|Home | About | Journals | Submit | Contact Us | Français|
A molecular biological survey on porcine norovirus (NoV) and sapovirus (SaV) was conducted in Toyama Prefecture, Japan, during fiscal year 2008. Both NoV and SaV were detected from swine fecal samples throughout the surveillance period, indicating that these viruses were circulating in this region. NoV strains detected in this study belonged to three genotypes that are known as typical swine NoVs. Although human NoVs were occasionally detected, it was unclear whether they replicated in pigs. As for SaV, genogroup VII (GVII) and other divergent genogroups were identified in addition to the dominant genogroup, GIII, which is the prototypic porcine SaV. In addition, 3 strains genetically related to human SaV were detected. Two of these 3 strains were closely related to human SaV GV. Our study showed that genetic diversification of porcine SaV is currently progressing in the swine population.
Norovirus (NoV) and sapovirus (SaV) belong to the genera Norovirus and Sapovirus, respectively, within the family Caliciviridae and are an important cause of acute gastroenteritis in humans. NoV is genetically diverse and currently classified into genogroup I (GI) to GV by molecular characterization based on the partial or complete capsid or RNA-dependent RNA polymerase (RdRp) sequence (1, 10, 47, 49). GI, GII, and GIV are known to infect humans (48). GI and GII are further divided into many genotypes, and this classification is in constant evolution with the discovery of new strains. NoV has also been detected in several animal species, including swine, cattle, mice, lions, and dogs (21, 28, 30, 32, 51). Among them, bovine and murine strains are classified as GIII and GV, respectively, and are genetically distinct from human strains (21, 37, 42). Swine are susceptible to NoV GII, the most prevalent cause of recent acute gastroenteritis in humans. Porcine strains are related to human strains genetically and antigenically (6, 43, 44, 45, 50, 51). In addition, a human NoV strain reportedly replicates and shows a mild pathogenesis in experimentally inoculated gnotobiotic pigs (5). This evidence raises the possibility of a swine reservoir for human NoVs; however, except for one recent report (32), strong evidence of common NoV strains in both humans and swine is still lacking.
SaVs have been thought to infect infants and young children, with an outcome of milder diarrhea than infection with NoV (38, 40); however, this perception might be changing (13, 54). SaVs are also classified into GI to GV, which are further subdivided into a number of clusters and genotypes, according to the genetic variation of their capsid or RdRp genes (8, 16, 36, 41). GI, GII, GIV, and GV have been detected in humans, while GIII has been detected in swine. Recently, GVI, GVII, and other divergent groups have been reportedly detected in the swine population (2, 26, 27, 31, 33, 52, 55, 56). Several strains among these porcine SaVs showed a genetic relation to human SaVs (26, 29). This suggests the possibility of zoonotic circulation of SaV between humans and swine.
In this study, we conducted a molecular biological survey of the swine population to identify the current prevalence and genetic diversity of porcine NoV and SaV in Toyama Prefecture, Japan, during fiscal year 2008. Swine herds in this region were found to harbor both NoV and SaV; while NoV detected in this study clustered with conventional NoVs, the detected SaV showed higher genetic diversity.
Fecal samples were collected at a veterinary diagnostic laboratory annexed to an abattoir in Toyama Prefecture, Japan. Finisher pigs, 6 months of age, came from different premises located throughout Toyama prefecture. All pigs were apparently healthy at the time of slaughtering. During each month from April 2008 to March 2009, 20 pigs were randomly chosen and their intestinal contents were collected. In August 2008 and January 2009, additional samples were collected from pigs at three specifically chosen farms. RNA was extracted from the supernatants of 10% (wt/vol) fecal suspensions in phosphate-buffered saline (PBS) using a QIAamp viral RNA kit (Qiagen K.K., Tokyo, Japan) according to the manufacturer's instructions.
Extracted RNA was treated with 5 U DNase I (Takara Bio Inc., Shiga, Japan), and cDNA was synthesized by SuperScript III reverse transcriptase (Invitrogen Japan K.K., Tokyo, Japan) with a random hexamer according to the manufacturers' instructions. Obtained cDNA was used for the subsequent PCR. The target gene and nucleotide locations of primers used for PCR to detect porcine NoV and SaV are summarized in Table Table1.1. Primer pair p290/p289 was designed to universally detect NoV and SaV (17) by targeting the RdRp gene. The expected size of the amplified fragment is about 320 bp. Primer pairs G1SKF/G1SKR and G2SKF/G2SKR were designed for the detection of GI and GII NoVs, respectively. They amplified the 5′ end of the capsid gene with an expected size of about 330 bp and are currently used for diagnostic reverse transcription-PCR (RT-PCR) in human outbreak cases (25). Primer pair PEC66/PEC65 specifically amplifies a 330-bp fragment within the RdRp gene of porcine SaV, especially GIII (53). For the general detection of human SaV, the primer pair SV-F11/SV-R1 was used for the 1st round of PCR followed by nested PCR using the primer pair SV-F2/SV-R2 (35). An approximately 430-bp fragment of the 5′ end of the capsid gene was amplified by this nested PCR. Conditions of PCRs were as follows. After initial denaturation at 94°C for 3 min, 35 amplification cycles were performed. Each cycle consisted of denaturation at 94°C for 30 s, primer annealing at 46°C for 30 s, and an extension reaction at 72°C for 1 min, followed by a final extension at 72°C for 7 min. Primers PEC68, PSV6, and PSV11 were designed on basis of the RdRp genes of divergent porcine SaV strains (53). These primers were used for additional nested-PCR experiments in combination with SV-R1 and SV-R2 to amplify the region of the 3′ end of the RdRp gene to the 5′ end of the capsid gene. This PCR was performed with modification of the extension reaction at 72°C for 2 min. The PCR products were separated by electrophoresis in 1.5% agarose and visualized under a UV lamp after ethidium bromide staining.
PCR amplicons of the expected sizes were cloned into the pGEM-T Easy vector (Promega K.K., Tokyo, Japan) according to the manufacturer's instructions. Plasmid DNA was extracted using an alkaline lysis method. At least three clones for each amplicon were subjected to DNA sequencing using BigDye Terminator kit, version 3.1, and the ABI 3130 automatic capillary sequencer (Applied Biosystems Japan Ltd., Tokyo, Japan). The obtained sequence information was edited by Sequencher (version 4.7) software (Gene Codes Co., MI). Multiple sequence alignments of NoVs or SaVs were generated by Clustal W, and the bootstrapped phylogenetic trees were constructed by the neighbor-joining method with 1,000 bootstrap replicates using Molecular Evolutionary Genetics Analysis (MEGA) version 3.1 software. Genetic distances between NoV or SaV strains were calculated by Kimura's 2-parameter method (24). The genotypes of the NoV strains detected in this study were determined according to the classification based on the partial capsid gene (20, 23). Reference strains and genotype numbering are cited from a report by Kageyama et al. (20). Prototypic porcine NoV strains QW48, QW101, and QW170 (51) were also used as porcine reference strains to construct the phylogenetic tree. The reference strains of SaV to determine the genogroup are cited from previously published reports (2, 11, 26, 29, 33, 52). The DDBJ accession numbers of the published representative strains are shown in Fig. Fig.11 to to33.
Real-time PCR was carried out to quantify human GII NoV as described by Kageyama et al. (19). Briefly, 50-μl reaction mixtures containing 5 μl cDNA, 25 μl TaqMan universal PCR master mix (Applied Biosystems Japan Ltd., Tokyo, Japan), 20 pmol each of primers COG2F, ALPF, and COG2R, and 11.4 pmol RING2AL-TP probe were prepared. Subsequently, PCR amplification using the ABI 7500 system (Applied Biosystems Japan Ltd., Tokyo, Japan) was performed as follows: 45 amplification cycles of 95°C for 15 s and 56°C for 1 min after the preheating step at 50°C for 2 min and 95°C for 10 min. The copy number of GII NoV was calculated based on standardized control samples (19).
Nucleotide sequence data have been deposited in the DDBJ database under accession numbers AB521754 to AB521770 (partial capsid genes of NoV GII strains), AB521771 to AB521791 (partial RdRp genes of SaV strains), and AB521771 and AB521772 (partial capsid genes of SaV strains).
Primers p290 and p289 were designed based on the conserved motifs found in the RdRp genes of caliciviruses. As this primer pair reportedly detects a broad range of caliciviruses from different animal species (7, 17, 28, 37, 51), we first employed this primer pair to detect NoV and SaV in the swine population. The primer pair detected 24 strains from 240 pigs tested, and 6 strains of NoV and 18 strains of SaV were identified by sequence analysis, respectively (data not shown). We next employed primer pairs G1SKF/G1SKR and G2SKF/G2SKR to capture additional NoV strains. These primer pairs have shown high sensitivity to detect human NoV GI and GII, respectively (25), and G2SKF/G2SKR was able to detect more NoV GII strains than p290/p289, while two pigs were positive for only p290/p289 (Table (Table22).
NoV GII strains were detected every month, while NoV GI strains were not detected throughout the surveillance period (Table (Table2).2). The relative number of NoV GII-positive pigs increased in November and December 2008 and January 2009.
Primer pair PEC66/PEC65, which is designed for the effective detection of porcine SaV, successfully detected SaV every month, except March 2009 (Table (Table2).2). Twelve SaV strains were detected only by p290/p289 (Table (Table2).2). No apparent change in the number of SaV-positive pigs was observed throughout the year (Table (Table2).2). Overall, 42 and 56 of 240 pigs were positive for NoV and SaV, respectively. Two pigs were simultaneously infected with both porcine and human NoVs, and a pig was simultaneously infected with both porcine and human SaVs (Table (Table2).2). Simultaneous infection with both NoV and SaV was observed in 13 pigs; therefore, 85 pigs were positive for at least one of the caliciviruses and the remaining 155 pigs were negative for both viruses.
In August 2008 and January 2009, we collected additional fecal samples from pigs from three specifically chosen farms to assess whether the prevalences of NoV and SaV differed among farms (Table (Table3).3). In August 2008, porcine NoV GII was detected at farms A and B while none of the pigs was positive at farm C. Porcine SaV was more frequently detected than NoV at farms A and B. None of the pigs at farm C was positive for SaV. Two pigs at farm A were simultaneously infected with porcine NoV and SaV. Neither human NoV nor human SaV was detected in this period. In January 2009, NoV GII was detected at farms A, B, and C with different frequencies. Human NoV GII was detected at farms A and C. A high positive rate (15/20) of SaV was observed at farm A, whereas no pigs or only one pig was positive for SaV at farm B or C, respectively. Five pigs at farm A were simultaneously infected with porcine NoV and SaV.
Since pigs were occasionally observed to be infected with multiple strains, the nucleotide sequences of NoV and SaV strains were determined after the molecular cloning of RT-PCR products to discriminate different strains in multiply infected pigs. As a result, nucleotide sequence data for 63 strains of NoV and 103 strains of SaV became available from both randomly and specifically chosen pigs. Genotypes of NoV and SaV strains were subsequently determined by phylogenetic analysis. All NoV strains detected by primer pair p290/p289 were porcine NoV (data not shown). A phylogenetic tree constructed based on the nucleotide sequence of the 5′ part of the capsid gene of NoV showed that 51 of 63 analyzed NoV strains belonged to the porcine NoV cluster. Among them, 24 and 26 strains belonged to clusters with QW101 (previously described as GII/18) (51) and QW48 (GII/11) types, respectively. One strain formed a cluster with the QW170 (GII/19) type (Fig. (Fig.1).1). Twelve strains with human NoV genotypes were also detected, mainly in winter (November 2008 to February 2009) (Tables (Tables22 and and3).3). Four strains, 7 strains, and 1 strain were classified into GII/3, GII/4, and GII/13 genotypes, respectively (Fig. (Fig.1);1); however, their genomes were only slightly amplified by RT-PCR, possibly because these human NoV strains had replicated in a very inefficient manner in the pigs. The copy numbers of these human NoVs analyzed by quantitative real-time PCR were consistently under the detectable level (data not shown).
The phylogenetic tree of SaV was constructed based on the partial RdRp gene. Seventy-three strains belonged to GIII, which originally prevailed in swine populations (11). Twelve strains belonged to the GVII cluster. Fifteen strains belonged to another cluster, reported recently (2, 26, 33) and tentatively designated GVIII here. These strains formed 2 and 3 subclusters within GVII and GVIII clusters, respectively. These subclusters were tentatively termed GVII/1, GVII/2, GVIII/1, GVIII/2, and GVIII/3 (Fig. (Fig.2).2). The mean calculated distance between GVII/1 and GVII/2 was 0.558 ± 0.056. The distances between GVIII/1 and GVIII/2, GVIII/1 and GVIII/3, and GVIII/2 and GVIII/3 were 0.542 ± 0.057, 0.428 ± 0.045, and 0.531 ± 0.053, respectively. These values were comparable to the reference distance between GI/1 and GI/2 (0.431 ± 0.052) or GII/1 and GII/4 (0.253 ± 0.035). One strain formed a cluster with strains with which genetic relation to human SaVs was previously reported (29). No GVI strain was detected in this study (Fig. (Fig.22).
Primer pairs SV-F11/SV-R1 and SV-F2/SV-R2 were used to amplify the capsid region of human SaV to assess whether human SaVs existed in the swine population. Two strains, swine/TYMPo31/07/JP and swine/TYMPo239/07/JP, were detected by nested PCR (May and December 2008) (Table (Table2).2). Sequencing and phylogenetic analyses of the partial capsid genes of these strains revealed that they formed a new branch that was closer to human SaV GV than those previously reported (Fig. (Fig.3).3). These strains were supposed to possess the RdRp gene, unlike conventional porcine SaV, because they were not amplified by the PEC66/PEC65 primer pair. Alternatively, these strains might be generated by genetic recombination at the RdRp gene-capsid gene junction region, which has been identified as the major breakpoint (12, 14, 52). Therefore, to characterize these strains further, the nucleotide sequences of the partial RdRp genes of these strains were determined and the phylogenetic pattern was compared with that based on the partial capsid gene. Since the RdRp sequences of swine/TYMPo31/07/JP and swine/TYMPo239/07/JP were unknown, sense primers p290, PEC68, PSV6, and PSV11, designed to amplify the RdRp gene regions of divergent SaVs (53), were used to amplify the region spanning from the 3′ end of the RdRp gene to the 5′ end of the capsid gene in combination with antisense primers SV-R1 and SV-R2. As a result, seminested PCR using primer pairs PSV11/SV-R2 and PSV6/SV-R2 could amplify the definite fragments. Sequence analysis revealed that PSV11/SV-R2 amplified the region corresponding to nucleotides (nt) 4835 to 5650 on strain GV Ehime475 (DQ366344). Unexpectedly, the PSV6 primer bound in both sense and antisense orientations, resulting in amplification of the region corresponding to nt 3575 to 4884 on strain Ehime475. These overlapping fragments were aligned, and a contiguous sequence was obtained. The phylogenetic pattern of the partial RdRp gene showed that the genetic locations of swine/TYMPo31/07/JP and swine/TYMPo239/07/JP strains were also related to human SaV GV (Fig. (Fig.2).2). Nucleotide alignments around the conserved RdRp gene-capsid gene junction region showed high similarity between strains swine/TYMPo31/08/JP, swine/TYMPo239/08/JP, and GV human/Ehime475/04/JP (Fig. (Fig.4).4). These results indicated that swine/TYMPo31/07/JP and swine/TYMPo239/07/JP were not recombinant but were a novel genotype or genogroup with genetic similarity to human SaV GV.
The genetic diversity of NoV and SaV is one of the major characteristics of these viruses. A great variety of NoV genotypes have been identified in the human population, whereas the genetic variation of porcine NoV is limited to only three genotypes thus far (51, 57). Because human and porcine NoVs are genetically related, the potential of zoonotic transmission between humans and swine has been noticed. Porcine SaV is divergent, and at least 4 genogroups have been proposed for their classification (2, 26, 27, 31, 33, 52, 55, 56). Recent studies showed that porcine SaV was genetically related to human SaV (26, 29), suggesting the future possibility of interspecies transmission of SaV. Our study showed that NoV and SaV were frequently detected in the swine herd throughout the surveillance period, indicating that both viruses are continuously prevalent in this region. The frequencies of detection were quite different among the specifically chosen farms, suggesting that NoV and SaV infection rates were influenced by the farms where pigs were kept, and especially by their sanitary conditions. Clean facilities may prevent caliciviruses from spreading in the swine herd.
Most NoVs detected by primer pair G2SKF/G2SKR were classified into typical swine genotypes, indicating that the primer pair was able to capture porcine NoV in addition to human NoV. The genetic variation of porcine NoV detected in this study was limited, as previously reported (51, 57). Porcine NoVs might have acquired genetic stability compared to human NoVs, which were also detected in this study. The number of NoV outbreaks in the human population increases in winter (15). In 2008, NoVs GII/3 and GII/13 in addition to GII/4 were included in the major causes of human outbreaks in Toyama prefecture (our unpublished observation). The occasional detection of these human NoVs from pigs might be associated with an increased number of human outbreaks; however, it is obscure whether human NoVs replicated in pigs because the targeted gene was only slightly amplified by RT-PCR. A previous serological study showed a high prevalence of antibodies against both GI and GII NoVs in pigs (6), indicating that pigs were frequently exposed to human NoVs. Nevertheless, they might be eliminated by dominant replication of porcine NoVs.
Primer pair PEC66/PEC65 successfully detected several porcine SaVs, most of which were classified into porcine GIII, because PEC66 and PEC65 were designed on the basis of the GIII SaV RdRp gene (53). On the other hand, the p290/p289 primer pair detected four divergent genogroups. Recently, the number of divergent genogroups of SaV detected in pigs has increased (2, 26, 27, 31, 33, 52, 55, 56). In this study, subclusters within GVII and GVIII were observed by phylogenetic analysis. The calculated distances revealed that these subclusters could be classified into distinct genotypes. These results suggest that the genetic properties of SaV are currently diverse in the swine population.
Two strains were detected as a novel genotype or genogroup. Although these strains were genetically similar to human SaV GV, it is presently unclear whether they originated from porcine or human SaV. Extended sequence analyses indicated that these strains did not arise by recombination because they maintained their genetic similarity through the RdRp and capsid genes analyzed. This novel type possibly emerged from a GV-like SaV as a result of the accumulation of genetic mutations. Such genetic similarity between porcine and human SaVs may suggest the potential for zoonotic transmission of SaVs.
An increasing number of reports have indicated that recombination is a major driving force in NoV evolution (3, 4, 9, 18, 34, 39, 46). Genetic recombination of SaV has also been noticed since inter- or intragenogroup recombination of SaV was reported (12, 14, 22, 52). In this study, multiple infections with different genotypes of NoV or different genogroups of SaV were observed in 6 or 7 pigs, respectively. Coinfection with NoV and SaV was also observed in 20 pigs. Such coexistence of different strains could render an opportunity for genetic recombination among strains. Although our study did not find evidence of recombination, the emergence of novel types of NoV or SaV by recombination remains possible in the future.
All swine fecal samples were collected from finisher pigs which were apparently healthy at the time of slaughtering, indicating that NoVs and SaVs were maintained in pigs without illness. Although the possibility that these pigs had developed clinical symptoms when younger cannot be ruled out, the pathogenicity of NoV and SaV in pigs is expected to be mild. Nevertheless, frequent infections with NoVs and SaVs in pigs increased the opportunity for the emergence of novel variants with changes in pathogenicity during virus replication.
In summary, we revealed the frequent detection of NoVs and SaVs in a swine herd in this region. Porcine NoV showed limited variability and clustered in typical swine genotypes, except for the uncommon detection of human NoV. Porcine SaV prevailing in this region showed high genetic diversity. A novel genotype or genogroup of SaV showing genetic similarity to human SaV GV was identified in addition to the 4 genogroups reported previously. Continuous surveillance for calicivirus in swine, humans, and other animals will enable the detection of as yet unidentified caliciviruses, leading to precise understanding of the evolution of these viruses.
This study was supported in part by a Grant for Research on Emerging and Re-emerging Infectious Disease from the Japanese Ministry of Health, Labor and Welfare.
We thank the staff members of Toyama Prefectural Meat Inspection Centre for collecting swine samples. We are grateful to M. Maekawa for her excellent technical assistance.
Published ahead of print on 17 February 2010.