Use of sequence analysis of the VP4 gene to classify recent Vietnamese rotavirus isolates

Introduction

Group A rotaviruses are members of the family Reoviridae and are a major cause of acute gastroenteritis in infants and young children. Rotaviruses cause c. 22% of childhood hospitalisations associated with diarrhoea, and are responsible worldwide for c. 611 000 childhood deaths annually [1]. The rotavirus genome consists of 11 segments of double-stranded RNA enclosed in a triple-layered capsid. The outer capsid layer is composed of two proteins: VP7, which defines G types (derived from glycoprotein); and VP4, which defines P types (derived from protease-sensitive protein) [2]. At least 15 different G genotypes and 27 P genotypes have been established, based on sequence analysis of the VP7 and VP4 genes, respectively [2–4]. Epidemiological studies worldwide have revealed that rotaviruses of types G1–G4, P(4) and P(8) are responsible for most infections, and four G–P combinations, G1P(8), G2P(4), G3P(8), and G4P(8), have been linked to 88.5% of the cases of rotavirus diarrhoea among children worldwide. However, these four G–P combinations accounted for <70% of rotavirus infections in South America, Asia and Africa [5].

Since VP7 is one of the two neutralising proteins, extensive studies on the molecular characteristics of the gene encoding this protein have been conducted. However, epidemiological studies of the distribution of P genotypes are also necessary for the development of a rotavirus vaccine, as sequence analyses of the VP4 gene could help to identify novel genotypes or sub-genotypes in the human population, including recombinant viruses containing genes of animal origin. Data from epidemiological studies have indicated that P(8) is the predominant type, and that types P(8) and P(4) account for >90% of all cases of rotavirus diarrhoea in most parts of the world [5]. Sequence analysis of available P(8) rotavirus strains has revealed that they can be divided into three lineages: the P(8)-1 or Wa-like lineage; the P(8)-2 or F45-like lineage; and the P(8)-3 or OP354-like lineage, with the last being a rare subtype comprising only a few rotavirus strains isolated from Malawi [6]. It has also been reported that, besides the dominance of the P(8) type, unusual P genotypes, e.g., P(19), which are normally predominant in animals, have been found increasingly in humans [7]. Further investigation of these strains could provide important insights into the evolution of rotaviruses, including the emergence of new P serotypes or evidence of human–animal recombinant viruses [8,9].

Vietnam is a developing country with a population of 80 million individuals. There are 1.52 million births annually, and a high mortality rate among children aged <5 years (42.2 deaths/1000/year), with 15.4% of deaths being related to diarrhoeal disease [10]. Several epidemiological studies of viruses causing acute gastroenteritis have been conducted, and these have revealed the significant impact of rotavirus on diarrhoeal disease in Vietnam [11–14]. The burden of rotavirus diarrhoea in Vietnam has also been estimated in order to consider the feasibility of a vaccine trial in the near future [15,16]. However, studies concerning the molecular characteristics of rotaviruses, especially the gene encoding the VP4 protein, have been very limited. The first study of the VP4 gene of Vietnamese rotavirus strains reported the presence of P(6) rotaviruses in Khanh Hoa Province, with these viruses probably originating from porcine rotavirus [17]. The present study determined the characteristics of the VP4 gene of rotaviruses belonging to the P(8), P(4) and P(19) genotypes that were detected during surveillance in Ho Chi Minh City in 2002–2003.

Materials and Methods

Results

The nucleotide sequences of a 687-bp fragment of the P gene were determined for 28 strains of P(8), four strains of P(4), and one strain of P(19) rotavirus. Results from the BLAST searches confirmed the P genotyping results obtained previously by nested type-specific PCR (data not shown). Table 1 summarises the clinical data of the patients studied. The age distribution of the patients ranged from 2 to 31 months. None showed dehydration, except for one patient, VN964, who was moderately dehydrated. Evaluation of diarrhoea severity, using a 20-point numerical score, showed a range between 7 and 15 points, with a mean score of 10.5 (Table 1).

Sequence analysis of the VP4 gene of P(8) rotaviruses

The phylogenetic tree of the nucleotide sequences demonstrated that the 28 Vietnamese P(8) rotaviruses clustered into two different lineages, lineage P(8)-2 or F45-like (21/28), and lineage P(8)-3 or OP354-like (7/28) (Fig. 1). The results of amino-acid (aa) comparisons, spanning aa  86–263, for seven Vietnamese P(8)-3 rotavirus strains, revealed that they had a high degree of homology with other strains within lineage P(8)-3 (98.3–98.8%), and a lower homology with other P(8) rotaviruses in lineages 1 and 2 (85.9–88.2%) (data not shown). Amino-acid alignments within the same region are shown in Fig. S1 (see Supplementary material), and revealed that the highly conserved proline molecules at residues 224 and 225 were seen in all strains studied. The seven Vietnamese P(8)-3 strains were, in general, identical with other rotaviruses in the OP354-like lineage, within the studied region, and had several unique residues that were characteristic of OP354-like viruses (aa  140-F, aa  149-S, aa  178-K, aa  182-G, aa  184-V, aa  187-G, SD at position 191–192, aa  211-I, aa  221-Y). However, the Vietnamese P(8) rotavirus strains did not have the same conserved residues as some Malawian strains at two positions (aa  194, I instead of T; and aa  258, A instead of T).

In view of the hypothesis that the rare OP354-like viruses might cause more severe diarrhoeal disease than other viruses in common lineages because they could escape protective immunity in communities with common P(8) rotaviruses, the severity of diarrhoea in the patients was investigated using Vesikari’s numerical score. The mean score for disease caused by rotaviruses in lineages P(8)-3, P(8)-2 and P(4) was 12.6 ± 1.0 (n = 6), 9.8 ± 0.7 (n = 16) and 9.5 ± 2.5 (n = 2), respectively. The difference between the mean score for the rare P(8)-3 lineage and that for the other lineages was statistically significant (p <0.05).

Sequence analysis of the VP4 gene of P(4) rotaviruses

All four Vietnamese P(4) rotavirus strains in this study were identical with each other, and clustered into a single lineage with other recent P(4) rotaviruses isolated worldwide (Fig. 2).

Sequence analysis of the VP4 and VP6 genes of a P(19) rotavirus

Nucleotide and amino-acid sequence analysis of the VP4 gene of the VN375/2003 isolate detected during 2002–2003 revealed that this strain had greater identity with P(19) rotaviruses (94.5–96.7% for nucleotides; 94.4–97.2% for amino-acids) than other P genotypes (24.2–68.1% for nucleotides; 44.6–72.2% for amino acids). The phylogenetic tree indicated that the Vietnamese P(19) strain clustered in the same group as other human and several porcine P(19) rotaviruses (Fig. 3). A deduced amino-acid comparison, from aa  88 to aa  273, showed that these P(19) strains were identical. However, the VN375 strain had several unique residues at positions aa  89-I, aa  147-N and aa  195-K. Interestingly, only the Vietnamese strain and the prototype porcine strain 4F had a valine at position 254 (Fig. S2, see Supplementary material).

A 462-bp nucleotide sequence of the VP6 gene of the VN375 strain was also analysed. The results revealed that this gene had closer identity with cognate genes of porcine rotavirus strains (91.1–93.7% for nucleotides; 98.7–99.3% for amino-acids) than with other human strains (82.4–83.7% for nucleotides; 91.5–92.2% for amino-acids).

Discussion

Genotype P(8) has been found to be the most frequent P type among rotaviruses worldwide, with a prevalence of 51% in Africa and 87.8% in Europe [5]. Reports from several surveillance studies in Vietnam have also indicated a high prevalence of P(8) rotavirus (70.7–71.8%) [11,14]. The present study analysed the VP4 gene sequences of these isolates and found that, surprisingly, seven of 28 randomly chosen P(8) strains clustered in the rare P(8)-3, or so-called OP354-like lineage. Since the first report of OP354-like viruses in Malawi during 1998–1999 [6], these rare rotavirus strains have not been isolated elsewhere. A BLAST search of P(8) nucleotide sequences available from GenBank showed that only a few P(8) rotavirus strains had a high identity with the OP354 virus, including some strains from Malawi [6] and India [23], and one each from Thailand and Finland. The detection of these viruses from various countries in different continents suggests that these viruses are widespread around the world. The rather low rate of detection of these rare strains could be explained by large-scale mutation in the nucleotide sequence of the VP4 gene at the primer-binding site of the primer, 1T-1, specifically for P(8), which may make recognition of these rotavirus strains more difficult [6].

Cunliffe et al. [6] reported that Malawian P(8) rotaviruses predominantly comprised strains belonging to a distinct G type, in which the G4 type for OP354, the G3 type for MW258, and the G1 type for OP601 were clustered. In the present study, the phylogenetic tree indicated clearly that Vietnamese P(8) rotaviruses belonged to two separate clusters, the OP354 cluster and the OP601 cluster. The strains in the OP601 cluster could be divided further into two variants, in which one variant had a tendency to belong to G types 1 and/or 9, and the second variant had the characteristics of types G9 or G4. However, the concordance between P and G type in the OP354-like viruses was unclear (Fig. 1).

A rotavirus vaccine, which includes a P(8)-1 rotavirus strain as a component, has begun to be used widely in some countries. The finding in the present study that viruses of the P(8)-3 lineage caused, on average, more severe diarrhoeal disease may be important for the development and application of rotavirus vaccines, especially in countries in which OP354-like rotaviruses are prevalent.

The P(19) rotaviruses are known to commonly infect pigs, with the prototype virus being the porcine 4F strain [24]. To our knowledge, there have been only two reports of the isolation of the P(19) rotavirus in humans worldwide, one from Thailand and one from India [7,9]. The detection of this virus in Vietnam emphasises that the P(19) rotavirus is circulating in Asia. Sequence analysis of the VP6 gene of the Vietnamese G9P(19) strain also showed closest identity with the cognate genes of other porcine rotaviruses, suggesting gene reassortment or a human infection by a virus similar to one that infects animals. Further investigation of this virus might provide interesting data concerning the relationship between human and porcine rotaviruses, especially in Asian countries in which farming is the main activity and close contact between humans and animals could promote inter-species transmission.

Accession Numbers

The VP4 nucleotide sequences of the Vietnamese isolates described in this study have been deposited in GenBank under the following accession numbers: EF673752–EF673779 for the P(8) strains; and EF673780–EF673783 for the P(4) strains. The accession numbers of the nucleotide sequences of the VP4 and VP6 genes of the VN375/2003 strain are EF063154 and EU042135, respectively.