Elsevier

Journal of Virological Methods

Volume 236, October 2016, Pages 87-92
Journal of Virological Methods

Large scale production of a mammalian cell derived quadrivalent hepatitis C virus like particle vaccine

https://doi.org/10.1016/j.jviromet.2016.06.012Get rights and content

Abstract

A method for the large-scale production of a quadrivalent mammalian cell derived hepatitis C virus-like particles (HCV VLPs) is described. The HCV core E1 and E2 coding sequences of genotype 1a, 1b, 2a or 3a were co-expressed in Huh7 cell factories using a recombinant adenoviral expression system. The structural proteins self-assembled into VLPs that were purified from Huh7 cell lysates by iodixanol ultracentrifugation and Stirred cell ultrafiltration. Electron microscopy, revealed VLPs of the different genotypes that are morphologically similar. Our results show that it is possible to produce large quantities of individual HCV genotype VLPs with relative ease thus making this approach an alternative for the manufacture of a quadrivalent mammalian cell derived HCV VLP vaccine.

Introduction

Hepatitis C Virus (HCV) infects 2% of the world’s population and is the leading cause of liver disease requiring transplantation. Recent advances in the treatment of HCV with directly acting antiviral agents (DAAs) have significantly improved sustained virological response rates. However, hepatitis C will continue to pose a significant and growing public health problem that will only be partially addressed with the introduction of new antiviral therapies. With up to 90% of HCV cases occurring in people who inject drugs (PWID) and as reinfection in this group is not infrequent (Sacks-Davis et al., 2013) the expectation of controlling of hepatitis C infection in the short term with antiviral drugs alone is not realistic (Sievert et al., 2014). In contrast simulation models of hepatitis C dynamics in high risk populations have predicted that the introduction of a vaccine even of modest efficacy will have a significant effect on reducing the incidence of hepatitis C even with the availability of DAAs (Hahn et al., 2009, Scott et al., 2015).

Numerous HCV vaccine candidates have been reported, including recombinant E1 and E2 proteins (Colombatto et al., 2014, Houghton, 2011), synthetic peptides (Chua et al., 2008, Torresi et al., 2007a, Torresi et al., 2007b), DNA (Grubor-Bauk et al., 2016, Gummow et al., 2015, Puig et al., 2006, Sallberg et al., 2009a), recombinant adenoviral and prime-boost strategies with MVA vaccines or recombinant E1E2 glycoproteins (Chmielewska et al., 2014, Folgori et al., 2006, Thammanichanond et al., 2008). However, few have progressed to Phase I and II clinical trials in humans including synthetic peptide-based vaccines (Klade et al., 2012), recombinant poxvirus (Habersetzer et al., 2011), adenoviral and MVA vaccines (Barnes et al., 2012, Swadling et al., 2014), and DNA vaccines (Castellanos et al., 2010, Sallberg et al., 2009b). These vaccines have resulted in the production of robust cross reactive HCV-specific CD4+ and CD8+ T cell responses and reductions in HCV viral load.

Hepatitis C virus-like particles (HCV VLPs) represent a vaccine approach that is able to produce both neutralising antibody (NAb) and cellular immune responses (Beaumont et al., 2013, Chua et al., 2012, Kumar et al., 2016, Patient et al., 2009). Furthermore, as HCV-specific NAbs recognize quaternary structures (Keck et al., 2005, Keck et al., 2008, Keck et al., 2011) (Giang et al., 2012), the ordered particulate structure of HCV VLPs makes them an attractive vaccine candidate (Beaumont et al., 2013, Chua et al., 2012, Elmowalid et al., 2007, Garrone et al., 2011, Kumar et al., 2016).

We have previously shown that mammalian cell derived genotype 1a HCV VLPs adjuvanted with TLR2 agonists are strongly immunogenic (Chua et al., 2012). In a subsequent study we demonstrated that mammalian cell derived genotype 3a HCV VLPs were also strongly immunogenic producing strong neutralising antibody and T cell responses in mice (Kumar et al., 2016). Most recently we have described the morphological and biochemical characterisation of the genotype 1a HCV VLPs and shown that they are representative of HCV virions (Earnest-Silveira et al., 2016). However, the success of HCV VLPs as a viable vaccine candidate will also rest on the ability to markedly scale up the production of VLPs to quantities sufficient to produce a commercial vaccine. Here we report a method for the large scale production of mammalian cell derived HCV VLPs and also describe some of the difficulties encountered with different protein purification methods. We also describe the large scale production of a quadrivalent HCV vaccine comprised of genotype 1a, 1b, 2a and 3a VLPs. We selected these four genotypes include in our vaccine as these genotypes constitute the most common HCV genotypes globally (Negro and Alberti, 2011).

The construction of a recombinant adenovirus encoding the HCV structural proteins (core, E1 and E2) of HCV 77H, genotype 1a and 3a has been reported previously (Chua et al., 2012). The genotype 1b gene sequence was derived from a clinical isolate and has been reported previously (Trowbridge and Gowans, 1998). The genotype 3a HCV core, E1 andE2 genes were amplified from a HCV-genotype 3a infected Indian patient (accession no. core: GU172376 and E1E2: GU172375) (Kumar et al., 2016). HCV VLPs of genotypes 1b, 2a and 3a were constructed using recombinant adenoviruses encoding the respective HCV structural proteins (core, E1 and E2). For genotype 1b VLPs, the core/E1/E2 genes of HCV genotype 1b were restricted from the plasmid pGem T Easy-G CoreE1E2 with Bgl II and Sal I, ligated in to pAdTrack CMV that had been linearised with Bgl II and Sal I and confirmed by DNA sequencing. The pAdTrack-CMV-Gt1bCoreE1E2 plasmid was digested with PmeI and transformed into AdEasier cells by electroporation (Bio-Rad Gene Pulser) (Chua et al., 2012). Positive clones were confirmed by restriction with Pac I, then transformed into Top 10 F′ cells.

For the production of genotype 2a VLPs the coreE1E2 gene was amplified from the JFH1 clone (a gift from Prof R Bartenschalger, Heidelberg, Germany). The forward primer (5′ gcgcggccgcgccaccatgagcacaaatcctaaacc 3′) was designed to introduce a Not I enzyme restriction site followed by a Kozak sequence and a start codon at amino terminal end of the core protein. The reverse primer (5′ cgaagctttcatcaggcgtccgctaagagcaggaataag 3′) introduced a Hind III at the 3′ end of E2 gene. The amplified core E1E2 genome was ligated in to pGem T Easy, subcloned into pAdTrack CMV linearised with Hind III and Not I and confirmed by DNA sequencing. The pAdTrack CMV-2aCE1E2 plasmid was restricted with Pme I and transformed into AdEasier cells (Chua et al., 2012).

For the production of gentoype 3a VLPs the core E1E2 genome was restricted out of pVaxGen3a CE1E2 (Kumar et al., 2016) with Bam HI and Eco R I and cloned into pBluescript SK+. The coreE1E2 genome was then restricted out of pBluescript SK+ with Hind III and Not I, cloned into pAdTrack CMV and confirmed by DNA sequencing. The pAdTrack CMV-3aCE1E2 plasmid was restricted with Pme I and transformed into AdEasier Cells (Chua et al., 2012).

To scale up HCV VLP production 1272 cm2 cell factories (Corning Inc, Life Sciences) were infected with the respective genotype specific rAdHCVcoreE1E2 s at a MOI of 1.0. To ensure a high infection efficiency cells were infected when they were 60% confluent. At the time of infection a cell factory contains 7.5 × 107 cells. Infection of Huh7 cells with rAdHCVcoreE1E2 was highly efficient resulting in transduction of 100% of cells (Fig. 1). Four cell factories were infected at one time, each representing one of the 4 vaccine genotypes (1a, 1b, 2a, 3a). Cell factories were harvested 96hr post-infection in 7 ml of cell lysis buffer (50 mM Tris. HCL pH7.5, 50 mM NaCl and 0.5 mM EDTA) and cells disrupted using a Polytron homogenizer. Cell lysates were clarified by centrifugation as described previously (Chua et al., 2012), layered on to a 30% sucrose cushion in 20 mM Tris pH7.4 and 150 mM NaCl and centrifuged at 43,800g for 4 h at 4 °C. The pellet was resuspended in 20 mM Tris pH7.4 and 100 mM NaCl and homogenised again with a Dounce homogenizer. Lysates were clarified by centrifugation and then layered on to a 10–40% continuous iodixanol gradient and centrifuged at 16° C for 14 h at 143,000g in an SW38 rotor followed by the collection of 12 fractions. The fractions containing HCV VLPs were identified using Western immunoblot with HCV VLPs localizing to fractions 7–9 (Fig. 2).

HCV VLPs were purified and concentrated in a Stirred Cell ultrafiltration chamber pressurised with nitrogen gas using an Ultracel 30 kDa ultrafiltration disc (PLTK04310 Ultrafiltration Discs, Merck PTY LTD, Millipore Australia). In brief, 18 ml of pooled iodixanol gradient fractions 7–9 was diluted to 50 ml with sterile PBS and transferred to the chamber. The chamber was pressurized with nitrogen gas to 55 psi (3.7 kg/cm2) and the VLPs concentrated 10 fold with constant stirring. Purified VLPs were quantitited by Bradford assay and the typical yield of VLPs from a single cell factory ranged from 4.75 to 8 mg. The concentration of VLPs was checked by SDS-PAGE and Coomassie staining of gels and compared against a titration of recombinant E2 (Fig. 3a). Recombinant HCV E2 protein was produced using pCDNA3.tpaE2-661myc plasmid as described previously (Torresi et al., 2007b). The concentration and recovery was comparable for HCV VLPs of genotypes 1a, 1b, 2a and 3a (Fig. 3a–d).

The concentration and recovery of the VLPs using Stirred Cell ultrafiltration was also compared to Amicon 100 ultrafiltration units. Concentration of the VLPs using Stirred Cell ultrafiltration resulted in a 80–90% recovery of protein with minimal loss in the chamber flow through. In contrast, although Amicon ultrafiltration resulted in adequate concentration of VLPs this was accompanied by up to 60% retention of VLPs in the Amicon ultrafiltration membrane making this an inefficient method for VLP concentration (result not shown).

We also explored other possible purification methods including gel filtration with Sephadex G columns and affinity chromatography. Filtration through Sephadex G 25, 75 and 100 columns (GE LifeSciences) resulted in a relatively poor recovery of VLPs in specific fractions and instead the VLPs were recovered throughout elution fractions (result not shown).

We attempted two methods of affinity chromatography with HiTrap ConA Sepharose 4B columns (GE LifeSciences) and Heparin Agarose columns (GE LifeSciences). In both instances the HCV VLPs were poorly bound in the columns and VLPs were recovered in the initial flow through (result not shown).

Finally, we examined two additional concentration methods. The VLPs were sealed in dialysis tubing and applied to a bed of PEG6000 or SpectraGel absorbent (Spectrum laboratories). In both cases the volume of the VLP suspension was reduced 5 fold but there was an associated 40% loss of VLPs, presumably as a result of the VLPs adhering to the dialysis tubing.

Having established that the yield of VLPs was greatest with Stirred Cell ultrafiltration chamber we next wanted to determine that the purified VLPs had retained their morphology. To confirm this we performed transmission electron microscopy and negative staining of purified HCV VLPs with 2% aqueous uranyl acetate. Transmission electron microscopy revealed polymorphic VLPs of all 4 genotypes (Fig. 4a). Furthermore, high magnification of the genotype 1a VLPs demonstrated a morphology of the VLPs consistent with HCV (Gastaminza et al., 2010) (Fig. 4b).

Section snippets

Discussion

We have developed a mammalian liver cell-derived quadrivalent HCV VLP vaccine that contains the E1 and E2 glycoproteins and core protein of genotypes 1a, 1b, 2a and 3a. For such a vaccine to be a potential commercial option it is essential to demonstrate that this vaccine can be produced on a large scale. As a proof of principle we have developed a method for the large scale production of our HCV VLP quadrivalent vaccine.

HCV VLPs offer a vaccine approach that is able to elicit both NAb and

Acknowledgments

This work was supported by research grants from the National Health and Medical Research Council (NHMRC) of Australia (Project grant APP1060436), Australian Center for HIV and Hepatitis Virology (ACH2), Commonwealth Department of Health and Aged Care, Australia; Australia-India Biotechnology Research Fund (BF040005), Department of Innovation and Industry, Australian Commonwealth Government. J. Torresi is supported by an NHMRC Practitioner Fellowship (APP1060433) and Eric Gowans by a NHMRC

References (52)

  • D. Thammanichanond et al.

    Fiber-modified recombinant adenoviral constructs encoding hepatitis C virus proteins induce potent HCV-specific T cell response

    Clin. Immunol.

    (2008)
  • E. Barnes et al.

    Novel adenovirus-based vaccines induce broad and sustained T cell responses to HCV in man

    Sci. Transl. Med.

    (2012)
  • E. Beaumont et al.

    Chimeric hepatitis B virus/hepatitis C virus envelope proteins elicit broadly neutralizing antibodies and constitute a potential bivalent prophylactic vaccine

    Hepatology

    (2013)
  • M. Castellanos et al.

    Immunization with a DNA vaccine candidate in chronic hepatitis C patients is safe, well tolerated and does not impair immune response induction after anti-hepatitis B vaccination

    J. Gene Med.

    (2010)
  • A.M. Chmielewska et al.

    Combined adenovirus vector and hepatitis C virus envelope protein prime-boost regimen elicits T cell and neutralizing antibody immune responses

    J. Virol.

    (2014)
  • B.Y. Chua et al.

    Hepatitis C VLPs delivered to dendritic cells by a TLR2 targeting lipopeptide results in enhanced antibody and cell-mediated responses

    PLoS One

    (2012)
  • P. Colombatto et al.

    HCV E1E2-MF59vaccine in chronic hepatitis C patients treated with PEG-IFNalpha2a and Ribavirin: a randomized controlled trial

    J. Viral Hepat.

    (2014)
  • Y.P. de Jong et al.

    Broadly neutralizing antibodies abrogate established hepatitis C virus infection

    Sci. Transl. Med.

    (2014)
  • I. Desombere et al.

    Monoclonal anti-envelope antibody AP33 protects humanized mice against a patient-derived hepatitis C virus challenge

    Hepatology

    (2016)
  • L. Earnest-Silveira et al.

    Characterisation of a hepatitis C virus like particle vaccine produced in a human hepatocyte-derived cell line

    J. Gen. Virol.

    (2016)
  • G.A. Elmowalid et al.

    Immunization with hepatitis C virus-like particles results in control of hepatitis C virus infection in chimpanzees

    Proc. Natl. Acad. Sci. U. S. A.

    (2007)
  • A. Folgori et al.

    A T-cell HCV vaccine eliciting effective immunity against heterologous virus challenge in chimpanzees

    Nat. Med.

    (2006)
  • P. Garrone et al.

    A prime-boost strategy using virus-like particles pseudotyped for HCV proteins triggers broadly neutralizing antibodies in macaques

    Sci. Transl. Med.

    (2011)
  • P. Gastaminza et al.

    Ultrastructural and biophysical characterization of hepatitis C virus particles produced in cell culture

    J. Virol.

    (2010)
  • E. Giang et al.

    Human broadly neutralizing antibodies to the envelope glycoprotein complex of hepatitis C virus

    Proc. Natl. Acad. Sci. U. S. A.

    (2012)
  • B. Grubor-Bauk et al.

    Intradermal delivery of DNA encoding HCV NS3 and perforin elicits robust cell-mediated immunity in mice and pigs

    Gene Ther.

    (2016)
  • Cited by (18)

    • Investigating virus–host cell interactions: Comparative binding forces between hepatitis C virus-like particles and host cell receptors in 2D and 3D cell culture models

      2021, Journal of Colloid and Interface Science
      Citation Excerpt :

      Therefore, differences in HCV VLP–host cell interactions as measured by atomic force spectroscopy are likely indicative of differences in the formation of the initial cell-VLP attachment complex between 2D and 3D cell cultures. HCV genotype 1a VLPs were produced by expression of HCV core, E1 and E2 proteins in Huh7 cells transfected with a recombinant adenovirus vector and purified by ultracentrifugation as previously described [46,47]. For VLP labeling with fluorescein isothiocyanate (FITC; Sigma USA.

    • Non-protein biologic therapeutics

      2018, Current Opinion in Biotechnology
      Citation Excerpt :

      In addition to the peptide and nucleic acid vaccines described above, primarily as oncology therapeutics, and the cell-based immunotherapies described below, traditional vaccines directed against infectious agents including emergent pathogens (e.g. Zika, SARS, etc.), mutating pathogens (e.g. influenza, HIV) as well as challenging targets such as hepatitis C continue to be an area of investigation. In particular, there has been a move toward producing virus-like particles (VLPs), primarily in insect cells, as vaccines as they are safer, can be assembled more rapidly and can provide an alternative to viruses that are difficult to culture [55–57]. Turning the concept of viral vaccination on its head, oncolytic viruses (OVs) are designed to kill cancer cells by direct injection into a cancerous lesion.

    • Microscale stirred-cell filtration for high-throughput evaluation of separation performance

      2018, Biochemical Engineering Journal
      Citation Excerpt :

      For example, one of the earliest reported studies elucidated the transport of lipoproteins through a UF membrane [6]. More recent studies have used stirred cell devices in applications such as exosome isolation [7], plasmid DNA separation [8], the study of biofouling behaviour [9], and the production of Hepatitis C virus-like particles [10]. While the stirred cell is not the only available membrane evaluation tool (tangential flow filtration cassettes and hollow fiber modules also exist), it is frequently used in evaluating filtration processes at laboratory scale.

    View all citing articles on Scopus
    View full text