How flies taste salt

Salt is one of those things that take too little and you’re dead, but too much will kill you. So animals have evolved very complex mechanisms to taste salt and control how much they eat, and that includes fruit flies. And what Mike Gordon has found is that flies have multiple salt-detecting nerve cells; some of these respond chiefly to low concentrations of salt in the diet and seem to encourage intake. The others respond to high concentrations and normally deter a fly from eating a foodstuff that might contain a toxic salt burden. But, as Mike explains to Chris Smith, the intriguing thing is that, if the animal is deficient in salt, it can temporarily override and ignore the aversive “toxic” signal…

Mike - Well we really set out to try to understand how the taste system of the fly interprets salt in their environment, because salt is a particularly interesting taste modality because it's attractive at low concentrations but it's actually aversive at high concentrations. And so, unlike most tastes, which elicit either attraction or avoidance, salt actually does both. And so we figured that the way that the taste system must interpret salt had to be more complicated than other tastes.

Chris - Is it complicated like that because you need a bit of salt, but too much is a bad thing?

Mike - Right. Exactly. There were models in the taste field that suggested that there are essentially two different responses that get layered on top of each other. So there's an attractive response that kicks in at low concentrations and then, as you get to higher concentrations, you'll get an aversive response that sort of overrides that. But what we found is that it's actually even more complicated than that, and there's actually two populations of cells that are attractive, and respond at low concentrations, and two distinct populations of cells that are aversive and respond at high concentrations.

Chris - And this sensation or this detection this is happening in the equivalent of a taste bud?

Mike - So flies are a little bit different than us; their taste system actually uses neurons to detect tastes. And so these neurons, that express the receptors that actually bind to the taste ligands, are in various body parts of the fly. They're present on the mouth of the fly. They're actually also present on the legs of the fly, which makes some sense because flies spend a fair amount of their time actually walking on their food! And so it's a good way of kind of deciding what the food tastes like before you actually start eating it!

Chris - The mind boggles of what would happen if we could do that! But how did you actually do the experiments, because if you want to unpick what's going on how did you actually put a label on what the different nerve cells are that normally detect things, and how they are individually then responding to salt, or not?

Mike - So what we started with was trying to make a fairly comprehensive map of the taste system using different markers. These are just genetic markers based on the expression of specific receptors. So, for example, a receptor that responds to sweet compounds, if we use the gene that encodes that receptor then we can label the neurons that respond to sweet. And so we did this, and compared them all against each other to try to make a map where we had different, discrete populations that we could then look at the activity out of.

Chris - I get it. So you've got a molecular label if you like - a chemical tag - that says "I am this sort of nerve cell with this sort of receptor on me," and you can then test how that particular class of receptor responds to salt and at what concentration?

Mike - Exactly. And so to do that what we actually do is to express another label in those particular neurons. And this particular fluorescent protein will actually become brighter when the neuron fires, and then we use a microscope to just look at the activity of these neurons while the fly is tasting different concentrations of salt.

Chris - But just to be clear: under all circumstances - even when there is an aversion to salt at high concentration - it's still activating those cells. It's just that, centrally, that's interpreted as "that's not very nice" versus when the same signal - the excitation - comes in from a different class of nerve cells it's interpreted as "yummy, that's nice, I like that".

Mike - Exactly. And so what actually happens is that those "yum" cells actually continue to fire even when the salt becomes a high enough concentration that it's aversive, but it's just that the "yuck" cells kick in on top of that and override the attraction of those other cells.

Chris - So that sort of aligned with what people had decided probably was going on previously. So where does the extra detail come in which you flushed out with the new way of unpicking what these different receptor classes are?

Mike - So there were two main things that we found. One is that we identified the second population of attractive neurons, which had not really been found or characterized before. But the potentially more interesting thing that we found is that the second population of aversive cells - the negative cells - had a really interesting characteristic, which is that the fly only paid attention to the activity of these cells when it had previously eaten salt. So, if the fly became deprived of salt, they then ignored the sort of bad signal coming from these cells, presumably because when you're salt deprived, even a little bit of a high concentration of salt is going to be beneficial.

Chris - That's intriguing isn't it. Where do you think that decoding is happening is that a central process then so that the fly brain is comparing how much salt it knows it's got on board as total body sodium, for example, and it's saying "well, I'm going to disregard this aversive input because I can tell that actually my salt reserves are very low across my whole body".

Mike - We think it must be, because we actually then used optogenetics, which is a method to essentially activate neurons using light, and we activated these different populations of neurons. And so, even when we force the activity of these negative neurons, if the fly had been deprived of salt, they didn't care about the activity from these neurons. So it can't be happening at the sensory neuron level - it must be something downstream. So we presume that in the higher order circuits of the fly brain that this is integrated somehow...