Hepatitis C virus (HCV) encodes two putative virion glycoproteins (El and E2) which are released from the polyprotein by signal peptidase cleavage. In this report, we have characterized the complexes formed between El and E2 (called ElE2) for two different HCV strains (H and BK) and studied their intracellular localization. Vaccinia virus and Sindbis virus vectors were used to express the HCV structural proteins in three different cell lines (HepG2, BHK-21, and PK-15). The kinetics of association between El and E2, as studied by pulse-chase analysis and coprecipitation of E2 with an anti-El monoclonal antibody, indicated that formation of stable ElE2 complexes is slow. The times required for half-maximal association between El and E2 were 60 to 85 min for the H strain and more than 165 min for the BK strain. In the presence of nonionic detergents, two forms of E1E2 complexes were detected. The predominant form was a heterodimer of El and E2 stabilized by noncovalent interactions. A minor fraction consisted of heterogeneous disulfide-linked aggregates, which most likely represent misfolded complexes. Posttranslational processing and localization of the HCV glycoproteins were examined by acquisition of endoglycosidase H resistance, subcellular fractionation, immunofluorescence, cell surface immunostaining, and immunoelectron microscopy. HCV glycoproteins containing complex N-linked glycans were not observed, and the proteins were not detected at the cell surface. Rather, the proteins localized predominantly to the endoplasmic reticular network, suggesting that some mechanism exists for their retention in this compartment.

Ten monoclones directed to the 42,000-dalton inner structural protein of rotavirus were analyzed. Eight monoclones reacted broadly with antigenic domains common to virtually all mammalian rotaviruses. Two monoclones had specificities similar or identical to previously characterized subgroup specificities. These subgroup monoclones were more efficient in detecting subgroup antigen than either hyperimmune or postinfection antisera. Using the subgroup monoclones, we determined that some animal as well as human rotavirus strains carry subgroup 2 specificity and that epizootic diarrhea of infant mice virus and turkey rotavirus are antigenically distinct from other mammalian rotavirus strains.

A series of monoclonal antibodies was isolated which reacted with one of two major surface proteins of rhesus rotavirus. Thirty-six monoclonal antibodies immunoprecipitated the 82-kilodalton outer capsid protein, the product of the fourth gene, the viral hemagglutinin. These monoclonal antibodies exhibited hemagglutination inhibition activity and neutralized rhesus rotavirus to moderate or high titer. Three monoclonal antibodies immunoprecipitated the 38-kilodalton outer capsid glycoprotein, the eighth or ninth gene product. These three monoclonal antibodies neutralized rhesus rotavirus to high titer and also inhibited viral hemagglutination.

An effort was made not to exceed passage 72. Cultures of primary AGMK were purchased from MA Bioproducts. Viruses. Rotaviruses for primary isolation were in the form of diarrheal or normal feces, either as approximate 10%o suspensions or as rectal swab fluids. Seventy-three such fecal samples containing rotavirus were kindly supplied by the following investigators: H. W. Kim and colleagues, Children'

The immune mechanisms involved in clearance of and immunity to rotavirus infection are poorly understood. Although mice with severe combined immunodeficiency (SCID mice) become chronically infected, nude mice have been reported to clear rotavirus infection similarly to immunocompetent controls. To better characterize the role of cytotoxic T lymphocytes (CTLs) in clearance of and immunity to rotavirus infection, we infected naive or previously infected ␤ 2-microglobulin (␤ 2 m) knockout mice with murine rotavirus. Naive ␤ 2 m knockout mice shed rotavirus antigen 2 days longer than did normal control mice but completely resolved primary infection. ␤ 2 m knockout naive mice treated with depleting doses of an anti-CD8 monoclonal antibody before infection shed viral antigen for an additional day. Upon rechallenge, ␤ 2 m knockout mice, either treated with the anti-CD8 antibody or not treated, were completely resistant to reinfection. Clearance of rotavirus infection in naive ␤ 2 m knockout mice correlated with the development of intestinal rotavirus-specific immunoglobulin A. Before rechallenge, ␤ 2 m knockout mice had high levels of intestinal rotavirus-specific immunoglobulin A. These findings suggest that CTLs mediate rotavirus clearance but are not required for this function and that CTLs are not necessary for development of immunity to rotavirus reinfection. To further characterize the effector mechanisms involved in clearance and prevention of rotavirus infection, similar studies were performed with B-cell-deficient J H D knockout mice. After primary infection, most naive J H D mice had similar virus-shedding clearance curves as did control mice and completely resolved primary infection. However, 2 of 29 became chronically infected. All J H D mice treated with anti-CD8 antibody became chronically infected with murine rotavirus. Upon rechallenge, J H D mice which had cleared primary infection were all susceptible to reinfection. These findings suggest that B cells also play a role in clearance of primary infection but are absolutely necessary for development of immunity against rotavirus reinfection.