Blogging about one's own research always feels good: the amount of your work has accumulated enough to at least provide sufficient material for a story and some figures. It has passed the first hurdle of scientific scrutiny, peer review. On the other hand, now an exciting time begins: what will the colleagues say? Will people find the one major flaw that neither you, your co-authors, the people who proof-read the drafts before submission nor the reviewers caught? Will the results lead to new, exciting collaborations, will it be cited or will it just be met with utter apathy and complete indifference? After all, with about 1.5 million scholarly publications in approx. 24.000 journals, it's not at all unlikely that your paper just simply never is found by anyone who might be interested in it.

All this of course applies to our latest paper as well. However, for those who do read it, I'm sure it will be a fascinating read. It is, without exaggeration, the craziest scientific story I've ever been involved in, an example of serendipity in science if ever there was one. Here's how it all happened. Shortly after we published our mathematical analysis of spontaneous turning behavior in stationary flying Drosophila, Bruno van Swinderen contacted me about a collaboration. I knew Bruno from before (see here and here), but we had never worked together on a project. So I was very excited to hear that he wanted me to test some of his flies in the Drosophila flight simulator, using the exact same techniques we just had published. The initial idea was to use these mathematical analyses to test his mutant flies for any abnormalities in their spontaneous behavior. However, it turned out that the mutant flies that he had sent me, radish, didn't really fly all that well, at least not well enough to generate enough data to run our mathematics on them. Neither did Bruno tell me about his results, we wanted that I should be blind as to the deficits he had found in radish. I knew radish was a memory mutant for olfactory conditioning, but that it could learn visual patterns just fine. Beyond that, nobody knew what other behavioral phenotypes this strain would exhibit.

So I went on to test a whole bunch of other things for which I had the experimental setups readily available. Of all these experiments, I picked the simplest one and sent Bruno the raw data back. I just measured the flies' spontaneous turning attempts without any visual stimulation for as long as I could get the flies to fly continuously, which was six minutes. As a control experiment, the flies' behavior was measured in a flight simulator-like situation, where they could control their flight direction with respect to four visual landmarks (but still tethered, of course). I sent Bruno the data in blind, which means he didn't know which group was the wildtype control and which was the mutant group. He immediately wrote back accurately identifying the mutant group. I had no idea how he could have figured out which group was which so quickly and the experiments were basically concluded, so he started to show me his data.

Bruno does something very few people on this planet are doing: he can put tiny little electrodes in the flies' brains and record their brain waves. Now with this particular mutant strain, he found that they had a peak in the power spectrum of their brain waves at around 1.6 Hz. Stunningly, when he computed the power spectrum of their turning behavior (i.e., my data), he also found a peak at about 1.6 Hz, but only when the flies were flying with the four visual landmarks. The peak was much less pronounced when there were no explicit stimuli in their environment. It was by this peak that he had recognized the mutants so quickly. But what could this peak mean? One thing it could mean is that the mutant flies become fidgety, if there's something in the environment they need to pay attention to. About one and a half times per second, the flies are initiating some turning maneuver, which can be seen as a peak in the power spectrum. In other words, the flies are hyperactive or fidgety, in this very well-defined, oscillatory kind of way. He then went on to tell me that they also were more easily distractable than the wildtype controls, both in behavior and by inferring from their brain waves when presenting them with various competing visual stimuli.

I thought this was really quite amazing. Flies which are known for their memory loss are both hyperactive and have an attention deficit. I immediately thought of people with attention deficit / hyperactivity disorder (ADHD). They also have learning problems. As a joke, I suggested we put the flies on Ritalin, the drug used to treat patients with ADHD. Bruno replied that these data were actually intended for a different publication, but they indeed would fit very well with this one, too. I was flabbergasted! He had already done the experiments with methylphenidate (Ritalin is just the trade name). To my utter astonishment, the flies on methylphenidate performed like their wildtype counterparts in almost all of the tests we subjected them to. This is even more amazing when you consider that the mutated gene, radish, is required during brain development (late during pupation) and not during the behavioral test. In other words, methylphenidate rescues a deficit in adulthood that even a healthy copy of the originally mutated gene cannot rescue any more.

I find it absolutely crazy to find a fly model for a human psychiatric disorder. On top of that, Ritalin, the drug used to treat ADHD in humans actually also successfully treated the flies. It's even more crazy to find all that by accident, without even looking for it! All we wanted was to study some interesting fly mutants to learn more about some basic brain function. How can it be possible to find something like this just by serendipity? My favorite hypothesis is that there are some fundamental principles about how all brains work and we have stumbled across one of them. We still don't know what it is or how it works, only that it has to do with how brain allocate attention to different processing streams. It is tempting to speculate that this process has to do with switching of activity between separate networks, but there currently is no data to tell either way.

I do have plenty of other interesting results from this mutant, both in flight and in walking. However, I cannot make much sense of them, yet, so a lot of further research is required before part two of this story can be presented.

Of course, as usual, I have a copy of the paper, together with all the supplementary material, on the download page.

---

van Swinderen, B., & Brembs, B. (2010). Attention-Like Deficit and Hyperactivity in a Drosophila Memory Mutant Journal of Neuroscience, 30 (3), 1003-1014 DOI: 10.1523/JNEUROSCI.4516-09.2010