Many clinical trials of mRNA-based drugs or vaccines have failed to successfully pass Phase I or Phase II. The reasons behind this are various, including low efficacy of candidate drugs and lower-than-expected clinical risk/treatment benefit profiles. Preclinical safety assessments aim to identify well-tolerated and efficacious LNP-mRNA formulations, and when toxicity is observed, in vivo, in vitro and ex vivo experiments aim to understand the underlying mechanisms and, ideally, improve the formulations design under development.

The main safety issues of LNP-mRNA preparations in preclinical development can be divided into immunopathogenicity and liver and spleen toxicity (only studies on modified and/or dsRNA-purified mRNA are considered). Understand the different formulations of LNP-mRNA preparations and mechanisms of adverse effects help us better design and optimize drug candidates that are well tolerated and effective.

Liver and Spleen Toxicity

Since LNP-mRNA has significant liver and spleen biodistribution, microscopy and histopathology of the liver and spleen are standard practice during preclinical development. Acute drug-induced liver injury is routinely assessed by measuring plasma levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (APT). Most current published studies of LNP-mRNA therapeutic applications report only mild pathological findings.

In one such study, modified mRNA encoding human methylmalonyl-CoA mutase (hMUT) was synthesized to treat methylmalonic acidemia/aciduria (MMA), a serious A hypomorphic mouse model of a rare metabolic disorder. At the highest intravenous dose of the LNP-mRNA formulation, there were no clinical chemistry findings, but 80% of mice had mild paracentral splenic artery lymphopenia. This effect, which was attributed to LNPs rather than hMUT mRNA or its expression, was also observed in LNPs carrying mRNA for green fluorescent protein (eGFP).

In another study, modified mRNA encoding human arginase was used to treat a mouse model of arginase deficiency. Although there were no biochemical or histopathological findings, electron microscopy of liver sections from the group receiving luciferase mRNA single control preparation revealed the presence of submicron-sized lipid droplets.

Furthermore, a single intramuscular injection of an LNP-mRNA preparation encoding the influenza hemagglutinin H3 antigen showed increased levels of AST, ALT, and C-reactive protein in a rabbit model. Histopathological findings of the liver included focal subscapular vacuoles, inflammatory cell and erythroid infiltrates, and pleocytosis (lymphocyte expansion) was observed in the germinal center of the spleen.

Immune Response

The human body’s adverse immune responses to nanomedicines include reactogenicity, hypersensitivity reactions, systemic complement immune responses, and cytokine-mediated reactions after vaccination. For LNP-mRNA-based therapies in preclinical development, such events may compromise their safety and reduce therapeutic efficiency. However, due to the inherent compositional complexity of LNP vectors, it is not easy to determine which components of a given LNP-mRNA complex elicit unwanted innate immune responses and which conditions may exacerbate them (dose, route of administration, pre-existing inflammation, etc.) thing.

• TLR Activation and Cytokine Secretion

Stimulation of TLRs by LNP-mRNA is considered to be the upstream pathway of cytokine production. In a related study, lipopolysaccharide (LPS) was used to treat mice to induce inflammation. The inflamed mice then received a single intravenous injection of LNP-mRNA and exhibited IL-6, a C-C motif chemokine ligand 2 (CCL2) and other proinflammatory cytokines (in serum) and C-X-C motif chemokine ligand 2 (CXCL2, in liver) are elevated. Similar immune responses were observed when mice were given empty LNP. Research points out that the ionizable cationic lipid ALC-0315 is the most important immune-stimulating ingredient. The inflammatory phenotype is ablated in macrophage-depleted mice as well as Tlr4−/− mouse models. Interestingly, similar findings were observed with mRNA preparations using two other ionizable cationic lipids (DLin-MC3-DMA or C12-200).

In an independent study, colocalization of TLR4 and LNP-mRNA in endosomes following in vitro LPS activation of mouse macrophages corroborates the hypothesis that LNP-mRNA preparations trigger innate immune system responses via TLR4 in response to pre-existing inflammation. Under these conditions, LNP endosomal escape using cKK-E12 ionizable cationic lipids is impaired, while phosphorylation of protein kinase R (PKR) downstream of TLR4 results in reduced translation of intracytoplasmic mRNA. Although LNP alone was not introduced for control, these TLR4-related findings point to lipid-mediated biological effects on delivery efficiency.

Overall, TLR activation and pro-inflammatory cytokine release are common innate immune system effects induced by LNP-mRNA, sometimes causing strong adverse reactions and affecting protein translation. Activation of TLR4 by ionizable cationic lipids is a possible upstream initiating event, although the exact molecular mechanism has not yet been elucidated. Other factors may also be necessary to cause or exacerbate the inflammatory effects, which may depend on the payload, dose, or route of administration, etc.

• Inflammasome Activation

Inflammasome activation was recently identified as a unique innate immune system effect induced by LNP-mRNA. The canonical pathway of pyroptosis requires an initiation signal to activate NF-κB and initiate the transcription of NOD-like receptor pyrin domain protein 3 (NLRP3) and pro-IL-1β, the latter of which may originate from TLR activation. Secondary signals initiate the assembly and activation of the NLRP3 inflammasome.

In an in vitro study, NLRP3 inflammasome activation resulted in reduced mRNA transfection efficiency after administration of DLin-MC3-DMA-based LNP-mRNA to bone marrow-derived macrophages in vitro. LPS serves to provide the initiation signal, but mRNA may also promote initiation through TLR activation. Lysosomal disruption and damage-associated molecular patterns serve as secondary signals during LNP-mRNA escape. NLRP3 activation can be inferred from IL-1β release, cleaved gasdermin D and caspase 1 expression, and cathepsin B maturation.

When human peripheral blood mononuclear cells (PBMC) are treated with LNP-eGFP mRNA configured with DLin-MC3-DMA or SM-102, IL-1β and other pro-inflammatory cytokines such as IL-6, CCL2, CCL4 and TNF, etc.) release increases dramatically. Notably, the SM-102-based empty formulation also triggered strong IL-1β secretion, suggesting that these lipids provide priming and activation signals for inflammasome activation.

• Complement Activation and Hypersensitivity Reactions

Complement proteins in plasma and cell surfaces are essential components of the innate immune system. Complement activation, a series of proteolytic events, supports the phagocytic clearance of pathogenic substances or particles. The emergence of biologics and nucleic acid therapies has rapidly elevated immunotoxicological studies related to complement activation into routine practice in drug development. Identification of pathogenicity is the upstream pathway that initiates complement activation, of which there are three pathways: the classical pathway, initiated by pattern recognition of IgG or IgM; the alternative pathway, initiated by the hydrolysis of the thioester bond in complement protein C3.

LNP-mRNA preparations have been shown to activate the complement pathway. In an in vivo study in a cynomolgus monkey model, intravenous administration of LNP-mRNA expressing hEPO resulted in a mild and reversible increase in plasma complement protein C3a and C5b-9 levels. In another study, C3b/c and soluble C5b-9 were also increased after incubation of CD40L-expressing LNP-mRNA in complement-active human serum. Anti-PEG IgM is required for complement activation and is associated with loss of LNP integrity. These findings are consistent with previous reports that immune cells mediate accelerated clearance of nanomedicine formulations due to immunoglobulin opsonization and complement activation.