. The amplified products were inserted into the pGEM-T vector (Promega, Madison, WI, USA) and sequenced by Sangon Biotech (Shanghai, China). The obtained BNoV sequence fragment results were assembled using SeqMan in DNAstar version 7 (DNASTAR, Madison, WI, USA).

The obtained gene sequences of BNoV were aligned with the reference sequences published in GenBank using the Clustal W algorithm. Subsequently, a phylogenetic tree was constructed using the maximum-likelihood method in MEGA version 7. The tree reliabilities were tested with 1,000 bootstrap replicates to yield a majority consensus tree.

The sequences were aligned using MEGA version 7, and multiple sequence alignment was performed by ESPript 3.0 (28).

A model was constructed using SWISS-MODEL with three-dimensional models of the VP1 proteins of BNoV (34). Sequence identity greater than 30% indicated successful model construction, and global model quality estimate values ranged from 0 to 1, with larger values indicating more reliable model quality. The three-dimensional model files were viewed using the Mol* 3D Viewer tool of the Research Collaboratory for Structural Bioinformatics Protein Database (30).

Infection with BNoV was detected by RT-PCR in 193 out of 974 calf faeces samples from six regions in Xinjiang (Table 1). The overall results showed that the infection rate of BNoV was 19.82% (193/974), and that positive cases were recorded in all six regions. The highest infection rate was noted in autumn (43.98%, 73/166) (Table 1).

The complete genome sequences of two BNoV strains from two regions were obtained with two replicates per region. One strain, from Tacheng in northern Xinjiang, was named Bo/XJ-TC/01/CHN, and the other strain, from Kashgar in southern Xinjiang, was named Bo/XJ-KS/02/CHN. Their complete genome sequences have been uploaded to the GenBank database and assigned accession numbers OM991839 and OM992315. The complete genome sequence length of OM991839 was 7,295 nucleotides, with a guanine/cytosine value of 57.5%, and that of OM992315 was 7,316 nucleotides, with a guanine/cytosine value of 57.4%.

The results of the phylogenetic tree showed that the obtained BNoV belonged to the GIII genogroup and GIII.2 genotype. Interestingly, the OM991839.1 Bo/XJ-TC/01/CHN strain from Tacheng and OM992315.1 Bo/XJ-KS/02/CHN strain from Kashgar belonged to the same branch. OM992315.1 and OM991839.1 clustered with three Chinese strains, the former being closest in affinity to the MZ573179.1 and MN480761.1 strains from China, and the latter being closest in affinity to the MN122335.1 strain from China. It is most distantly related to the Belgian FJ946859.1 strain and the British AY126468.2 strain (Fig. 1).

Fig. 1.

The nucleotide identity between the two BNoV strains was greater than 92%, and the amino acid homology was greater than 90%. The nucleotide identity and amino acid homology of genogroup GIII of BNoV was analysed, and Bo/XJ-TC/01/CHN was found to be closely related to MN122335.1, with a nucleotide identity of 93.48%. Further analysis showed that the ORF1 gene and VP1 gene of Bo/XJ-TC/01/CHN were most similar to those of MN122335.1, while the VP2 gene of this strain was most similar to that of MZ573179.1 (Table 2). The highest nucleotide identity between the other strain and a reference strain was between it and MZ573179.1, at 93.46%. The genes of Bo/XJ-KS/02/CHN which were most similar to their counterparts in MZ573179.1 were ORF1 and VP1, and the VP2 gene of Bo/XJ-KS/02/CHN was most similar to that of MK159169.1 (Table 3).

Unexpectedly, the amino acid homology between Bo/XJ-TC/01/CHN and MZ573179.1 was 97.6%. Further analysis showed that the genes of Bo/XJ-TC/01/CHN which had the highest amino acid homology with MZ573179.1 were ORF1 and VP2, whereas the VP1 gene of this strain had the highest homology with MN122335.1 (Table 2). In addition, the amino acid homology between Bo/XJ-KS/02/CHN and MZ573179.1 was 98.37%. In this instance the ORF1 gene was not one of the pair with the highest homology with the same genes in with MZ573179.1, but rather the VP1 and VP2 genes were that pair. The ORF1 gene of Bo/XJ-KS/02/CHN had the highest homology with this gene in strain MN122335.1 (Table 3).

Mutations in the amino acids of the norovirus VP1 protein are important factors in the evolutionary mechanism of norovirus (18). Therefore, the amino acid sequences of the Bo/XJ-TC/01/CHN and Bo/XJ-KS/02/CHN VP1 proteins were compared with the VP1 proteins in the respective reference sequences with the highest amino acid homology: MN122335.1 and MZ573179.1. Differences between the Bo/XJ-TC/01/CHN sequence and consensus sequence of the other subbranch (Bo/XJ-KS/02/CHN and MZ573179.1) were revealed at one site in the S domain, two sites in the P1 subdomain and nine sites in the P2 subdomain. Between the Bo/XJ-KS/02/CHN sequence and the consensus sequence of the other subbranch (Bo/XJ-TC/01/CHN and MN122335.1), there were one site of difference in the S domain, five sites in the P1 subdomain and four sites in the P2 subdomain. Interestingly, among all variations, 11 sites of difference in the same subbranch were found: 5 sites in the subbranch of Bo/XJ-TC/01/CHN and MN122335.1 and 6 sites in the subbranch of Bo/XJ-KS/02/CHN and MZ573179.1. The differences among the four strains are shown in detail in Fig. 2, which reveals the rapid evolution of the novel BNoV genocluster.

Fig. 2.

The possible three-dimensional structures of the structural proteins of the representative epidemic strains (MN122335.1 and MZ573179.1) were predicted and compared with those of the Bo/XJ-KS/02/CHN and Bo/XJ-TC/01/CHN strains. The superposition of Bo/XJ-KS/02/CHN and Bo/XJ-TC/01/CHN protein structures showed that the α-helix at 419–421 amino acids (aa) and 50–54 aa of Bo/XJ-TC/01/CHN had been replaced by a random coil (Fig. 3a).

Fig. 3.

The Bo/XJ-TC/01/CHN and MN122335.1 proteins had the same three-dimensional structures. Unique to Bo/XJ-KS/02/CHN was the random coil at 50–54 aa in the P1 subdomain (Fig. 3b). Comparing the VP1 structures of the Xinjiang strains with the VP1 structure of GIII.1 BNoV, there were significant differences in the P2 subdomain, which also verified the finding of a previous study (23) that the NoV genetic variation can cause differences in the P2 subdomain. (Fig. 3c). It was shown that amino acid mutations in the P structural domain affect the ability of the virus to bind to histo-blood group antigen, in addition to affecting the virus’ propensity to develop gastroenteritis in the host (when the mutations are in the P2 subdomain (4).