Therapeutic Delivery of mRNA: The Medium Is the Message

Messenger RNA (mRNA) has several advantages over DNA for gene transfer and expression, including the lack of any requirement for nuclear localization or transcription and the nearly negligible possibility of genomic integration of the delivered sequence. However, the labile nature of mRNA and its capacity to elicit innate immune responses are important limitations to its in vivo application. Now, an article by Kormann et al.¹ published in Nature Biotechnology addresses these drawbacks and provides new hope for the potential use of mRNA as a source of therapeutic gene product in vivo.

The advent of RNA interference,² along with that of antisense and catalytic RNA, has provided molecular tools for inhibition of gene expression in the treatment of dominant hereditary diseases, cancer, and infectious diseases, whereas mRNA has found utility in the form of genetic vaccines, through both in vivo administration³ and loading of dendritic cells for the generation of cellular cancer vaccines.⁴ However, as noted above, the use of mRNA also brings the challenge of eliciting innate immune responses that can compromise the effectiveness of nucleic acid delivery. Double- and single-stranded RNAs interact with certain pattern-recognition receptors (PRRs), including Toll-like receptors (TLRs), that detect pathogen-associated molecular patterns as a first-line defense against microbial invasion.^5,6 Subsequent downstream signaling through the MyD88 and TRIF pathways⁵ activates cellular and inflammatory reactions that are an important part of the antiviral response.

Endogenous RNA molecules are distinguished from those of invading microbes by a preponderance of nucleotide modifications that affect PRR engagement.⁷ The central development tested by Kormann et al. is the use of mRNA that reflects these modifications by incorporating nucleotide analogs (2-thiouridine and 5-methylcytidine, in particular) during in vitro transcription. Incorporation of such nucleotide analogs was previously demonstrated to prevent signaling through TLR pathways and block cytokine expression in dendritic cells exposed to mRNA.⁷ The authors show that incorporating both pyrimidine analogs inhibited engagement with TLRs 3, 7, and 8 and the retinoid-inducible gene-1 product (another PRR) in pull-down experiments, and also abrogated cytokine expression (interferons IFN-g and IFN-a, and interleukin-12 (IL-12)) in transfected peripheral blood mononuclear cells. Innate immune reaction to the transfected RNA was thus substantially inhibited.

This transfected RNA had increased stability in vitro and in vivo, such that intramuscular injection of message encoding murine erythropoietin (Epo) raised hormone levels sufficiently to elevate the hematocrit at 2 and 4 weeks after infusion. Similarly, the authors report substantially increased levels of luciferase reporter activity in the lungs of mice after high-pressure intratracheal spraying of the cognate mRNA. They further applied this approach to mRNA therapy in a mouse model of surfactant protein B (SP-B) deficiency. Complete deficiency of SP-B is lethal in neonatal mice, so animals engineered for inducible expression of SP-B were used in these experiments. After withdrawal of SP-B induction, control SP-B^−/− mice that received luciferase-encoding mRNA died within 5 days. However, two doses of modified mRNA encoding SP-B extended the survival of SP-B^−/− mice up to 10 days, and continuous (twice-weekly) treatment resulted in nearly 90% survival in the 1-month study, with essentially normalized lung histology and lung function. The authors emphasize that modification of mRNA substantially reduced the amount of cytokines (IFN-g and IL-12) measured in bronchial lavage fluids soon (8 hours) after administration.

The development of mRNA as a therapeutic faces the same challenge as any nucleic acid: delivery, delivery, delivery. Extension of the approach described by Kormann et al.¹ to other targets will probably require some degree of compaction and complex formation. On the other hand, the effectiveness of intravascular delivery under pressure (“hydrodynamic” delivery) of uncomplexed DNA might extend to modified mRNA delivery into liver,^8,9 muscle,¹⁰ and other organs. One limitation of modified mRNA delivery compared with integrative gene transfer¹¹ is the transient duration of expression, as exemplified in the Epo studies that Kormann et al. report. Moreover, erythropoiesis is exquisitely sensitive to even slight increases of Epo in the circulation. For many other therapeutic gene products, such as clotting factors, enzymes, and even cytokines, a much greater amount of gene product will be necessary. It is therefore likely that substantial improvements will be required in the efficiency of mRNA delivery and translation into protein product to reach a level that is of more general therapeutic utility.

In these days of heightened safety concerns, it seems as though every advance in gene transfer and expression needs to somehow provide a solution to the problem of leukemogenesis.¹² Therefore, one of the incentives for testing the in vivo effectiveness of modified RNA was to provide an alternative to the potential risk of insertional mutagenesis associated with integrative DNA gene transfer and expression. However, adverse events associated with integrative gene transfer have thus far been limited to circumstances in which continued expression of the gene product is required in cellular progeny after extensive proliferation and differentiation.¹³ Modified mRNA is unlikely to be maintained at a level sufficient for corrective expression of gene product in these circumstances, so alternative means of supporting maintained expression (i.e., corrective gene integration or chromosomal modification) will be required in these cases.

At present, applications of this technology will include those in which a short or intermittent burst of gene product is anticipated to have a beneficial effect. This extends beyond protein replacement—for example, in the use of modified mRNA for reprogramming in stem cell generation and differentiation.¹⁴ Modified mRNA may also be used to express a recombinase that mediates genome modification^15,16 or nucleases for site-specific chromosomal modification,¹⁷ because a short duration of expression may be sufficient while avoiding unwanted gene integration and long-term expression. The results reported by Kormann et al. thus provide new inspiration for the therapeutic testing of mRNA.