Multiple immunizations are usually required for most vaccines to be successful in protecting against pathogens. For instance, five-dose series of Tetanus, Pertussis and Diphtheria (DTaP) vaccine is required during childhood and an adolescent booster dose to elucidate the aimed protection (1). It is not entirely clear why some vaccines require more additional boosters than others; however, it is well accepted that multiple immunizations are essentials for better responses, in particular for non-replicating vaccines (2). An annual dose of influenza vaccine is also recommended for persons who may be at increased risk for complications from influenza infections.

Additionally, heterologous prime-boost regimens in particular when using vectored vaccines have previously shown better immunogenicity; for example, priming with DNA and boosting with viral vectors (3, 4). The need to change the vector may be explained by the induction of neutralizing antibodies against the vector, which compromises further boosting with the same vector. Vaccine-specific antibodies have also been reported to suppress the cytotoxic (CTL) response when using the non-infectious virus-like particles (VLPs) derived from human papilloma virus (HPV) or Qβ-p33 for immunization (5, 6).

COVID-19 pandemic has overwhelmed the healthcare systems worldwide and several countries have prioritized the development of SARS-CoV-2 vaccine to contain the virus. Currently, 30 vaccines are approved in various countries and around 80 vaccines are in clinical development, of which 19 are in phase 3 (7). Several leading vaccines have been shown to confer protection and induction of neutralizing antibodies against the wild type SARS-CoV-2. However, with continuous emergence of SARS-CoV-2 variants and the relatively rapid drop in antibody titers induced in vaccinated individuals, concerns are raised by public health organizations regarding the duration and efficacy of antibody responses by current vaccines (8).

Next-generation COVID-19 vaccines which have been approved for emergency use, namely mRNA vaccines such as BNT162b2, mRNA-1273 and adenovirus-based vaccines, have their own advantages and disadvantages. Obviously, the main advantage is their remarkable efficacy of 95% for Pfizer^® mRNA vaccine, 94.1% for Moderna^® and around 80% for Ad26.COV2.S (vector based vaccine) respectively (9–11). On the other hand, their adverse reactions has raised concerns with respect to their safety. The incidence of local and systemic adverse reactions (AR) has been shown to be relatively high for mRNA and adenovirus-based vaccines with Local AR reaching 40%-88.9% and systemic AR between 44% and 86%, respectively (12).

Virus-like particles (VLPs) are considered traditional vaccine platforms. They are virus-derived structures and have the ability to self-assemble and mimic the parental virus in size and shape. One important aspect is that VLPs lack any genetic materials; accordingly, they are not capable of infecting the host cell or replication (13). Such traditional VLP-based vaccines have been approved for human use against different viruses decades ago. For example, vaccines against Hepatitis B and E viruses (HBV and HEV), Human Papilloma Virus (HPV) and the newly developed malaria vaccine (14). There are currently three VLP-based vaccines which have also been tested for COVID-19 in preclinical and clinical trials (15). The traditional vaccine platforms have shown lower incidence of adverse side effects (16); however they typically also showed inferior immunogenicity in comparison to the next-generation mRNA vaccines (17–19).

In line with the above, heterologous prime-boost vaccination strategies using different vectors should be tested to elicit broader and more efficient protective immune-responses with better and improved safety profiles. This strategy may meet the emergency needs in the current pandemic. In the current study, we evaluated different prime and boost regimens combining mRNA-1273 with a VLP-based vaccine based, namely mCuMV_TT-RBM. We have previously designed and developed the scalable and immunogenic VLP-based COVID-19 vaccine mCuMV_TT-RBM, which has been shown to induce RBD-specific IgG and IgA antibodies with strong neutralizing capability (17). We show that priming and boosting with mRNA-1273 or mCuMV_TT-RBM induced high levels of high avidity antibodies. Interestingly, highest antibody responses were measured, when vaccination was performed by heterologous administration of mRNA for prime and VLP as boost and vice versa, priming with VLP and boosting with mRNA. Thus, these findings, heterologous prime boost strategies may be able to improve efficacy and economics of current vaccine protocols.

The highly contagious SARS-CoV-2 virus has put a heavy toll on the worlds public health systems and caused a disruption of the world economy (25). The use of masks, physical distancing, contact tracing as well as isolation of sick or infected people are important but insufficient measures to reduce the spread of the corona virus disease (26). Accordingly, vaccine campaigns are needed to curb the spread of SARS-CoV-2 and thereby reducing the mortality and morbidity associated with the disease.

The commercially available mRNA vaccines require stringent storage conditions during transport and at vaccination sites as they have a short half-life (27). According to manufactures information, mRNA-1273 is stable up to 6 month when kept at -20°C, whereas to store BNT162b2 for the same time-period, temperatures between -60°C to -80°C are needed (9, 10). With regard to stability at room temperature, mRNA-1273 is stable for up to 12h; in contrast BNT162b2 has to be administered within 6h after thawing (28). These strict storage conditions represent a major obstacle for efficient use in less affluent parts of the world. Beside the challenging storage conditions, a further obstacle represents the production-costs of RNA-based vaccines, limiting their global access. BNT162b2 has a production cost of $19.50 and mRNA-1273 with even higher costs at $32-37 (12, 29). This can cause a major problem in low-income countries, to afford these types of vaccines. Therefore, cost effective and efficient vaccine candidates are of high importance considering the continuous persistence of the virus and the obstacles in eradicating it worldwide. We have previously shown high stability of our newly developed VLP-based vaccine-candidate mCuMV_TT-RBM which was stable for at least 14 months at 4°C without signs of degradation (17). This has also been confirmed with other preclinical studies showing the stability of mCuMV_TT -based vaccines for up to 12 months when kept at 4°C. Thus indicating a significant advantage for such VLP-based vaccines (30). Additionally, the production of mCuMV_TT-RBM is highly scalable allowing the production of millions of the doses in a single 1000L fermenter run (15). The rapid pace of mRNA vaccine development accompanied with fears of potential long-term adverse effects have raised some concerns in global community. VLP-based vaccines are therefore of high interest as they display a safe platform, due to the lack of replicating genetic material which is also been confirmed in the approved and marketed vaccines (HBV-HEV-HPV and recently Mosquirix™) (13). In addition, the here employed mCuMV_TT technology does not need use of an adjuvant.

Our data show that both vaccine-types, mRNA-1273 and mCuMV_TT-RBM are capable of inducing high RBD-specific antibody titers 7 days following the priming dose. Interestingly, mRNA-1273 shows a superior induction of antibody response in comparison to the VLP-based vaccine, at least at the high dose chosen. However, after the administration of the 1^st homologous boosting, this difference was not statistically significant anymore (Group I and III). The same was true for the heterologous boosts (Group II and IV), indicating the ability of VLPs to efficiently boost the B cells primed by mRNA. We have previously hypothesized that natural infection with SARS-CoV-2 virus induces short-lived neutralizing antibodies due to the unusually large distance between RBD-epitopes embedded in the membrane surface (31, 32). This obstacle may be overcome by grafting RBM or RBD onto highly repetitive nanoparticles such as the here used mCuMV_TT VLPs, resulting in successful display of the virus-epitopes at optimal distances of 5-10nm. This pathogen-associated structural pattern (PASP) may also play a role in the induction of higher avidity antibodies as shown after the 2^nd booster dose.

With the continuous emergence of SARS-CoV-2 VoCs, it may be essential to administer repetitive booster doses, as higher antibody levels may need to be maintained for sustained protection. Indeed, our data demonstrate that a 2^nd boost can maintain antibody titers. Furthermore, the 2^nd boost in the different heterologous regimens in groups II, III and IV have significantly enhanced the avidity of the antibodies, likely a key attribute for neutralization of emerging new variants.

The neutralizing capacity of the induced antibodies following the homologous or heterologous vaccination strategies showed that both approaches are efficient against the wild-type SARS-CoV-2. No statistical difference in the ability to neutralize different VoCs has been detected between groups I, II and IV. Furthermore, heterologous boosting of mice primed with mRNA-1273 or mCuMV_TT-RBM revealed similar outcome. These data confirm the ability of mCuMV_TT-RBM to efficiently boost antibodies primed with mRNA-1273 vaccine in a similar way to the standard followed vaccination strategy (mRNA-1273 prime/boost).

Collectively, our data support premise of using a heterologous prime/boost vaccination regimen against COVID-19. CuMV_TT-RBM VLPs may therefore constitute an efficient platform for boosting previously primed B cell responses; here we demonstrate this for mRNA primed responses, but it is likely that this may extend to B cell responses primed by other modalities, including vector- or virus induced immune reactions.

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

The animal study was reviewed and approved by Swiss Federal Veterinary Office.

A-CV, LJ, PK, XC, AZ, MM, and MB: Design of experiments, acquisition of data, interpretation, and analysis of data. A-CV, MM, MV, and MB: Writing, revision, and editing of manuscript. LJ, MV, and AZ: Technical, material, and tool support. MM and MFB: Study supervision. All authors read and approved the final manuscript.

This work was supported by Saiba AG and Inselspital Bern. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication. Open access funding was provided by the University Of Bern.

Author BM was employed by company Artemis Bio-Support. MB is a board member of Saiba AG and holds the patent of CuMVTT-VLPs. MM received payments by Saiba AG to work on the development of vaccines. MB and MM are shareholder of Saiba AG.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The reviewer SL declared a shared affiliation with the author MB to the handling editor at the time of review.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.