Summary

Lipid nanoparticles, encapsulating mRNA offer a flexible platform for enabling a variety of advanced therapeutic approaches such as mRNA vaccines, gene editing, and protein replacement. These approaches are transforming how diseases like cancer, Alzheimer’s, and even the flu will be treated in the near future. Naturally, LNPs deliver mRNA to the liver when administered intravenously. While this has been demonstrated as efficacious in a wide range of disease models, it may still limit the vast potential of mRNA therapies. For instance, cancer treatments would be more efficacious and better tolerated if mRNA or the encoded protein were limited to acting in malignant cells. Particularly, in liver cancer where the affected tissue contains a mix of healthy and malignant cells.

Several strategies have been adopted to increase specificity of delivery. These include strategies to deliver the LNP to target sites by engineering their properties, for instance by conjugating targeting molecules to the surface of LNPs (e.g. Kedmi et al. 2018). Moderna Therapeutics, who have several mRNA-LNP therapeutics in clinical trials, have taken a different approach. Their strategy, published in the October 2018 issue of Nucleic Acid Therapeutics, takes advantage of natural gene regulation pathways to dictate in which cells the therapeutic mRNA is expressed. Their approach leverages differences in expression levels of naturally occurring small RNAs called microRNAs that bind to recognition sites on mRNA to mark it for degradation. Recognition sites are naturally found on the 3’ untranslated region (3’-UTR) of mRNAs. This paper tests whether the differential expression of microRNAs between cancer cells and healthy cells can be used to control which cells express the exogenous gene. Importantly, they also test whether this mechanism would work with modified mRNAs. Modifications such as substituting uridines are necessary to make mRNAs more stable and hence practical to work with as an active therapeutic.

To test this approach, they analyzed the microRNA expression profiles between healthy hepatocytes and hepatocellular carcinoma cell and found miR122 is expressed in healthy liver but not in hepatocellular carcinoma. They encoded an apoptotic protein in an mRNA substituting uridine with pseudouridine for stability, and bearing up to 3 binding sites for microRNAs in the 3’-UTR. They encapsulated this mRNA into an LNP containing the ionizable cationic lipid DLin-MC3-DMA (a.k.a. MC3 for short) using NanoAssemblr technology. They tested the formulation in mice bearing a subcutaneous Hep3b xenograft via intratumoral injection. They also had a variety of controls including empty LNP, LNPs containing the therapeutic mRNA without miRNA binding sites, and untreated controls. To determine activity, they compared histology between the tumour tissue and healthy liver.

The mRNA without miR122 binding sites showed necrosis in both the tumour tissue and the liver, indicating that the LNPs made it to the liver and both tissues had expressed the apoptotic protein. On the other hand, mRNAs that had either 1 or 3 miR122 binding sites showed noticeably greater necrosis in the tumour tissue than the liver. They also concluded that having 3 miR122 sites was more effective than 1 in terms of selectively targeting the tumour. These results indicate that the miR122 binding sites are effective in preventing the expression of the mRNA in healthy cells. This proof of principle for using microRNA expression differences to target cancer cells with therapeutic mRNA offers an innovative way to target specific cells. Importantly, the approach is effective even when using mRNA bearing stabilizing modifications such as pseudouridine. This proof of concept is particularly impactful because our knowledge of microRNA expression profiles for various cells continues to grow, and new miRNA targets continue to be discovered, which will lead to better targeting in the near future.

Abstract

The advent of therapeutic mRNAs significantly increases the possibilities of protein-based biologics beyond those that can be synthesized by recombinant technologies (eg, monoclonal antibodies, extracellular enzymes, and cytokines). In addition to their application in the areas of vaccine development, immune-oncology, and protein replacement therapies, one exciting possibility is to use therapeutic mRNAs to program undesired, diseased cells to synthesize a toxic intracellular protein, causing cells to self-destruct. For this approach to work, however, methods are needed to limit toxic protein expression to the intended cell type. Here, we show that inclusion of microRNA target sites in therapeutic mRNAs encoding apoptotic proteins, Caspase or PUMA, can prevent their expression in healthy hepatocytes while triggering apoptosis in hepatocellular carcinoma cells.