Modified mRNAs in the Cardiovascular System: A New Platform for Gene Therapy

Main Text

Gene delivery to the cardiovascular system as a therapeutic strategy has exploded recently, especially in the setting of inherited cardiomyopathies.1, 2, 3 In the last few years, a novel non-viral vector system based on modified mRNA (modRNA) has emerged. In this issue of Molecular Therapy, Sultana et al.⁴ report on the optimization of cardiac delivery of modRNA both in vitro and in vivo. Several recent reports, including Sultana et al.,⁴ make it clear that modRNAs have become an exciting platform for cardiac gene therapy. The rapid onset and transient nature of expression, off-the-shelf availability, excellent safety profile, and ability to be injected multiple times renders them ideal vectors for cardiac gene transfer.

The efficiency of transduction continues to be a major obstacle to successful gene delivery in the heart.¹ The vectors for gene delivery can, in general, be divided into viral or non-viral platforms. Different types of viral vectors have been used to deliver genes to a variety of tissues, including heart. Viral vectors have both advantages and limitations in the context of cardiovascular gene therapy. Recombinant adenoviral vectors have been quite popular in experimental models both in vitro and in vivo.⁵ Even though they drive very robust transgene expression, the inflammatory response they elicit limits their use in vivo. Lentiviral vectors can infect non-dividing cells within the myocardium and lead to long-term expression. The main limitations of such vectors are their integration into the host genome and poor transduction efficiency. More recently, recombinant adeno-associated vectors (AAVs) have become the most widely used gene delivery system in the cardiovascular field. AAVs are derived from the Parvoviridae family and consist of a capsid with a single-stranded DNA genome. Different AAV serotypes infect various organs with differing efficiencies and can induce long-term transgene expression without eliciting an immune response. Many recent clinical trials in cardiovascular diseases have made use of AAV vectors and have reported excellent safety profiles. However, like all viral vectors, AAVs can only be administered once because they elicit neutralizing antibodies that preclude re-injection.

Until recently, non-viral vectors have centered on DNA plasmids and their derivatives (liposome-DNA complexes [lipoplexes] and polymer-DNA complexes [polyplexes]), oligonucleotides, and their analogs, either alone or in complexes. Even though such vectors are easy to produce, they have proven to be inefficient in transducing myocardium and can cause inflammation at the site of injection. mRNA formed synthetically has been used for gene transfer in vivo but is limited by poor transduction efficiency owing to activation of the innate immune system (via activation of Toll-like receptor 7/8) and rapid degradation by RNase.

In 2005, Kariko and Weissman6, 7 tested several modifications of mRNA and showed that replacement of uridine with pseudouridine rendered the modRNA more stable and less sensitive to RNases and resistant to Toll-like receptors while retaining high translational capacity. These modifications included changing uridine to 5-methyluridine (m5U), 2-thiouridine (s2U), and pseudouridine (Ψ).⁷ The resulting synthetic mRNAs had much higher efficiency of translation and have been used for various applications, including vaccines, induction of pluripotent stem cells, and cancer therapy. Recently, Zangi et al.⁸ showed that delivery of vascular endothelial growth factor (VEGF-A)-encoding modRNA in a myocardial infarction (MI) model induced cardiovascular regeneration, reversed cardiac dysfunction, and improved survival. Additionally, Turnbull et al.9, 10 showed that nanoparticle-formulated modRNA delivery induced highly efficient rapid and short-term expression in rat and pig hearts when introduced by intravascular approaches.

In this issue, Sultana et al.⁴ focused on optimizing delivery in vitro and in vivo by changing two sets of parameters: (1) the composition of the modRNA by changing uridine to modified uridine and (2) the transfection reagent used to carry the modRNA. The authors show that 100% substitution of uridine with 1-methyl-pseudo-uridine (mψU) results in the highest expression in vitro and in vivo when compared to other uridine modifications. They also found that positively charged transfection reagents are superior for in vitro transfection of cardiomyocytes by modRNA. However, in vivo naked DNA in sucrose citrate buffer produced the best transduction profile in the heart when directly injected into the myocardium. Interestingly, expression appeared very quickly (within 5 min) in vivo and lasted for 7–10 days. The fast and reversible expression of modRNA can be an advantage when the delivered genes need to be expressed transiently and with early onset, such as in cases of gene transfer for acute disease. There was no apparent damage to the heart and no inflammatory responses following the intramuscular injection. These results are valuable because they provide a roadmap for investigators in the cardiac gene therapy field on how to use modRNA.

However, unanswered questions remain regarding the use of modRNA in the heart. The authors tested 100% replacement of uridine to 1-methyl-pseudo uridine. Partial replacements or mosaic replacements of the various modifications may also lead to transduction benefits at lower cost. For widespread clinical application to the heart, it will be important to inject the modRNA through coronary vessels or other arteries/veins. The current study was performed in mice, which are not a suitable animal model for intracoronary injection, owing to their small size, and therefore do not inform us about the potential efficiencies of the various vehicles in larger animals and humans. Sucrose-citrate buffer would probably not be useful in the clinic because circulating RNases would degrade the modRNA within a few minutes. The studies by Turnbull et al.¹⁰ revealed that charged nanoparticles used with modRNAs do result in reasonable cardiac transduction; although, as was shown by Sultana et al.,⁴ the use of nanoparticles reduced the translation to 5% of that observed with naked modRNA delivered in sucrose-citrate buffer. modRNAs are currently designed to express transgenes, and it will be important to generate designs able to knock down gene expression. Finally, the use of modRNAs in vivo seems to target a wide variety of cell types, and it will be important to design modRNAs that are cell specific by designing expression cassettes that take advantage of specific expression profiles within a specific cell type or organ. This will create a big advantage over currently used adenoviral vectors because modRNA are a safe, transient, local, non-immunogenic cardiac gene delivery method.