WEBVTT 00:00:00.017 --> 00:00:05.497 This is Paul Foucher, the Conversion Science Network, talking to Bill Hansen. 00:00:06.437 --> 00:00:12.777 Bill organized a conference on evolution of olfaction. We happen to be on Christmas Island. 00:00:13.697 --> 00:00:20.577 So Bill, why did you choose olfaction as an entry point into the evolution of sensory systems? 00:00:22.217 --> 00:00:25.437 I think it's a cool system. 00:00:25.537 --> 00:00:30.297 That's the first thing in general. And I also think it's extremely important 00:00:30.297 --> 00:00:33.197 for many of the animals that we have chosen to work on. 00:00:33.277 --> 00:00:39.477 So I mainly work on insects and then they are very smelled ribbon as far as I see it. 00:00:39.497 --> 00:00:47.797 And this has made their olfactory sense evolve in different ways. 00:00:47.917 --> 00:00:53.677 And we find different themes, both peripherally and centrally, 00:00:53.737 --> 00:00:56.717 where they have gone in different directions. 00:00:57.317 --> 00:01:03.957 And as I also mentioned in my talk, that we also find these systems that exploit 00:01:03.957 --> 00:01:07.037 the insect olfactory system in different ways. 00:01:07.157 --> 00:01:11.817 And there we have like co-evolution going on, which is also very interesting 00:01:11.817 --> 00:01:21.297 how exquisite one system can exploit another one to dupe other organisms to 00:01:21.297 --> 00:01:24.077 do what they want them to do. 00:01:24.077 --> 00:01:26.797 Right, but now you're jumping ahead of it, right? 00:01:26.837 --> 00:01:31.897 Because in your talk, you made this point that actually the way we look at olfactory 00:01:31.897 --> 00:01:33.537 systems has sort of changed, 00:01:33.757 --> 00:01:39.737 that we have initially taken more of this view of, okay, let's just look at 00:01:39.737 --> 00:01:42.037 these receptors, let's see what they tell us about the world, 00:01:42.237 --> 00:01:45.817 let's see what they tell us about the stimuli that we're processing, 00:01:45.977 --> 00:01:49.957 and now this whole process seems to have reversed, if you want. 00:01:50.057 --> 00:01:54.017 So can you say something about this change? change yeah that there was also 00:01:54.017 --> 00:01:56.877 a little bit about the methodologies that we use so we have. 00:01:58.137 --> 00:02:01.877 We we started off with having a lot of we knew 00:02:01.877 --> 00:02:04.757 something about the olfactory system but not very much so 00:02:04.757 --> 00:02:09.837 what we knew was more or less that we could we could record a signal and we 00:02:09.837 --> 00:02:14.137 could use this signal to tell us something about what insects or other organisms 00:02:14.137 --> 00:02:22.457 met so therefore we We used the system as such as a detector for interesting odors. 00:02:22.457 --> 00:02:25.537 And we used it to identify the odors. And this way we could see, 00:02:25.557 --> 00:02:29.217 for instance, how have pheromone systems evolved. 00:02:29.497 --> 00:02:34.217 And then we could identify what one moth smells, and then the closely related 00:02:34.217 --> 00:02:37.137 moth smells, and the less closely related moth smells. 00:02:37.197 --> 00:02:40.337 And we could draw phylogenetic and evolutionary conclusions. 00:02:41.512 --> 00:02:46.192 Now, I think during the last decade, this has turned around a bit. 00:02:46.272 --> 00:02:52.632 So now we know a lot of the natural things that insects and other organisms detect. 00:02:52.952 --> 00:02:56.712 And now we use that to probe the olfactory system. 00:02:56.992 --> 00:03:00.232 So we take the reverse entry in. 00:03:00.332 --> 00:03:03.992 So we go in from the chemicals where we know that, for instance, 00:03:04.052 --> 00:03:08.752 a fruit fly likes a certain substrate. Okay, then we can use that substrate 00:03:08.752 --> 00:03:12.192 by using state-of-the-art chemistry. 00:03:12.652 --> 00:03:17.472 We can identify what these compounds are, and then we can go into the olfactory 00:03:17.472 --> 00:03:22.612 system, and we can look at what part of the olfactory system is tuned to detect 00:03:22.612 --> 00:03:23.992 these different compounds. 00:03:24.332 --> 00:03:31.212 So we can sort of dissect the system by odors before we dissect the odors by 00:03:31.212 --> 00:03:36.252 the system. Yeah, but this seems almost circular now, because this only seems 00:03:36.252 --> 00:03:38.732 to make sense if you really know what these odors are about. 00:03:39.232 --> 00:03:43.412 How do you then ground this understanding? I mean, it doesn't mean that you 00:03:43.412 --> 00:03:47.612 have a full understanding of the chemical composition of what we call an odor, 00:03:47.732 --> 00:03:49.552 and therefore you can interpret this. 00:03:49.752 --> 00:03:52.972 That's what's good about insects, because we know a lot of different things. 00:03:53.072 --> 00:03:56.932 So the first thing that people started dissecting were the pheromones. 00:03:57.552 --> 00:04:02.152 So in 1959, Butanant identified the first silk moth pheromone, 00:04:02.152 --> 00:04:05.092 what we call by using half a million females. 00:04:05.532 --> 00:04:10.392 Today, we can use more or less one female and one male and get to the chemical 00:04:10.392 --> 00:04:11.852 identity of this pheromone. 00:04:12.012 --> 00:04:16.892 And by then using that one, we can go into the receptor families and try to 00:04:16.892 --> 00:04:20.452 pull out which receptor is tuned to these different compounds. 00:04:21.572 --> 00:04:26.772 But also, as I said before, one of my favorites is to use deceptive systems. 00:04:27.112 --> 00:04:31.832 Because if you want to deceive someone and you want that to be a stable deception 00:04:31.832 --> 00:04:37.252 evolutionarily, you have to be really, really good because it's a constant Red Queen race. 00:04:38.112 --> 00:04:40.092 So if you're a flower... 00:04:41.098 --> 00:04:46.178 And you don't want to pay your pollinator with nectar, which is pretty expensive. 00:04:46.458 --> 00:04:47.998 It's sugar, constant production. 00:04:48.478 --> 00:04:52.998 Then you have to smell so good that the other guy, he cannot say no, he will come. 00:04:53.238 --> 00:04:56.558 So therefore, in many, many flowers, this has developed. 00:04:56.698 --> 00:05:02.218 We think of it sometimes as deception, but I find now the more I look at it, 00:05:02.278 --> 00:05:06.638 that if you really go into the literature, one of the biggest families, 00:05:06.758 --> 00:05:08.758 for instance, that we have of plants is orchids. 00:05:08.898 --> 00:05:11.438 And one third of the orchids are deceptive. 00:05:11.938 --> 00:05:14.218 So they have all invented ways 00:05:14.218 --> 00:05:17.858 of fooling insects to think that they are something else than they are. 00:05:19.298 --> 00:05:23.538 And now also in your own research here, you went through a number of stages, right? 00:05:23.578 --> 00:05:29.538 Because I think you were the first who highlighted this deception using, 00:05:29.678 --> 00:05:31.898 let's say, the smell of rotting meat. 00:05:32.918 --> 00:05:36.218 This was sort of the entry point in this whole issue of deception. 00:05:36.578 --> 00:05:41.258 And now that story has advanced quite a bit, also using very different kinds of technologies. 00:05:41.758 --> 00:05:46.018 So can you say something about that evolution of your own work in this area of deception? 00:05:47.258 --> 00:05:51.538 Well, what we actually started with was orchids. That was when we were still 00:05:51.538 --> 00:05:56.498 working on pheromones a lot and trying to understand the olfactory background to pheromones. 00:05:56.778 --> 00:06:02.918 And then we found these flowers that are really super mimicking bee pheromones 00:06:02.918 --> 00:06:05.518 to dupe bees to pollinate them. 00:06:06.218 --> 00:06:11.658 This was the collaboration with our Austrian colleagues. And that system is 00:06:11.658 --> 00:06:13.638 just as amazing as anything else. 00:06:13.678 --> 00:06:17.818 And it was a good starting point on this whole journey because these flowers, 00:06:17.918 --> 00:06:23.158 they do very intelligent insects, intelligent within Cytacean Mark. 00:06:23.278 --> 00:06:29.458 But I mean, a male bee of this species, the female will only mate with the male once. 00:06:30.218 --> 00:06:34.978 And each female has a unique odor. order. So when the male has mated with this 00:06:34.978 --> 00:06:38.998 female, he learns the order of that female and he will not go back to her because 00:06:38.998 --> 00:06:41.558 he knows that he will not be allowed to mate once more. 00:06:44.393 --> 00:06:48.053 Analogously if every flower that 00:06:48.053 --> 00:06:50.933 were mimicking these odors smelled the 00:06:50.933 --> 00:06:53.633 same the male would also learn the odor of that 00:06:53.633 --> 00:06:56.513 flower and would never go back so what does 00:06:56.513 --> 00:06:59.613 the flower do it mimics exactly the variation that 00:06:59.613 --> 00:07:03.353 you have between the females within these bees and in 00:07:03.353 --> 00:07:06.193 this way dupe the males to to go back 00:07:06.193 --> 00:07:09.233 and and to try complex how complex is 00:07:09.233 --> 00:07:12.493 this variation chemically but it's a 10 compound found 00:07:12.493 --> 00:07:16.213 paramount and and there are three 00:07:16.213 --> 00:07:19.513 or four of these that vary in proportions to 00:07:19.513 --> 00:07:23.553 to a certain extent and in this way they they get these individual signatures 00:07:23.553 --> 00:07:28.373 so that's more or less where we started but then we got into these flowers that 00:07:28.373 --> 00:07:34.713 smell that's what we call sexual mimicry right but then we have the the brood 00:07:34.713 --> 00:07:36.973 site mimicry and that's where we got more and more 00:07:37.013 --> 00:07:40.893 interesting because in the first one you dupe males to 00:07:40.893 --> 00:07:43.973 go for sex in the second one you dupe females 00:07:43.973 --> 00:07:47.733 to go for a place to lay their eggs and that's 00:07:47.733 --> 00:07:53.413 where we started on the the sardinian flowers that smell like rotten meats and 00:07:53.413 --> 00:07:58.133 and we first we observed the system and we saw that flesh-eating flies were 00:07:58.133 --> 00:08:02.733 attracted and what we found in the end was that these flowers mimic exactly 00:08:02.733 --> 00:08:05.993 in the order of rotting flesh to attract flies. 00:08:06.313 --> 00:08:12.373 But not only that, but they also mimic the increased temperature of a rotting body. 00:08:13.309 --> 00:08:18.509 So the flower increases more than 15 degrees above ambient temperature to actually 00:08:18.509 --> 00:08:20.309 attract the flies into itself. 00:08:20.769 --> 00:08:25.569 So the odor is a long-range attractant, and then the temperature is the short-range 00:08:25.569 --> 00:08:27.989 attractant that makes the fly really go into it. 00:08:28.089 --> 00:08:32.049 So it's a multi-sensory deception. 00:08:32.729 --> 00:08:38.609 Okay. But then in this case, this is all assessed at the behavioral level. 00:08:38.609 --> 00:08:42.589 Well, basically sitting there and looking at this long enough and seeing what 00:08:42.589 --> 00:08:44.789 the flies do, what the insects do. 00:08:45.089 --> 00:08:49.669 No, we could never have done that without having access to the olfactory system. Okay. 00:08:50.229 --> 00:08:55.049 First, we saw the behavior in the field, but then we used the olfactory system 00:08:55.049 --> 00:09:02.729 of the bee, of the fly, and so on and so forth, to really pinpoint which odors are the active ones. 00:09:02.929 --> 00:09:07.009 So we could never have done that without the biological system as a detector. 00:09:07.009 --> 00:09:09.649 So which aspects of the system did you look at at that stage? 00:09:09.949 --> 00:09:14.329 This was the first stage, so we only looked at the sound potentials, 00:09:14.329 --> 00:09:15.869 which is called an electroantenogram. 00:09:16.109 --> 00:09:22.209 So you look up the antenna of an insect, and then you kind of record like an 00:09:22.209 --> 00:09:25.669 EKG or an EEG, but here we call it an EEG. 00:09:26.089 --> 00:09:29.489 And on the antenna, essentially, you have the receptors that would be sensitive 00:09:29.489 --> 00:09:33.509 to different aspects of these molecules. Just like in your nose, 00:09:33.569 --> 00:09:38.249 you have thousands and millions of receptors on an insect antenna. 00:09:38.589 --> 00:09:41.309 An insect antenna is like an inside-out nose. 00:09:41.869 --> 00:09:46.309 So there are the receptors sitting on the outside, constantly exposed to the surrounding. 00:09:47.189 --> 00:09:51.029 So how is that work advanced now? Because sometimes also in your presentation, 00:09:51.369 --> 00:09:55.889 you try to show that you have now much more also electrophysiological access 00:09:55.889 --> 00:09:58.889 to the system, and to also more specifically pinpoint, 00:09:59.229 --> 00:10:04.209 let's say, the neuronal correlate, if you want, of the chemical structure of 00:10:04.209 --> 00:10:07.449 the specific volatile molecules that you might be interested in. 00:10:07.629 --> 00:10:10.249 So what has been the advance there? 00:10:11.049 --> 00:10:17.329 So, EAGs have been the methodology of choice for many, and it's something that anyone can do. 00:10:17.489 --> 00:10:23.789 So, it's today a general method in any lab working on insect semiochemicals. 00:10:23.929 --> 00:10:25.589 But then we have now progressed. 00:10:25.789 --> 00:10:29.549 So, now we go in and make recordings from single neurons on the antenna, 00:10:30.409 --> 00:10:34.749 and then you gain sensitivity and specificity. 00:10:34.749 --> 00:10:41.249 But at the same time of course you lose the wide scope of the EEG because in 00:10:41.249 --> 00:10:45.349 the EEG you will get peaks more or less for every compound that the antenna 00:10:45.349 --> 00:10:49.429 detects but the question is then, 00:10:50.289 --> 00:10:55.869 do you get every compound and that's probably not true but if you go for single 00:10:55.869 --> 00:11:01.449 neurons then you increase the sensitivity to such a degree that you really detect 00:11:01.449 --> 00:11:04.369 more or less everything that that single neuron detects 00:11:04.749 --> 00:11:11.249 but you also gain a lot of information on the system per se because you gain 00:11:11.249 --> 00:11:15.529 information exactly on what a specific receptor is tuned to detect it. 00:11:16.889 --> 00:11:24.769 But in this case, you're still stuck, if you want, at the level of single receptor neurons. 00:11:25.129 --> 00:11:28.849 Yes. So if the encoding of the compound, the overall compound, 00:11:28.929 --> 00:11:32.829 would require multiple receptor neurons, how can you access that information? 00:11:32.829 --> 00:11:36.989 Information, can you perform multiple recordings simultaneously on this system now? 00:11:37.229 --> 00:11:42.609 The first way to get at that is, of course, to make exhaustive studies of the whole antenna. 00:11:42.789 --> 00:11:46.529 And that's what we've actually been doing. So you record from hundreds of neurons. 00:11:47.629 --> 00:11:52.289 Consecutively, not at the same time, but you probe and you run and you probe and you run. 00:11:52.449 --> 00:11:57.469 And in this way, we have actually mapped out more or less the full system of an insect, 00:11:57.469 --> 00:12:01.269 like the mosquitoes for instance we have really done very 00:12:01.269 --> 00:12:04.709 very exhaustive screenings but what i 00:12:04.709 --> 00:12:07.989 guess you're getting at is the next level where we can where 00:12:07.989 --> 00:12:13.589 we can open the brain of the insect and we can image the first relay center 00:12:13.589 --> 00:12:17.989 in the brain which is the antenna lobe where you have the glomeruli and each 00:12:17.989 --> 00:12:23.129 glomerulus represents input from a certain receptor type and there we can and 00:12:23.129 --> 00:12:27.069 stimulate the antenna with the output of the gas chromatograph, 00:12:27.349 --> 00:12:32.869 and then we can see how different glomeruli light up in the brain of Drosophila, for instance. 00:12:33.169 --> 00:12:36.329 And in several species, we have such a good map of the brain, 00:12:36.689 --> 00:12:43.489 a constant map that is invariant, that we can map certain receptors to exactly 00:12:43.489 --> 00:12:47.809 the specific glomerulus that is activated. 00:12:48.209 --> 00:12:53.189 So we know that if this glomerulus lights up, Well, that means that the receptor 00:12:53.189 --> 00:12:56.509 69A is activated on the antenna. 00:12:56.829 --> 00:13:00.289 But now, let's take Drosophila as our example, right? 00:13:00.389 --> 00:13:06.649 So what's the odour space that these insects live in? How complex is this odour space? 00:13:08.565 --> 00:13:14.985 I think it is not as complex as ours. We have 200-300 active receptors. 00:13:15.185 --> 00:13:17.745 They are playing around with 50-60. 00:13:18.165 --> 00:13:24.325 So, I mean, it will be more reduced, but still they can encode an enormous amount of odors. 00:13:24.885 --> 00:13:32.965 And, I mean, if you look at all these receptors and the combinatorial coding 00:13:32.965 --> 00:13:37.125 that you can get between them and by getting a few or more activated and so 00:13:37.125 --> 00:13:39.305 on, still already by having 60, 00:13:39.465 --> 00:13:42.185 you can code an immense amount. 00:13:42.445 --> 00:13:49.065 So in terms of real, let's say, olfactory objects, how big would that space be for Drosophila? 00:13:49.465 --> 00:13:51.285 Dozens, hundreds, thousands? 00:13:51.725 --> 00:13:57.165 No, come on. We say that with our receptors, 200, 250 active ones, 00:13:57.225 --> 00:14:00.885 we can code an innumerable number of orders. 00:14:00.885 --> 00:14:04.205 And so it's more or less the same answer 00:14:04.205 --> 00:14:06.985 for drosophila they can more or less 00:14:06.985 --> 00:14:11.245 code any odor that they are exposed to yeah the base in theory right if we say 00:14:11.245 --> 00:14:15.745 all possible combinations of receptor neurons are being used and that's that 00:14:15.745 --> 00:14:20.125 space would be very large but do we know behaviorally that also that that's 00:14:20.125 --> 00:14:25.485 that size of of encoding space is really being used or. 00:14:26.625 --> 00:14:32.565 Well you're getting into very uh interesting also evolutionary questions there 00:14:32.565 --> 00:14:38.765 because i mean of course they have a specific repertoire of behaviorally relevant 00:14:38.765 --> 00:14:43.305 odors that mean something to them and that's where that's where we got an instep 00:14:43.305 --> 00:14:45.225 into this system by using our latest. 00:14:46.365 --> 00:14:51.105 Devious flower as well because we find that this flower is targeting different receptor 00:14:51.305 --> 00:14:54.225 populations that mean specific things to the fly to 00:14:54.225 --> 00:14:56.905 the fly so which which flower is that that's the 00:14:56.905 --> 00:15:00.085 solomon's lily in the north of israel aaron palestina 00:15:00.085 --> 00:15:06.305 and this one dupes drosophila to pollinate it without reward and what we found 00:15:06.305 --> 00:15:12.105 is that it's really hitting three types of receptors one type is saying that 00:15:12.105 --> 00:15:18.065 here is fermentation going on one type says that it's It's fruit going on here. This is fruit. 00:15:18.345 --> 00:15:22.305 And then the third one says that this is exactly your type of fruit that you 00:15:22.305 --> 00:15:24.965 like. So it's really like a three-step rocket. 00:15:25.165 --> 00:15:27.345 And it's probably not mimicking one specific thing. 00:15:28.604 --> 00:15:32.884 Uh substrate but it's collecting super attractive 00:15:32.884 --> 00:15:36.004 parts of several different stimuli 00:15:36.004 --> 00:15:39.204 to build up an image of 00:15:39.204 --> 00:15:43.864 something that the fly cannot say no to no because why am i asking why i wanted 00:15:43.864 --> 00:15:49.404 us to agree a little bit on on let's say the the magnitude of this of this olfactory 00:15:49.404 --> 00:15:54.564 space because this this should constrain a bit our questions around the the 00:15:54.564 --> 00:15:57.544 antenna lobe itself Because in the end, 00:15:57.564 --> 00:16:00.944 if we go from our receptor neurons through our antenna lobe, 00:16:01.024 --> 00:16:02.684 sort of in that interaction, 00:16:02.904 --> 00:16:07.824 the encoding of this effective olfactory space should happen, I would assume. 00:16:08.124 --> 00:16:13.204 So then what does the antenna lobe really contribute to this process? 00:16:13.304 --> 00:16:16.904 Because now you're saying, I can just look at these combinations of receptor 00:16:16.904 --> 00:16:22.284 responses and actually with that already I can tell you what I'm sensing out 00:16:22.284 --> 00:16:27.444 there in the world. So what's the antenna lobe really adding to this process? 00:16:28.424 --> 00:16:33.904 Well, I think the antenna lobe adds balance. So we definitely can tell what 00:16:33.904 --> 00:16:38.884 the receptor neurons detect by looking at activations of glomeruli. 00:16:39.064 --> 00:16:44.284 But as you know, with rosophila, we have the possibility to encode genetically 00:16:44.284 --> 00:16:47.104 our markers at any level. 00:16:47.104 --> 00:16:53.404 So we can encode one marker at the input part of the antenna lobe, 00:16:53.404 --> 00:16:56.224 and we can encode another marker at the output. 00:16:56.484 --> 00:16:59.124 And we can compare these two activations. 00:16:59.724 --> 00:17:04.464 And then we see that there are definitely things happening in the antenna lobe. 00:17:04.644 --> 00:17:10.444 So if, for instance, we have done a very large screening of synthetic odors that we know are. 00:17:11.725 --> 00:17:14.825 Positive for the fly so they will go for them in 00:17:14.825 --> 00:17:17.805 a bioassay and then we have other orders that 00:17:17.805 --> 00:17:20.705 are that are negative so they will not definitely not 00:17:20.705 --> 00:17:23.685 they will go away from them and what we find when 00:17:23.685 --> 00:17:26.805 we look at the input only we don't find any 00:17:26.805 --> 00:17:33.265 correlation among patterns and the balance of the orders but when we look at 00:17:33.265 --> 00:17:40.005 the output of the antenna lobe we found very nice clustering of of of patterns 00:17:40.005 --> 00:17:46.145 that are encoding attractivity and patterns that are encoding repulsion. 00:17:46.625 --> 00:17:53.125 So there is definitely something going on in there that is sharpening the balance 00:17:53.125 --> 00:17:56.125 of the message that is coming. So not the discrimination then? 00:17:56.305 --> 00:18:00.765 So in this case you're clustering data from the projection neurons or from the 00:18:00.765 --> 00:18:02.405 glomeruli, from optical imaging? 00:18:03.265 --> 00:18:08.745 Yes, we are doing optical imaging of both the input, so the receptor neuron 00:18:08.745 --> 00:18:15.365 input into the glomeruli, and then we're doing optical imaging of the projection neuron patterns. 00:18:15.705 --> 00:18:21.605 So in the same fly, you can do both, but using different kinds of dives. 00:18:22.385 --> 00:18:24.885 Or you can do it in different fly circles. Sure, of course. But I mean, 00:18:24.925 --> 00:18:32.645 you can clearly separate what is encoded in the input and what is going out of the system. 00:18:32.805 --> 00:18:38.345 And then we know that in between there are these large local interneurons that 00:18:38.345 --> 00:18:41.345 are shuffling messages between the glomeruli. 00:18:41.745 --> 00:18:46.005 So are you saying that then the glomeruli are just performing some transformation 00:18:46.005 --> 00:18:48.525 from an odor space into a balance space? 00:18:49.688 --> 00:18:53.328 In some way, that's what we're seeing right now, that there seems to be something 00:18:53.328 --> 00:18:57.008 going on that gives meaning to these orders. 00:18:58.788 --> 00:19:03.988 This is really very interesting, but also a bit, in some sense, counterintuitive, no? 00:19:04.068 --> 00:19:08.128 Because we have this projection that's going to the mushroom bodies that many 00:19:08.128 --> 00:19:10.868 people have talked about, that have been, let's say, really very much involved 00:19:10.868 --> 00:19:14.828 in associative learning and especially also valence-driven learning. 00:19:14.988 --> 00:19:18.868 So how should we now compare this interpretation to this more standard view 00:19:18.868 --> 00:19:21.408 on this happening at the level of the mushroom bodies? 00:19:21.688 --> 00:19:27.808 No, but I think there is an initial sorting, and then this gets refined as they move up the system. 00:19:28.048 --> 00:19:33.068 And then you have the two pathways. If you know the insect system, 00:19:33.168 --> 00:19:36.748 you know that there is one pathway that goes from the antenna lobe to the mushroom 00:19:36.748 --> 00:19:39.748 body, and then it goes out into the lateral protocephalogram. 00:19:40.308 --> 00:19:44.828 But then we have one that totally bypasses the mushroom bodies and goes direct 00:19:44.828 --> 00:19:50.128 to the lateral horn or the lateral crotocerebral. 00:19:50.468 --> 00:19:55.888 And there we think that these two pathways are mediating different things. 00:19:56.028 --> 00:20:05.428 One is taking the road over the conscious center of the insect brain and then 00:20:05.428 --> 00:20:07.808 going out to maybe the executive center. 00:20:07.988 --> 00:20:10.968 But one path is going directly to the executive. 00:20:11.548 --> 00:20:14.608 And i think our feeling is that maybe 00:20:14.608 --> 00:20:17.788 this maybe this is extremely important especially for 00:20:17.788 --> 00:20:20.608 this reflexive pathway so that you need a 00:20:20.608 --> 00:20:23.828 sorting already here to tell the reflexive pathway 00:20:23.828 --> 00:20:27.488 what is going on yeah it would still be strange i mean i'm a bit confused now 00:20:27.488 --> 00:20:32.548 because in some sense it would mean that if if my output from my from my antenna 00:20:32.548 --> 00:20:37.728 load processing is just if you want good or bad um the balance of of the stimuli 00:20:37.728 --> 00:20:40.868 i'm dealing with and i lose my discrimination determination abilities i cannot 00:20:40.868 --> 00:20:42.928 say anymore whether it was an apple or an orange. 00:20:43.588 --> 00:20:46.448 You would expect that also for let's say your 00:20:46.448 --> 00:20:49.668 behavior this might be rather unspecific yeah but 00:20:49.668 --> 00:20:54.908 i don't i that's going too far i think because because it yeah it seems like 00:20:54.908 --> 00:21:00.908 you have clusters so so within each cluster of of there might be 15 glomeruli 00:21:00.908 --> 00:21:05.068 so one saying apple one saying orange one saying pineapple one saying banana 00:21:05.068 --> 00:21:08.008 and and i mean but but it seems like what 00:21:08.048 --> 00:21:10.788 we're finding is that these that say something good, 00:21:11.807 --> 00:21:16.087 and sort of take part in the combinatorial coding of something that is good. 00:21:16.227 --> 00:21:21.587 They seem to be in some way clustered and have become clustered during evolution. 00:21:21.907 --> 00:21:26.707 And the glomeruli, maybe also 10 or 15, that are saying that something is bad, 00:21:26.807 --> 00:21:29.007 they have also become clustered in another way. 00:21:29.187 --> 00:21:35.387 And this makes it possible for us by principal component analysis to say, to predict. 00:21:36.307 --> 00:21:41.287 Was this pattern, was this activation pattern in the antenna lobe, 00:21:41.807 --> 00:21:44.767 for a an attractive odor or was 00:21:44.767 --> 00:21:47.467 it for a repulsive all right so you're saying 00:21:47.467 --> 00:21:50.467 it's like a multidimensional code it's something like on the 00:21:50.467 --> 00:21:55.727 one hand it will tell you what it is like this is a banana and it's good yeah 00:21:55.727 --> 00:22:00.207 or this is some acid and it's bad so it will tell you both yeah all right so 00:22:00.207 --> 00:22:04.407 but why at the level of the antenna also from an evolutionary and behavioral 00:22:04.407 --> 00:22:10.327 perspective it seems counterintuitive that already at this early processing stage we start to label. 00:22:10.987 --> 00:22:13.687 Stimuli with respect to valence or not do you 00:22:13.687 --> 00:22:16.687 find this intuitive i i don't find 00:22:16.687 --> 00:22:20.587 it so counterintuitive because if we we 00:22:20.587 --> 00:22:26.307 think that the system has evolved by by glomeruli splitting and forming new 00:22:26.307 --> 00:22:31.067 ones you know you have the evolution of a receptor by a mutation of a previous 00:22:31.067 --> 00:22:37.287 existing one and this adds a new glomerulus in in the antenna so this means that 00:22:37.327 --> 00:22:40.447 maybe you had one detecting something, 00:22:40.647 --> 00:22:45.767 a good fruit of some kind, but then you got a refinement that was evolutionarily 00:22:45.767 --> 00:22:50.187 advantageous, that they could detect another kind of fruit, 00:22:50.767 --> 00:22:55.007 that might be good to separate from the ones they knew already before. 00:22:55.407 --> 00:22:59.867 So then you got the glomerulus added, and these glomerulus will very likely 00:22:59.867 --> 00:23:03.887 be added close to the one that it was split off from. 00:23:04.487 --> 00:23:07.647 And in this way, I see that these clusters have formed. 00:23:07.927 --> 00:23:14.487 I don't see that they really, they might not really form a functionally important cluster. 00:23:17.452 --> 00:23:22.972 Functionally important character, but I find that it's evolutionarily interesting, 00:23:23.572 --> 00:23:31.592 because these clusters have been built maybe from some proto-glomerulus that they all stem from. 00:23:32.012 --> 00:23:34.052 So there has been a few bad ones, 00:23:34.152 --> 00:23:37.212 a few good ones, but then these have evolved and become more and more, 00:23:38.012 --> 00:23:42.952 If the way it's derived now sounds a bit like I can keep on adding glomeruli 00:23:42.952 --> 00:23:48.152 until my skull is filled up or my exoskeleton explodes because I have no more 00:23:48.152 --> 00:23:51.392 space if I look at the mouse the mouse has 1200, 00:23:52.692 --> 00:23:58.572 but they are much much smaller so you can I think you can keep on adding glomeruli 00:23:58.572 --> 00:24:03.512 but I think there is also an end to when you need more, 00:24:04.392 --> 00:24:09.392 that was the whole issue of the effective stimulus space but then in Drosophila 00:24:09.392 --> 00:24:14.892 to what extent is this really genetically predefined or to what extent and isn't 00:24:14.892 --> 00:24:17.772 really open to the stimuli the animal is exposed to. 00:24:19.492 --> 00:24:22.992 What is predefined? The number of glomeruli that they will express. 00:24:24.172 --> 00:24:29.852 I think it is definitely predefined in the drosophila that we work on today. 00:24:30.132 --> 00:24:36.832 I mean, we can go from individual to individual to individual, and it will be constant. 00:24:37.312 --> 00:24:41.672 And we know that glomerulus D, it's always there. 00:24:41.792 --> 00:24:44.612 We can find it. right so so that 00:24:44.612 --> 00:24:48.072 that one is is pretty fine but then i'm also sure that 00:24:48.072 --> 00:24:51.352 there can be during evolution a mutation that is 00:24:51.352 --> 00:24:54.412 advantageous enough to be become fixed in 00:24:54.412 --> 00:24:57.892 the population that creates a new a new receptor but 00:24:57.892 --> 00:25:01.332 then we also know that receptors can pseudogenize and they 00:25:01.332 --> 00:25:04.052 become inactive so so i'm i mean the 00:25:04.052 --> 00:25:07.412 over evolutionary over evolutionary time the 00:25:07.412 --> 00:25:10.592 the system can change it can go backwards it 00:25:10.592 --> 00:25:14.132 can go forwards it can go left or right so 00:25:14.132 --> 00:25:17.152 so i'm sure i'm sure that things are happening there but 00:25:17.152 --> 00:25:21.132 but they take a little too long for us to observe them right what we can do 00:25:21.132 --> 00:25:26.332 is to go and compare species and that's what we're doing so we can for instance 00:25:26.332 --> 00:25:33.532 look at all the 12 species of the melanogaster group and and see okay did we 00:25:33.532 --> 00:25:37.012 have for instance looked at what does diet change mean? 00:25:37.292 --> 00:25:42.492 So we have one species that have started eating only one single fruit out on the Seychelles. 00:25:43.372 --> 00:25:46.732 And this fruit smells very, very strongly of a pineapple odor, 00:25:46.912 --> 00:25:49.512 ethyl hexanoate, and of acids. 00:25:49.912 --> 00:25:53.632 And the acids make the other species more or less die when they eat it. 00:25:53.892 --> 00:25:59.532 And here we find a very, very strong effect on the olfactory system of the fly. 00:25:59.752 --> 00:26:10.832 So this fly has more or less sacrificed three or four other cells that are active in detecting the. 00:26:12.933 --> 00:26:16.193 The sort of normal drosophila melanogaster odors and 00:26:16.193 --> 00:26:19.713 instead into those and see like it has put cells that only 00:26:19.713 --> 00:26:23.933 detect this new type of fruit so so 00:26:23.933 --> 00:26:27.113 it has zoomed in it has totally zoomed 00:26:27.113 --> 00:26:30.313 in its its peripheral system towards this 00:26:30.313 --> 00:26:33.573 fruit and at the same time it has increased the 00:26:33.573 --> 00:26:37.053 center of the brain the glomeruli that takes care 00:26:37.053 --> 00:26:41.013 of this input by 200 percent so so 00:26:41.013 --> 00:26:43.813 it's it's really the first example we know this from 00:26:43.813 --> 00:26:46.893 sex detection in moths that the 00:26:46.893 --> 00:26:49.833 male has increased the number of detecting units enormously 00:26:49.833 --> 00:26:55.353 and at the same time expanded his brain area taking care of female input enormously 00:26:55.353 --> 00:26:59.433 but this was the first time we see such a thing happening with when it comes 00:26:59.433 --> 00:27:05.113 to food and diet but then is it is it really the case that for these insects 00:27:05.113 --> 00:27:07.293 just allocating more real 00:27:07.373 --> 00:27:11.873 estate to the processing of the signal increases their discrimination ability? 00:27:12.733 --> 00:27:17.513 Is that all it takes? I would say that it increases their detection ability. 00:27:17.773 --> 00:27:22.093 Because these flies, they're a bit weak, and they have one competitor. 00:27:22.353 --> 00:27:26.153 And if they are not there first, they will be out-competed by the other guy. 00:27:26.433 --> 00:27:31.973 So they're very dependent on detecting the fruit at an early stage of ripeness. 00:27:32.693 --> 00:27:35.653 At least that what i told you of the of this 00:27:35.653 --> 00:27:38.513 amplification by numbers is not the only thing because 00:27:38.513 --> 00:27:41.353 they have also boosted the sensitivity of each neuron 00:27:41.353 --> 00:27:50.333 by about a thousand times so so they have they have maybe 200 more neurons and 00:27:50.333 --> 00:27:54.713 then each neuron is a thousand times more sensitive so we get down to a detection 00:27:54.713 --> 00:28:00.093 limit that very very clearly competes with moth pheromone detection, 00:28:00.213 --> 00:28:02.493 which is among the most sensitive we know before. 00:28:02.873 --> 00:28:08.253 So we're down to picogram, nanogram of compound being detected. 00:28:08.653 --> 00:28:15.413 In terms of these detection thresholds, what's the sensitivity for drosophila 00:28:15.413 --> 00:28:18.133 at the periphery for these kinds of stimuli? 00:28:18.433 --> 00:28:23.933 And is that sensitivity also boosted by the processing that happens in these 00:28:23.933 --> 00:28:26.673 glomeruli in some way? Or you don't know? 00:28:29.191 --> 00:28:34.331 What we know is that we go down to levels of moth pheromone communication, 00:28:34.671 --> 00:28:40.211 where we can be playing with 10 molecules, which is really homeopathic concentrations. 00:28:41.631 --> 00:28:48.131 Then what we have observed in our investigations before is that we sometimes 00:28:48.131 --> 00:28:52.971 see a boost in amplification at the next level. 00:28:53.811 --> 00:28:59.471 And this boost we still cannot explain. And we see sometimes a similar thing in Drosophila. 00:28:59.611 --> 00:29:05.931 So there is something going on after the receptor neuron has detected with its sensitivity. 00:29:06.431 --> 00:29:12.071 And then to the output levels of the antenna load, we see an augmentation of 00:29:12.071 --> 00:29:14.751 sensitivity that we still cannot really explain. 00:29:15.071 --> 00:29:18.851 Okay. So if we talk about things that cannot be explained, so would you claim 00:29:18.851 --> 00:29:21.711 that to a large extent you understand the system now? 00:29:21.791 --> 00:29:24.131 Could you really make that claim? 00:29:24.391 --> 00:29:28.151 Oh, no. Okay. Okay, so where are we with respect to our understanding of this system? 00:29:29.951 --> 00:29:34.531 Despite the very big investigations, both from our lab, 00:29:34.671 --> 00:29:38.731 but even more from other labs that have come out lately on these local interneurons, 00:29:38.871 --> 00:29:45.551 we still don't understand what they do that shuffle the message between glomeruli. Secondly... 00:29:46.845 --> 00:29:50.965 We lack one very basic thing, and that is how things are hooked up. 00:29:52.265 --> 00:29:56.285 And that's why we're now entering into very, very detailed electron microscopic 00:29:56.285 --> 00:29:59.505 investigations of single neurons, 00:29:59.665 --> 00:30:05.945 of single glomeruli, and trying to understand how are really receptor neurons 00:30:05.945 --> 00:30:07.465 hooked up to local neurons, 00:30:07.645 --> 00:30:09.545 how are hooked up to projection neurons. 00:30:09.745 --> 00:30:13.065 Are there things feeding back in the system? 00:30:14.645 --> 00:30:21.065 Why do you need to know this? Well, you can never understand the system unless you know how it's wired. 00:30:22.165 --> 00:30:27.225 We don't even know the basics. No one has ever taken the trouble of looking 00:30:27.225 --> 00:30:28.625 at the very, very basics. 00:30:28.845 --> 00:30:34.385 There's one study of cockroaches 10 years, 15 years ago, where Dagmar Malun 00:30:34.385 --> 00:30:39.965 in Jürgen Birch's lab actually took the trouble of looking at some of the connectivity 00:30:39.965 --> 00:30:41.945 of the antenna load. Okay. 00:30:42.405 --> 00:30:47.625 But, I mean, you have neurons that are coming in, and you don't even know what they hook up to. 00:30:47.985 --> 00:30:52.805 We hypothesize that they might hook up to this neuron or they might hook up 00:30:52.805 --> 00:30:55.525 to that neuron, but we still don't know. 00:30:55.765 --> 00:30:59.825 Yeah, but look, on the other hand, you do have a lot of electrophysiology of this system. 00:30:59.885 --> 00:31:03.285 You know what these projection neurons that are reading out the glomeruli, 00:31:03.465 --> 00:31:05.665 how they are responding to different kinds of stimuli. 00:31:05.965 --> 00:31:09.785 Yeah, but why do they respond like that? No, but you also know the dynamics of the glomeruli. 00:31:09.985 --> 00:31:15.105 So that means apparently something is missing from the pure physiological perspective. 00:31:15.105 --> 00:31:16.265 So what is missing there? 00:31:16.305 --> 00:31:18.965 And what's the ambiguity exactly you're trying to resolve now? 00:31:19.105 --> 00:31:20.845 But there are still different schools. 00:31:21.245 --> 00:31:26.585 There is the Nobel Prize winner, Richard Axel School, who say that nothing is happening. 00:31:27.445 --> 00:31:31.865 Things are coming in and things are going out. And it's the same. 00:31:32.963 --> 00:31:38.183 And then there is the more Gilles Laurent school that says that things are coming 00:31:38.183 --> 00:31:41.883 in, then they are modified, and then they go out. 00:31:42.403 --> 00:31:46.563 And here, in this case, I more agree with the Laurent school. 00:31:46.643 --> 00:31:48.823 I think more and more are doing that. 00:31:49.343 --> 00:31:53.903 Of course, things are happening in the antenna lobe. It's not just a relay station. 00:31:53.903 --> 00:31:59.163 Why would we have hundreds and sometimes thousands of local neurons shuffling the message, 00:32:00.043 --> 00:32:02.903 if you wouldn't need a modification in there 00:32:02.903 --> 00:32:05.843 well you could have as i say a simple interpretation i will 00:32:05.843 --> 00:32:09.343 need some gain control because concentrations can vary and 00:32:09.343 --> 00:32:13.943 i should sort of regulate now the amplitude of my responses and that's it but 00:32:13.943 --> 00:32:19.083 that's nothing to do with encoding this olfactory space itself of course you 00:32:19.083 --> 00:32:23.263 could have but we don't see that that's not what we observe as i told you we 00:32:23.263 --> 00:32:25.903 We see sharpening of images. 00:32:26.263 --> 00:32:32.923 We see what we think is contrast enhancement, where one odor can more or less 00:32:32.923 --> 00:32:34.863 turn off all the other glomeruli. 00:32:35.043 --> 00:32:40.983 It increases the response in one glomerulus enormously, and it depresses the 00:32:40.983 --> 00:32:43.943 response in all other glomeruli in the whole antenna lobe. 00:32:44.143 --> 00:32:49.743 Okay. So there are definitely things similar to the visual system going on in there. 00:32:49.943 --> 00:32:56.303 All right. So in terms of how we encode this olfactory world. 00:32:58.035 --> 00:33:02.215 Are you also there looking more at, let's say, the Laurent interpretation, 00:33:02.775 --> 00:33:05.975 like we have some kind of possible temporal coding coming out of the system? 00:33:06.455 --> 00:33:11.495 Or do you think you see a more spatial kind of coding or rate kind of coding? 00:33:11.615 --> 00:33:12.915 So where are you going there? 00:33:13.435 --> 00:33:17.235 Because in the end, of course, for the rest of the brain, what matters is not 00:33:17.235 --> 00:33:20.895 what the glomeruli do, it's what the projection errors are telling the rest of the system, right? 00:33:21.975 --> 00:33:25.015 I think we have to look in both ways all the time. 00:33:25.015 --> 00:33:32.615 I definitely think there is a spatial code of some kind, but what that really 00:33:32.615 --> 00:33:37.015 means in the readout, I mean, what really means something is which projection 00:33:37.015 --> 00:33:38.195 neuron will get activated. 00:33:38.195 --> 00:33:41.775 It and and i mean is this 00:33:41.775 --> 00:33:45.535 is the spatial code just a product of 00:33:45.535 --> 00:33:48.635 of that we we have glomeruli in there and it's parcel 00:33:48.635 --> 00:33:51.495 in this way so that what what we 00:33:51.495 --> 00:33:57.295 what we choose to see as as a spatial map is is a product of of the connectivity 00:33:57.295 --> 00:34:01.435 of the antenna lobe i i don't think that's so important i think the main thing 00:34:01.435 --> 00:34:06.755 is that certain receptor neurons come into one location and the message of those 00:34:06.755 --> 00:34:10.975 receptor neurons are mainly picked up by projection neurons at the same location. 00:34:12.095 --> 00:34:15.095 So in that way, you could say that you have some kind of spatial map. 00:34:15.375 --> 00:34:20.375 But then at the same time, I think it's really important to not forget the temporal 00:34:20.375 --> 00:34:24.175 aspects, but you also have to choose the way that you look at it. 00:34:24.675 --> 00:34:28.335 I get very confused by some of the studies that have been done, 00:34:28.435 --> 00:34:30.855 and I still struggle to understand them really. 00:34:31.855 --> 00:34:34.655 For what reason? I mean, what's the confusing bit? 00:34:36.635 --> 00:34:37.255 Took, 00:34:39.049 --> 00:34:47.649 I mean, you get very, very intricate analysis going on of the oscillations going 00:34:47.649 --> 00:34:52.109 on, for instance, in the antenna lobe and the mushroom body and so on. 00:34:52.309 --> 00:34:58.469 And I'm still not, I still haven't got my brain all ready with that. 00:34:58.989 --> 00:35:02.349 Right. It also has been fairly, it's not that this is all clear cut, 00:35:02.469 --> 00:35:03.589 right? There's still quite some debate. 00:35:03.909 --> 00:35:07.329 Oh, there's a lot of debate. This is the right way to look at it, yeah. Yeah. 00:35:07.769 --> 00:35:11.529 But I also have this feeling that we have more to learn about, 00:35:11.689 --> 00:35:13.289 for instance, coincidence detection. 00:35:14.569 --> 00:35:18.689 I think that is one way that you could augment the sensitivity of the system 00:35:18.689 --> 00:35:22.849 in the way that we don't understand right now, so that we see the higher sensitivity 00:35:22.849 --> 00:35:29.309 at the second level that we could explain by theory from just a mere convergence. 00:35:29.309 --> 00:35:35.189 And this higher increase could be explained by, for instance, 00:35:35.269 --> 00:35:38.549 cells going into synchrony on 00:35:38.549 --> 00:35:42.389 the antenna, and then this synchrony being detected in the antenna lobe. 00:35:42.469 --> 00:35:46.229 And that's exactly what we're trying to get at now by doing multiple recordings 00:35:46.229 --> 00:35:50.309 from neurons on the antenna. So have you seen this kind of synchrony? 00:35:50.669 --> 00:35:54.749 The analysis, we did it a few months back. 00:35:54.749 --> 00:36:00.569 But the analysis to get at, to get really, to get, okay, here the stimulus reaches 00:36:00.569 --> 00:36:05.749 both of these neurons exactly at this millisecond, and then go into the analysis 00:36:05.749 --> 00:36:09.569 and say that, okay, after 10 milliseconds, they start spiking together. 00:36:09.909 --> 00:36:15.889 Right. That takes some time. Okay. So now, we'll talk about the evolution here, 00:36:15.969 --> 00:36:17.069 right, certainly in this meeting. 00:36:17.809 --> 00:36:21.329 And it was sort of interesting, maybe also ironic, that you said, 00:36:21.409 --> 00:36:25.329 well, olfaction is a bit like a visual system. But from an evolutionary perspective, 00:36:25.589 --> 00:36:28.749 you could also make the argument that actually the visual system is like the olfactory system. 00:36:29.449 --> 00:36:32.529 And so where do you stand on that issue? 00:36:32.609 --> 00:36:38.209 Do you see really that the olfactory system is like a proto-sensory system? 00:36:39.193 --> 00:36:43.013 Whose principles have been generalized towards other modalities, 00:36:43.093 --> 00:36:44.473 or what's your take on that? 00:36:46.953 --> 00:36:51.213 I think all these sensory systems are created from cilia. 00:36:51.933 --> 00:36:58.273 And cilia were there from the beginning. We have cilia in different forms on all organisms. 00:36:58.693 --> 00:37:01.833 And as Heather Eisen was saying in her talk here at the conference, 00:37:01.833 --> 00:37:10.453 for instance, probably chemoreception and mechanoreception were the two first sensory systems. 00:37:11.213 --> 00:37:19.153 And if we then compare, probably the cilia of the visual system might have occurred 00:37:19.153 --> 00:37:22.673 from the mechanosensories and cilia in some way. 00:37:22.773 --> 00:37:25.713 And you have got the rhodopsin integrated and so on. 00:37:26.533 --> 00:37:31.833 So I think it's very hard to talk about the proto-sense of some way. 00:37:33.213 --> 00:37:38.233 I find the old theory... That is interesting because you're sidestepping the 00:37:38.233 --> 00:37:41.573 issue a bit by focusing so much at the sensory periphery, right? 00:37:41.633 --> 00:37:46.533 You could still argue that the processing that has been performed, 00:37:46.733 --> 00:37:50.973 is being performed by the antenna lobe, combined with their projection neurons, 00:37:51.253 --> 00:37:55.833 and maybe a little bit downstream from that, that provides, let's say, 00:37:55.873 --> 00:37:59.433 the proto-processing principles of any sensory modality, 00:38:00.613 --> 00:38:04.293 irrespective of how the periphery gets you the specific signals. 00:38:04.633 --> 00:38:06.973 Would you buy that? Also that you would not buy necessarily. 00:38:08.793 --> 00:38:13.073 I mean, you can compare it. If you look at audition hearing, 00:38:13.333 --> 00:38:18.533 then you have this, that you have certain frequencies being taken care of in 00:38:18.533 --> 00:38:21.453 certain barrels in our auditory cortex and so on. 00:38:21.453 --> 00:38:28.233 So I think that this parcellation definitely occurs in different sensory systems. On the other hand, 00:38:29.588 --> 00:38:33.528 The glomerulus formation we see all over the place. 00:38:35.468 --> 00:38:41.668 We see some organisms that have an agglomerular input from receptor neurons, 00:38:41.768 --> 00:38:43.308 but I see them as the exception. 00:38:43.548 --> 00:38:47.808 I don't agree with some of what we heard earlier here at the conference, 00:38:47.948 --> 00:38:49.268 that there are many, many. 00:38:49.448 --> 00:38:51.808 I think there are a few where we don't see this. 00:38:53.088 --> 00:38:57.248 But take an antenna and move it to the place of a leg of a fly, 00:38:57.248 --> 00:39:02.888 and glomeruli will be formed where the nerves hit the nervous system. 00:39:03.068 --> 00:39:08.468 So I mean, olfactory receptor neurons have the capability of forming glomeruli 00:39:08.468 --> 00:39:11.848 where they hit the olfactory system in more or less any organism. 00:39:12.148 --> 00:39:17.108 So this seems to be a general principle that has evolved probably independently 00:39:17.108 --> 00:39:22.588 in several different lineages during evolutionary time. 00:39:22.588 --> 00:39:28.308 So I think that is a very basic way of organizing chemo detection. 00:39:29.228 --> 00:39:34.528 But then if you look at how visual, then you don't see this kind of architecture. 00:39:35.788 --> 00:39:40.108 Well, but still you would have some patterns of convergence and divergence that 00:39:40.108 --> 00:39:45.568 are especially convergence initially that are regulated in some form. 00:39:45.568 --> 00:39:52.108 So you might speculate to say, well, the formation of the glomerulus is like 00:39:52.108 --> 00:39:53.648 your proto-receptive field. 00:39:54.168 --> 00:39:58.068 And other sensory modalities that's thrown in, let's say, a thalamus, 00:39:58.068 --> 00:40:03.748 an intermediate processing stage, just developed a more complicated form of a glomerulus. 00:40:03.988 --> 00:40:09.248 But now it really became a mechanism to define more complex receptive fields. 00:40:09.388 --> 00:40:13.008 I mean, I'm just sucking this out of my thumb right now, but you could tell 00:40:13.008 --> 00:40:15.228 a story along these lines that you could try to defend. 00:40:15.668 --> 00:40:20.508 But that's not necessarily one you would buy right now. You would really see 00:40:20.508 --> 00:40:23.128 it as separate principles for separate modalities. 00:40:24.961 --> 00:40:30.201 Yes, basically. But at the same time, I agree that you can definitely have receptive 00:40:30.201 --> 00:40:32.761 fields and that it gets organized, of course. 00:40:33.121 --> 00:40:36.121 Organization is needed for sensory detection. 00:40:36.441 --> 00:40:40.821 Right. But I don't see direct parallels when I look at the olfactory system 00:40:40.821 --> 00:40:42.441 and the visual system, for instance. 00:40:43.581 --> 00:40:47.321 So I think more and more science is needed. 00:40:47.461 --> 00:40:53.881 But before we can say that, there has been a lot of evolution going on since they were born. 00:40:53.881 --> 00:40:57.001 Yeah that that's what you guys tell me so now 00:40:57.001 --> 00:41:00.501 um so another thing you're 00:41:00.501 --> 00:41:04.701 doing here on the island is to look at specific species of crabs or specific 00:41:04.701 --> 00:41:09.121 particular one the rubber crab right so so why is that really how is this helping 00:41:09.121 --> 00:41:14.301 us to understand your factory system well here we are really interested in understanding 00:41:14.301 --> 00:41:17.941 how does the sensory system adapt during evolutionary time 00:41:18.181 --> 00:41:24.341 to go from detection in water to detection in air and this has happened during 00:41:24.341 --> 00:41:29.961 the last five million years it's a pretty fast process and we know also interesting 00:41:29.961 --> 00:41:31.641 new facts about different kinds of 00:41:31.641 --> 00:41:36.601 receptors detecting different kinds of stimuli we know that this the old. 00:41:38.033 --> 00:41:43.053 Evolutionarily, probably old ionopropyl receptors are very active in detecting 00:41:43.053 --> 00:41:44.973 waterborne compounds and so on. 00:41:45.533 --> 00:41:50.733 So what we're trying to look at is basically all the way from the antenna into 00:41:50.733 --> 00:41:52.753 the brain, what has happened to these guys. 00:41:53.113 --> 00:41:57.473 And it's not only in this species, but then we're comparing to other species. 00:41:57.733 --> 00:42:02.213 And what we find really interesting is that different species of crustaceans 00:42:02.213 --> 00:42:05.333 have taken totally different evolutionary pathways. 00:42:05.333 --> 00:42:11.693 Pathways so out of the five lineages that went down to land we have chosen mainly 00:42:11.693 --> 00:42:18.773 two to look at and that's the the hermit crabs and the isopods the woodlice and they have taken, 00:42:19.553 --> 00:42:22.473 diametrically different pathways so the the 00:42:22.473 --> 00:42:27.913 hermit crabs they have expanded the brain area that take care of olfactory input 00:42:27.913 --> 00:42:35.573 to occupy almost 50 percent of their brain while isopods have totally decreased 00:42:35.573 --> 00:42:41.133 almost to zero the part of their brain that takes care of olfactory input. 00:42:41.493 --> 00:42:48.593 So I find this very interesting. Why do two different types of the same type 00:42:48.593 --> 00:42:52.533 of animal, crustacean basically, when they go on to land. 00:42:53.513 --> 00:42:57.373 Go in so diametrically different evolutionary directions? 00:42:57.813 --> 00:43:01.833 And that's what we're trying to understand. To understand it, we do, 00:43:01.833 --> 00:43:06.653 do we have we are studying the antenna we're studying all the receptors set 00:43:06.653 --> 00:43:11.513 up so far we have only found gustatory receptors and ionotropic receptors we 00:43:11.513 --> 00:43:18.673 cannot find our type of olfactory or the insect type right and at the same time 00:43:18.673 --> 00:43:21.073 we're doing the brain anatomy we're looking at 00:43:21.313 --> 00:43:26.413 how it's built how things project how the different centers are constructed 00:43:26.413 --> 00:43:32.413 comparing comparing between isopods and crustaceans and the hermits and so on. 00:43:33.766 --> 00:43:37.906 We were in the middle of that project, but evolutionarily I find it extremely 00:43:37.906 --> 00:43:43.086 interesting because we have sort of an experiment that has been going on during 00:43:43.086 --> 00:43:44.326 the last five million years. 00:43:44.586 --> 00:43:49.546 Right. But now it seems a bit counterintuitive, right? Because aren't you sort 00:43:49.546 --> 00:43:52.086 of swimming against the current now? 00:43:52.266 --> 00:43:56.566 Because, you know, we have preparations like Drosophila, like the mouse, 00:43:56.746 --> 00:44:02.266 which are in some sense becoming now the standard preparation because we know so much about them. 00:44:02.266 --> 00:44:07.326 And what we have is molecular access to them that they can become highly controlled 00:44:07.326 --> 00:44:08.406 experimental preparations. 00:44:08.806 --> 00:44:14.286 And now you come and you find some really very odd crab species to start to 00:44:14.286 --> 00:44:16.246 look at, about which we know very little. 00:44:16.486 --> 00:44:18.986 So isn't it a high-risk operation? 00:44:19.946 --> 00:44:23.666 Of course, but at the same time, I think it's highly dangerous to get stuck 00:44:23.666 --> 00:44:25.186 only in the model species. 00:44:25.726 --> 00:44:30.966 So we should use the model species to the utmost, and I really like working 00:44:30.966 --> 00:44:34.966 on Drosophila, But at the same time, of course, we need a comparison with other systems. 00:44:35.886 --> 00:44:40.886 Otherwise, we get extremely narrow in our view of the world. 00:44:41.186 --> 00:44:46.026 And we get totally Drosophila-centric or Musa-centric or whatever you would say. 00:44:46.526 --> 00:44:50.946 And I mean, we need comparison. And to understand evolution specifically, 00:44:51.266 --> 00:44:54.646 you can never understand evolution only by looking at Drosophila melanogaster, 00:44:54.706 --> 00:45:02.046 an animal that has been totally associated with humans during the last 15,000 or 25,000 years. 00:45:02.526 --> 00:45:07.546 So it's really, really needy that we also keep our eyes open, 00:45:07.626 --> 00:45:09.246 and that's what I really like to do. 00:45:09.286 --> 00:45:15.646 I like to use the model system, but I also really like to go outside of that 00:45:15.646 --> 00:45:19.586 and to try to understand what has really gone on in nature. 00:45:20.226 --> 00:45:24.306 Right. So then how many years will it take for us to understand this crab now? 00:45:25.466 --> 00:45:28.346 Oh they you know when you start that 00:45:28.346 --> 00:45:31.186 on a new system like this one first you have 00:45:31.186 --> 00:45:34.026 to nail down the the ecology of the 00:45:34.026 --> 00:45:37.086 species to try to understand what they are doing and what 00:45:37.086 --> 00:45:41.286 you should basically do is just to sit down in your ass for one year and watch 00:45:41.286 --> 00:45:46.406 the crab but we don't have time to do that most of us so are you good in this 00:45:46.406 --> 00:45:52.126 we i now have one postdoc on the island and she's doing massive marking of the 00:45:52.126 --> 00:45:55.946 crabs and we're really trying to get the impression of what's going on. 00:45:56.146 --> 00:46:01.306 And we're marking them with GPS loggers and so on to try to understand their 00:46:01.306 --> 00:46:04.586 movements and basically how they interact. 00:46:04.766 --> 00:46:07.146 But all of this is supposed to form the base for...