N-terminal truncations on L1 proteins of human papillomaviruses promote their soluble expression in Escherichia coli and self-assembly in vitro

Abstract

Human papillomavirus (HPV) is the causative agent in genital warts and nearly all cervical, anogenital, and oropharyngeal cancers. Nine HPV types (6, 11, 16, 18, 31, 33, 45, 52, and 58) are associated with about 90% of cervical cancers and 90% of genital warts. HPV neutralization by vaccine-elicited neutralizing antibodies can block viral infection and prevent HPV-associated diseases. However, there is only one commercially available HPV vaccine, Gardasil 9, produced from Saccharomyces cerevisiae that covers all nine types, raising the need for microbial production of broad-spectrum HPV vaccines. Here, we investigated whether N-terminal truncations of the major HPV capsid proteins L1, improve their soluble expression in Escherichia coli. We found that N-terminal truncations promoted the soluble expression of HPV 33 (truncated by 10 amino acids [aa]), 52 (15 aa), and 58 (10 aa). The resultant HPV L1 proteins were purified in pentamer form and extensively characterized with biochemical, biophysical, and immunochemical methods. The pentamers self-assembled into virus-like particles (VLPs) in vitro, and 3D cryo-EM reconstructions revealed that all formed T = 7 icosahedral particles having 50-60-nm diameters. Moreover, we formulated a nine-valent HPV vaccine candidate with aluminum adjuvant and L1 VLPs from four genotypes used in this study and five from previous work. Immunogenicity assays in mice and non-human primates indicated that this HPV nine-valent vaccine candidate elicits neutralizing antibody titers comparable to those induced by Gardasil 9. Our study provides a method for producing a nine-valent HPV vaccine in E. coli and may inform strategies for the soluble expression of other vaccine candidates.

Fig. 1. Expression analysis of the N-terminal-truncated L1 proteins and purified L1 proteins of HPV 33, 45, 52, and 58.

a Whole-cell lysates, expressing different N-terminally truncated L1 proteins of HPV 33, 45, 52, and 58, were subject to SDS-PAGE and western blotting with corresponding type-specific antibodies. L1 proteins are denoted by black arrows. b SDS-PAGE and western blotting of purified HPV 33, 45, 52, and 58 L1 proteins. The wide-spectrum HPV L1 linear mAb 4B3 was used to probe the L1 proteins

Fig. 2. Molecular weight, secondary structure and thermal stability characteristics of the HPV 33, 45, 52, and 58 L1 virus-like particles (VLPs).

a The molecular weight of HPV L1 proteins was measured using MALDI-TOF MS. The monomer molecular weight of HPV 33, 45, 52, and 58 L1 proteins was ~55–57 kD. b Circular dichroism spectra of HPV 33, 45, 52, and 58 L1 virus-like particles (VLPs) showed similar secondary structural compositions of the antigen across different HPV types. c Differential scanning calorimetry profiles of HPV L1 proteins show the different transition temperatures. The Tm values of HPV 33, 45, 52, and 58 L1 proteins were 72.3, 80.2, 64.2, and 68.7 °C, respectively

Fig. 3. Size and morphology distribution of the HPV 33, 45, 52, and 58 L1 VLPs.

a High-performance size-exclusion chromatography profiles of HPV 33, 45, 52, and 58 L1 VLPs. b Dynamic light scattering analysis of HPV 33, 45, 52, and 58 L1 VLPs. c Analytical ultracentrifugation sedimentation profiles of the HPV 33, 45, 52, and 58 L1 VLPs

Fig. 4. Structural characterization of HPV 33, 45, 52, and 58 L1 particles in solution.

(a–d, upper) Micrographs of negatively stained (left) and vitrified (right) HPV 33, 45, 52, and 58 L1 VLPs samples. Scale bars, 100 nm; (a–d, lower) left, reconstructed 3D cryo-electron microscopy (cryo-EM) maps of HPV 33, 45, 52, and 58 L1 VLPs, color-coded by diameter from 450 to 600 Å, and viewed along the icosahedral twofold axis; right, the same as left but with the closest half of their density map removed to reveal internal features of VLPs of the four types. e The diameter distribution of E. coli-based L1 VLPs of HPV 6, 11, 16, 18, 31, 33, 45, 52, and 58 measured from their corresponding cryo-EM structures. f The resolutions of the cryo-EM reconstructions of E. coli-based HPV 6, 11, 16, 18, 31, 33, 45, 52, and 58 L1 VLPs

Fig. 5. Antigenicity of HPV 33, 45, 52, and 58 L1 VLPs.

Binding capacity of mAbs to HPV 33, 45, 52, and 58 L1 VLPs, as determined by ELISA. The reactivities of HPV 33, 45, 52, and 58 L1 VLPs against a panel of genotype-specific neutralizing mAbs were measured by indirect ELISA and are denoted as reactivity titer, which reflects the maximum threefold dilution time to show a positive ELISA reading. All experiments were repeated thrice, and histograms reflect the mean values and standard deviations

Fig. 6. Immunogenicity of HPV 9-valent vaccine.

a, b Comparison of the neutralizing antibody response against HPV 6, 11, 16, 18, 31, 33, 45, 52 and 58 induced by E. coli-based HPV 9-valent vaccine with Gardasil 9 vaccines in mice (a) and monkeys (b). Six groups of mice (n = 5) were immunized with 20-, 200-, or 2000-fold diluted dosages of our 9-valent vaccine or Gardasil 9 vaccine at 0, 2, and 4 weeks. Serum samples were collected 2 weeks after the third immunization and titrated using a pseudovirion-based neutralization assay. For monkey trials, two groups of monkeys (n = 4) were immunized with 1:4 dilution of our 9-valent vaccine or Gardasil 9 vaccine at 0 and 4 weeks. Serum samples collected at 6 weeks post immunization were used to evaluate the neutralization response. c Kinetics of serum neutralization titers in Cynomolgus Macaques after immunization with the HPV 9-valent vaccine. Twelve animals were divided into two groups (n = 6) and were immunized with 270 μg of HPV 9-valent vaccine and hydroxyl aluminum hydroxide adjuvant as control at 0, 4, and 27 weeks. Immune sera were collected monthly after immunization, and serum neutralization titers were determined and plotted