Epidemiological study of hepatitis E virus infection of dogs and cats in Japan.

HEPATITIS E virus (HEV) is the causative agent of human hepatitis E in many developing and some industrialised countries (Aggarwal and Krawczynski 2000, Meng 2000a, Schlauder and Mushahwar 2001, Huang and others 2002, Mizuo and others 2002, Takahashi and others 2003). It is widely known that in humans HEV is primarily transmitted by the faecal-oral route through contaminated water, and occasionally causes large epidemics in endemic areas (Arankalle and others 1994, Aggarwal and Krawczynski 2000). However, in non-endemic areas, the reservoir for sporadic human cases with no history of travel to HEV-endemic areas has remained unknown for many years (Harrison 1999, Schlauder and Mushahwar 2001, Mizuo and others 2002, Takahasi and others 2002). In 1997, the first animal HEV strain, swine HEV, was discovered in a pig in the USA (Meng and others 1997). Subsequent epidemiological studies indicated that, for example, in Japan, most farmed pigs had been exposed to swine HEV (Takahashi and others 2003). There is a growing consensus that HEV is a potential zoonotic agent and that pigs can act as a reservoir for humans. HEV has also been detected in chickens (Haqshenas and others 2001) and wild rodents (Kabrane-Lazizi and others 1999, Favorov and others 2000, Arankalle and others 2001, He and others 2002, Hirano and others 2003a), and anti-HEV antibodies have been found in cattle, water buffaloes, sheep, goats, non-human primates, dogs and cats (Tsarev and others 1993, Meng 2000b, Arankalle and others 2001, Hirano and others 2003b, Usui and others 2004). In an unusual case, an incident was reported in which a cat was suspected to be a reservoir for human infection (Kuno and others 2003). This short communication describes the examination of blood samples and rectal swabs from dogs and cats for evidence of HEV infection. An ELISA using the purified, empty virus-like particles (VLPs) of HEV (Li and others 1997) was performed with 424 canine and 202 feline serum samples. Samples were collected from animals at animal hospitals over a five-year period, from 2000 to 2004, in over 30 prefectures covering northern Hokkaido and south-western Okinawa, Japan, and stored at –20°C. The ELISA method used was the same as that described by Li and others (2000). The sera were examined simultaneously using VLP-coated and VLP-uncoated (blank) plates, and the antibody titre was expressed as an optical density at 492 nm (OD492) by subtracting the OD492 value of the blank plate from that of the VLP-coated plate. OD492 values of greater than 0·1 were obtained from 10 canine and eight feline serum samples (Table 1). These samples were subsequently absorbed with the same VLPs used as the antigen in the ELISA to confirm the specificity of the reaction. The serum sample (100 μl at a dilution of 1:200) was mixed with 1 μg of VLPs, and the mixture was incubated at 37°C for one hour, and then re-examined. A reduction of the OD492 value of the sample by greater than 50 per cent after the absorption was considered to indicate an anti-HEV antibody-positive result, as described by Arankalle and others (2001). On this basis all the canine serum samples were considered to be anti-HEV antibody-negative, and four feline serum samples (1·98 per cent) were considered to be anti-HEV antibody-positive (Table 2). A reverse transcriptase-PCR (RT-PCR) assay, described for the detection of swine HEV (Huang and others 2002), was used to test 100 canine and 66 feline rectal swabs from animal hospitals located in Tokyo and its environs obtained over a six-year period from 1999 to 2004. In addition, the 18 sera showing an OD492 value of greater than 0·1 (Table 1) were also examined by RT-PCR. When a questionable PCR product was obtained, it was analysed by sequencing. No specific PCR products were amplified from any of the canine or feline samples examined. Only a small number of cases of HEV infection in dogs and cats have been reported (Tien and others 1997, Arankalle and others 2001, Usui and others 2004). In the present study, the samples were collected nationwide in Japan and anti-HEV antibody was found in a very small percentage of the cats and in none of the dogs tested by the ELISA. These results were in contrast to a previous report. Among 135 cats visiting an animal hospital in a provincial capital in Japan, 44 (33 per cent) possessed anti-HEV antibody but no HEV RNA was recovered from the 135 sera (Usui and others 2004). No convincing explanation for the discrepancy in the seroprevalence rates between the previous and the present reports was possible. It should be noted that there is a possibility that the recombinant antigen itself could have non-specific, cross-reacting epitopes among antibodies in animal sera. According to a report describing HEV infection in dogs from India (Arankalle and others 2001) and Vietnam (Tien and others 1997), 22·7 per cent to 27 per cent of the dogs were anti-HEV antibody-positive. It may be true that dogs and cats in such HEV-endemic areas have been exposed to HEV more frequently than animals in non-endemic areas. In non-endemic areas, there may be some common infectious source affecting both human beings and their companion animals, as they share mostly the same living environment. One probable reservoir, especially for companion animals, is rodents, and it has recently been reported in Japan that wild rats have been found to be infected with HEV (Hirano and others 2003a). Cats are rodent hunters, which is inherent to their nature, but this is not so much the case in dogs. This causal relationship between disease prevalence and a behavioural characteristic has already been seen in other viral diseases in cats, such as Borna disease virus (Berg and Reduction rate of ELISA optical density Place of Age at 492 nm after HEV Cat residence (years) Sex Clinical signs absorption (%) antibody*