1 Introduction

Undoubtedly, vaccines are one of the cornerstones of modern medicine, keeping dozens of life-threatening diseases in check and preventing hundreds of millions of cases of illness, thus saving millions of lives every year [1]. A long way from Jenner’s groundbreaking experiments, the development of genetic vaccines is a current hot topic of biomedical studies, made possible by such advances of the previous century as nucleic acid sequencing, amplification, and synthesis. Considered to be one of the pioneering first steps, a report by Wolff et al. showed that protein is readily expressed by intramuscularly injected, naked plasmid DNA or mRNA, thus ushering in a new era of vaccine research [2]. And understandably so: genetic vaccines have many potential benefits over their conventional counterparts; these will be briefly discussed later.

DNA-based vaccines have shown great promise in preclinical studies, inducing potent immune responses in small animals [3, 4]. However, large animal experiments and clinical trials have mostly been less successful so far, with immunogenicity remaining below expectations [5,6,7]. This is currently one of the major issues genetic vaccines need to overcome, and RNA-based vaccines are sure to benefit from the knowledge and experience already accumulated in DNA vaccine studies in this and other regards as well.

RNA-based vaccines are a relatively new addition to the field of genetic vaccines—dismissed for a long time as non-feasible mainly because of stability issues. While they share some basic features with DNA vaccines, RNA vaccines may offer solutions to some of the concerns and limitations of the former. In this review, we will attempt to summarize the advancements that make RNA-based vaccines a viable alternative to DNA, pointing out possible advantages of each over the other, and of both over conventional vaccines. Furthermore, we will elucidate on the current and possible future applications of RNA vaccines and on the challenges that need to be overcome in this regard.

2 The Emergence of Genetic Vaccination

Conventional vaccines, despite their clear value, face limitations in areas of safety (live attenuated vaccines), lack of broad and/or durable protective efficacy against many pathogens, and ease and time of production. As recombinant DNA techniques developed and became widespread, delivering genetic material as a vaccine seemed feasible. Paoletti et al. [8] modified vaccinia virus to express hepatitis B virus surface antigen or Herpes simplex virus glycoprotein D, then inoculated rabbits and observed antigen production along with the appearance of high-titer antibodies [8]. This study can be considered as a stepping stone between conventional whole virus vaccines and genetic vaccines. In 1990, pure genetic material—both DNA and RNA, as a means of protein expression—was delivered by Wolff et al. into mouse muscle, and protein production was detected [2]. Comparing plasmid DNA and mRNA encoding firefly luciferase , the authors found that peak activities reached similar levels for both, however, mRNA-encoded luciferase activity rapidly diminished, whereas substantial amounts of luciferase were detected for 60 days in the case of plasmid DNA. Concerns about the stability, difficult delivery, and overall low translatability of RNA-steered gene therapy and genetic vaccine research toward the study and development of DNA-based approaches. In 1993, Ulmer and colleagues used plasmid DNA to express a conserved internal influenza protein in mice, and protection from lethal heterologous influenza virus challenge was demonstrated [9]. Multiple medical fields, like vaccine and cancer immunotherapy research, were eager to adapt the DNA platform thereafter. Although, in the same year, a study showed that antigen-specific cytolytic T cells could be induced by delivering influenza virus nucleoprotein-encoding mRNA formulated into liposomes [10], mRNA platforms were not intensively pursued for some time.

One of the most important milestones of vaccination, genetic or conventional, was the leap in gene and whole organism sequencing. In 1995, the genome sequence of Haemophilus influenzae was published [11], and many more have followed in the years. Sequencing and annotating a genome can provide a large number of potential targets for recombinant protein vaccines or genetic vaccines. Candidate antigens can then be screened and tested. One of the notable achievements of this “reverse vaccinology” method was the development of a universal vaccine against serogroup B meningococcus using recombinant proteins [12]. With advances in whole genome sequencing technologies, pathogens can be sequenced in a matter of days, enabling almost instant access to prospective targets for genetic vaccination.

Recent technological innovations enabled RNA to be a feasible alternative to both conventional and DNA-based vaccines. Nucleoside modification, stringent purification, and optimization of the coding and untranslated regions of RNA were demonstrated to drastically improve in vivo translatability by allowing synthetic RNA to evade toll-like receptors and a variety of other RNA sensors that trigger inflammatory reactions and consequently block mRNA translation [13,14,15,16]. While the majority of genetic vaccine research focused on DNA vaccines in the 1990s and 2000s, RNA vaccines emerged as a viable alternative in the last decade, and today both approaches are vigorously pursued.

3 Pros and Cons of DNA and mRNA Vaccines

One of the reasons why genetic vaccination (whether DNA or RNA) holds such promise is that it generally shows much greater ability to induce CD4+ and CD8+ T-cell responses than other noninfectious vaccine formats, e.g., protein subunit and whole inactivated pathogens [9, 17,18,19,20]. An important factor in determining the quality of the T-cell response appears to be the route by which antigens are taken up and presented by antigen-presenting cells (APCs). This is particularly well-documented in the case of CD8+ T-cell activation, where direct presentation of antigens on Major Histocompatibility Complex (MHC) class I that are translated within an APC can be much more efficient [21, 22] in stimulating CD8+ T-cell proliferation and effector functions compared to “cross” presentation [23] of proteins or pathogens that are endocytosed from the extracellular space by the APC. While it is conventionally thought that CD4+ T cells are mainly activated via an exogenous, endolysosomal pathway, there is now mounting evidence that CD4+ T cells are also activated most efficiently by direct, endogenous presentation, though the relevant pathways are incompletely understood [24]. One possible explanation for these phenomena is the greater magnitude and half-life of peptide/MHC complex presentation that is expected in the case of continuous, direct presentation [25, 26]. It should also be noted that T cells activated by exogenous pathways may not recognize epitopes that are presented on cells infected by the target pathogens, since the different proteolytic activities of different cellular compartments (e.g., cytosolic proteasome vs. late endosome) likely select for separate populations of peptides that can be presented on MHC molecules. Both DNA- and RNA-based vaccines may therefore promote high-quality T-cell responses in part by mimicking intracellular infections.

Still, there are some important differences in antigen presentation between DNA and RNA vaccines, and there are still many unknowns. DNA delivered into the muscle by electroporation is believed to transfect mainly myocytes as opposed to professional APCs such as dendritic cells (DCs) and macrophages [2, 27], and in this way may mostly rely on the less efficient cross-presentation pathways [28] for T-cell activation. However, DCs can also be directly transfected by DNA and prime CD8+ T cells [29], and DNA-transfected myocytes can upregulate MHC class I and co-stimulatory molecules, thus priming them to activate naïve CD8+ T cells [30]. The contribution of these various antigen presentation pathways to DNA vaccine immunogenicity is not well understood. mRNA, on the other hand, is generally efficient in transfecting DCs, macrophages and other cell types and is thus, in theory, well positioned to induce robust, direct antigen presentation. To be sure, more research is needed to investigate the antigen presentation and T-cell activation pathways at play for both of these categories of vaccine.

Stemming from their closely tied function and role in gene expression, DNA and RNA share other advantages over conventional live attenuated, killed, or subunit vaccines, the most obvious being safety: delivery of a single protein component of a pathogen encoded as a piece of nucleic acid carries no risk of the vaccine itself converting to a pathogenic form and causing illness, which is a concern in case of live attenuated vaccines, especially in immunocompromised patients [31]. Secondly, production of nucleic acids is undoubtedly simpler and quicker than either large-scale protein production and purification, or live pathogen proliferation technologies. For example, current influenza virus vaccines take about 6 months to produce (which may cause the deployment of a given batch to miss the peak of the number of incidents in a given influenza season) [32, 33], but nucleic acid-based vaccines could be deployed in a fraction of that time, as in vitro manufacturing is rapid, cost-effective and scalable. Moreover, once the manufacturing process is in place for a specific type of genetic vaccine, it can theoretically be used to produce any antigen-encoding vaccine of the same format—a level of simplicity and versatility that does not exist for the production of viral or protein subunit vaccines. This is also an important point in the case of emerging pathogens, as quick response is critical in preventing an epidemic. Thirdly, using the host’s own transcription–translation machinery may be beneficial in some cases to allow natural protein folding or posttranslational modifications, including the potential generation of defective ribosomal products [34], thus increasing T-cell epitope availability compared to in vitro antigen production. It also needs to be mentioned that, for obvious reasons, both DNA and mRNA vaccines are only able to deliver protein antigens. Although in the majority of cases proteins serve as targets for recognition during pathogen infection , sometimes other molecules, e.g., bacterial polysaccharides, are also important antigens [35].

Stability is a critical requirement of pharmaceutical products, since long periods of time may pass between production and administration, not to mention the desirable option of stockpiling certain seasonal, large-demand vaccines. Another serious consideration is the possible need for cold-chain transport and storage, especially in economically developing countries where necessary infrastructure is limited. In this regard DNA has a clear advantage, as it is certainly more stable both in vivo and in vitro, RNA being very sensitive to contamination—ever-present RNases are able to degrade RNA in minutes. However, RNA nucleoside modifications, addition of suitable 5′ and 3′ untranslated region (UTR) elements, cap structure and a sufficiently long poly (A) tail can stabilize cytoplasmic RNA and ensure high levels of translation in vivo and in vitro. Additionally, appropriate manufacturing conditions can ensure that synthetic mRNA is free of RNase contaminants, and encapsulation of RNA by a wide variety of new methods [36] can effectively protect RNA from degradation by RNases in vivo. Instability, however, can also be turned into an advantage, in cases where more control over the duration of protein expression is desired. DNA, once delivered, can persist for months [2, 37, 38] without the option for removal, whereas with mRNA, controlled short intervals or bursts of expression can be achieved by modifying the above-mentioned stabilizing elements.

Persistence of DNA brings up a serious concern for safety, as it is theoretically possible for DNA to integrate into genomic DNA. This is a particular concern for virus-derived DNA, but plasmid DNA has also been rarely associated with genomic integration [39]. DNA-based expression constructs need to have, in principle, transcription-driving elements (i.e., a strong promoter). When integrated near e.g., proto-oncogenes, these elements can cause oncogenic transformation, leading to cancer, like in some cases of non-plasmid DNA therapeutics [40, 41]. Simple knock-out of a tumor suppressor can have similar consequences. Transformation of the host’s germline genome is even more of an issue, as this may be inherited by offspring [42, 43]. It needs to be emphasized that the possibility of DNA integration in the case of DNA-based vaccines is, as of yet, theoretical, with data pointing to no significant integration in some studies [37, 39, 40]. However, formulation and individual DNA sequences may show different behavior in this regard, thus requiring strict investigation for each DNA-based therapeutic [39]. RNA, on the other hand, has no known way of integrating into DNA, unless reverse-transcribed into DNA first. Again, there is a theoretical possibility of reverse transcriptase being present in a human cell (e.g., in case of retroviral infection or from retrotransposons), but this necessary (and unlikely) extra step taken together with short persistence of RNA make RNA-based integration a highly improbable event. Also, since mRNA molecules do not need to contain transcription-driving elements, integration of their cDNA would not cause overexpression of genes near the integration site. Investigation may be needed to clarify whether RNA-based integration in special cases (e.g., retroviral infection ) can occur.

Furthermore, while DNA needs to be localized to the nucleus, translation of RNA occurs in the cytoplasm, simplifying delivery strategies.

Large-scale manufacturing of DNA constructs is straightforward and well-established under Good Manufacturing Practice (GMP) conditions [44, 45]. Production of mRNA also requires production of a DNA template, in addition to several extra steps, namely in vitro transcription of mRNA, formation of a cap structure (which may be accomplished during transcription or done separately) [46,47,48], and removal of aberrant transcription products, such as short, aborted mRNA and double-stranded RNA fragments [14, 15]. Furthermore, most RNA-based vaccines are protected from RNase and delivered to the cytoplasm by lipid or polymer-based carriers, which also must be produced under GMP conditions. Although this extended pipeline carries additional costs and infrastructure relative to DNA purification alone, it is important to note that in vitro transcription of mRNA is a highly processive enzymatic reaction, in which a given amount of DNA template can yield one or two orders of magnitude more mRNA. Therefore, extremely large quantities of mRNA can in theory be produced from a relatively small amount of plasmid DNA, decreasing costs associated with the upstream DNA production. Advantages and disadvantages of DNA and RNA vaccines are summarized in Table 1.

Table 1 Comparison of beneficial and detrimental features of DNA and RNA vaccines
Full size table

4 mRNA Vaccine Types and Delivery

Current RNA vaccines are engineered in two basic forms: nonreplicating and self-amplifying [13, 49]. In principle, DNA vaccines are more closely related to nonreplicating mRNA vaccines, and thus are compared to them in this review. However, it is useful to elucidate on the differences of these two types of RNA vaccines. Nonreplicating mRNA vaccines use an RNA molecule containing the open reading frame of interest flanked by 5′ and 3′ UTRs, a cap structure and a poly (A) tail. Self-amplifying RNA vaccines carry additional elements derived from positive-strand RNA viruses [50]. The sequences encoding viral structural proteins are replaced by the gene of interest, while components of the RNA replication machinery are retained. When the vaccine is delivered into cells, the RNA-dependent RNA polymerase complex is expressed and begins replicating the antigen-encoding RNA. The obvious advantage of this mechanism is that a lower dose of the vaccine is required to achieve the same or even higher level of protein expression. While expression is delayed compared to nonreplicating mRNA vaccines because of an initial lower number of molecules present, overall expression is more durable [51]. Nonreplicating mRNA vaccines on the other hand produce a short, strong burst of protein expression, and may be more advantageous if tighter control of persistence is required. Nonreplicating mRNA vaccines can be produced with unmodified or modified ribonucleotides, while self-amplifying mRNA vaccines use the unmodified nucleotides present in cells for their replication.

As with any therapeutic, delivery is a crucial aspect of genetic vaccine applications. In the case of mRNA vaccines, delivery has proven to be a challenge, bottlenecking efficiency for a long time. Although the most straightforward method—naked mRNA administration by intramuscular injection—was used by a number of studies [2, 52], susceptibility to RNases and low overall efficacy necessitated the development of new approaches. One of the early attempts was the use of the “gene gun” —mRNA-coated gold particles penetrating the cell membrane with high velocity [53,54,55]. Protection of mRNA and increased delivery efficiency can also be achieved by protamine condensation, with the added benefit of higher levels of immune response stimulation [56,57,58,59,60]; however, protamine was found to bind RNA too tightly, which inhibited translation [61, 62], thus, this complexing agent is no longer commonly used for the delivery of protein-encoding mRNA. Electroporation has been used for DNA vaccine delivery, and it is also an option for mRNA vaccines, both in vivo and in vitro [63]. Current delivery strategies favor nanoparticles with cationic lipids (lipid nanoparticle , LNP) or polymer-based approaches. Many of these have been used for over a decade for siRNA administration [64], and have been adapted to mRNA in recent years [65]. Cationic lipids have been used in a plethora of nucleic acid delivery studies, and have provided promising results [13, 36, 66]. They may also be combined with polymers for different formulation strategies [67]. Cationic polymers/dendrimers form complexes with negatively charged mRNA spontaneously, are easy to manufacture and modify, and hold great potential, but their use is not as widespread as that of LNPs in clinical applications. Current efforts focus on creating biodegradable formulations with minimal cytotoxicity [36, 68].

5 Current Standing of the Field of mRNA Vaccines

Currently, mRNA vaccines are mainly used in two therapeutic approaches: cancer immunotherapy and infectious disease vaccines. Several detailed reviews have recently been published on this topic [13, 36, 51, 66, 69,70,71], thus, we will only briefly discuss the current standing of these fields and highlight the most critical findings.

5.1 mRNA-Based Cancer Vaccines

RNA can be used in two quite different strategies in cancer immunotherapy: total tumor RNA or single tumor-associated antigens can be delivered (reviewed in [72]). Total RNA can be extracted from patients’ tumors and used to generate anticancer immune responses against the set of antigens expressed by the particular tumor. Advantages of this approach include the lack of need to identify individual tumor antigens and lower chances of immune evasion by the tumor via immune escape mutations. On the other hand, autoimmunity is a possible issue, as total RNA includes targets also expressed by healthy cells. Multiple clinical trials on different cancer models have been completed using this approach [68, 73,74,75,76,77,78,79]. However, this strategy is very different from the principle of DNA vaccines, which serve as a basis for comparison in this review. Thus, we will focus on tumor-associated antigen RNA cancer vaccines, highlighting the most important results in this vast and intensively investigated field.

Dendritic cells are ideal targets for cancer vaccine delivery, since they play a key role in inducing antigen-specific immune responses, and strategies for their therapeutic use in conjunction with mRNA vaccines have been evaluated in recent decades. DCs can be directly purified from patients’ samples, or monocytes can be extracted and stimulated with granulocyte-macrophage colony stimulating factor and interleukin-4 to differentiate into DCs [80, 81]. These are then transfected with mRNA ex vivo, and finally re-infused into the host. Direct in vivo delivery of RNA cancer vaccines is also possible. While it is more cost effective and rapid than ex vivo DC transfection, targeting of desired cells can be challenging [82].

In 1996, Boczkowski and coworkers electroporated DCs with in vitro synthesized chicken ovalbumin (OVA)-encoding mRNA or OVA protein, and mRNA proved to be more effective at stimulating CTL responses against OVA-expressing tumor cells in vitro [83]. In 2002, a phase I clinical trial was performed to evaluate an mRNA cancer vaccine for safety, feasibility, and efficacy in the treatment of metastatic prostate cancer. Thirteen study subjects received different doses of DCs transfected with prostate-specific antigen (PSA)-encoding mRNA; no adverse effects were observed. PSA-specific T-cell responses were detected in all patients, and 6 out of 7 evaluable patients exhibited a significant decrease in serum PSA levels. Analysis of circulating tumor cells of 3 patients showed transient molecular clearance in all 3 cases [84]. Discovery of several immune regulatory proteins has contributed to better efficacy of DC cancer vaccines: electroporation of DCs with mRNA encoding costimulatory molecules such as CD83, tumor necrosis factor receptor superfamily member 4 (TNFRSF4, also known as OX40) or 4-1BB ligand (4-1BBL) led to measurable increase in immune stimulation via DCs [75, 85,86,87]. mRNA can also be used to modulate the function of DCs by delivering pro-inflammatory cytokines or trafficking-associated molecules. TriMix, developed by Etherna, is a three-component adjuvant containing CD70, CD40L and constitutively active toll-like receptor 4 (TLR4) encoded as mRNA. It can be readily combined with antigen-encoding mRNAs and delivered into DCs via electroporation [88]. In one study, this formulation was successfully used to treat stage III and IV melanoma patients, resulting in the regression of tumor in 27% of treated individuals [89]. Completed cancer clinical trials using DC vaccines include metastatic prostate cancer, metastatic lung cancer, renal cell carcinoma, multiple types of brain cancer, melanoma, acute myeloid leukemia, and pancreatic cancer, among others [70, 90, 91].

As individual cancer cases can vary widely in terms of the phenotype (and of course genotype) of the affected cells, a universal vaccination method is not feasible. Thus, it has been a long-standing goal in the field to develop methodologies that can produce personalized therapeutics. A current leader in mRNA pharmaceuticals, BioNTech, has developed a platform called Individualized Vaccines Against Cancer (IVAC) that either uses a combination of pre-manufactured mRNA-encoded antigens based on the RT-qPCR profile of the tumor, or, alternatively, a vaccine that is tailored to an individual’s unique tumor antigens resulting from tumor-specific mutations identified by next generation sequencing of the tumor [92]. These studies have been crucial in demonstrating the feasibility of GMP-grade, large scale manufacturing and safety of mRNA vaccines and have pioneered the way for the application of mRNA to other therapeutic indications.

5.2 Infectious Disease mRNA Vaccines

The field of infectious disease vaccines offers useful points of reference for judging the effectiveness of new mRNA platforms, since they can be directly compared to conventional vaccine formats—a benchmark that is absent in the cancer vaccine field. Although fewer in number than mRNA cancer vaccines as of today, there are multiple ongoing or completed preclinical and clinical studies with mRNA infectious disease vaccines, and the field is rapidly expanding. mRNA vaccines have been developed against multiple viruses and several bacterial and parasitic targets. A detailed review on infectious disease mRNA vaccines has recently been published [66]. Here, we will briefly summarize the most relevant findings and mention the advantages of particular mRNA vaccines compared to current vaccine formats in the field.

HIV is one of the most difficult vaccine targets, which no approach has been able to effectively tackle so far. Several mRNA vaccine studies used DC vaccines, where DCs were electroporated with HIV antigens encoded as mRNA and injected into HIV-1 infected patients on antiretroviral therapy. Although the treatment gave rise to antigen-specific CD4+ and CD8+ T-cell responses, no clinical benefit was observed [93,94,95,96,97,98]. Attempts at directly injectable HIV vaccines have resulted in similarly suboptimal outcomes. Multiple studies using different formulations of mRNA vaccines have demonstrated HIV-specific T- and B-cell responses; however, antibodies induced by these vaccines could almost exclusively neutralize tier 1 (easy-to-neutralize) viruses and none of these vaccines could induce durable neutralizing antibody responses [99,100,101,102,103].

Influenza virus causes frequent, worldwide epidemics, but despite intensive research and decades of experience, conventional vaccines are limited in their ability to respond to rapidly changing viral strains. Most currently used influenza vaccines are produced in eggs, which may not be ideal in case of certain virus strains (e.g., H5N1, which affects chickens), and can also give rise to egg-adaptive mutations, further decreasing the efficacy of currently used vaccine formats [104, 105]. As mentioned above, conventional influenza vaccines could take about 6 months to deploy, whereas in 2013, Hekele et al. produced an mRNA-based influenza vaccine within 8 days after the coding sequences became available for the hemagglutinin (HA) and neuraminidase (NA) of the H7N9 strain causing an outbreak in China. Mice treated with the vaccine exhibited protective HA inhibition titers after two injections in 4 weeks [106]. The first demonstration of protection from influenza virus infection involving an mRNA vaccine was described by Petsch et al. in 2012. Immunogenicity at a similar level to a licensed inactivated virus vaccine was observed in mice, ferrets, and pigs [107]. Recently, in a publication by Pardi et al., nucleoside-modified, HPLC-purified mRNA encoding full length HA elicited durable antibody responses against the more conserved, but generally immunosubdominant stalk region of HA that is one of the targets of “universal” influenza vaccine research. A single immunization was protective against both homologous and heterologous influenza viruses, and two immunizations protected against heterosubtypic virus challenge in mice [108]. Two human clinical trials have been completed with LNP-complexed, nucleoside-modified HA mRNA influenza vaccine. 100 μg H10N8 HA mRNA was injected intramuscularly (IM), and 100% and 87% of 23 study subjects showed measurable levels of HAI and microneutralization titers, respectively, with side effects similar to conventional adjuvanted vaccines. For H7N9, 25 and 50 μg IM doses achieved measurable HAI titers in 96.3% and 89.7% of participants, and microneutralization titers reached 100% and 96.6%, respectively [109]. While these results are promising, antibody titers were significantly lower than in previous animal models, an issue that requires attention.

Following the outbreak and publicity of Zika viral infections in 2015–2016, efforts were made by many groups to quickly develop a vaccine, and the mRNA platform has shown more than promising results. Two groups independently demonstrated the efficacy of their own LNP-formulated, nucleoside-modified mRNA vaccines delivering Zika virus pre-membrane and envelope (prM-E)-encoding mRNA. Richner et al. observed high levels of neutralizing titers and protection from Zika virus challenge after 2 immunizations in mice [110]. Pardi et al. achieved similar results in both mice and NHPs after administration of a single dose [111]. Evaluation of the vaccine by Richner et al. has moved into the clinical phase (NCT03014089).

Animal studies of a rabies virus glycoprotein mRNA vaccine had shown promise, inducing protective immune responses against lethal challenge in mice, and raising significant levels of antibody titers in pigs [112]. Based on the promising preclinical data, a human study was initiated, using different doses and delivery routes. The vaccine failed to live up to expectations when delivered via needle injection either intradermal or intramuscular independent of dosage. Needle-free administration induced adequate antibody responses, but they declined after 1 year [113]. Optimization of the delivery system has been shown to increase durability of this vaccine. An LNP-formulated version of the vaccine was administered to NHPs, and adequate neutralization titers were observed, and remained stable for at least 5 months [114]. Testing of the vaccine has recently moved to phase I clinical trial (NCT03713086).

A limited number of mRNA vaccines against bacterial pathogens have also been evaluated. Lorenzi and colleagues intranasally administered naked, unmodified mRNA encoding Mycobacterium tuberculosis heat shock protein 65 (Hsp65) into mice and challenged them 30 days later. Bacterial loads were measured 4 weeks post-challenge, and a significant decrease was observed in mRNA-immunized animals [115]. Maruggi and colleagues demonstrated the efficacy of a self-amplifying mRNA vaccine approach against Streptococcus group A and B [116]. Streptolysin-O (group A) and pilus 2a backbone protein (group B) were administered, and the treatment conferred partial protection in a mouse challenge model.

Three reports have recently demonstrated that multiple mRNA-encoded antigens can be encapsulated into a single vaccine formulation and induce equally potent immune responses against each antigen [117,118,119]. This critical feature of mRNA vaccines offers the advantage of hitting a pathogen at multiple points, which could result in improved breadth and protective efficacy. Awasthi et al. used a trivalent nucleoside-modified mRNA-LNP vaccine against HSV-2, and compared it to an adjuvanted trivalent protein vaccine containing the same antigens the mRNA vaccine coded for [117]. The mRNA vaccine completely protected mice from HSV-2 challenge. No virus in vaginal cultures and no HSV-2 DNA in dorsal root ganglia was detected after vaginal infection with a viral dose at which trivalent protein vaccine was not effective. Chahal et al. reported on a multivalent mRNA vaccine against Toxoplasma gondii, an intracellular pathogen that expresses variable antigens throughout its life cycle. Six self-replicating mRNAs were co-formulated and administered, inducing protective immunity in mice [118]. John et al. combined five conventional mRNAs encoding the five subunits of the human cytomegalovirus (HCMV) pentameric complex, and robust immune responses were observed in NHPs. Addition of a sixth mRNA, one against HCMV glycoprotein B, did not interfere with responses, suggesting that mRNAs encoding different antigens can be efficiently combined [119]. This six-component vaccine is currently investigated in a phase I clinical trial (NCT03382405).

Eukaryotic unicellular parasites represent an important group of human pathogens, causing deadly diseases such as malaria, leishmaniasis, or trypanosomiasis. Malaria kills approximately one million people each year, and conventional approaches to create a vaccine have not been successful so far [120]. A self-amplifying mRNA vaccine was demonstrated to be effective in disabling an immune evasion mechanism of the malaria parasite in which the parasite produces macrophage migration inhibitory factor (PMIF) to suppress memory T cells. mRNA-encoded PMIF was delivered to mice, and both cellular and humoral responses were observed, and the animals were protected against reinfection [121]. A self-amplifying mRNA vaccine against Leishmania was tested by Duthie and colleagues. A heterologous prime-boost regimen using LEISH-F2 RNA and protein significantly reduced Leishmania infection in the liver of mice compared to homologous immunizations by either RNA or protein (two administrations in each case) alone [122].

6 Challenges and Future Directions

To date, genetic vaccines have only been approved for veterinary products [69]. One of the main issues with human applications, as stated above, is that clinical genetic vaccine studies so far have reported more modest immunogenicity in humans than in preclinical animal models. Three recent phase I clinical trials using mRNA vaccines against influenza and rabies viruses reported mediocre results [109, 113, 123], an issue shared with previous DNA vaccine trials [7, 124]. Since the field of RNA vaccines is relatively new, there is currently not enough clinical data to compare and evaluate the multiple RNA platforms available today in this regard. This is sure to change in the near future, as new formulations with different expressional and immunostimulatory profiles are tested. Further studies will need to focus on adapting the very encouraging results of preclinical trials to human applications. For DNA vaccines, some of the strategies for improving responses currently being tried are: increasing protein expression, addition of different adjuvants and immunostimulants, various prime-boost combinations (e.g., DNA prime with viral vector or protein boost), optimization of plasmid components (e.g., promoter, codon optimization), different delivery routes and delivery methods [69]. Similar strategies may be attempted for the improvement of RNA vaccines, in addition to the optimization of the RNA-specific structures (cap, 5′ and 3′ UTR, poly (A) tail) [13, 71, 125, 126].

Efficient delivery of mRNA as vaccines is of the utmost importance [36]. Recent advances in lipid and polymer biochemistry opened up new possibilities of delivery and targeting, benefiting RNA vaccines greatly. Optimization of RNA vaccines must involve the development and testing of as many different carrier molecules as possible, since different applications may require different delivery protocols.

Excitement about the prospects of RNA vaccines can be seen by the establishment of multiple pharmaceutical companies investing in RNA vaccine research, development and manufacture. BioNTech GmbH, CureVac AG and Moderna Therapeutics, partnering with such giants as Roche, Bayer, Sanofi, and Merck, are the current leading developers of the field, with billions of dollars channeled into RNA vaccine development.

Undoubtedly, RNA vaccines offer a more than promising alternative to conventional vaccines. While development of a truly effective genetic vaccine seems to be proving more difficult than originally imagined, the possible benefits are obvious. As both DNA and RNA vaccines are relatively novel in comparison to conventional platforms, and some infectious diseases have been unable to be tackled by conventional vaccines even with a significantly longer history of attempts, there is little reason to be discouraged by the as of yet unsatisfactory results of human trials. Genetic vaccine design and formulations offer a multitude of options for improvement.