A Love Letter to the Fruit Fly Brain

Finally, the holiday break has started! For me personally this means I will put on those kilos I lost during my climbing training thanks to intense gustatory stimulation, which goes with the appearance (and consumption) of amazing holiday food. Just the other day, as a bit of a prequel to the Christmas Day lunch, my immensely satisfied tummy and I had some roasted haloumi on fennel, topped with pistachios and pomegranate seeds with a honey – balsamico dressing… Sorry, I should stop drooling and get back to what I wanted to write about today: Behavior and neurobiology! No, the Pavlovian reflex is not going to concern us today, despite my talk of food and drooling. I am going to focus on one of my favorite topics to talk about when I´ve had a wee bit too much Christmas punch. To be honest, it´s my favorite topic no matter when you have the misguided courage to ask me what it is I do for a living, regardless of how much punch was involved. I am speaking of my obsession with the fruit fly brain.

First of all: those little buggers are a helluva lot smarter than you might think. Definitely smarter than what I accredited them to be the first time I heard about learning tests for fruit flies. Nonetheless the red-eyed nuisances, who seem to enjoy the close proximity to my fruit bowl, were also most certainly NOT at the top of my list of animals I would love to work with in the lab one day. But I really wanted to mess with neurons while simultaneously studying the effect of said messing around on an animal’s behavior. I soon realized: you just can´t do that with many model organisms these days.

To start with, ethic committees do not always agree with your grand scheme of fiddling around with a human’s brain. Cutting out entire parts of it is generally frowned upon. In the instances that it isn’t (poor mice), there is one major problem. A manual knockout (i.e. sticking a needle into a brain in order to physically destroy a single neuron) is very hard to achieve because of the neuron size. I can´t even pull a thread through the eye of a needle, much less stick said needle into something smaller than aforementioned thread without damaging the surrounding area. So there’s missing the single neuron target as one problem in large brains. A completely different matter is the fact that in general vertebrate neurons form clusters in birds and sheets in mammals. In neuroanatomy these clusters are referred to as a “nuclei”, whereas the sheets are termed “lamina”. When you’re trying to destroy an entire nucleus/lamina, in which all neurons fulfill approximately a single function, it’s easy to see that it’s very likely a few functional survivors will be left behind. It´s like trying to carve Michelangelo’s “David” from a block of marble using only a chainsaw. As disastrous as the outcome would be, the surviving neurons could potentially, however unlikely, still be able to live up to their role in the brain. Leftovers are messy and if there´s one thing science sadly deals with a lot but nevertheless hates passionately, it´s messy experimental set-ups.

In a fruit fly, you don´t have to worry about damaging the surrounding tissue. Now, this may seem counterintuitive. Their brains are even smaller, which should make it even harder to poke out a single neuron. But what makes the fruit fly so superior in comparison with other brain model organisms is the Split-Gal4 system, which has also been dubbed “The Drosophila scientist’s Swiss army knife”. With this technique, it is possible to delete single neurons, based on the Gal4/UAS system. I´ll explain Gal4/UAS, the principle behind the Swiss Army knife first:

Gal4 is a yeast protein and left to its own devices, it doesn´t do much to the cells in which it is produced if its preferred binding partner UAS is missing. UAS (upstream activator sequence) is an activator – hence the name – of whatever gene it has been placed in front of (“upstream”). The genes can encode things like GFP (green fluorescent protein), which can make the cell glow in a lime-green color, or halorhodopsin, which shuts off neurons when a light is turned on. Put together, we have a cell which produces Gal4, which binds to a, let´s say, GFP – linked UAS in the same cell. If the Gal4/UAS system was successful, we see the distinctive lime-green glow of GFP.

There are problems with this technique though. It´s very hard to guarantee for the Gal4/UAS system to work in only one type of cell. In our case, this would be one type of neuron. This is when the Split-Gal4 system comes into play. For this method, the Gal4 gene has been split into two parts. The female fruit fly will carry one half and her male mate the other half. The buzzing offspring then inherit the two genetic puzzle pieces for Gal4 and are able to produce the complete protein. How does this system guarantee higher precision than the simple version? Say Gal4/UAS is expressed in cell type (1) and (3) in the female, and in cell type (2) and (3) in the male. The offspring will exclusively express Gal4/UAS in cell type (3), since that is where the production of our proteins overlaps. We can switch off single neurons. I´m lacking the words to say how much I adore this method for its simple elegance. Brilliance at its finest!

To bring this blog post to a close, another advantage of the fruit fly is, that it has only approx. 300.000 neurons. In comparison, a cat has approx. 760.000.000 neurons. Forget about deleting every single neuron in a cat. No one would ever pay a PhD student army large enough to win that war, nonsensical as they all are. But for a fruit fly, it´s actually in the range of possibilities to knockout each and every one of them and then study how this changes the animal’s behavior.

This is what I did, using reward and punishment learning in fruit flies. How does one exactly punish/reward miniscule creatures in order to make them learn? I´ll tell you some other time, but let´s just say: During the Christmas season carols could be seen as stimulus onset, and the Christmas dinner as reward. And though they make my skin crawl (especially in techno versions), the sound of carols and the established memory link to food is a surefire way to make me look forward to another round of Christmas.

 

Reference

I´m sorry, most of this is information I picked up over the years. I don´t keep track of references in my brain, I´m too busy storing content. But if I had to mention references, it would be these:

GAL 4 System in Drosophila : A Fly Geneticist’ s Swiss Army Knife, Duffy (2002)

https://www.janelia.org/lab/rubin-lab/our-research/gal4-driver-lines/split-gal4-lines