Induction of broadly neutralizing antibodies using a secreted form of the hepatitis C virus E1E2 heterodimer as a vaccine candidate

Hepatitis C virus (HCV) is a global disease burden, and a preventive vaccine is needed to control or eradicate the virus. Despite the advent of effective antiviral therapy, this treatment is not accessible to many patients and does not prevent reinfection, making chronic hepatitis C an ongoing global health problem. Thus, development of a prophylactic vaccine will represent a significant step toward global eradication of HCV. HCV exhibits high genetic variability, which leads frequently to immune escape. However, a considerable challenge faced in HCV vaccine development is designing an antigen that elicits broadly neutralizing antibodies. Here, we characterized the immunogenicity of a vaccine based on a soluble, secreted form of the E1E2 envelope heterodimer (sE1E2.LZ). Sera from mice immunized with sE1E2.LZ exhibited an anti-E1E2–specific response comparable to mice immunized with membrane-bound E1E2 (mbE1E2) or a soluble E2 ectodomain (sE2). In competition-inhibition ELISA using antigenic domain-specific neutralizing and nonneutralizing antibodies, sera from sE1E2.LZ-immunized mice showed nearly identical or stronger competition toward neutralizing antibodies when compared with mbE1E2. In contrast, sera from mice immunized with sE2, and to a lesser extent mbE1E2, competed more effectively with nonneutralizing antibodies. An assessment of neutralization activity using both HCV pseudoparticles and cell culture–derived infectious HCV showed that immunization with sE1E2.LZ elicited the broadest neutralization activity of the three antigens, and sE1E2.LZ induced neutralization activity against all genotypes. These results indicate that our native-like soluble glycoprotein design, sE1E2.LZ, induces broadly neutralizing antibodies and serves as a promising vaccine candidate for further development.

Hepatitis C virus (HCV) is a global disease burden, with an estimated 71 million people infected worldwide (1, 2). Roughly 75% of HCV infections become chronic (3–5), and in severe cases can result in cirrhosis or hepatocellular carcinoma (6). Viral infection can be cured at high rates by direct acting antivirals (DAAs), but several issues have blunted their effectiveness in eradicating HCV. In particular, multiple public health and financial barriers (7, 8) restrict access to DAAs in areas with high incidence of infection and DAAs do not prevent reinfection. Moreover, HCV infection is largely asymptomatic and often does not generate sterilizing immunity, thereby contributing to reinfection or continued disease progression (7, 9, 10). Collectively, these issues have resulted in a continued rise in HCV infections.Acute HCV infections can be cleared by host immunity in ∼25% of cases. Among individuals who clear their first infection, the rate of clearance rises to 80% for subsequent infections, indicating an effective immune memory response (11–14). This type of natural protective immunity to HCV requires the induction of broadly neutralizing antibodies to E1E2 ectodomains and T cell responses to the structural and nonstructural proteins (15–17). The above clinical observations suggest that, if a vaccine candidate could induce broadly neutralizing antibody and cell-mediated immune responses equivalent to that seen in spontaneous clearance, such a vaccine would be highly effective at preventing HCV infection. An HCV vaccine therefore remains an essential proactive measure to protect against viral spread, yet vaccine developments against the virus have been unsuccessful to date (17, 18).A number of challenges exist that have thus far limited progress toward developing a prophylactic vaccine against HCV. One major challenge in developing a successful vaccine for HCV has been the remarkable genetic diversity of the virus which has six major genotypes (genotypes 1 to 6), in addition to two less-common genotypes (19) (genotypes 7 and 8), and intragenotypic diversity resulting in 90 total subtypes (20). Moreover, shielding of important neutralizing epitopes with glycans (21, 22), and the presence of immunodominant nonneutralizing epitopes (23–26) deflect the immune response from conserved regions that mediate virus neutralization. Multiple studies in chimpanzees and humans have used E1E2 formulations to induce a humoral immune response, but their success in generating high titers of broadly neutralizing antibody (bnAb) responses has been limited. In particular, immunological assessment in chimpanzees of an E1E2 vaccine produced superior immune responses as compared with E2 administered alone and resulted in sterilizing immunity against homologous virus challenge (27, 28), but with less cross-neutralization capacity against heterologous isolates (29). In addition, an E1E2 formulation tested in humans is well-tolerated (30). However, due to the limited neutralization breadth observed in the human clinical trial (31, 32), using native E1E2 as a vaccine is not likely to provide sufficient protection from HCV infection. Rather, optimization of E1E2 to improve its immunogenicity and capacity to elicit bnAbs through rational design appears to be the preferred path for developing an effective B cell-based vaccine (33).An additional bottleneck contributing to the difficulty in generating protective B cell immune responses required for an effective HCV vaccine is preparation of a homogeneous E1E2 antigen. HCV envelope glycoproteins E1 and E2 form a heterodimer on the surface of the virion (34–36). Furthermore, E1E2 assembly has been proposed to form a trimer of heterodimers (37) mediated by hydrophobic C-terminal transmembrane domains (TMDs) (36, 38, 39) and interactions between E1 and E2 ectodomains (40–42). These glycoproteins are necessary for viral entry and infection, as E2 attaches to the CD81 and scavenger receptor type B class I (SR-B1) coreceptors as part of a multistep entry process on the surface of hepatocytes (43–46). Neutralizing antibody (nAb) responses to HCV infection target epitopes in E1, E2, or the E1E2 heterodimer (25, 47–52). A significant impediment to the uniform production of an immunogenic E1E2 heterodimer that could be utilized for vaccine development is the association of the antigen with the membrane via the TMDs (36, 53). Progress has been made in the production and purification of the membrane-bound E1E2 complex via immunoaffinity purification (54, 55) or the use of tags that allow protein A (56) or anti-Flag (57) chromatography. While these methods produce high-quality samples, they all involve harsh elution conditions. How such conditions might influence sample quality at a scale required for vaccine trials is unclear. Furthermore, intracellular expression and membrane extraction limits the ability to produce large quantities of sufficient homogeneity required for both basic research and vaccine production.In contrast, viral glycoproteins of influenza hemagglutinin (58), respiratory syncytial virus (RSV) (59), SARS-CoV-2 (60), and others (61, 62) have been stabilized in soluble form using a C-terminal attached foldon trimerization domain to facilitate assembly. In addition, HIV gp120-gp41 proteins have been designed as soluble SOSIP trimers in part by introducing a furin cleavage site to facilitate native-like assembly when cleaved by the enzyme (63, 64). Recent efforts have made strides toward liberating the E1E2 complex from the membrane in its native form (65, 66). In particular, our previous work (66) showed that a soluble E1E2 (sE1E2) using the Fos/Jun leucine zipper (LZ) coiled-coil as a scaffold (sE1E2.LZ) is antigenically intact, as the protein is recognized by E1E2-specific mAbs AR4A and AR5A (67). Moreover, sE1E2.LZ elicited nAbs in mice immunized with the antigen, making this scaffold a promising potential platform for engineering of additional HCV vaccine candidates.Here, we describe the immunogenicity of our native-like secreted E1E2 construct sE1E2.LZ and compare it with the membrane-bound E1E2 complex (mbE1E2) and a secreted form of the E2 ectodomain (sE2). Immunization of mice with sE1E2.LZ produced sera possessing anti-E1E2 antibodies at levels comparable to mice immunized with mbE1E2 or sE2. Moreover, the antibody response in sE1E2.LZ-immunized mice is skewed more toward nAbs relative to non-nAbs than the other two antigens. Remarkably, sera from sE1E2.LZ-immunized mice exhibited broader neutralization activity than either mbE1E2 or sE2 when assessed using both pseudotyped HCV particles (HCVpp) and cell culture-derived HCV (HCVcc), suggesting that this sE1E2 platform represents a favorable starting point for developing scaffolded E1E2 vaccine candidates.