Acute hepatitis A virus infection is associated with a limited type I interferon response and persistence of intrahepatic viral RNA

Abstract

Hepatitis A virus (HAV) is an hepatotropic human picornavirus that is associated only with acute infection. Its pathogenesis is not well understood because there are few studies in animal models using modern methodologies. We characterized HAV infections in three chimpanzees, quantifying viral RNA by quantitative RT-PCR and examining critical aspects of the innate immune response including intrahepatic IFN-stimulated gene expression. We compared these infection profiles with similar studies of chimpanzees infected with hepatitis C virus (HCV), an hepatotropic flavivirus that frequently causes persistent infection. Surprisingly, HAV-infected animals exhibited very limited induction of type I IFN-stimulated genes in the liver compared with chimpanzees with acute resolving HCV infection, despite similar levels of viremia and 100-fold greater quantities of viral RNA in the liver. Minimal IFN-stimulated gene 15 and IFIT1 responses peaked 1-2 wk after HAV challenge and then subsided despite continuing high hepatic viral RNA. An acute inflammatory response at 3-4 wk correlated with the appearance of virus-specific antibodies and apoptosis and proliferation of hepatocytes. Despite this, HAV RNA persisted in the liver for months, remaining present long after clearance from serum and feces and revealing dramatic differences in the kinetics of clearance in the three compartments. Viral RNA was detected in the liver for significantly longer (35 to >48 wk) than HCV RNA in animals with acute resolving HCV infection (10-20 wk). Collectively, these findings indicate that HAV is far stealthier than HCV early in the course of acute resolving infection. HAV infections represent a distinctly different paradigm in virus-host interactions within the liver.

Fig. 1.

Persistence of hepatic viral RNA in HAV-infected chimpanzees. Two chimpanzees (4x0293 and 4x0395) were inoculated intravenously using 5,000 chimp infectious doses of HAV HM175 containing 3 × 10⁷ GE (a gift of Robert Purcell, National Institutes of Health, Bethesda, MD). A third animal (4x0396) became infected naturally while housed with infected animals. The exact day of natural infection was not known, but alignment of the three infection profiles by peak viremia, seroconversion for HAV IgM, and peak ALT suggests that infection of 4x0396 occurred 4 wk after inoculation of his cage mate, 4x0293; thus, this week was designated as week 1 for 4x0396. HAV RNA was quantified by TaqMan RT-PCR assays. The limit of detection varied in each compartment due to differences in the samples processed for the assay: serum (red line: linear limit of detection 10³ GE/mL serum), liver (green line: linear limit of detection 10¹ GE/μg total RNA), and feces (blue line: linear limit of detection 5 × 10³ GE/gm feces). Samples from all three compartments were analyzed throughout the study, but only positive samples are shown in the line graphs. The number to the right of each line represents the calculated t_1/2 for clearance from that compartment. ALT levels are indicated by the gray shaded area. IgM anti-HAV and total anti-HAV was determined by ELISA and are shown as positive or negative (+ or −) at the top of the graphs. The duration of follow-up differs for each animal.

Fig. 2.

Comparative features of acute HAV and acute resolving HCV infection in the chimpanzee. (A) Maximum viral RNA copy numbers in liver tissue (GE/μg total RNA) and serum (GE/mL). HCV infections were with genotype 1 (open circle) or genotype 3 (solid circle, 4x0296) virus. Dashed lines indicate mean values. (B) Maximum serum ALT activities. The shaded areas at the bottom indicate the mean baseline ALT activity plus one and two SDs. (C–E) Mononuclear cell infiltration, cell death, and proliferation in acute HAV infection. Formalin-fixed, paraffin-embedded liver tissue obtained 4 wk after HAV challenge (animal 4x0395) was (C) stained with hematoxylin and eosin, (D) immunostained with antibody to activated caspase 3 as a marker of apoptosis, or (E) labeled with antibody to Ki67 as a marker of cell proliferation. Liver tissue samples from preinfection and weeks 3, 4, and 10 after infection are shown in Fig. S2.

Fig. 3.

Intrahepatic ISG responses determined by whole-genome microarray analysis. Total RNA was extracted from liver biopsies collected at the times indicated from animals with acute HAV infection (4x0293 and 4x0395) or acute resolving HCV infection (4x0296, genotype 3) and subjected to Affymetrix GeneChip (U133 Plus 2.0) analysis. Numbers represent the fold-change in the Affymetrix signal from uninfected baseline samples. CXCL10 and ISG20 responses, which are known to be driven by IFN-γ as well as by type I IFN-α/β, are shown in red. Viral RNA copy numbers in liver tissue (GE/μg total RNA) are represented by the shaded area. The results of individual animals are presented in Fig. S4, and results from a second HCV-infected animal are shown in Fig. S5. The entire Affymetrix data set from HAV-infected chimpanzees is shown in Dataset S1.

Fig. 4.

Comparative analysis of ISG15 expression in HAV- and HCV-infected chimpanzees. (A) ISG15 transcript levels, quantified by TaqMan qRT-PCR and normalized to GAPDH transcript abundance, in liver tissue from (Left) three chimpanzees with acute HAV infection versus (Right) eight animals with acute resolving HCV infection. Results shown represent the fold-change from preinfection biopsy tissue. (B) Immunohistochemical staining for ISG15 expressed in liver tissue before and 1–4 wk after challenge with HAV (Left panels: 4x0395) or HCV (Center and Right panels: 4x0313 and 4x0295). Chimpanzee 4x0313 was treated with the immunosuppressive agent FK506 and had no evidence of an adaptive immune response at the time of biopsy (4 wk); this animal is not included among the data shown in Figs. 2 A and B and 4A.