Evaluation of antibody-based and nucleic acid-based assays for diagnosis of hepatitis E virus infection in a rhesus monkey model

Abstract

We have evaluated four hepatitis E virus (HEV) specific antibody assays, using sequential samples taken from 86 rhesus monkeys at intervals for up to 86 weeks after they had been infected with different doses of HEV. The animals are a common experimental model of hepatitis E. The large collection of sequential samples used avoids uncertainties encountered in previous studies regarding the precise infection status of study subjects and minimizes bias due to the individuality of response to infection. One assay (YES IgG) was produced with synthetic peptides; the others (E2 IgM, E2 IgG, and GL IgG) were produced with recombinant antigens. The results were compared with the viral RNA contents of the serum and stool samples and the occurrence of these virological and immunological markers in the course of the infection was temporally related to the development of hepatitis. Diagnostic utility of the markers was assessed according to their response rates and prevalence at different times in the course of infection. All the animals produced E2 IgG and developed viremia and all but one also produced E2 IgM and excreted the virus in stool, whereas response rates for the other antibodies were lower and decreased with virus dose. Hepatitis occurred over a period of 4 weeks between 3 and 7 weeks after infection. Virological activity occurred mainly during the incubation period and the prevalence of viral markers declined rapidly after the onset of hepatitis. Production of the E2 antibodies immediately preceded the onset of hepatitis, and this was followed about one week later by production of the other antibodies. Seroprevalence E2 IgM reached a peak value 3 weeks after the onset of hepatitis, whereas seroprevalence of GL IgG and YES IgG peaked after the disease had subsided. E2 IgG persisted in all animals for the entire duration of the experiment of up to 86 weeks and possibly beyond and, thus, can serve as a useful epidemiological marker of HEV infection. (C) 2003 Wiley-Liss, Inc.