Wednesday, 28 March 2012
No flies were tormented in the making of this blog post
It was memorable not for the result but because it was a monumental failure: the poor flies didn’t take kindly to the burning wax required to fix them to the splint, and the few whose wings had not been singed stubbornly refused to behave, probably in protest. Unsurprisingly, and reassuringly, that practical is no longer run, but I will never forget the experience.
So it came as some surprise to stumble across a new paper in the journal Neuron this month on the response of the fruit fly Drosophila melanogaster to varying sugar solution concentrations, as measured by proboscis extension. And a rather nifty experiment it is too, expanding the framework of my undergraduate practical to a much smaller organism and on a far smaller level. While such a paper might not grab headlines, it struck me as an excellent model from which to introduce some of the techniques that they used. Plus, it just happens to be on the organism I work with daily.
Wondering if proboscis extension is controlled by the same systems that control food intake, Sunanda Marella et al., first tried to identify the neurons responsible for this behaviour. The brain — even the fly brain — is full of thousands of nerves, each with different roles and different properties. Deducing the neurons responsible for a specific behaviour is the first step in not only working out how they work, but also in testing their ability to respond to change (an ability that we call plasticity).
The group employed some of the genetic tools that make Drosophila such a useful organism to work with. They needed to find a way of switching off certain nerves, reasoning that by doing so they would eventually find the neurons that prevent the proboscis extension response to sugar. The tricky bit was being able to switch off specific neurons. For this they used a system called — jargon alert — the Gal4/UAS system. Let’s unpack that jargon!!
Every gene is composed of several parts, each with crucial functions. First, there’s the protein-coding part — to some, this is the gene proper — that makes the protein product of a gene that then goes on to perform the role of the gene in the body. Indeed, if anybody tells you that a gene does something specific, what they really mean is that the gene makes a protein that performs that role; genes, on their own, don’t do very much. Second, and perhaps more importantly, there are parts to a gene either side of the protein-coding region, and sometimes very far away, that influence the expression of the gene. This includes the promoter region, to which a complex cellular machinery binds in preparation for making the gene’s protein product. Promoters are unique to each gene.
In theory, if you put a gene promoter next to the protein-coding region of another gene, you would cause the second gene to be made wherever the first is normally made. You are promoting the second gene in the context of the first. The Gal4/UAS system makes use of this logic, but adds an additional level of control. First taken from yeast, the system has become a staple of fruit fly genetics. The Gal4 protein binds to UAS (which stands for Upstream Activation Sequence), a short stretch of DNA found just before protein-coding gene regions. UAS then acts like a promoter to whatever it is next to.
So, for example, if you fuse the promoter of a completely unknown gene with the protein-coding region of Gal4, and at the same time fuse the UAS sequence with a gene that glows, you will boost the signal of your glowing gene wherever your mystery gene is normally active. You will be able to physically see where the gene is used.
Gal4/UAS provides an additional level of control because the promoter and activated gene are separated. If I wanted to boost the signal of one particular gene but in lots of different contexts, it would simply be a nightmare to build dozens of constructs combining promoters with my gene of interest: much better to build one construct — UAS+gene-of-interest — and cross these flies with any of the hundreds of commercially available Gal4 flies that have been made over decades. These include promoters for unique genes as well as genes common to specific cells or tissues. This makes it really easy to drive whatever you want in whatever tissue you wish.
Apple + Lemon + Lemon = £10 reward
Crank that lever and the symbols will spin, but the order of events stays the same. The parts in bold above can be swapped around, and the output will adjust accordingly. And so it is, sort of, with Gal4/UAS:
Promoter of gene A + Coding region of Gal4 → UAS + Coding region of gene B
You can mix and match promoters and swap what you combine with UAS in order to get the result you want.
In the paper, the scientists use the Gal4 system to drive the expression of Kir2.1. Again, more jargon, but it is enough to explain that this is a component of nerves that keeps them switched on permanently. This has the opposite effect to what you may think: a nerve that is permanently switched on is as useless as one that is switched off. A flat battery, after all, is as useless as a disconnected, fully charged, battery. And so Kir2.1 silences whichever nerves it is expressed in.
This, you can imagine, could completely mess up the development of the fly. Nerve firing is needed at many times, not just feeding, so to cause something this dramatic can have consequences far beyond your desired effect. The team had one further trick up their sleeves, however. There exists a second form of Gal4, called Gal80ts, which is switched off until the temperature exceeds 30°C. This allows the flies to be bred and grown normally, up until the point of experiment when the temperature can be cranked up. Gal80 then switches on, as does Kir2.1. By doing this, they found that one particular promoter altered proboscis behaviour dramatically. The promoter, in its natural context, drives a gene that makes an enzyme responsible for the production of dopamine, the neurotransmitter required for, among other things, pleasure and reward-driven learning.
The paper goes on to test the sensitivity of dopaminergic neurons, showing that overexpressing these nerve cells increases proboscis extension. Furthermore, this effect and these neurons are only sensitive to sugar, not water or other compounds such as those with a bitter taste.
It would be easy to start concluding that flies are similar to humans in this regard, with unique systems to detect specific flavours (our salt, sweet, sour, bitter and umami), but really that is not what this study is about. For me, here, it has provided an excellent example of how we as fruit fly biologists use genetic tricks to further our research — and I could go further into the paper where they use Gal4/UAS and more to pinpoint the precise dopamine neurons responsible. For the scientists involved, it is the beginning of a pursuit of further answers. How, for example, do these nerves operate to compute taste, and how do they communicate with the proboscis? What crosstalk might there be between the areas that encounter taste and these neurons that then influence the movement of the organ? What systems exist for the other tastes? Does dopamine make anything else happen, in readiness for the forthcoming intake of sugar? There are many questions that stem from such a standalone result. As ever, an answer has opened many further questions.
I never realised my undergraduate practical would open up so many avenues. Welcome to the world of a scientist!
Marella, S., Mann, K. & Scott, K. Dopaminergic Modulation of Sucrose Acceptance Behavior in Drosophila. Neuron 73, 941-950 (2012)