Enveloped virus-like particle platforms: vaccines of the future?

Abstract

The techniques to produce effective vaccines have evolved, and the early vaccines (live, inactivated, subunit…) are no longer considered as the most appropriate for new vaccine development. We question here what will be the future vaccines, and argue that virus-like particle (VLP)–based vaccines are promising candidates. In addition to being effective vaccines against analogous viruses from which they are derived, VLPs can also be used to present foreign epitopes to the immune system. The achievement of this strategy can be illustrated by the recent development of malaria candidate vaccine. We point out recent VLP-based vaccine developments and discuss future perspectives.

Keywords:

• antigen platform
• retrovirus
• RTS,S
• vaccine development
• virus-like particles

The development and widespread adoption of vaccines has been considered a public health triumph and has dramatically improved human health. Since the late 18th century, techniques to produce effective vaccines have evolved, and the early vaccines (live, inactivated, subunit) are not anymore considered as the most appropriate for new vaccine development. We question here what will be the future vaccines, and argue that recombinant virus-like particle (VLP)-based vaccines are promising candidates. Indeed, among the recent vaccines, VLP-based vaccines are currently commercialized worldwide. GlaxoSmithKline’s Engerix^® (HBV) and Cervarix^® (HPV), and Merck and Co., Inc.’s Recombivax HB^® (HBV) and Gardasil^® (HPV) have paved the way. Other VLP-based vaccine candidates are currently in development and include nine vaccine candidates, which are in Phase II study, in particular against influenza (NCT01991561, NCT02307851) or Norovirus GI.1/GII.4 (NCT02038907).

The key to the success of virus-like particle (VLP) vaccines resides in their accurate mimicry of wild-type viral particles. VLPs are supra-molecular assemblages incorporating the key immunologic features of viruses [1] which include repetitive surface patterns and particulate structures with potential for induction of innate and adaptive immunity through activation of pattern recognition receptors and antigen-specific B-cell receptors, respectively. Interestingly, they carry no replicative genetic information and can be produced in heterologous expression systems at a large scale. As a result, VLPs represent safe and effective vaccines capable of inducing strong B- and T-cell responses. The recognition of the potent immunogenicity and commercial potential of VLPs has greatly accelerated research and development activities. However, it is quite surprising that VLP vaccines have not been selected for being tested in humans in response to the outbreaks of Ebola virus disease, highlighting some limitations in the production of complex VLPs.

In addition to being effective vaccines against analogous viruses from which they are derived, VLPs can also be used to present foreign epitopes to the immune system. This can be achieved by genetic fusion or chemical conjugation between target antigens and structure viral proteins that can self-assemble into VLPs, offering new perspectives in vaccine development.

The achievement of this strategy can be illustrated by the recent development of malaria candidate vaccine, composed of the Plasmodium falciparum circumsporozoite protein antigen fused to hepatitis B surface antigen that form chimeric recombinant VLPs (rVLPs). This vaccine candidate, referred to as RTS,S, has reached Phase III clinical trial and induced approximately 30–50% protection [2]. Despite this moderate efficacy, which may be explained by the poor antigen expression on the surface of the VLPs, RTS,S is the first vaccine showing a significant protection against malaria, where other strategies such as synthetic peptides and viral vectors have failed until now. Similarly, other platforms have also been used to display foreign antigens, using simple structures made of self-assembling proteins such as HBcore protein, alfalfa mosaic virus coat protein, HPV L1 protein or Qβ bacteriophage, and have been shown to efficiently stimulate target-specific immune responses [3–6]. Notably, the choice of the pathogen from which the structural proteins are derived for VLP formation is crucial since it influences the overall immunogenicity of the antigens. Altogether, these studies demonstrate the great potential of rVLPs, as illustrated by the fact that many of them have already reached clinical trial stage [7]. Thus, we believe that rVLPs represent the future of modern vaccines and postulate that the commercialization of the Malaria RTS,S vaccine could give rise to a new era of vaccinology.

With the advance in the knowledge of more complex viruses, it is today feasible to exploit the rVLP strategy to design particles presenting more sophisticated structures. rVLPs derived from enveloped viruses offer the unique opportunity to integrate target antigens displayed onto the particles by their integration in the envelope (i.e., cell membrane). Pseudotyping serves two important functions: a means for ordered presentation in the native conformation of the target antigens and a means to enhance or target vector uptake by APCs. The cores of retroviruses have proven to be exceptionally accommodating for this type of ‘make-up’. We pioneered the development of rVLPs based on simple oncoretroviruses (retroVLPs), the murine leukemia retrovirus. Cellular expression of the Gag protein suffices to generate roughly spherical, pleomorphic, membrane-enclosed particles, with a diameter of 80–120 nm, that can be pseudotyped with glycoproteins from retroviruses and from other families of enveloped viruses, including lymphocytic choriomeningitis virus, spleen necrosis virus, vesicular stomatitis virus, hepatitis C virus, yellow fever virus, West Nile virus, influenza virus, HIV and CMV [8,9], without harming their functionality. Standard cell expression systems, including 293T or CHO cell lines, can be used to produce retroVLPs that can then be purified with sucrose density gradient or by chromatography.

Interestingly, retroVLPs can be adapted by recombinant technology to incorporate high density of displayed antigens on their surface, which constitutes a significant optimization improving their immunogenicity. We, therefore, designed modified glycoproteins in which we swapped the transmembrane domain and cytoplasmic tail for one of the G envelope proteins of vesicular stomatitis virus, which is efficiently pseudotyped on both retrovirus- and lentivirus-derived pseudoparticles. This led to considerably increased HCV- or HIV-glycoproteins’ pseudotyping efficiency and immunogenicity ([10] and [Vazquez T. et al. Pers. Comm.]), which is consistent with the importance of the high density of repetitive epitopes to induce an efficient antibody response [11]. We also demonstrated that capside proteins could be easily modified to display vaccine antigens inside the particles [12], favoring the antigen cross-presentation to CD8+ T cells. Obviously, this makes retroVLPs excellent vaccine candidates for a wide range of diseases of clinical interest.

We took major efforts for HCV and HIV vaccine development based on this strategy. Combining retroVLPs pseudotyped with either HCV-E1E2 or -E1, we observed that such rVLPs induced high-titer antibodies, including neutralizing antibodies, in both mouse and macaque [10,12,13]. Importantly, HCV-E1–specific antibodies were detected while neither recombinant adenovirus nor measles vectors were able to elicit significant levels of anti-E1 antibodies, demonstrating the superior potential of the retroVLP-based platform as compared to recombinant viral vectors. We also designed HIV-specific retroVLPs displaying at their surface HIV-GP140 antigens. In prime-boost immunization protocol, we observed a significant impact on the level and quality of HIV-specific immune responses as compared to ‘standard’ protocol using GP140 proteins, highlighting the advantage to display antigens onto rVLPs [Manuscript In Preparation]. Also, we showed that retroVLPs induced long-lasting specific antibody responses that remain stable for more than 1 year, similar to the outstanding results previously reported with naked VLPs [14,15]. RetroVLPs have also been developed as vaccine candidates against human CMV. It has been shown that these rVLPs produced in mammalian cells induced efficient neutralizing antibody responses in mice [16], and CMV-retroVLPs should enter clinical development in 2015 [17].

More generally, in addition to retroVLPs, other enveloped VLP systems could be used as antigen platforms (rVLPs). The use of rVLPs offers considerable advantages for the design of new vaccines. First, it allows using core proteins that are derived from well-known viruses, which permits an easy genetic engineering leading to a total safety and an efficient display of target antigens. Moreover, rVLPs bring a unique tool to produce VLP-based vaccines against pathogens unable to form pseudoparticles. Also, it offers the possibility to design polyvalent vaccines by incorporating proteins from several different viruses. This strategy could be particularly effective to fight against prevalent co-infections such as HIV/HCV or HIV/tuberculosis.

While rVLPs are attractive candidates to generate protective immune responses, their production and purification may represent a limitation, especially for enveloped rVLPs. Indeed, in vitro production of enveloped rVLPs requiring expression of multiple proteins can only be achieved in cell system in contrast to non-enveloped rVLPs that can easily self-assemble. Regardless of the expression systems used to produce the rVLPs, the particles are often contaminated with residual host cell components such as lipids, nucleic acids and proteins. These contaminants that may stimulate the innate immunity and augment the adaptive immune response, however, represent a limitation in vaccine development. To circumvent these limitations, alternative strategies have been proposed. For example, specific sequences encoding all components of rVLPs can be in vivo injected as plasmid DNA or recombinant viral vectors. Interestingly, these strategies combine the advantages of both vaccine formulations (genetic vaccines and VLPs) and are very cost-efficient as laborious ex vivo production of rVLPs is not required. We have developed retroVLP-expressing plasmids and demonstrated that those vaccine candidates improved significantly the antigen-specific responses and the antiviral immune protection in comparison with mutated plasmid DNA preventing retroVLP assembly [18]. In this line, GeoVax Co. has developed innovative human vaccines using DNA or modified vaccinia ankara vectors containing sufficient HIV genes to support the production of VLPs [19].

In conclusion, VLPs are a particular class of subunit vaccines that differentiate themselves from non-particulate recombinant antigens by stronger protective immunogenicity associated with the VLP structure. Thus, the rVLP strategy using these unique immune properties of the viral particles questions the classification of VLPs as adjuvants or as other particulate carriers (virosomes, nanoparticles).

Currently, the rVLP strategy has greatly broadened the scope of their use, from immunizing against microbial pathogens to immunotherapy for chronic diseases. rVLPs have been used to induce autoantibodies to disease-associated self-molecules. For example, Cytos Biotechnology (Switzerland) has developed a number of rVLP vaccine candidates, based on RNA bacteriophage VLPs, directed against allergies, neurodegenerative (Alzheimer’s disease) and autoimmune disorders (Type II diabetes mellitus), cancer (melanoma) and hypertension. In the same way, we have recently developed retroviral rVLP-based vaccine candidates for allergy prevention that gave some really promising results in preclinical studies [Manuscript In Preparation].

Altogether, rVLP vaccines will undoubtedly emerge as potent future vaccines, and should easily answer to major infectious threats, that is, worldwide pandemic alerts.

Financial & competing interest’s disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.