It’s easy to forget that there are other viral diseases loose in the world, but I’m glad to highlight some real progress against one that’s been trouble for a long time: dengue. This one’s notorious because there are four rather distinct serotypes out there, which has made vaccine development rather difficult. Indeed, a dengue vaccine led to one of the best-known examples of antibody-dependent enhancement (ADE), when vaccinated children who had never been previously exposed to the disease were found to then be at higher risk for severe infection. The vaccine was less efficacious against one of the four serotypes, and it turned out that in the dengue-naive patients the antibody response from the vaccine actually led to enhanced disease if they were later infected by that one. This is of course exactly what you don’t want, of course – and now’s the time to mention (yet again) that we are seeing absolutely no sign of this occurring with the current coronavirus vaccines. If you’d like to argue otherwise, I will refer you to this post, because the comment restrictions outlined it in still apply. Go do that somewhere else.
But back to dengue. The same four-serotype situation doesn’t help much with antiviral drug discovery, either. There are no approved drugs for dengue infection. The existing antivirals have of course all been tried, and efforts to make new dengue-specific ones haven’t worked out well, either (I did a little of that myself at one point). But now there’s an interesting new paper that details a surprisingly potent small molecule that works well across all known variants of the dengue virus, and does so via a mechanism that until now hasn’t been described.
The authors had identified some interesting leads in their previous dengue screens, and this new compound (JNJ-A07) emerged from a chemistry follow-up library around one of those structures. In cell assays, it was active at nanomolar levels and below against 21 different isolates of the virus, basically covering everything that is seen across the four serotypes. At the same time, though, it really wasn’t active at all against a whole panel of other viruses, showing that the mechanism, whatever that might be, was apparently specific to dengue. The team went about identifying that target in the classic way: by exposing the virus to sub-optimal levels of the compound for an extended period and trying to generate resistance. You then sequence the virus and find out what proteins had to mutate in order to escape the drug’s activity, because those have a very high chance of leading you to the target.
In this case, generating resistant virus was not easy (a point in the compound’s favor, of course). And in fact, the resistant strains showed greatly impaired ability to replicate in cells, which is all the better (and which also explains the difficulty in generating such variants to begin with). Sequencing showed a number of mutations in the viral NS4B protein, and the compound was found to block the interaction between that protein and the viral NS3 protein. That step had already been shown to be crucial for viral replication, but no one had ever seen a small molecule that shuts it down. The exact role of the NS4B/NS3 pairing is still being worked out – neither protein shows enzymatic activity per se, but they seem to be important for the activity of the replication enzymes that unwind and copy the virus’s RNA.
What’s more, the compound has shown (so far) perfectly reasonable safety and pharmacokinetics in rodent models, and has very strong antiviral effects in them (sending viral loads in many cases down below the limits of detection). It works prophylactically if given before exposure to the virus, and therapeutically if given even up to several days after the start of the infection. This compound (and this whole class of compounds for that matter) clearly show great promise as dengue therapeutics, and should be definitely be moved towards human trials.
What will it take to do that, and what could go wrong? There will need to be more extended tox testing (in at least two animal species), and more work on the pharmacokinetics, to make sure that there will be high enough blood levels in human subjects. The compound will need to be profiled across other known drug targets, just to make sure that there’s not something obvious that could cause trouble. There are no obvious warning signs in its chemical structure, so that’s good. Chemically, the scale-up and process people will have to start taking a look at the synthesis to make sure that there’s a solid route to the compound on large scale, and they’ll have to start producing it under the regulatory framework needed for human trials. Chemists in the crowd will have already noticed that it’s a single enantiomer, so that will be part of the synthesis work as well.
Clinically, infectious disease trials are relatively straightforward: you go to the areas where the disease is endemic and run prophylactic and treatment trials in various populations (different ages, risk levels, pre-existing conditions, etc.). Even with the low propensity for resistance, it’ll also be interesting to see if this compound can be combined with any other possible dengue therapy to hammer the virus down even harder. I hope that this compound or one of its successors progress that far and prove worthy out in the real world. Dengue is a scourge and it’s getting worse every year (the warming climate isn’t helping, either). I’ll be cheering this one on.
