Novel monkey mAbs induced by a therapeutic vaccine targeting the hepatitis B surface antigen effectively suppress hepatitis B virus in mice

Abstract

Background: We have previously obtained a mouse anti-hepatitis B surface antigen (HBsAg) antibody E6F6 with long-lasting serum HBsAg clearance effects. The E6F6 epitope-based protein CR-T3-SEQ13 (HBsAg aa 113-135) vaccination therapy in cynomolgus monkeys induced long-term polyclonal antibodies-mediated clearance of HBsAg in the HBV transgenic (HBV-Tg) mice.

Methods: We isolated monoclonal antibodies from CR-T3-SEQ13 vaccinated cynomolgus monkeys, compared their therapeutic effects with E6F6, identified their epitopes on HBsAg, determined the pharmacokinetics and studied their physical property.

Results: A panel of anti-HBsAg mAbs was generated through memory B cell stimulatory culture. Two lead monkey-human chimeric antibodies, C1-23 and C3-23, effectively suppressed HBsAg and HBV DNA in HBV-Tg mice. The humanized antibodies and humanized-mouse reverse chimeric antibodies of two antibodies exhibited comparable HBsAg clearance and viral suppression efficacy as those versions of E6F6 in HBV-Tg mice. Humanized antibody hu1-23 exhibited more efficacy HBsAg-suppressing effects than huE6F6-1 and hu3-23 in HBV-Tg mice at dose levels of 10 and 20 mg/kg. Evaluation of the binding sites indicates that the epitope recognized by hu1-23 is located in HBsAg aa 118-125 and 121-125 for hu3-23. Physical property study revealed that hu1-23 and hu3-23 are stable enough for further development as a drug candidate.

Conclusions: Our data suggest that the CR-T3-SEQ13 protein is a promising HBV therapeutic vaccine candidate, and hu1-23 and hu3-23 are therapeutic candidates for the treatment of chronic hepatitis b. Moreover, the generation of antibodies from the epitope-based vaccinated subjects may be an alternative approach for novel antibody drug discovery.

Keywords: B cell culture; antibody-mediated HBV suppression; chronic hepatitis B infection; cynomolgus monkey antibody; hepatitis B surface antigen.

Figure 1

Characterization of HBsAg-SEQ13 specific antibodies from two vaccinated cynomolgus monkeys. (A) The strategy used to generate HBsAg-specific monkey mAbs from vaccinated cynomolgus monkeys. (B) Isolation of CD20⁺CD27⁺ memory B cells from cynomolgus monkeys. (C) Analyses of the sequences of nine anti-HBsAg antibodies. The antibodies named with ‘1-’ were from monkey #1, and those named with ‘3-’ were from the other monkey #3. (D) Evaluation of the binding activity of anti-HBsAg antibodies against HBsAg, CR-T3-SEQ13 and SEQ13-polypeptide. (E) Blocking effects of anti-HBsAg mAbs on the binding of huE6F6-1 with HBsAg. The data were expressed as the mean ± SD.

Figure 2

Humanization of C1-23 and C3-23 and the comparation of their viral suppression effects to HBIG in HBV-Tg mice. (A) Humanized 1-23 and 3-23 mAbs maintained binding activity similar to that of the parental chimeric antibodies in ELISA-based assays. (B) Representative SPR sensorgrams for the binding of hu1-23 and hu3-23 to HBsAg. (C) Dose–response analyses of the binding activity of hu1-23, hu3-23, and HBIG. Based on the correlation analyses, the binding activity of 4 000 IU/kg HBIG was equal to 1.5 mg/kg hu1-23 or 1.9 mg/kg hu3-23. (D) The comparison of therapeutic efficacy among equivalent doses of hu1-23, hu3-23 and HBIG in HBV-Tg mice (N = 5). The values represent the mean ± standard deviation from at least three experiments.

Figure 3

Therapeutic effects of 1-23 and 3-23 suppress HBV in HBV-Tg mice. (A) Different types of engineered mAbs used in this study: monkey–human chimeric mAb, cmAb, humanized mAb and humAb. Humanized-mouse reverses chimeric mAb, which contained the mouse IgG2a constant region, rc-mAb. (B) Study design for this figure. (C) and (D) Serum HBsAg and HBV DNA profiles of HBV-Tg mice (N = 5) after single infusions with different concentrations of hu1-23 or hu3-23. C16G12 was a mouse-human chimeric isotype control. Antibodies were used at a dose of 10 mg/kg or 20 mg/kg. The data were expressed as the mean ± SD Immunohistochemical staining for HBsAg and HBcAg in the liver of HBV-Tg mice (N = 3) after mAb infusion. Assays were performed 3 days after rc-mAb infusion. (E) Liver HBsAg and HBcAg profiles of HBV-Tg mice (N = 3) on day 3 after rc-mAb treatment. The data were expressed as the mean ± SD; ^**p < 0.01 compared to control; ns, not significant. (F) Serum HBsAg, HBV DNA profiles of HBV-Tg mice (N = 5) after different single rc1-23 and rc3-23 infusion. 16G12 was a mouse isotype control. Antibodies were used at dosage of 10 mg/kg g. The data were expressed as the mean ± SD.

Figure 4

Detailed binding of hu1-23 and hu3-23 to HBsAg-SEQ13. (A) Evaluation of the EC50s of the antibodies for SEQ13 with single-site alanine-scanning mutagenesis. The cysteines at positions 121 and 124 were mutated to serine. (B) Critical amino acids contribute to the antibodies interacting with SEQ13. (C) Molecular docking of the interaction of hu1-23 and 3-23 against SEQ13.

Figure 5

Physical property and in vivo PK profiles of hu1-23 and hu3-23. (A) Degree of humanness, viscosity-concentration and pI profile of hu1-23 and hu3-23. (B) DSC scan of mAbs in 25 mM Histidine (PH 6.0), 5% sucrose and 0.02% PS80. Scan rate: 200°C/h. DSC was performed with Microcal VP-DSC(GE). (C) Serum antibody concentrations (μg/mL) of hu1-23 and hu3-23 following a single i.v. infusion of 20 mg/kg mAb into cynomolgus monkeys (n = 3 animals per antibody). The data were expressed as the mean ± SD.