Wild R. norvegicus rats were live-trapped by vector-control personnel in urban Los Angeles. The rats were anesthetized, and age and species was determined. Reproductive status and weight were recorded. Blood was obtained by cardiac puncture, and the serum was stored at −70°C.

We performed transmission studies with genotype 1 strains Sar-55 (15), Kashi-87 (16), Akluj-90 (17); genotype 3 strain Meng swine HEV (18); and genotype 2 strain Mex 14 (19). All strains were in 10% fecal suspensions, diluted as described in the Results, and all but 1 had been titered for infectivity in nonhuman primates or pigs (Table 1).

Because infectivity of HEV in nonhuman primates is ≈10,000-fold less when administered orally than when administered parenterally, commercially acquired, outbred, Sprague-Dawley (R. norvegicus) or athymic nude hooded laboratory rats (Harlan, Indianapolis, IN, USA) or rhesus monkeys (Macaca mulatta) that were bred and raised in captivity were infected intravenously with serum or homogenized 10% fecal or tissue samples in saline. The animals were housed and maintained at Bioqual, Inc. (Rockville, MD, USA). Housing and care of the animals complied with all relevant guidelines and requirements, and the animals were housed in facilities that are fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. All protocols were reviewed and approved by the Institutional Animal Care and Use Committees of the National Institute of Allergy and Infectious Diseases of the National Institutes of Health (Bethesda, MD, USA) and Bioqual, Inc.

Blood samples were obtained weekly and feces were obtained daily to 3×/wk. Serum samples were tested for liver enzyme levels by using standard methods (AniLytics, Inc., Gaithersburg, MD, USA). Postmortem liver tissue was fixed in formalin, embedded, sectioned, and stained with hematoxylin and eosin (American Histo Laboratories, Inc., Gaithersburg, MD, USA) and read under code by one of the authors (S.G.). Samples were scored for liver pathologic changes by the histologic activity index method.

Serum samples were tested for IgG and IgM isotypes against HEV by using a peroxidase-based ELISA as reported (12). The antigen used was recombinant open reading frame 2 protein of genotype 1. Serum samples were tested at 10-fold dilutions, and the highest dilution exceeding the cutoff value of optical density was taken as the endpoint titer of the serum.

RNA was extracted from 270 µL of serum by using Trizol LS (Invitrogen, Carlsbad, CA, USA), and purified RNA was resuspended in 20 μL of water. Nested reverse transcription PCR (RT-PCR) was performed with the same primers, enzymes, and thermal profiles as described (11). Nested PCR products were separated by electrophoresis on ethidium bromide–stained agarose gels, extracted from the gel, and sequenced to provide the consensus sequence.

RNA was extracted from 50 μL of serum, tissue suspension, or filtered (0.22 μm, UltrafreeMC; Millipore, Billerica, MA, USA) 10% fecal suspension by using the QIAamp Viral RNA Mini Kit, (QIAGEN, Valencia, CA, USA), and total RNA was eluted in a volume of 60 μL. A TaqMan assay was performed by using the 7900HT Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s recommendations. The primers (for a 332-base amplicon) consisted of 900 nmol/L forward (5′-ATG GTG CTT TTA TGG CGA TTG-3′) and 900 nmol/L reverse (5′-CAA ACT CAC TGA AAT CAT TCT CAA AAA C-3′), and 250 nmol/L probe (5′-6FAM-TAT GTT CAG GAG AAG TTG GAA GCC GCT GT-TAMRA-3′). One-step quantitative RT-PCR (qRT-PCR) cycling conditions were 15 min at 48°C, a 10-min incubation at 95°C, and 50 cycles for 15 s at 95°C and 1 min at 60°C. Rat TaqMan cycle threshold values were indirectly quantified against an in-house HEV genotype 1 quantity standard line that represented a 6-log dynamic range.

Because isolation of mammalian HEV strains from rats had been reported, we attempted to transmit to laboratory rats 6 well-characterized mammalian HEV strains (genotypes 1, 2, and 3) that can infect primates or pigs (Table 1). Adult Sprague-Dawley rats were injected intravenously with 0.1 mL of inoculum through the tail vein. Rats were bled weekly for 16 weeks and monitored for HEV RNA by real-time PCR with genotype-specific primers and for development of antibodies against HEV by ELISA. None of the animals had any evidence of infection.

We had reported that wild rats trapped in Baltimore, Maryland, and the Hawaiian Islands had prevalences of antibodies against HEV of 77%–94% (12). We tested 134 serum samples from R. norvegicus rats trapped in urban Los Angeles. Donor rats were a mixture of male and female animals and adults and juveniles weighing 26–508 g. Of these animals, 105 (78.4%) were positive for IgG against HEV (with or without IgM against HEV), 2 (1.5%) were positive for IgM against HEV only, and 27 (20.1%) were seronegative when tested by ELISA with antigen derived from human HEV (14). As we described (12), prevalence of antibodies against HEV increased with weight as a measure of age, and ≈50% of the youngest rats were already positive for antibodies against HEV (Figure 1).

Because HEV is neutralized by antibodies against HEV, seronegative serum samples or IgM-positive serum from animals early after infection offer the greatest chance of recovering infectious virus. Therefore, 6 pools of serum were prepared from 27 seronegative rats, and 250 μL of serum from each pool and individual serum samples from 2 IgM HEV-positive rats and 5 IgG/IgM–positive rats weres used to infect Sprague-Dawley rats. All 13 rats remained negative for HEV RNA, and only 3 rats (peak ELISA titers 100–400) seroconverted. Two of the seroconverted rats had received pooled seronegative serum, and the other had received IgM-positive serum.

Feces obtained during the first 4 weeks postinfection from the 3 seroconverting rats (B76, B79, and B84) were homogenized and pooled, and 200 µL was used to inject 4 additional rats each. Only 3 of 12 rats injected with feces from rats B76 or B84 seroconverted. Serum was less efficient at transmitting virus, and 0 of 8 rats were infected. To determine the duration of shedding, infectious virus was identified by seroconversion of rats that had been infected intravenously with 200 µL of 10% suspensions of individual serial fecal samples obtained every other day from rats B76 and B84 during the first 4 weeks postinfection; fecal suspensions were also tested by RT-PCR. Feces from rat B76 were positive for >11 days and feces from rat B84 were positive for >9 days (Table 2). Periods of PCR positivity coincided with intervals of transmissibility on the basis of seroconversions in recipient rats. However, none of the recipient rats became viremic.

Because serum and feces were poor sources of transmissible virus, we tested other clinical materials. Groups of rats injected with fecal pools from rat B76 and rat B84 were exsanguinated on various days, and serum, liver, and intestinal contents were harvested. Serum from these rats was injected into individual rats, which were tested for seroconversion. Only 2 serum samples (from rats B300 and B182) transmitted virus to a new rat. The liver of rat B182 was used for further transmission studies.

To establish a more robust infection, we injected nude rats, which lack a functional adaptive immune system. Nineteen nude rats were injected with 200 µL of a 10% liver homogenate from rat B182 at a dilution of 10⁻¹ and 1 rat was killed daily (days 2–20). We then used 200 µL of a 10⁻² dilution of serum from the killed rats to infect Sprague-Dawley rats. Only 3 of these rats seroconverted, indicating that only 3 of the nude rats (killed on days 13, 15, and 19) had infectivity titers >10². One of these 3 nude rats, rat B350, was further studied.

To determine the infectivity titer of rat HEV in liver, serum, and feces of selected infected rats, reverse titrations were performed with Sprague-Dawley rats and were monitored for seroconversion (Table 3). In Sprague-Dawley rats, 50% rat infectivity doses (RID₅₀) of 10⁴–10⁵/g of liver tissue were observed; in nude rat B350, a titer >10^6.0/g of liver and a titer of 10^3.7 in serum were detected. Titers of virus in feces and intestinal contents of Sprague-Dawley rats were <10¹ and <10³, respectively. Feces from nude rats were not tested. These samples were also titered for PCR positivity by qRT-PCR (Table 3). PCR titers of rat HEV paralleled infectivity titers but averaged an ≈10–100-fold higher titer.

A 327-nt product was amplified from the liver of rat B350 by nested RT-PCR and directly sequenced to yield the consensus sequence. The B350 rat virus sequence was as genetically similar to the 2 rat sequences from Germany as they were to each other at the nucleotide and amino acid levels (Table 4).

We have shown that some mammalian HEV strains show a dose response: higher doses (>10⁴ infecting virus) are more likely to be associated with higher serum liver enzyme levels. To determine whether this phenomenon was true also for rat HEV, we infected 6 Sprague-Dawley rats with 200 µL of liver homogenate from rat B350 that contained 10^4.5 RID₅₀ of rat HEV. Animals were bled 2×/wk, and levels of alanine aminotransferase, γ-glutamyl transpeptidase, and isocitrate dehydrogenase were measured for 3 months. All 6 animals seroconverted 2.0–3.5 weeks (mean 3.0 weeks) postinfection (Figure 2). As reported, liver enzyme levels varied considerably, but seroconversion and liver enzyme levels were not temporally associated. Thus, these infections were biochemically inapparent infections.

Two HEV-infected rats (B182 and B300) and 2 uninfected Sprague-Dawley rats were examined under code for histologic evidence of hepatitis. The 2 uninfected rats had essentially normal livers. Rat B182 had parenchymal foci of necrosis and aggregates of lymphocytes and Kupffer cells in hepatic lobules and had mild portal inflammation (Figure 3). Rat B300 had similar but less obvious lesions. This mild hepatitis was consistent with normal liver enzyme levels measured in serum of these animals.

Because rats and humans often share the environment, especially in inner cities, we tested whether rat HEV was transmissible to nonhuman primates. Seronegative rhesus monkeys, which are surrogates of humans, were injected intravenously with rat liver homogenate containing 10^3.5 RID₅₀ of rat HEV from rat B182 or with 10^5.2 RID₅₀ of rat HEV from rat B350. The animals were monitored for 15 weeks for seroconversion by ELISA and for genomic RNA by qRT-PCR. The animals remained negative (Table 5). Thus, rat HEV does not appear to be transmissible to rhesus monkeys.