Rat Hepatitis E Virus Linked to Severe Acute Hepatitis in an Immunocompetent Patient

Abstract

Hepatitis E virus (HEV) is a major public health concern in developing countries where the primary transmission is via contaminated water. Zoonotic HEV cases have been increasingly described in Europe, Japan, and the United States, with pigs representing the main animal reservoir of infection. We report an unusual acute hepatitis infection in a previously healthy man caused by a rat HEV with a considerably divergent genomic sequence compared with other rat HEV strains. It is possible that rat HEV is an underrecognized cause of hepatitis infection, and further studies are necessary to elucidate its potential risk and mode of transmission.

hepatitis E virus, Hepeviridae, phylogenetic analysis, Western blot, immunocompetent

Hepatitis E virus (HEV) is transmitted mainly by the fecal-oral route and causes large outbreaks in developing countries with suboptimal sanitary conditions related to contaminated water. In recent years, sporadic, locally acquired cases have been increasingly occurring in many countries with good sanitary standards where the transmission is foodborne, mainly through consumption of undercooked meat from domestic pigs [1, 2]. The disease is often self-limiting, although it can become chronic in immunocompromised patients [3]. HEV transmission through blood transfusion is also becoming a concern in industrialized countries [4].

The virus is classified as a member of the family Hepeviridae, genus Orthohepevirus A [5]. At least 4 HEV genotypes cause disease in humans; HEV genotypes 1 and 2 (gt1 and gt2) infect only humans and are usually associated with waterborne infections, whereas gt3 and gt4 have been detected in humans, domestic pigs, wild boar, and deer. A single human case of a new HEV genotype (HEV gt7) found in camels has been recently described [6]. Other HEV-like viruses have been identified in moose, ferrets, fox, rats, chicken, and bats. Rat HEV was first discovered in Germany [7] and later found in the Unites States, Vietnam, Indonesia, and China [8]. Genetically, rat HEV is distantly related to other mammalian HEV, with approximately 55%–60% sequence homology to HEV gt1-4, and it is classified in a different genus, Orthohepevirus C [5].

The host range of rat HEV is poorly characterized. Based on the inability to infect rhesus monkeys with a North American rat HEV isolate, it has been considered an unlikely human pathogen. We performed virological, immunological, and partial epidemiological analyses of a Canadian case with acute hepatitis caused by a novel rat HEV strain.

METHODS

Case Description

A 49-year-old man who worked with the United Nations (UN) was admitted on 8 April 2017 to Queen Elizabeth II Health Sciences Centre, Halifax, Canada, for severe acute hepatitis. His medical history was unremarkable aside from dyslipidemia, for which he took atorvastatin (20 mg/d taken orally). Before presentation, he had been based at different UN facilities in the Democratic Republic of Congo and Gabon (January to 12 March 2017). Shortly after returning home to Canada, he vacationed in the Caiman Islands (29 March to 4 April 2017).

The patient presented to care on 7 April 2017 with 48 hours of urticarial rash, jaundice, steatorrhea, nausea, and decreased appetite. He was afebrile and hemodynamically stable but profoundly jaundiced. Initial investigations showed severe transaminitis (Supplementary Table 1). Doppler ultrasound of the liver demonstrated extensive gallbladder wall thickening compatible with reactive changes related to acute hepatitis, no ductal dilatation, and no signs of Budd-Chiari syndrome. Results of diagnostic testing for infectious causes, including hepatitis A, B, and C viruses, cytomegalovirus, parvovirus B19, and malaria, were all negative, as were results of subsequent testing for antinuclear antibodies and antimitochondrial antibodies were negative; α₁-antitrypsin and ceruloplasmin levels were within normal limits.

The patient underwent percutaneous liver biopsy, which showed severe acute hepatitis with inflammation, interface activity, and significant hepatocyte damage (Figure 1), with no histochemical evidence of cytomegalovirus (immunostain), herpes simplex virus, Epstein-Barr virus (in situ hybridization), spirochetes, fungi, or bacteria. The cause of the patient’s hepatitis was not clear, and he was managed expectantly, with no specific therapy aside from statin discontinuation. He was discharged home on 12 April with resolving transaminitis and with plans for outpatient infectious diseases and hepatology follow-up. He was subsequently seen in May, June, September, and November 2017, at which times he was asymptomatic with no biochemical evidence of hepatitis. The patient gave his consent for publication of the data.

Figure 1.

Liver histopathology. A, An inflamed portal tract is seen on the upper left. The inflammation is predominantly lymphoplasmacytic and spills out into the parenchyma (interface activity). The lobule shows disarray, with swollen hepatocytes, lobular inflammation (including ceroid-laden macrophages in areas of hepatocyte dropout [arrows]), and acidophilic bodies (arrowheads). The hepatic venule is seen on the right (hematoxylin-eosin, original magnification ×200). B, Lobular activity. This image shows an acidophilic body (apoptotic hepatocyte [white arrow]), areas of hepatocyte dropout (prior necrosis [black arrow]), swollen hepatocytes with focal regeneration (mitotic figure [black arrowhead]), lobular lymphoplasmacytic inflammation, and ceroid-laden macrophages (hematoxylin-eosin, original magnification, ×400).

Polymerase Chain Reaction and Phylogenetic Analysis

Nucleic acid extraction from patient’s serum sample was performed using the NucliSENS easyMAG nucleic acid automated extraction system (BioMerieux) and further tested for the presence of HEV RNA with a commercial real-time polymerase chain reaction (PCR) assay (RealStar HEV RT-PCR; Altona-Diagnostics) and a broadly reactive heminested PCR assay that detects largely divergent HEV variants [9]. Amplicons were gel purified before cycle sequencing, using an ABI 3730XL DNA Analyzer (Applied Biosystems). Sequences were assembled and analyzed with Lasergene sequence analysis software (DNASTAR). The full-length HEV genomic sequence was aligned with other complete genome HEV rat sequences for phylogenetic analysis. HEV PCR amplification of the patient’s paraffin-embedded liver biopsy was also performed (Supplementary Data).

Characterization of the Humoral Immune Response to HEV Rat and Human Recombinant Capsid Antigens

Laboratory testing for routine HEV serology was performed at the National Microbiology Laboratory using a Wantai anti-HEV immunoglobulin G (IgG)/immunoglobulin M (IgM) enzyme-linked immunosorbent assay (Wantai Biological Pharmacy). Further serologic testing was done with several recombinant HEV proteins. Briefly, capsid protein-encoding sequences for human HEV gt1 and gt3 and rat gt1 (GenBank accession Nos. M73218, AF082843, and GU345042, respectively) and the newly identified in this study rat HEV were tagged with 6x-His and 8x-FLAG tags at their C-termini, and expressed in a vaccinia virus system, according to per a protocol described elsewhere [10].The HEV capsid proteins collected from the recombinant vaccinia virus–infected cell lysate were purified with a Dynabeads His-tag pull-down kit (Invitrogen). The purified HEV antigens were examined using a conventional Western blotting procedure with acute-phase serum samples from this patient (patient 1683) and 3 other patients infected with HEV gt1 (patients 0972 and 2287) and gt3 (patient 4692). Odyssey Fc Imaging System—LI-COR Biosciences was used for Western blot digital imaging (Supplementary Data).

RESULTS

A blood sample collected on 10 April 2017 was sent to the National Microbiology Laboratory for HEV serology. The anti-HEV IgG was positive, and anti-HEV IgM (a marker for current infection) was borderline positive. Testing using a commercial real-time HEV PCR assay was negative; however, an in-house broadly reactive heminested PCR yielded an amplicon, which, when sequenced, showed 85% homology with a European HEV rat strain (accession No. MF480317), but was highly divergent (62.3% homology) from the human HEV gt1 prototype strain SAR-55 (accession No. M80581).

We used a conventional PCR-based strategy for genome walking to extend the HEV rat sequence obtained from the heminested in-house PCR and amplify its flanking regions by designing a set of downstream/upstream consensus primers based on conserved regions from an alignment of the 10 known complete HEV rat genomes. The full-length HEV genomic sequence of this particular viral strain, labeled HEV H17-1683 (GenBank accession No. MK050105), was obtained by conventional Sanger sequencing of 12 overlapping PCR fragments (Supplementary Figure 1). Excluding the poly(A) tail the genome is 6939 nucleotides long, and its sequence identity to HEV gt1 and gt3 is only 57%. The highest nucleotide homology (range, 80.5%-81.1%) was observed with a group of rat HEVs from Europe and United States provisionally classified as genetic group 1 and with rat HEVs from Indonesia and Vietnam proposed to cluster in genetic groups 2–3 (range, 76.2%-77.2%). The genome encodes the typical open reading frames (ORFs) for HEV 1–3 of 1636, 644, and 102 amino acids, respectively. In addition the putative ORF 4 found in rat HEV was also predicted with the expected size of 183 amino acids.

Results of phylogenetic analyses based on the nucleotide sequence comparison of HEV H17-1638 strain with other rat HEVs suggest that it may represent a new genetic group (Figure 2A). It is genetically distinct (28.1% divergence) from the recently described strain from a Hong Kong transplant recipient infected with rat HEV [11]. Examination of the protein sequence data inferred by ORFs 1-3 also supports the putative segregation of HEV 17-1638 from the rest of the rat HEVs (Supplementary Figure 2).

Figure 2.

Phylogenetic tree of the novel hepatitis E virus HEV 17-1683 strain and its capsid protein reactivity as shown by Western blot. A, Genetic relationship based on 16 rat HEV strains whose entire genomic sequences are known, including those causing human infection (HEV 17-1683 [black dot] and MG813927 marked [black square]) in a maximum likelihood phylogenetic tree together with members of the other 3 genera. Each rat HEV strain label includes the accession number followed by the strain name. The 3 phylogenetic clusters provisionally designated as genotypes 1–3 (gt1–gt3) are indicated by braces, and the HEV 1–4 clades are marked by a vertical dashed line. Only bootstrap values >70% are included. The scale bar indicates the number of nucleotide substitutions per site. B, Western blotting detection of different HEV recombinant capsid antigens with serum samples from HEV-infected patients. Gel 1 shows the FLAG-tagged recombinant HEV gt1 and rat HEV 17-1683 capsid proteins. For gel 2, an acute-phase serum sample from patient 0972 was incubated with human HEV gt1 and its homologous rat HEV 17-1683 antigens, respectively; for gel 3, an acute-phase serum sample (1:200) from a patient (patient 0972) infected with HEV gt1 was used. Note the cross-reactivity, but also the stronger signal between the homologous antigen-antibody pairs. Black and white arrowheads indicate the sizes of the rat HEV and HEV gt1 capsid antigens, respectively. With gels 4–6, additional recombinant antigens were used, human HEV gt3 and rat HEV gt1 (GU345042). The membrane from gel 4 was incubated with an acute-phase serum sample from another patient (patient 2287) infected with HEV gt1, and for gel 5 an acute-phase serum sample from patient 4692 infected with HEV gt3 was used. As with the previous 2 gels, band reactivity was stronger between homologous antigen-antibody pairs. Note that, as previously observed, HEV gt1 and 3 cannot be differentiated using this technique, however band intensity for rat antigens is weaker. The above-mentioned acute-phase serum samples were all positive for both anti-HEV immunoglobulin G and immunoglobulin M. The HEV recombinant antigens in gel 6 were exposed to a convalescent-phase serum sample from a patient with previous HEV gt3 infection positive for anti-HEV IgG only (9 months after onset of disease); note the weak reactivity with HEV gt1 and 3 antigens, but not with the rat HEV antigens.

The etiologic role of this HEV strain was further confirmed by the identification of HEV viral RNA in the liver. RNA was extracted from paraffin-embedded liver tissue, and HEV was amplified by means of in-house HEV real-time PCR and 2 single-stage reverse-transcription PCR assays based on the 5’ untranslated region and the 3’ end of ORF 2 (Supplementary Data). The HEV sequences obtained from the liver biopsy and plasma samples were identical.

The specificity of the patient’s antibody response was assessed using Western blot analysis; we compared the immunoreactivity of acute-phase serum samples from the patient in the current study (patient 1683), patients 0972 and 2287 with acute HEV gt1 infection, and patient 4692 with acute HEV gt3 infection, using recombinant antigens derived from human HEV gt1, gt3, HEV 17-1683, and rat HEV gt1 (Figure 2B). A convalescent-phase serum sample (anti-HEV IgG⁺/IgM⁻) from patient 2199 with resolved HEV gt3 infection was also used. The acute-phase serum sample from our patient (patient 1683) showed stronger reactivity with its homologous recombinant antigen than that of human HEV gt1 (Figure 2B). The opposite was true when an acute-phase serum sample from a patient with HEV gt1 infection was used. This pattern of cross-reactivity—having stronger band reactivity with homologous and weaker with heterologous antigens—was observed with other HEV gt1 and gt3 acute-phase serum samples. Interestingly, an anti-HEV IgG⁺/anti-HEV IgM⁻ convalescent-phase serum from patient 2199 with HEV gt3 infection reacted only with the recombinant antigens gt1 and gt3, but not with this patient’s or HEV rat gt1 antigen. Similar results were observed when anti-IgM Western blotting was performed (Supplementary Data).

In view of the unusual HEV infecting virus strain, our patient was reinterviewed and specifically asked about contact with rats, rat droppings, or any other live animal exposure while in Africa, which he denied. He ate at the UN base, at his hotel, and twice at a local Indian restaurant. Only the UN base would have had any approved sanitary or cooking standards. His diet consisted mostly of eggs, fruit, meat, and coffee. He drank only UN-approved bottled water. In late February 2017, the patient had experienced 1 episode of gastrointestinal upset/diarrhea.

DISCUSSION

We describe a severe case of acute hepatitis caused by a novel strain of rat HEV. The diagnosis might have been missed if not for the use of a broadly reactive PCR assay that detects largely divergent HEV variants. The case progression begins with modest transaminitis in the prodromal period and urticarial rash, followed by severe clinical presentation and full recovery. Rashes are not uncommon in acute or chronic hepatitis B or C virus infection, but have not been described in HEV infection [12]. Apart from this, the clinical features of the case did not differ from the standard acute viral hepatitis presentation: anorexia, nausea, fatigue, and jaundice. Generally, acute HEV infections resolve spontaneously within 4-6 weeks, and this case was no exception.

Recent serologic data suggest that humans can be infected with rat HEV. Anti-HEV IgG antibodies in some forestry workers in Germany showed higher reactivity to rat HEV than the typical pig-related HEV gt3 [13]. Likewise, IgG and IgM antibodies to rat HEV antigen were detected in a febrile patient from Vietnam with hepatitis [14]. Anti-HEV specific antibodies from the patient in the current study also reacted more strongly with the homologous rat capsid antigen than with the HEV gt1 and gt3 antigens. However, interpretation of these findings is impeded by the antigenic cross-reactivity among rat and human HEV antigens [15]. Thus, the detection of the rat HEV genome in acute-phase serum samples from the patient is essential for the recognition of rat HEV infection.

This is the second recording of human infection with rat HEV, just months after the first world’s ever case was identified in Hong Kong [11]. Our patient contracted the infection while working in Gabon and Democratic Republic of Congo. The HEV strain in this study is genetically distant from the Hong Kong HEV strain, as well as from all other rat HEV isolates, and tentatively could be assigned to a new genotype within the HEV C1 clade. Preliminary analyses could not identify common genomic features between the two viruses, associated with interspecies transmission, leaving open the possibility that zoonotic transmission of rat HEV might also occur by other members of the HEV C1 clade .

The Hong Kong liver transplant recipient lived in rat-infested housing estate and was immunosuppressed, which may have facilitated the interspecies transmission [11]. The rat HEV infection described in the current article occurred in a previously healthy, immunocompetent patient with no obvious contacts with rodents. Therefore, apart from the sequence relatedness, there is no direct evidence linking the patient to rats, so a secondary, still-unknown route of transmission may be involved. Further studies are needed to assess the zoonotic potential and mode of transmission of rat HEV in humans.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. The posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. We thank the patient for a retrospective re-interview and for providing epidemiologically related information relevant to this study.

Financial support. This work was supported by the Public Health Agency of Canada.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.