With the broad use of the mRNA coronavirus vaccines, it might be tempting to think of the mRNA therapeutic delivery problem as having been solved. But there’s a lot more to it. Keep in mind (for one thing) that a vaccine dose is very low because of the natural amplification that the immune response provides. For therapeutic use, you’re going to have to bring in more mRNA and do it over and over. Another big point, which has always been a big point in therapeutic oligonucleotide ideas, is tissue selectivity.
Injecting a small amount of mRNA into the deltoid muscle works fine if you just want it to mostly drain into the lymphatic system, and if you don’t care much about where the rest of it goes. But it mostly goes into the liver, just like most things mostly go into the liver. If you inject directly into the bloodstream instead, your dose is pretty much all going to wind up there, which is why the people working on such therapies are mostly concentrating on liver diseases. They’re making lemonade out of the pharmacokinetic lemons that we have to work with.
Now, it’s not that tissue selectivity is impossible – there are surely ways to achieve it, but we just don’t know enough (yet) about what those might be. This new paper is a good illustration of that. The authors (a large team from Penn) are studying a different RNA formulation than the one used in the current vaccines. Instead of four lipid components, they have just one more complicated molecule involved, an “ionizable amphiphilic Janus dendrimer” (IAJD). A reasonable response that that statement is “What the heck are those?”, and that is one of those pictures worth a thousand words, even if you’re a chemist. And unfortunately you’ll have to be a bit of a chemist for the picture to completely make sense.
OK, there they are, and the color-coding helps you to see the layout of these beasts. These sorts of structures have been investigated in the past, and this groups has been working on them for some time as potential RNA delivery vehicles. Down at the bottom is an aryl group with two different long-chain alkyl substituents coming off those phenol oxygens. And up at the top we have a piperazine, the basic amine part of things standing off from the aryl part a bit, with various substituents on its far nitrogen. In the lower right-hand corner of the figure, you can see another class where that whole piperazine side chain part has been replaced by another multisubstituted aryl – and in those, the basic amines are dangling off the end of even longer sides chains from that second aryl group. Even a brief reflection on these structures will tell you that you can make variations on things like this until the flippin’ cows come home, just hundreds and thousands (and hundreds of thousands) of different combinations of ever-branching mixtures of polar basic amines and big greasy side chains of various lengths and complexities. The question is, do you need to do that, and do you want to?
The answer in this paper is that you very well may. The authors use all sorts of these structures (one at a time) to deliver a luciferase mRNA to light up the destination tissues in mice, with special attention to the lungs, liver, spleen, and lymphatic system. Without going into the details too much, it turns out that what appear to be rather minor changes to the IAJDs make big differences in how much mRNA gets delivered and expressed, and in which tissues. For example, going from a symmetrical hydrophobic species to a given nonsymmetrical one could increase the activity by up to 90-fold (!) Adding those polyethylene-glycol-like spacers tended to send more of the activity to the lung tissue, for reasons yet unknown. Of the eleven IAJD vectors with the highest activity, two show up best in lung, three in the liver, and six in the spleen and lymph, but it’s interesting to note that none of the nonsymmetrical ones really went for the liver (all the winners there were symmetric). Meanwhile, there are other IAJD structures that are notably worse than their close analogs, again for reasons that are not yet understood.
The authors say that “Most probably, this report provides also the largest number of synthetic vectors from the literature producing specific delivery to such a diversity of organs“, and I’m not going to dispute that! The paper announced an intention to survey this landscape as widely and intensively as possible to try to figure out what’s going on and to try to generate some rules, and I definitely wish them luck in that large task. What I think this illustrates in general is that there are a lot of potential formulations that have not yet been tried and that these can have completely unexpected effects (good and bad) on mRNA delivery in vivo. It is indeed a wide open field, a huge scientific space stretching off into the distance in every direction. Every step could put you on top of a mound of solid gold or send you down a trap door, and until we get some clues (which could take a while) there’s only one way to find out: go run the experiments.
