*8.1. Immune Activation*

While both DNA and mRNA vaccines are often thought of as simply an expression system for the desired protein, neither is immunologically inert. Both DNA vectors (which are based on bacterial plasmids) and in vitro transcribed mRNA activate the innate immune system. DNA plasmids do so via their CpG motifs, which stimulate TLR9. While CpG was successfully used as an adjuvant [46] for a recombinant protein-based Hepatitis B vaccine licensed in 2018, the impact on the immunogenicity of DNA vaccines by increasing the number of CpG motifs in the plasmid has been less clear. In fact, for certain DNA vaccine efforts, notably those of Steinman's group, as therapies for autoimmune diseases, they specifically switched CpG motifs for GpG motifs (guanine connected via a phosphodiester bond to another guanine; these compete with CpG motifs for binding to TLR9 receptors) in an effort to specifically decrease the Th1 help for their human clinical studies (see below) [47]. The double-stranded structure of the DNA plasmid is also thought to be an immune stimulant [48] through non-TLR mechanisms. In fact, plasmid DNA also acts on the TBK1-STING pathway through cytosolic receptors [49,50]. This results in the generation of Type 1 interferons, which then act as adjuvants for the generation of immune responses against the antigen(s) encoded by the plasmid DNA vaccine.

As noted above, the use of modified nucleosides for the construction of mRNA is one method of decreasing the reactogenicity of the in vitro transcribed mRNA. However, mRNA acts via multiple pathways, including the innate system (via TLR3, TLR 7, and TLR8) and via cytoplasmic proteins (PKR, OAS, RIG-I, and MDA5) [11,51]. The multiple routes of activation result in several effects in addition to inflammation and include inhibition of mRNA replication (both via TLR7 through an MYD88 pathway affecting interferon, and via TLR3 through TRIF), stalled translation, and RNA degradation [5]. Some of these various activities could decrease the potency of the mRNA by a net decreased protein production, as was seen pre-clinically for an HIV mRNA vaccine complexed in cationic lipids [52]. This also raises the issue of how effective repeat dosing of mRNA will be if previous injections result in an environment with decreased translation or increased RNA degradation, although simply changing an injection site may potentially circumvent this particular issue.

Other molecular entities that are introduced or generated during the manufacture of the in vitro-transcribed mRNA and then remain (left-over contaminating nucleoside triphosphates, DNA templates, and dsRNA) are also quite immunostimulatory and therefore need to be purified following production of the mRNA [53].

The potential issues due to the various inflammatory effects of mRNA vaccines upon clinical efficacy and safety are summarized in Table 2 and are discussed below. The possible utility of RNA-induced inflammation for vaccines is demonstrated by the fact that one of the first uses of RNA for vaccines was to include non-coding RNA in human clinical trials as an adjuvant for a rabies vaccine (composed of an inactivated virus) [54], although this effort has been replaced by a rabies vaccine that utilizes mRNA that itself encodes the rabies antigen [55], as is discussed below. The continued evaluation of non-coding RNA as an adjuvant is ongoing in clinical testing for various cancers without the provision of an antigen (see below).

**Table 2.** Issues to be addressed for clinical efficacy and safety of mRNA related to inflammation.

	- Several pathways of RNA-induced inflammation: TLR 3, 7, 8, plus cytoplasmic pathways
	- Known toxicities of drugs containing unnatural modified nucleosides
	- Potential mitigation or enhancement due to formulation of the mRNA

## *8.2. Toxicities of mRNA*

The flip side of the possibly beneficial adjuvant inflammation, however, is potential toxicity of the mRNA vaccines. Toxicities are seen with antivirals and anti-cancer drugs that contain unnatural nucleoside analogues [56–58]. Such toxicities, not predicted by pre-clinical studies due to species differences between humans and the animals used for pre-clinical safety testing, have been seen with drugs that contain unnatural modified nucleosides. The clinical adverse effects have included myopathy (caused by mitochondrial toxicity), lactic acidosis, pancreatitis, lipodystrophy, liver steatosis, and nerve damage; certain ones have been fatal.

Indeed, some toxicity has been reported for mRNA pre-clinically along with limited human adverse events. Liver toxicity was observed in pre-clinical studies with one potential mRNA therapeutic delivered in lipid nanoparticles for Crigler-Najjar syndrome, selected as a "lowest-hanging fruit" target because very low doses of protein were needed. These were serious enough to apparently halt the work with this particular entity, or at least that formulation [59]. The formulation of the mRNA was thought to potentially play a role in the toxicity [60], and repeat doses were used. Nevertheless, this observed toxicity may be concerning for vaccines as well, since even live replicating viruses and viral vector vaccines (which generally are more immunogenic than subunit vaccines) need repeat dosing. In addition, most of the mRNA vaccines in clinical trials appear to need formulation. The mRNA vaccines in clinical trials against infectious diseases from this same company are described as formulated in lipid nanoparticles, but whether they are the same formulations as those used for the Crigler-Najjar study is not publicly known.

Self-limited local and systemic adverse events (AEs) seen in a human clinical trial for an mRNA rabies vaccine, although summarized as still indicating the vaccine was generally safe (described below in the clinical trials section), may also reflect the inflammatory nature of the mRNA [55]. These results highlight the potential toxicity downside of the inflammatory activity of mRNA vaccines, adverse effects not seen to this extent with plasmid DNA. Also note that, for providing monoclonal antibodies [61] (whether for preventing or for treating infectious diseases for other therapeutic applications), this would likely require repeat administration of mRNA, which might not only increase the potential for toxicities, but may also have an impact upon potency due to effects of the mRNA upon decreasing translation, etc., via the other inflammatory effects.

Thus, it may still be a work in progress to find the best balance of inflammation and any deleterious toxicities via harnessing adjuvant activities of mRNA while limiting or suppressing inherent toxicities for vaccines and immunotherapeutics. This will involve optimizing nucleoside substitutions, the design of other elements of the mRNA construct, any included immunostimulants, and/or specific formulations, delivery devices, and routes of administration. The mechanisms of mRNA inflammation that are relevant to their potential efficacy and safety as vaccines are also reviewed elsewhere [34], where they are aptly referred to as the "yin and yang of innate immunity".
