Killing flies with bacteria and other things I couldn’t possibly explain to my grandfather
written by Dr. Justin Nodwell
image by Andrew Zhai
My grade 11 math teacher, Mr. Dimass, was a fan of short, elegant proofs. Conversely, he had an aversion for anything that was unnecessarily complicated. To indicate such excesses, he used to say, “it’s like you’re putting a mosquito in a hydraulic press.” That vivid image really stuck with me, though not exclusively in the way he intended. The thing is, my first thought at the time and on many occasions since was, “a mosquito in a hydraulic press? Let’s do it!” The point being that sometimes the most interesting thing involves taking something you do all the time and applying it to something that seems counterintuitive or even nonsensical.
Which brings me to the most recent paper from the Nodwell lab: “Chemical entrapment and killing of insects by bacteria” – a real masterpiece by a recent PhD student, Dr. Louis Ho. We started looking for antibiotics many years ago – mostly our search centres around the ‘specialized metabolites’ of the Streptomyces bacteria. As often happens, we developed a sort of stock approach that we’ve applied in varying ways many times. It goes like this: apply a perturbation to a set of potential antibiotic producing environmental microbes > make small molecule extracts from them > test the extracts for antibiotic activity against a pathogen > purify active compounds and solve their structure > figure out how they work. Our usual target is a Gram positive pathogen like Staphylococcus aureus (e.g. Daniel-Ivad et al., 2017) though sometimes we go a little wild and screen against a Gram negative like Burkholdaria cenocepacia (McKenzie et al., 2010) or even a unicellular eukaryote like Saccharomyces cereviciae (Pimentel-Elardo et al., 2015). It’s easy to explain why we do this work: there is a pressing need for new antibiotics and, by-and-large, the pharmaceutical companies are doing nothing. Against this successful backdrop, Louis’ paper is a decidedly odd fit.
In all honestly, I think it was Mr. Dimass’ voice, lurking in my head since 1979, that drove me to suggest to Louis that he apply our stock approach to fruit fly embryos rather than bacterial pathogens. It was the sort of speculative idea that faculty sometimes give to new trainees. It gives the student a chance to learn a few things about instrumentation and techniques and maybe it will lead somewhere. But if not then no harm done – the student has still learned something new and they can apply it to another problem down the line. It’s worth adding that trial balloons like this, while very interesting, don’t easily fit the mould of most of the Federal granting agencies. Every now and then however, these ideas land in the right place. And so, Louis being Louis, he went for it.
The result included three major surprises. The first was that at least 10% of streptomycetes make small molecules that kill insect larvae dead. Sometimes the effect is instantaneous as in the case of avermectin, other times it’s a slow, wasting process as we observed with cosmomycin. The second surprise was that the spores of the producing organisms could deliver these toxic molecules, with devastating effect. This flies in the face of the prevailing assumption that, aside from the occasional potato pathogen, the streptomycetes are benign, filamentous place holders in the dirt. This went on to become figure 2a in Louis’s paper and it is among the most exciting discoveries in the 22 year history of my lab. Finally, there was the discovery that many producing bacteria also make volatile compounds that can attract unsuspecting adult flies to this chemical peril. Suddenly it started to seem as though this might actually be a widespread factor in terrestrial ecosystems. The streptomycetes are ubiquitous on this planet…maybe this really matters?
There were many interesting digressions along the way. At one point we had a collaboration with an actual mosquito lab (Mr. Dimass….still thinking of you sir) as we thought we might have discovered the next great bug repellant (no). The price of admission for Louis in that endeavour was sticking his naked and oh-so vulnerable-looking forearm into a giant cage full of swirling, hungry mosquitos (there are cell phone videos). We also learned a great deal about things like dissecting out insect digestive tracts and assessing cell death at locations deep within them (Figure 3). Many thanks to a stellar collaboration with the Hurd, Smibert and Boullienne labs for their vital contributions to this work.
What to make of this digression? As scientists in this most wretched of centuries, we are forever trying to convince our politicians and seemingly disinterested public that what we do will help them at some point in the not-too-distant future. That we will cure diseases and solve economic problems. And of course, often we do and that really matters – this has never been more clear than at this particular time. But none of us got to be scientists because of that. We all got into this work because we are curious about nature. We look at things that, to most people, appear simple and solved and yet, all we see are questions. To put it another way, we have a perverse desire to put a mosquito in a hydraulic press and see what happens.
We should give in to that impulse more often.