I’ve been meaning to write about this paper from the RIder group at MIT’s Lincoln Labs, which shows some very interesting approaches to killing off a wide variety of viruses. They’ve dubbed these new agents DRACOs, for Double-stranded RNA Activated Caspase Oligimerizers, which is certainly one of those acronyms with a lot packed into it.
So now to unpacking it. The first key point is the double-stranded RNA (dsRNA) part. For a long time, that was thought to be a form that isn’t wasn’t found in human cells (as opposed to single-stranded stuff). We now know that short dsRNAs (up to twenty-odd base pairs) are part of human biology, but viruses produce much longer strands of it during their replication process – or, more accurately, they hijack human cellular machinery to produce it. (Viruses, as a rule, don’t do anything for themselves that they don’t absolutely have to).
Naturally enough, cells have evolved ways to recognized long dsRNAs as a sign of infection – there’s a whole list of proteins that recognize these things and bind to them. Some of them inhibit its downstream processing directly, by just hanging on and gumming up the works, while others set off responses further downstream. One of those is apoptosis, programmed cell death, a brutal but effective fall-on-your-sword pathway that gets initiated by all sorts of unfixable cellular problems. (When a cell’s internal controls give a “Fatal Error” message, it’s taken literally). And naturally enough, viruses have evolved ways to try to evade these defenses, both by targeting the dsRNA detection proteins and by inhibition of apoptosis pathways. (As a side note, it’s always been interesting to untangle these counter-counter-countermeasure situations whenever a new cellular pathway relating to infection is worked out. You find, invariably, that hundreds of millions of years of evolutionary pressure have built up crazily elaborate frameworks around all of them).
This approach tries to speed up the dsRNA-means-apoptosis connection. A DRACO turns out to be a good-sized protein with two functions: one end recognizes and binds to dsRNA, and the second contains a signal to induce apotosis. If multiple copies of the DRACO protein stick to the same viral dsRNA strand, that should be enough to initiate cell death and interrupt the viral replication process. The team tried out a whole range of possibilities for both those functional domains, with the best (so far) using either Protein Kinase R (PKR) or RNAaseL domains to recognize viral RNA and an Apaf caspase recruitment domain for apoptosis signaling. Another key modification was the addition of a PTD (protein transduction domain) tag, which allows large proteins like these entry into cells through active transport. (Cells only take in whole proteins through gatekeeping transport mechanisms; otherwise they just sort of bounce off – this effect was confirmed with DRACOs that lacked the PTD tags).
So, basically, this is the sort of protein that you might expect evolution to stumble onto eventually, but now the connecting line has been drawn by hand instead. It’s worth noting at this point, though, that this general idea has occurred to others before: here’s a paper from Boston University trying the same sort of strategy. That one was published online in 2009, but didn’t make it to print until May of this year, which makes you wonder if that’s a typical delay for that journal (FASEB J.) or not. It’s also worth noting that, for whatever reason, this new MIT paper does not cite the one from BU.
How did they work? Pretty well. The PTD tags did what they were supposed to, taking the proteins into cells rapidly. Once inside, the DRACOs themselves hung around for several days before being degraded, which is another big hurdle. And they did indeed protect against infection by an impressively wide range of viruses in cell culture: rhinovirus, encephalomyelitis, adenoviruses, arenaviruses, bunyaviruses, flaviviruses, reovirus, and flu. A lot of nasty pathogens fall into those bins.
But that’s in cell culture, which is a long way from a living organism. To their credit, the team went on to try out their idea in live mice, and they show some encouraging results. Administering their best DRACO candidates to mice and then exposing them to influenza virus took the survival rate (at ten days) from under 10% in the control groups up to 60-70% survival for the PKR version and up to 100% survival for the RNAaseL version.
It’s an impressive graph, but there are some things to note about it. For one, the DRACO proteins were administered by injection – these are probably never going to be feasible as oral medications, since they’re large proteins which will just get digested. But again to their credit, the MIT group also tested dosing via intranasal injection (yep, squirting the protein solution up the noses of mice, truly the glamorous end of science). That also showed a strong protective effect after influenza virus exposure, which is a good sign.
Now comes the next concern. You might have already wondered about my mention of the injection route, since we already give millions of people a year injections to combat viral infection: flu shots. Those, though, are vaccines meant to last the whole season (and beyond). DRACO proteins get cleared out in mice on a time scale of days; they wouldn’t be expected to have any long-range immune effects. (Of course, their broad antiviral effects, versus the sometimes way-too-specific nature of a vaccine, is a strong point in their favor). But this brings up another issue that’s going to have to be addressed: when you look at the graphs of the mice experiments, you note that the DRACOs were given either on Day 0 or Day -1 compared to the exposure to virus.
That’s actually a big deal in this field. The problem with antiviral therapies has always been that you don’t usually know that you’ve been infected until, well, after you’ve been infected. Sometimes that lag time is rather long, and it’s always long enough for the virus to get a good running start. Symptoms, after all, don’t occur until things are well under way. In the real world, the two opportunities for antiviral therapies are (1) something that you can take long before you’re even exposed, and that lasts for a long time (like a vaccine) or (2) something that you can take after you’ve already realized that you’re sick (like an antiviral drug). So far, the DRACO proteins fall in between these two, and the next challenge for these agents is to see if they can stretch into one or the other. The authors, no fools, realize this:
Based on these encouraging initial animal trials, future work should be done to test and optimize antiviral efficacy, pharmacokinetics, and absence of toxicity in vitro and in vivo. Future experiments can further characterize and optimize dsRNA binding, apoptosis induction, cellular transduction, and other DRACO properties. More extensive trials are also needed to determine how long after infection DRACOs can be used successfully, or if DRACOs are useful against chronic viral infections without producing unacceptable levels of cell death in vivo.
It’s going to be very interesting to see how this field develops. It’s a promising start, for sure, but there are still a lot of ways for things not to work out. Just getting this far along in the “promising start” phase is a real accomplishment, though, and more than many people have ever been able to manage.