Rat hepatitis E virus (Rocahepevirus ratti) exposure in cats and dogs, Hong Kong

ABSTRACT

Hepatitis E virus (HEV) variants infecting humans belong to two species: Paslahepevirus balayani (bHEV) and Rocahepevirus ratti (rat hepatitis E virus; rHEV). R. ratti is a ubiquitous rodent pathogen that has recently been recognized to cause hepatitis in humans. Transmission routes of rHEV from rats to humans are currently unknown. In this study, we examined rHEV exposure in cats and dogs to determine if they are potential reservoirs of this emerging human pathogen. Virus-like particle-based IgG enzymatic immunoassays (EIAs) capable of differentiating rHEV & bHEV antibody profiles and rHEV-specific real-time RT–PCR assays were used for this purpose. The EIAs could detect bHEV and rHEV patient-derived IgG spiked in dog and cat sera. Sera from 751 companion dogs and 130 companion cats in Hong Kong were tested with these IgG enzymatic immunoassays (EIAs). Overall, 13/751 (1.7%) dogs and 5/130 (3.8%) cats were sero-reactive to HEV. 9/751 (1.2%) dogs and 2/130 (1.5%) cats tested positive for rHEV IgG, which was further confirmed by rHEV immunoblots. Most rHEV-seropositive animals were from areas in or adjacent to districts reporting human rHEV infection. Neither 881 companion animals nor 652 stray animals carried rHEV RNA in serum or rectal swabs. Therefore, we could not confirm a role for cats and dogs in transmitting rHEV to humans. Further work is required to understand the reasons for low-level seropositivity in these animals.

KEYWORDS:

• Hepatitis E virus
• hepatitis antibodies
• rats
• cat diseases
• dogs

Introduction

Hepatitis E is an important cause of viral hepatitis in humans accounting for an estimated 20 million cases per year globally [1]. Hepatitis E virus (HEV) variants that cause human infection belong to two species of the Hepeviridae family: Paslahepevirus balayani (bHEV) and Rocahepevirus ratti (rHEV). bHEV genotypes are the most common cause of hepatitis E and may either be transmitted faecal-orally (genotypes 1 & 2) or as a foodborne zoonosis / bloodborne infection (genotypes 3 & 4) [2]. Although most hepatitis E infections are self-limiting, severe disease may occur in pregnant women (bHEV genotype 1) and persons with chronic liver disease (bHEV genotypes 3 & 4) [3]. In addition, immunocompromised individuals often develop chronic hepatitis E due to bHEV genotypes 3, 4 and 7 [4–6].

rHEV, also known as rat hepatitis E virus, is a common pathogen of commensal rats globally [7]. rHEV was considered unable to infect humans due to considerable phylogenetic divergence from bHEV. However, in 2018, we reported the world’s first case of human rHEV infection in Hong Kong [8]. By establishing a city-wide surveillance programme, we have since identified 19 more cases of rHEV infection in Hong Kong [9–11]. rHEV infections have also been reported in Canada, Spain, and France, confirming that this is a globally prevalent zoonosis [12–14]. rHEV seroprevalence studies in Hong Kong, Germany and Vietnam also indicate wide exposure in human populations [15–17]. Although an effective vaccine against bHEV is available in China, it only offers partial protection against rHEV based on preclinical studies and antibody-binding assays [18–20].

We have demonstrated that human rHEV infections are spatiotemporally linked to epizootics in street rats [9]. Rats serve as a vast, readily infected natural reservoir of rHEV [7,21]. However, the exact route of transmission from rats to humans is still unknown. Surveillance of rHEV RNA in swine and meat products in Hong Kong are negative so far, pointing against a foodborne transmission route similar to bHEV genotypes 3 & 4 [9,22]. We hypothesized that companion animals, particularly cats and dogs, might serve as reservoir hosts and sources of human rHEV infection. Cats and dogs are common urban animals that potentially interact with street rats or their droppings. Cats and dogs are susceptible to other rodent-borne diseases like plague and leptospirosis. Furthermore, viruses within Hepeviridae frequently exhibit inter-species transmission [2]. Although cats and dogs have never been shown to harbour active HEV infection, many serological studies showed that they carry anti-HEV antibodies [23–28] (supplementary table 1). However, these studies are often hampered by reliance on bHEV-based RT–PCR or antibody assays that either miss or cannot differentiate animals with rHEV exposure. In this study, we probed bHEV and rHEV seroprevalence in cats and dogs using specific quantitative real-time reverse transcription PCR (RT-qPCR) assays and an enzymatic immunoassay (EIA) system that is validated to detect and differentiate bHEV and rHEV antibody signatures in sera [17]. We have previously used this EIA system to report on rHEV seroprevalence in humans [17]. We aimed to investigate sero-reactivity in cats and dogs to rHEV infection and measure species-specific HEV exposure rates in these animals.

Materials & methods

Animal and human samples

Companion and stray cats (Felis catus) and dogs (Canis lupus familiaris) were sampled for this study. The companion animal cohort comprised cats and dogs undergoing blood taking for diagnostic purposes by veterinary clinics located in multiple districts across Hong Kong between January 2022 and December 2022. Blood samples were sent to the Department of Microbiology, Queen Mary Hospital for PCR screening of veterinary pathogens (Babesia spp., Ehrlichia spp., and feline coronavirus). Leftover archived sera from these blood samples were stored at −20°C prior to use in this study. The stray animal cohort comprised rectal swab samples from captured stray cats and dogs, which were sent to us in virus transport medium by the Agriculture, Fisheries and Conservation Department of Hong Kong over a 3-year period between January 2016 and January 2019. The starting period of 2016 was chosen because the first human cases of rHEV infection in Hong Kong had symptom onset in early 2017 [8,9]. Samples were stored at −80°C prior to analysis. The collection of these samples was approved by the institutional animal ethics committee (CULATR Ref. No: 3330-14 and 4526-17). Archived human bHEV and rHEV RNA-positive samples were used to demonstrate the specificity of immunoblots. The collection of human samples was approved by the Institutional Review Board of The University of Hong Kong/ Hospital Authority West Cluster (UW 18-074).

Expression and characterization of HEV-A4-P239 and HEV-C1-P241 peptides

We have previously described the expression of HEV-A4-P239 and HEV-C1-P241 peptides, which incorporate the major neutralizing epitopes of the Open Reading Frame 2 (ORF2) capsid protein of bHEV and rHEV respectively [18]. Briefly, truncated ORF2 fragments of bHEV genotype 4 (HEV-A4-P239, amino acid residues 368–606) and rHEV (HEV-C1-P241, amino acid residues 357–597) were overexpressed in E. coli inclusion bodies and purified using nickel affinity chromatography. Both peptides were characterized using sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and control antisera immunoblots. Detailed protocols of protein expression, SDS-PAGE, and immunoblots are described in the supplementary material.

In-house EIAs

The in-house EIAs were coated with HEV-A4-P239 (0.5 μg/mL) or HEV-C1-P241 (0.5 μg/mL) in the solid phase to produce bHEV and rHEV EIA plates respectively [17]. We have previously shown good performance of this assay system in differentiating acute bHEV and rHEV infection in human patient sera (supplementary table 2). bHEV and rHEV antibodies in companion animal sera were assessed in parallel in bHEV and rHEV EIAs. Each serum sample was diluted 1:100 and incubated at 37°C for one hour in wells. The plates were then washed with 1× phosphate-buffered saline (PBS) containing 0.3% (vol/vol) Tween 20 (Thermo Fisher Scientific, Waltham, USA) (PBS-T) and probed with 1:4000 horseradish peroxidase (HRP) conjugated goat anti-canine IgG or anti-feline IgG (SouthernBiotech, Birmingham, USA) as appropriate. Plates were incubated at 37°C for 30 mins and washed six times with PBS-T. Tetramethylbenzidine (TMB) (Thermo Fisher Scientific) was used as a substrate for the enzymatic reaction, which was stopped with 0.3 M sulphuric acid after ten mins. Optical densities (OD) were recorded at 450 nm. Mean and standard deviations (SD) of cat and dog sera in both bHEV and rHEV EIAs were calculated.

To verify the performance characteristics of these EIAs when using a dog or cat serum matrix, total IgG was purified from bHEV-antibody positive and rHEV-antibody positive human patient sera using Nab Protein G Spin Kit (Thermo Fisher Scientific). The concentration of IgG was measured using the BCA protein assay (Thermo Fisher Scientific) and prepared at a final concentration of 1 μg/μL followed by serial dilution (1, 0.5, 0.25, 0.125 and 0.0625 μg/μL). These IgG preparations were spiked into pools of seronegative dog and cat sera. bHEV and rHEV EIAs were then performed as described above except that 1:8000 horseradish peroxidase (HRP) conjugated goat anti-human IgG (Thermo Fisher Scientific) was used as the secondary antibody.

Commercial EIAs

Any serum sample that produced ODs exceeding three standard deviations from the mean OD of the entire cohort in either bHEV or rHEV EIA was further tested using the commercial Wantai IgG assay (Wantai BioPharm, Beijing, China) [29]. Fifty serum samples that were negative in both EIAs were randomly selected. Ten of them were used as negative controls, and the remaining negative samples were selected for the purpose of generating cutoffs in the Wantai assay. Serum samples were diluted 1:10 and added to wells, followed by incubation, and washing procedures suggested by the manufacturer. Dog-specific or cat-specific secondary antibodies were then added to the plates as previously described [30,31]. Plates were washed with PBS-T, followed by the addition of 50 μL chromogen solution A and 50 μL chromogen solution B as per manufacturer instructions. The enzymatic reaction was stopped after 4 mins and ODs were measured at 450 nm. An OD of 3 standard deviations above the mean of the 40 negative samples was adopted as the cutoff.

HEV-A4-P239 and HEV-C1-P241 immunoblots

Samples testing positive in rHEV EIAs were further assessed by immunoblots based on HEV-C1-P241 peptides as described in the supplementary material. Sera of dogs and cats were diluted 1:200 with 10% skim milk in PBS-T and added to membranes in the MiniProtean II Multiscreen Apparatus (BioRad, Hercules, USA) and incubated for one hour. Sample dilutions were then adjusted further (ranging from 1:200 to 1:400) to reduce background reactivity. Also, the membrane was exposed to 1:5000 diluted His-tag antibodies (Bio-Station, Hong Kong, China). The MiniProtean II Apparatus enables lane-by-lane separation of the immunoblot without cutting into strips. The membranes were washed for 30 mins and then incubated with species-specific HRP-conjugated secondary antibodies for 30 mins. The immunoreaction was detected using Alliance Q9 Advanced (Uvitec, Cambridge, UK) with WesternBright ECL HRP substrate (Advansta, San Jose, USA), with an exposure time between 60 and 90 sec.

bHEV and rHEV RT-qPCR and pan-HEV conventional RT–PCR

Total nucleic acids (TNs) from companion and stray animal samples were extracted as described in the supplementary material. Companion animal samples testing positive for HEV IgG in either EIA or stray animal rectal swab samples were individually extracted. Sera testing negative for IgG in both EIAs were combined into pools of five for extraction (supplementary material). We have previously demonstrated the utility of minipool testing for HEV RNA detection by RT-qPCR [9]. Total nucleic acids were stored at −80°C prior to RT-qPCR. bHEV and rHEV RT-qPCR assays were performed on minipools, individual sera and rectal swab TN samples. In addition, seropositive animals were tested with a pan-HEV consensus primer RT–PCR assay capable of detecting all genera within the subfamily Orthohepevirinae. These assays have been described by us previously [8]. Primers, probes, and detailed protocols are listed in the supplementary material and supplementary table 3.

Statistical analysis

Mann–Whitney U tests were performed to compare the median ODs of companion dog and cat samples in bHEV and rHEV EIAs. Mean ages were compared using Student’s t-test and proportions of genders were compared using the chi-square test. Graphs were plotted using GraphPad Prism software (San Diego, USA). P-value < 0.05 was considered statistically significant.

Results

Characteristics of animal samples

Archived sera of 751 companion dogs and 130 companion cats attending veterinary clinics in Hong Kong were retrieved. The median age of sampled dogs was 9 years (interquartile range (IQR): 2–11 years). 383 (53.0%) of 723 dogs with available gender data were male. The median age of companion cats was 2.4 years (IQR: 7 months-3 years). 75 (60.0%) of 125 cats with available gender data were male. 138 (18.4%) and 28 (3.7%) dogs were infected with Babesia and Ehrlichia respectively (diagnosed by PCR of blood), while one cat (0.8%) was infected with feline coronavirus. The stray animal cohort comprised 328 stray dogs and 324 stray cats sampled across the city as shown in supplementary Figure 1. 169/328 (51.5%) dog samples and 119/324 (36.7%) cat samples were collected in the densely populated Kowloon area where most human rHEV cases in Hong Kong have been reported [10,11].

Figure 1. Characteristics of HEV-A4-P239 and HEV-C1-P241 peptides. (A) Genome map of hepatitis E virus (HEV) showing Open Reading Frame 2 (ORF2) region corresponding to HEV-A4-P239 and HEV-C1-P241 peptides. S: shell, M: middle, P: protruding domains of ORF2 are highlighted in pink. (B) Sodium dodecyl sulphate-polyacrylamide gel electrophoresis of HEV-A4-P239 and HEV-C1-P241 showing bands at 40–55 kDa. (C) Immunoblots based on HEV-A4-P239 and HEV-C1-P241 showing that Paslahepevirus balayani (bHEV) infected patient sera (lane 1: bHEV genotype 3, lane 2: bHEV genotype 4) reacted in HEV-A4-P239 immunoblot, but not HEV-C1-P241 immunoblot. Rocahepevirus ratti (rHEV) infected patient sera (lane 3) reacted in HEV-C1-P241 immunoblot, but not HEV-A4-P239 immunoblot. N: blank lane; P: anti-His antibody control. (D) Optical density (OD) values of bHEV-positive and rHEV-positive patient-derived IgG spiked into dog and cat serum pools using rHEV and bHEV enzymatic immunoassays. Purple lines represent rHEV-patient IgG spiked into cat serum pool. Orange lines represent rHEV-patient IgG spiked into dog serum pool. Pink lines represent bHEV-patient IgG spiked into cat serum pool. Brown lines represent rHEV-patient IgG spiked into dog serum pool. Black lines represent HEV-negative patient IgG spiked into HEV-negative dog serum pool. Green lines represent HEV-negative patient IgG spiked into HEV-negative cat serum pool. Red lines represent IgG purified from the HEV-negative dog serum pool. Gray lines represent IgG purified from the HEV-negative cat serum pool.

Characterization of HEV-A4-P239 and HEV-C1-P241 peptides

HEV-A4-P239 and HEV-C1-P241 peptides are homologous truncated fragments of the ORF2 capsid proteins of bHEV and rHEV respectively (Figure 1(A)). These fragments contain most important antibody epitopes of HEV and self-assemble into virus-like particles improving conformational antigen presentation [32,33]. The amino acid identity between HEV-A4-P239 and HEV-C1-P241 is 54.8%. Both peptides formed bands at 40–55 kDa (Figure 1(B)), which matched the characteristics of peptide dimers described by us previously [18]. Immunoblots showed that plasma from bHEV genotype 3 and genotype 4 human patients reacted strongly in the HEV-A4-P239 blot, but not the HEV-C1-P241 blot (Figure 1(C)). Similarly, plasma from a rHEV patient formed bands in the HEV-C1-P241 immunoblot, but not the HEV-A4-P239 blot. HEV-A4-P239 and HEV-C1-P241 peptides were used to coat bHEV and rHEV EIA plates respectively. We have previously demonstrated that this EIA system is capable of differentiating bHEV and rHEV antibody signatures in human sera [17]. To investigate performance of these EIAs using dog or cat serum matrices, we spiked serially diluted IgG isolated from plasma of bHEV and rHEV positive patients into pools of 10 dog or cat sera that produced near-blank level ODs in a pilot evaluation in both bHEV and rHEV EIAs. Spiked samples were then tested in bHEV and rHEV EIAs. Elevated ODs were observed with good linearity when spiked samples were tested with their cognate EIAs (Figure 1(D)). As expected, bHEV IgG-spiked samples produced low ODs in rHEV EIA and vice versa.

bHEV and rHEV EIA of companion animal samples

Companion animal sera were tested in parallel in bHEV and rHEV EIA plates. OD values of 13/751 (1.73%, 95% CI 0.80%–2.66%) dog sera exceeded the cut-off set at mean + 3SD of the entire cohort in either or both EIAs (Figure 2(A)). These samples had robust OD values, ranging from 0.718 to 1.976 in bHEV and 1.029–2.145 in rHEV EIAs. Four dogs (0.53%) only had positive signals in the bHEV EIA, while five dogs (0.67%) only had positive signals in the rHEV EIA. Four dogs (0.53%) had positive signals in both EIAs. The rHEV EIA ODs of these four animals were higher than the bHEV EIA OD even though the bHEV EIA generally gave higher background signals when positive samples were excluded (supplementary figure 2). Median ODs of dog samples testing positive for DNA of Babesia or Ehrlichia were not significantly elevated compared to uninfected animals indicating no cross-reactivity with these pathogens (supplementary figure 3).

Figure 2. OD values of (A) dog and (B) cat serum samples tested in bHEV and rHEV EIAs. Blue dots and lines: sera testing positive only in bHEV EIA; red dots and lines: sera testing positive only in rHEV EIA; green dots and lines: sera testing positive in both EIAs. Horizontal black lines represent respective assay cutoffs (mean OD + 3 standard deviations of the group).

As shown in Figure 2(B), 5/130 (3.85%, 95% CI 0.54%–7.15%) cat sera had ODs exceeding the cut-off (mean + 3SD of the entire cohort), These samples had OD values ranging from 0.8677 to 1.938 in bHEV EIA and 0.8677–1.2635 in rHEV EIA. Three cats (2.31%) were only positive in the bHEV EIA, while one cat (0.77%) was only positive in the rHEV EIA. One cat (0.77%) was positive in both EIAs.

Confirmation with commercial EIA and HEV-C-P241 immunoblots

Serum samples testing positive in either bHEV or rHEV EIA were then tested using the Wantai kit protocol with modifications. This kit uses a recombinant antigen based on bHEV genotype 1 ORF2 (pE2 region). Samples testing negative in bHEV and rHEV EIA produced low OD values in the Wantai kit. For companion dog samples testing positive in the bHEV EIA (either rHEV EIA + or rHEV EIA-), 6/8 produced signals exceeding the cutoff (Figure 3(A)). However, only one of five samples solely testing positive in the rHEV EIA (bHEV EIA-) produced positive signals in the Wantai assay, which was expected as bHEV-based EIAs are known to frequently miss rHEV antibody responses in human sera [18]. Three companion cat samples including both rHEV-reactive samples gave positive results in the Wantai assay (Figure 3(B)).

Figure 3. Optical Density (OD) values bHEV and rHEV EIA positive companion (A) dog and (B) cat sera in Wantai assay. Gray dots: bHEV and rHEV EIA negative control samples (n = 10); blue dots: sera testing positive only in bHEV EIA; red dots: sera testing positive only in rHEV EIA; green dots: sera testing positive in both EIAs. Horizontal lines represent cutoffs derived from mean OD + 3 standard deviations of 40 companion animal sera testing negative in both bHEV and rHEV EIAs.

To further confirm the specificity of the rHEV EIA results, all 11 cat or dog samples testing positive in rHEV EIA (either bHEV EIA + or bHEV EIA-) were evaluated in an immunoblot based on HEV-C1-P241. As shown in Figure 4, all samples produced bands confirming a specific interaction between animal sera and the HEV-C1-P241 peptide. On the other hand, rHEV EIA sole positive sera tested negative in an immunoblot based on HEV-A4-P239 (supplementary figure 4).

Figure 4. HEV-C1-P241 immunoblot of rHEV EIA-positive companion animal sera samples. (A) Immunoblot of rHEV EIA-positive dog sera. (B) Immunoblot of rHEV EIA-positive cat sera. N: blank lane; P: anti-His antibody control. Blot assays were performed in a MiniProtean II Apparatus, which enables lane-by-lane separation of the immunoblot without cutting into strips.

Characteristics of seropositive animals

Characteristics of seropositive animals are presented in Table 1. The mean age of seropositive dogs with data available was 7.0 years while that of seropositive cats was 6.0 years. The mean ages of seropositive animals were not significantly different to that of seronegative animals (P-value = 0.918 for the dog cohort and P-value = 0.126 for the cat cohort). Gender distributions of seropositive and seronegative animals were also not statistically different. Although the exact residential address of the companion animals was not available, referring veterinary clinic address was taken as a surrogate of the residential district. Notably, nearly all rHEV EIA-positive animals were either from or adjacent to districts reporting human rHEV infections between January 2017 and June 2023 (supplementary figure 5).

Table 1. Characterization of bHEV and rHEV EIA seropositive companion animals, Hong Kong.

Animal code	EIA positivity	Age (years unless otherwise specified)	Gender	Residential district
D-A-1	bHEV only	10	M	Kowloon City
D-A-2	bHEV only	12	M	Wan Chai
D-A-3	bHEV only	1 month	F	Not provided
D-A-4	bHEV only	6	M	Not provided
D-C-1	rHEV only	5	F	North
D-C-2	rHEV only	7	F	Kowloon City
D-C-3	rHEV only	6	F	Sham Shui Po
D-C-4	rHEV only	2	F	Kowloon City
D-C-5	rHEV only	9	F	Tai Po
D-AC-1	bHEV & rHEV	9	F	Yau Tsim Mong
D-AC-2	bHEV & rHEV	8	M	Wan Chai
D-AC-3	bHEV & rHEV	Not provided	M	Sha Tin
D-AC-4	bHEV & rHEV	9	M	Yuen Long
C-A-1	bHEV only	1	M	Yau Tsim Mong
C-A-2	bHEV only	8	M	North
C-A-3	bHEV only	11	M	Tai Po
C-C-1	rHEV only	5 months	Not provided	Yau Tsim Mong
C-AC-1	bHEV & rHEV	8	M	Sha Tin

Dog and cat RT-qPCR and pan-HEV conventional RT–PCR

TN individually extracted from seropositive animal sera, serum minipools from companion animals and rectal swab samples from stray animals were tested using bHEV and rHEV-specific RT-qPCR. All samples tested negative. Pan-HEV conventional RT–PCR, which is capable of detecting all species within Orthohepevirinae, was additionally performed on the seropositive animals and was negative as well. Therefore, animals were not actively infected with HEV at the time of blood or rectal swab collection.

Discussion

The Hepeviridae family comprises several virus variants with the ability to cross species barriers. This is best characterized by bHEV genotypes jumping to humans from swine (genotypes 3, 4), deer (genotype 3), rabbits (genotype 3ra), and camels (genotype 7) [34]. The genus Rocahepevirus was considered restricted to rodents, ferrets, and shrews. However, since 2018, we have reported endemic rHEV transmission in humans in Hong Kong with population seroprevalence indicating low-level endemicity [10,17]. Therefore, the distribution of Rocahepevirus variants in animals may be underestimated. Identifying rHEV-susceptible animals with human contact is essential because transmission routes from rats to humans are currently undetermined.

We investigated potential exposure to rHEV and bHEV in cats and dogs. Several studies have examined HEV seroprevalence in dogs and cats, although nearly all of these studies used methods inappropriate for rHEV antibody detection (supplementary table 1). Some investigators found higher HEV seroprevalence in pet owners and small-animal veterinary practitioners, but this is not supported in all studies [26,35–39]. Contact with dogs has been identified as a risk factor for HEV seropositivity in a recent meta-analysis, but significant publication bias renders any conclusion doubtful [40].

We provide evidence using multiple assays that companion cats and dogs mount antibodies to bHEV and rHEV or related antigens in Hong Kong, a city where both variants are endemic. A major strength of our study was the use of separate bHEV and rHEV-specific EIAs for initial screening, whereas previous studies have relied on approaches that were biased towards bHEV antibody detection and did not differentiate bHEV and rHEV antibody responses. The EIA systems used in this study have been extensively validated using human infection sera and use a high serum dilution to limit non-specific reactivity [17]. We have previously shown that cross-reactive responses are common in rHEV-infected patients who might have been dually exposed to the more prevalent bHEV in the past [18]. In this study, we again demonstrated both single and dual reactivity patterns to bHEV and rHEV EIAs. This finding is supported by a recent study in Cordoba, another city reporting sporadic rHEV human cases, which found that a subset of cats and dogs testing positive in bHEV-based EIAs produced rHEV-specific antibody responses on western blot [41]. However, we found that many animals only mount detectable responses in a rHEV EIA (prevalence similar to bHEV mono-exposure), which would have been missed with a bHEV-based screening approach.

However, no study to date has detected HEV RNA in cat or dog samples (supplementary table 1). We confirmed that this is the case even when rHEV-specific RT-qPCR primers are used. There are three possible explanations for this. Firstly, HEV infections in cats and dogs may be sporadic, subclinical, or transient (as is the case with most human infections); RT-qPCR testing outside this narrow window may miss these infections. It is also possible that cats and dogs rapidly mount immune responses to HEV without productive infection of hepatocytes. There is some indication that this is the case with bHEV in dogs; Liu et al. attempted to infect two dogs with bHEV and elicited an antibody response without detectable virus shedding [30]. On the other hand, canine liver cancer cells support replication of the Kernow-C1 p6 clone of bHEV genotype 3 [23]. Another possibility is that these animals may be susceptible to an as-yet-undiscovered HEV variant or related antigenic agent. Extensive prospective investigations in seroconverting cows and goats have failed to identify HEV infection [42,43]. Red foxes, also members of the family Canidae, have been putatively linked to a Rocahepevirus-like HEV variant that is distantly related to vole HEV (Rocahepevirus eothenomi) [44,45]. However, these variants have mostly been detected in faeces that might be explained by RNA contamination due to the consumption of voles. Lastly, cats and dogs could seroconvert to HEV variants due to exposure via the oral route without productive infection. rHEV is enzootic in rodent populations and ground-level contamination by rHEV in rodent excreta is probably quite heavy. Interestingly, rHEV seropositive animals in this study were often from parts of Hong Kong reporting rHEV infections in humans. Support for a dietary source comes from studies finding that seroprevalence in dogs is higher in animals fed kitchen waste or other foods compared to pet food [46,47].

A limitation of this study is that we did not have access to serum samples from stray animals. Stray animals may be more heavily exposed to rHEV compared to household pets. Also, faecal samples from stray and companion animals were not collected and assessed in the study; this may have lowered our ability to detect rHEV RNA compared to rectal swabs. We acknowledge that our immunoassay required a relatively high sample dilution that may have affected seroprevalence estimates. We lacked a virus neutralization assay, which is considered a gold-standard serology assay because of the difficulty in growing rHEV in vitro and the large serum volumes required to demonstrate neutralization [48]. We compensated by including confirmatory commercial EIAs and immunoblots, which confirmed a specific interaction between antibodies and ORF2 protein of respective HEV species. Another limitation of our study was that we did not have access to antisera from cats and dogs immunized with the HEV-C1-P241 peptide.

In conclusion, we demonstrate that dogs and cats harbour antibodies reactive against rHEV or a related antigen. Further studies are required to confirm the susceptibility of dogs and cats to rHEV. This has public health significance given the emerging role of rHEV in human health.

Acknowledgements

We thank the Agriculture, Fisheries, and Conservation Department of the Hong Kong Special Administrative Region Government for providing us with the stray animal samples used in this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by the Health and Medical Research Fund, the Food and Health Bureau, the Government of the Hong Kong Special Administrative Region under grants CID-HKU1-12 & 20190362, the HKUMed Research Fellowship Scheme for Clinical Academics, and the Partnership Programme on Enhancing Laboratory Surveillance and Investigation of Emerging Infectious Diseases and Antimicrobial Resistance for Department of Health.