 Oh, let's go live. And it appears we are officially on. So hello everybody and welcome back to another session of our Sussex Vision Seminar Series, as always within the Worldwide Neuroinitiative. I'm George Caffetzis, a former master's student in Thomas Euler's lab, and currently a PhD student with Tom Baden. And as your host for today, I would like once again to begin by thanking Tim Vogels and Panos Bozellos for putting forward this ever-growing and developing initiative towards a greener and much more accessible seminar world. Having said that, please allow me of course to get back to the reason we all gathered here for today and introduce our guest from Ljubljana University, Professor Gregor Belusic. Following his bachelor's studies in biology, Gregor went on and obtained in 2003 his PhD in biological sciences from the same university. The topic, TRG in transgenic Drosophila, with some work conducted in Germany and some in the Netherlands. In 2013, he was appointed as an assistant professor in Ljubljana University and he has been located there ever since, nowadays holding the title of Associate Professor of Animal Physiology. At the forefront of technological development, including among others a multispectral light synthesizer in his lab, they use a plethora of techniques ranging from ERG to single cell IFIS or spectrophotometry and to behavioral assays in attempts to understand how insects see the world. And today we will be hearing about their latest findings on butterfly vision with a phylogenetic aspect as a bonus as they have studied directly of about 10 different butterfly species. So without any further ado from my side, please all welcome Professor Belusic for his token title opponent processing in the expanded retinal mosaic of Nymphaline butterflies. Gregor, this stage is all yours officially. Thank you, George. Hello, good people of the internet. So let me share my screen with you and yeah, can you see it? You can see the power point and the laser pointer. Okay, so today I will tell you something about the processing of color by means of direct interphotoreceptor of ponency in the largest family of the urinal butterflies, which is the Nymphaline butterfly or brush footed butterflies. The welcome slide shows you the reflection from the eye of the blue morpho which belongs to this great butterfly family and the colorful pattern is called the eye shine phenomenon that can be readily observed in many of the butterflies, virtually all Nymphalids and some other families of butterflies and that can be exploited to infer about the function of the retinal photoreceptors that make this eye shine so colorful. We will hear today about the redomatidia and the non-redomatidia which are apparently green, yellow, blue and all colors due to all kinds of screening and visual pigments in theomatidia and we, as we'll hear, have managed to decipher a bit of this phenomenon and now we can more or less say what's inside hidden below this beautiful mosaic. So first a few words on color vision. Color vision based on chromatic sets of spectral photoreceptors was first I think proposed by from Helmholtz and he just suggested that human vision should be based on three types of spectral photoreceptors but that is definitely true and valid in nowadays but this alone cannot explain some more complex phenomenon in color vision like the after effects, the non-existing colors like combination of red and green, et cetera. So that's why hearing German psychologists proposed in the same century the theory of opponent processing of color where he suggested that inputs from different spectral photoreceptors, spectral channels are compared in some kind of interneurons that weigh the different inputs with different polarities and then report to the brain the outcome which is usually presentable in as a family of curves of sensitivity that oscillate between the positive and negative parts and tell us that our neurons that convey signals from the retina process color in the opponent manner. So a neuron can weigh blue versus yellow color which is a sum of green and yellow or it can weigh red against green. Bushbaum and Gottschalk in 83 proposed and proved that the opponent system should be accompanied by an acromatic signal channel reporting the absolute luminance and thereby the information transferred to the brain is optimal so two opponent axis and one non-opponent or acromatic axis. What is new is that now we can measure the opponent signals not at the level of the interneurons but in the retina, in the photoreceptors themselves. The first reports in the literature date from like 83 where Tomislav Matić has published an article in comparative physiology on spectral sensitivity in Papilla butterfly where you can see that the spectral sensitivity curves go to negative parts and he showed that this is due to electrical interactions in the retina. The problem with this was that it is quite difficult to discern this hyperpolarizing responses of like bioreceptors hyperpolarizing in the green or blue receptors hyperpolarizing in the green orange from field potentials from ERG from extracellular artifacts and usually this has been attributed to bad preparation, poor electro penetration but now we know that it's a true phenomenon that can be treated and even exploited to measure the sensitivity of opponent units not only on the primary units. Subsequently, it has been also recorded in other aspects of animal vision like polarization vision. Roger Harley has published an article on dorsal rim receptors in the flies where you can see that the responses from polarization sensitive receptors can have negative going tails so which are from the antagonistic polarization detectors oriented at 90 degrees. This has been conveniently shown with calcium imaging in receptor terminals in the dorsal rim by Peter Weir and Dickinson in 2016. Now, in the recent years, we have a number of key publications that convince us that the opponent processing is indeed taking place in the receptors themselves without the intervention of interneurons such as the case like in Drosophila where color vision is conveyed by the central photoreceptors, UV blue or UV green sensitive and in their terminals, you can measure with calcium imaging negative signals in the spectrally opponent parts of their sensitivity. We have published also an article together with the lab of Kentaro Aricava on opponent signals in the accents of photoreceptors in Papillio where spectrally antagonistic receptors project synapses and utilize histamineergic chloride channels to convey inhibitory inputs into the spectrally opponent photoreceptors. Another class of histamineergic channels is used to convey signals from the photoreceptor to the LMC interneurons for downstream processing. So the same neurotransmitter histamine opens two different kinds of histamineergic channels. One are fast with the sharp dependence in the LMCs and the other are slightly less sensitive in the photoreceptor accents. Another striking example are the photoreceptors of zebrafish as published by my host lab today where you can see signals with opposed polarity in different parts of the spectrum. So like blue plus, red minus, UV plus, I think it's also green minus receptors at a certain level. I think this is mediated by interneurons. So it's not totally direct between the photoreceptors but nevertheless the signal is present in the photoreceptors. So the opponent processing is taking place at the very first stage of vision. We focused in our research on diurnal butterflies, specifically nymphalic butterflies. So diurnal butterflies have nothing but excellent color vision which can be shown beautifully with behavioral experiments. So they are able to discriminate colors from grace which means that they possess true color vision which is the ability to discriminate spectral composition independent of the intensity and they can be nicely conditioned and demonstrated that they are able to show that to discriminate colors as far as, or as little as one nanometer apart from each other, wavelengths one nanometer apart from each other. This is all based on either simple sets of photoreceptors like in the monarch where you have three different options, UV blue and green options or expanded sets of obscenes like in Papillio. And this together with lots of color filters, results usually in expanded sets of photoreceptors where receptors and non-visual pigments mutually produce a very complicated spectral sense of detectors in the retina. Brush food butterflies have been traditionally thought to be all trichromatic insects similar to the honeybees which possess UV blue and green photoreceptors. And this has notoriously been showing up in all physiological studies so far. Parantica, the night related to the monarch or Polygonia or Sasakia, they all appear to have trichromatic sense of photoreceptors. But behavioral evidence has been accumulated in the years that these butterflies, although they possess only three spectral channels, presumably or at least three options, they are able to distinguish parts in the red part of the spectrum. So like red from orange or red from deep red. This has been demonstrated for the monarch and for Heliconius by Adriana Grisco and Collab. So even though there is no extra option here, they still can see colors in the red part. So to identify the red receptors, we have studied a number of these beautiful butterflies and recorded from the retina with single cell electrode technique with sharp microelectrodes, recorded their eye shine. We studied their anatomy and we found the corresponding substrate for color vision in the red part of the spectrum in some of them, but not all of them. So we collected some at the faculty, some on the beach of the Adriatic Sea and some were from Costa Rica, a butterfly supplier. So we could do experiments all year long. To study them, we lean very much on our own produce from our garage, which is the so-called net synth, a spectral combiner that utilizes the diffraction rating to combine the output of the LEDs into a tunable spectral sequence of stimuli, with which you can very quickly record the spectral sensitivity of impaled photoreceptors. So as expected in Apatura Ilya, which is one of the infallid butterflies that we studied, we have found the green photoreceptors that are actually very broad bed sensitive, the UV receptors and the blue receptors, which are showing signs of opponent input from the green photoreceptors. So actually the UV receptors can be called UV plus, green minus and blue plus, green minus for the blue receptors. We are very sure that this is due to the direct interphotoreceptor-opponent interactions via histamine-ergic chloride channels. The green receptors are not opponent in this case. So Apatura possesses a basic infallid retina. We studied a number of animals found many, many cells and they all belong to three well-definable spectral classes, UV, blue or green sensitive. The UV and blue receptors always have a maximal sensitivity to vertically polarized lights so doors are ventrally oriented. And we know that these are the massive receptors R1 and R2 at the distal part of the butterfly raptom that are long visual fibers that convey color vision directly to the medulla to the second optical ganglion. The rest is the green receptors which are either oriented horizontally or diagonally. So this can be told by measuring their polarization sensitivity. And these are mostly dedicated to acromatic motion vision but of course they also contribute to the discrimination of colors in the green part of the spectrum. Now this opponents in the UV and blue receptors is of course dependent on light intensity used but somehow in the blue receptors, for example, you can see that the positive and negative going part are well matched so that the gain in the blue and the green receptors is optimized that the opponent system works at a huge range of light intensity and allows for color discrimination and deem to very bright light. It is also surprisingly not very dependent on the aperture of the stimulus. So either if you illuminate light, only one or a number of matidia, the effect is very slight, more matidia creates slightly bigger negative response but this suggests that the opponent process is taking place in a single photoreceptor. The opponent response is of course slightly delayed in the blue receptor because there is a synapse between the green and the blue receptor but for example if you look in the ultraviolet receptor the latency to UV response is slightly longer than the latency to a green receptor which is probably due to the very high transduction gain in the UV receptor which then becomes slightly slower than the massive green opponent counterparts which can strike the opponent signal even before the UV receptor gets depolarized. Mapping of receptive fields show that everything comes indeed from a single omatidium and these receptors like blue and green receptors combine into combinations with either high polarization sensitivity in the main part and low or high polarization sensitivity in the opponent part. So omatidia can either contain all green receptors converging onto single UV or blue receptors or pairs of horizontal converging onto vertical and one type of omatidia is suitable for color vision the other is supposedly suitable for polarization vision because you have also polarization of poignancy between the blue and green photoreceptors. Apatura contains the basic mosaic of nymphalids which is based on the long visual fibers that are either UV or blue sensitive and short visual fibers that are green sensitive and always opponent to the long visual fibers. The opposite is never true or can never be observed in the short visual fibers so that the long visual fibers would feed opponent synapses onto the short visual fibers. Together these are one and two blue or UV sensitive form three types of omatidia that are randomly distributed across the retina combinations blue, blue, blue, UV or UV, UV. This forms the retinal mosaic that has a predictable pupil sensitivity. So we also measured the pupillary response of these receptors of these omatidia by illuminating them with monochromatic iso-quantal stimuli and we observed how is the shine from the reflected light from the omatidia decreased upon stimulation with monochromic light. Thereby we could record the action spectra of distinct omatidia. And you can see that a dark adapted eye has a uniform eye shine. The UV or blue adapted eye has a different pattern of attenuation of pupil excitation. And in Vanessa, we noticed that indeed we could confirm that the omatidia best responded to green stimuli, but some omatidia responded strongly to UV and blue, some to UV and not to blue and some to blue and not to UV, which corresponds to the three omatidial types randomly distributed across the retina, like shown here in this patch of omatidia that has been studied with pupillometry or optical retinography. Then in some nymphalates, we nevertheless found some additional spectral classes, a novel spectral class that we found particularly exciting because these are receptors that we found in, for example, in arginis, the fritillary that are depolarized in green and hyperpolarized upon red flashes. The question, of course, is whether the hyperpolarization in the red is an ERG artifact or true synaptic interaction. Therefore, we injected current through the electrode. And first, with positive current, we depolarized the cell, which led to the decrease of the depolarizing responses in the green. And it also increased the hyperpolarizing responses in the red. So we then turned to negative current, which hyperpolarized the cell, increased the responses in the green, and reversed the hyperpolarizing responses in the red. So the responses in the red are due to the current through histamineurgic chloride channels. And this chloride current can be reversed if you are injecting sufficient current close to the synaptic site, which is in the axons in the lamina. In the retina, this is quite difficult to do. But if you are close to the synaptic site and it works, you can show that your current can be reversed and it's a chloride inhibitory current. And then we used selective adaptation with the monochromatic light. First, if we adapted the cell, which was green plus, red minus with red light, we killed the response in the red and we isolated the green unit. Then we also shown monochromatic green light. We silenced the green part and we isolated the red unit. And therefore, we showed that two spectral sensitivities of the units in the opponent pair could be measured independently using selective chromatic adaptation. So we took some large animals, like this archaeopropona, which is half a centimeter high. And we could measure for a longer period in those receptors. We found those green plus, red minus receptors. And we could measure the spectral sensitivity of the red unit using selective adaptation. And it appeared to be a beautiful, spiking, very sharply tuned red receptor with a maximum at around 620 nanometers. Of course, we also measured the sensitivity of the green unit with red adaptation. And these are the green phases. A receptor maximum is sensitive to green light at 520 nanometers. So this type of receptor has been found in all of these butterflies, nine species, the NAUS, prepona, heliconeus, Haraxes, the two-tailed Pasha, fritillaries, arginis, and spaharia, melitea, and morpho. They all show these negative-going responses in the red, but not in Apatura or Vanessa. The telltale sign for the discrimination between those two butterflies versus the rest is their eye shine. So if you look at the reflected light, you can see that in Apatura and Vanessa, the eye shine is uniform with no red color to matidia, while in Haraxes, the eye shine has non-redo matidia, which are yellow to green, even blue sometimes, and red or matidia. It has been noticed also before that species with mosaic eye shine with red or matidia, like heliconeus, are indeed able to discriminate red from orange or red from deep red, but not Vanessa, which has a uniform eye shine. It can only discriminate like red from blue, but not red from orange. So it's not in terms of the independent color vision. Of course, we took heliconeus from the butterfly supplier, and we immediately shown in the same species the existence of this red minus unit. The cellular identity of this green plus red minus cell was surprisingly R1 or R2. We showed it with Lucifer yellow dye injection. So in the location of the usual UV or blue receptor, in the long visual fiber, we found the green sensitive cell, which means that these cells express long wavelength opsin, and this has been sitting in the bioarchive for a while now, found by McCulloch and Adriana Briscoe, that in heliconeus, a class of R1 and 2 receptors express a green opsin. Indeed, they even co-express a green and blue opsin. So we could confirm the co-expression, also in heliconeus, because this green cell is actually picking in the blue with a tail in the green. So a more difficult question was the cellular identity of the red unit. We were never able to directly record from the red unit. We found many hundreds of cells. Our number is something like 500 recorded cells, and we don't have a single recording from the red unit directly. We made multiple correlative experiments to show that this unit is actually probably the basal cell R9 sitting in every butterfly or matidium with a so far unknown role. This unit has a very small dynamic range. It operates a bright light, because it's at the base of the retina shaded with visual and screening pigments, and it's a small cell with a few microvillage. So therefore, its sensitivity is not very high. If you make an anatomical section to the eye, to her axisi, which I've shown before, half a millimeter of eye, and only 12 microns at the bottom is this biloved R9 receptor heavily pigmented. But the lucky coincidence is that it's sitting directly on top of a mirror made from tracheoles that actually doubles its optical path. So its optical length is about 25 microns. So it's still very small, but we know from mosquitoes that cells with such size can operate and can convey vision or color vision. Second, the latency of the response was in the red unit very, very long, compared to the latency of opponent response of green units in the UV plus and blue plus, green minus receptor. Here, the difference is on the order of one or two milliseconds, but the red unit becomes very slow at low light intensity, and even at high light intensity, the delay is long, which means, again, that it's a small cell with a very high transduction gain that makes everything very slow. And finally, we measured its polarization sensitivity with rotating polarizer. And we found that both the green unit and the red unit best responded to a vertical polarizer. So the excitation of the green unit and the inhibition by the red unit were maximal at vertical polarizer. And the only candidate cells with vertical microvilla in butterfly retina are R1 and 2, which is the green unit, and the basal R9, which is the red unit. Finally, we mapped the receptive fields of both units, and we found that there is no pooling across the retina. All signals emerge from a single omatidium. So the green and the red unit sensitivity spatially overlap. We created then a very simple optical model. We projected supercontinuum white light through the random, and we calculated light absorption by each micrometer slab of the random. And we know from published articles that R9 expresses a green opsin. But due to the filtering with overlaid random years of green sensitives for the receptors, most of the green light is absorbed by them, and the sensitivity of the red unit is shifted to red part. But then the subclass of the material contains this red pigment besides the random. And these are the omatidia that are shining red in the eye shine. And this pigment suppresses the sensitivity in the green part, thereby making our selective adaptation experiments at least possible. Because if you project green light, it will not excite the red unit because of this red screen pigment. Surprisingly, very, very similar sensitivities of cells, of basal cells, have been published for wild honeybees in the article of Page and Menzel from 1992. So some wild honeybees, which have mine receptors, probably contain red-sensitive basal receptors, which, of course, opens new avenues for honeybee research, at least for the wild honeybees. Then we analyzed not only electrophysiological, but also eye shine data across the new phallid phylogeny. And we found that across new phallids, the red omatidia with red receptors occur really sporadically. So closely related tribes of subfamilists can have either only green receptors, no red receptors, or red receptors. These all have only UV blue-green sets of receptors. These have red receptors. So it's surprisingly spread along the phylogeny, which means that all these traits, the green sensitive R1 and 2, the basal R9, the screening pigment, they all occur upon a simple genetic switch that can be switched on or off. If you look carefully in the tribe Arginini, the fritillaries, the retina is even sexually dimorphic. So only the males have the red receptors. Now I'll explain this a bit later. Back to the page in Menzel article, many hymenoptera show red receptors, which is most likely made possible by the basal R9. So again, this can be switched on or off among across the species. So together, these omatidia that are red shining create an expanded new phallid retina, expanded mosaic with expanded receptor sets. We decided to present the sensitivity of such animals like the two-tailed pasha with basic receptors, which are UV plus, green minus, blue plus, green minus, and green non-opponent receptors. Along with the expanded set, which is UV plus, yellow minus, blue plus, yellow minus, then green plus, red minus, and red receptors, plus the yellow plus, which is due to the polarization sensitivity, probably the diagonal receptors 5 to 8 in the proximal tier of the retina in the red omatidia. So receptors actually function as broadband detectors, serving quite possibly motion vision, expressing very likely a green opsin, but with sensitivity slightly deformed due to the red screen pigment and the filtering in the raptor. What is very important is that the polarization sensitivity of these green plus cells is super small. So they are almost insensitive. It can be measured that it's vertical, but it's very low. This will be important later in our pupillometry. Again, in this expanded retina, you can see blue and UV receptors combining with opponent units that are usually with very low polarization sensitivity. In a few cases, there are units with polarization sensitivity in the opponent part. But the blue receptors can be either blue plus, green minus or blue plus, yellow minus. And to prove that this is formed by the green or yellow receptors in non-red or red omatidia, we made a simple simulation where we subtracted the sensitivity of a green-adapted blue receptor, which is the steel curve in both cases, from the sensitivity of a green receptor or a yellow receptor, weighed by a certain linear factor. And the result is the dashed curve, which perfectly fits to the recorded sensitivity of a blue plus, green minus, and blue plus, yellow minus receptor. So we see that these opponent combinations can occur in both types of omatidia and perhaps contribute to color vision if this negative going tail has any significance in the subsequent processing of color in the medulla. Whether it has or not is a very difficult question, but still relevant, as we will see later. The spectral sets of receptors in the expanded retina are highly conserved across Nifalidae. We could find identical sets of receptors in Charaxes, the Puteil Pasha, also in the Monarch. Note these orange receptors, the red receptor, the blue plus receptor. Then in Prepona, again, an orange receptor, red receptor, also morpho, orange receptor, red receptor. So this is quite a surprising find that is so conserved, despite the eyeshines being so differently looking. Probably the eyeshines is more influenced by the eye size and by the dynamics of the pigment than by the spectral sets of the photoreceptors. Similar sensitivities have been, by the way, published also by McCulloch and Adriana Briscoe in Helikonium. So they are sitting in the article elsewhere. Now, accidentally we came across the same sister species of Arginis in a study together with the lab of Kentaroa Ricava. We recorded from Arginis Pafia, a polyarctic species, and Kentaroa and Peru Chen have recorded from a few years ago from Arginis Sagana. Their main aim was to show the putative sexual dimorphism in the retina because females have such black, prominent wings. Pafia has orange wings in both sexes. There are some black moths also around in Europe, but I think more importantly, the males have very, very large entreconia, which are structures for dispersing the pheromones, and therefore the female search for males may be more dependent on olfaction than on vision. When we looked into the retina with the eye shine with the oftalmoscope, we have seen immediately that the females of both species have a uniform non-mosaic eye shine, which is the panel A and C, while the males have dark omatidia, if shown with monochromatic light, that appear red if shown with white light. So the eye shine in the males is mosaic with red omatidia and in the females, non-mosaic. We made histological analysis and we found this red screening pigments that I've shown before in the male eyes and not in the female eyes. And these are the screens that are able to screen to filter incident light into the basal photoreceptor if the omatidium is shining red. Again, electrophysiological analysis has indicated that the basic set is present in the females, UV blue-green receptors, while the males have, in addition to the basic set, UV blue-green, also blue plus, yellow minus, green plus, red minus, orange receptors and red receptors, all due to the special filtering pigments in the red shining omatidia and the basal red-sensitive R9 cell. Such an expanded retinal mosaic has been found already in Heliconius, which belongs to the same subfamily as Arginis. So it's both are Heliconini, but Arginis is a tribe of this. So in Heliconius, the sexual dimorphism is of two types. One is due to the application of the UV pigment, which has co-evolved with the UV pigmentation on the wings, and two UV pigments then create six types of omatidia because the combination can be blue-blue, UV-1 blue or UV-2 blue and UV-1, UV-1 and UV-2, UV-2, et cetera. Six types and similar in Heliconius with red omatidia and red receptors, the opsin in R1 and 2, which is now green, co-expressed with blue, leads to six omatidial types, UV-blue, blue-blue, UV-UV, green-UV, green-blue and green-green. So the whole retinal mosaic is like duplicated in order to get another extra spectral channel for seeing in the red. That is quite a high price to pay for an extra channel, but this is apparently the developmental way to go in the butterflies. So to summarize the opulent processing in infallid butterflies, the basic mosaic is based on UV and blue long visual fibers, being fed opulent signals from the long, from the short visual fibers sensitive to green. The red omatidia additionally contain this green sensitive R1 and 2 and having opulent signal fed from the basal R9, which has light filtered with red screen pigments. It takes all these factors to create a red receptor. An additional factor that is required to create red receptor is also a shift of the main green opsin from 520 to about 540, 45 nanometers, which has been observed in all these species that we have studied. So if you want to put red sunglasses to create a red receptor, which expresses the green opsin, it's not sufficient to have a simple short wavelength green, 520, because then the red filter will be effectively an infrared filter. You must go to a longer wavelength green receptor in order to exploit the property of the red filter. The predicted pupillary response of these cells is now shown below. So there will be high response to blue and to green in a blue, blue omatidia. Response to UV and blue in blue, UV omatidia, response to UV light and green in UV, UV omatidia, and the new red omatidia will be presumably sensitive to blue and green, UV and green, or green and green. So no UV response. And those wavelengths symbolize the response to a rotating polarizer. We've seen before that these green receptors are very, very insensitive to polarized light. So the pupillary response to polarized light will be very independent from the polarizer orientation in the green receptor, which is not true for the blue or UV receptors. So we then illuminated the eye of the two-tailed Pasha with monochromic light and the principal component analysis of the pupillary response yielded six classes that perfectly corresponds to the prediction. Blue-blue receptors had strong response in the blue and weak in the green. UV blue, strong response in the UV, and in the blue, UV UV, strong response in the UV. And the three, this rectangle symbolize the size of oscillation upon polarizer rotation. Those three types in the lower row have very small response to polarizer in the green. So we believe that these are the blue-green omatidia, green-green omatidia, and green-UV omatidia. So we found six types of omatidia that were then mapped to the original eye shine. So all the top row omatidia, ABC, which are BB, UB, and UU, were mapped to the non-red omatidia, while ESG, the omatidia with the green receptors were mapped to the red omatidia. So together, the non-red and the red created this matrix, which corresponds to the slightly boosted image of the original eye shine. So we have proven that the green receptors, together with the red receptors, reside in red-colored omatidia. This is now how we understand the eye shine, which is, again, shown on the right-hand side. We record this from the central part, but as usual, the eye is regionalized as in all other butterflies the dorsal part contains very few red omatidia and the ventral part is enriched because the ambient in the ventral hemisphere is much richer in the long-wavelength colors than the dorsal hemisphere. Similarly, convincing results was found in the paper kite, which belongs to the danaïne. Again, pupillary response was strong in the green in three types of omatidia and the response to polarized light was very independent of polarizer orientation, which is not true for the bottom row, where you have UV blue and blue blue omatidia, which strongly responds to polarized light at different orientations. So again, six omatidial types in a species with a mosaic eye shine with red omatidia. We can now consider the butterfly eye shine explained, but of course it's slightly more complicated than just saying that it has non-red and red omatidia because you can see that Harax eye shine in a dark-adapted state has non-red omatidia, which can even appear blue and light-adapted appear green or yellow. This is due to the massive conversion of the main visual pigment from rhodopsin to metrodopsin. This can be triggered with white light and this photoisomerization creates a large differential spectrum and then a sum of rhodopsin and metrodopsin can be fitted to this differential spectrum and you can immediately get the main rhodopsin, which is here, 544 nanometer rhodopsin, the green rhodopsin in this species, which has red receptors. Apatura with a non-mosaic eye shine has a 520 nanometer opsin, so a convincing shift of the opsin to create red receptors. So meanwhile, we have contributed a bit to the study of Papillio lamina conducted by the lab of Kentaro-Rikawa, Kentaro and the co-workers have painstakingly analyzed the synaptic connectome of the lamina. They have traced all the cells in these atomatidia and reconstructed all the synapses, nine photoreceptors in each matidium, four LMC neurons and we have contributed a bit of modeling of the responses based on the synaptic strain. What we can learn from this matrix of connections in Papillio is first that spectrally identical cells, R3, R4, or R5, R2, R9 are not opponent to spectrally identical counterparts, which makes perfect sense, except in some parts, R5 and R6, which are probably spectrally identical, but with microvilli oriented at different angles, which forms then polarization analyzers. Second, in Papillio, long visual fibers, R1 and R2 can be presynaptic to the short visual fibers, R5 to R8 and R9, which is very different from lymphalates, where this never happens, at least as we could observe with electrophysiology. The result is then a stunning complexity of the retina that is underlying this fascinating ability of the animal to discriminate one nanometer apart, monochromic light, but we could also explain the actual physiological measurements, for example, the process dominating the spectral sensitivity of the LMC neuronesis convergence. So the spectral sensitivity can be explained just by a linear sum of the spectral sensitivities of the overlaying photoreceptors, and this can be explained either without knowing or with knowing the connectome. So the connectome perfectly explains that you have three types of formatidia and three types of spectral LMCs that differ in the UV and the red part slightly and possibly could convey the long wavelength information to the medulla, although the sensitivity is relatively broad and all the spectral richness of the photoreceptors is somehow lost. So the question remains, how is then the color information conveyed into the medulla here in Papillio? One possible answer is that it goes through the opponent tail of the long visual fibers that are red, UV plus, green minus, blue plus, green minus, violet plus, broadband minus, or blue plus, red minus. And these sensitivities are again, nicely explained by the linear weights of the synapses and by the physiological results. So this is again showing the previous measurements in Papillio lamina. To conclude, we have now seen the opponent processing phallus and in Papillio, and we can of course ask ourselves, what is the adaptive significance of this spectral design and cellular design of early opponent processing circuitry in butterflies? The first answer that has been proposed by many authors before us is that the spectral opponents leads to a de-correlation of spectral channels and it narrows down the spectral sensitivity of the photoreceptors because they are log amplifiers. And even if the opsin has only one per mile or even less absorbance in a very remote spectral part, the log amplification machinery in the cell will create a large response, for instance, to red light in a green sensitive cell. And with the opponent, this is then reduced, it's narrowed down. Second, the opponent circuitry, as shown now, appears to form an achromatic channel and two opponent channels which conform to the theory proposed by Bushbaum and Gottschall for optimal information transfer. This has been suggested, for example, by Rudi Behnia, he just last year in Drosophila. Another finding by the host lab is that the spectral tuning of opponent channels actually maximizes the variance of information across spectral channels. So Tom and co-workers have went to the nature, measured natural spectra in the natural habitat of the zebrafish and found that the principal component analysis of this spectra nicely matches the spectral characteristics of the achromatic brightness channel and the two opponent channels. We can add a little bit of our insight into this mosaic. So the question here posed is, how is the transduction gain in the photoreceptors adjusted so that the strength of the opponent is just right, which means that the opponent's responses do not kill the main cell or vice versa. So first in Ivalida, the short visual fibers are always opponent to the long visual fibers. Never the opposite happens. And the long visual fibers are short wavelength receptors in principle, UV and blue. And short visual fibers are a long wavelength receptors, green or further. Under the natural illuminance, for example, the skylight, the daylight spectrum, the quantum catch is very low in the UV receptors, slightly higher in the blue receptors and the highest in the green receptors. Therefore, the transduction gain is very high in UV receptors, high in blue receptors, and low in green receptors. This has been noted, for example, by von Helversen, who behaviorally measured the absolute sensitivity in the honeybee, a very high sensitivity in the UV, smaller in the blue, and smaller in the green. We have measured sensitivity in the mouse, for example. In UV receptors are one and a half log units more sensitive than blue, and blue are more sensitive than green receptors. So their gain is adjusted to the composition of the daylight so that all receptors get equally depolarized by exposing themselves to the average light, which brings them into the optimal range of depolarization for optimal signal-to-noise ratio. Now, a tendency from UV and blue receptors to green receptors would, presumably, very much negatively, detrimental influence the signal in the green receptors, because these cells, upon slight exposure to UV or blue patches, would strongly suppress the signal in the green receptors, which is detrimental for motion vision. Therefore, we've seen that in the expanded retina, the green versus red channel has found its own bypass through the motion vision in the means of the long visual fiber, which is sensitive to green, receiving opponents from the red. This is an exception where R9 is a point to a long visual fiber, but R9 is heavily shaded, it has low photon catch, and the green cell is not a part of the motion vision circuitry. So an obvious answer to the hypothesis is, of course, opponent circuitry has evolved to function under natural illumination. Not surprising. And perhaps another hypothesis stemming from our finding is whether the option co-expression and filters and all the sophisticated tricks that the butterflies use in the retina evolved to perhaps adjust finally the gain in opponent pairs to detect colors in some particular parts of the spectrum, where, for example, the signals from the wings of conspecifics or from the flowers are the strongest. So aside from this generalistic opponent circuitry, which has been found across species, butterflies apparently have also more specifically evolved channels where these optical tricks are perhaps used for getting the game just right so that the opponents works perfect. I shall conclude this lecture with small advertising for the articles that describe all the stories that I've told right now. One has been published in the proceedings of the Royal Society B on the finding of the red receptor and the green receptor. The retinal mosaic explained is now the review in philosophical transactions and the sexual dimorphism in fritillaries in the same journal. The article on Papillot is in preparation and should be submitted quite soon. So you will be able to read all about it in detail probably next year. I would like to thank you for your attention and to my co-workers from the lab, especially Primoz Spiri, who developed all the fancy optical measuring devices, the pro-pillometry. Marco Illich has contributed a lot to the analysis, to the modeling. Andrei is our anatomical expert. Aleš has contributed to the electrophysiology. Marco to the optical measurement. The butterflies were supplied by Costa Rica entomological supplies and by a colleague Mark Jals from the UK who supplied Apatura. They were determined also by our colleagues from the zoology department here in biology and the graduate students that have worked on the butterflies at Lucia and Iost. All done in collaboration with Kintaro Aikawa and Dukas Teringa, financed mostly by the Air Force Office of Scientific Research and a little bit by the Slovenian Research Agency. Thank you very much. I have not been able to see your questions because I'm actually in quite a Zoom and your questions are on YouTube. So this will be moderated by George if there are questions. Yes, here is where I come in again. And thank you very much Gregor for this impressive work like both in wealth and in breadth. And I really enjoyed personally how your work in non-model species can inform work in others like as the Hanebis as you mentioned already. There are already a number of questions appearing in the chat so I will keep mine for later on. I will restrict myself to being the host and the moderator of this discussion. And I would like to remind our audience that they can either post your question in the chat or join us in 10 minutes from now in the Zoom room that we are currently sitting in to ask in person or continue in a more informal get together. So given that you touched on it towards the end of your talk, I will start with the last question that appears on the chat currently from Charlotte Lewis. Why do they bother making the red receptor in the first place? And you already mentioned something on it. It seems very energy and cost inefficient. So if you could design it, would you design it differently, let's say? Yes, the solution to the red receptor in most other butterflies is very different. So the piered butterflies and papillionate butterflies, they utilize their diagonal proximal tier of receptors R5 to 8, which are equipped with a long wavelength sensitive opsin and a bit of filters. And they create a very powerful, sensitive system of red receptors. And this solution in infallid butterflies appears indeed slightly clumsy, energy consuming, but filters are involved all over the place because red opsin is seemingly not trivial to evolve. So you need some kind of long wavelength opsin plus the filters to create a red receptor. And of course the selective pressure to design this receptor must be enormous because as we have seen the phylogeny, these receptors occur and appear and disappear all across the infallids, sorry for clicking this. But yeah, they can be switched off among, on and off among the sexes and present in very closely related sister species or absent in those species. So if I would design them, I would definitely go the papillionate and piered way because it's so much stronger system. But then maybe those receptors, the diagonal red receptors can influence the motion vision system because as we have seen in the publications from Papillio, motion system in Papillio motion detection system is not independent from color contrast. So it can be triggered with color contrast only. And maybe that's the price to pay that your motion detection becomes sensitive to color variation. And I think there will be like some really nice insights from Kit Longden early next January, like when we continue the series about this specific topic that we just mentioned. Next question is from Luisa Ramirez. Do you think that a reason for having an interphotoreceptor mechanism instead of an interneural mechanism such as horizontal cell connections could be again in the image resolution? And if not, do you have other explanation for such a difference across species? So the main difference that I've shown here was for example, the difference between Drosophila and the butterflies. In Drosophila, you can only see the opponents in the receptor terminals. That's why they have never been recorded in the receptor somata except for the dorsal rim. And we hypothesized that the Drosophila laminar has been secondarily reduced due to the development of the neural superposition. So this very complicated developmental sequence that leads to the convergence of different material to common interneurons. Then somehow abolishes any possibility to create this fancy connectome that allows for color opponent circuitry. But direct opponents between the photoreceptors is probably the most economical way to do it because avoiding one interneuron is a good way to save on energy and resources. So if you have the laminar dedicated to other tasks like motion vision, high sensitivity motion vision, this is the path that the flies have taken, then color vision comes second. And this is very nicely seen in all behavioral tests of color vision in flies which are just off the scale if you compare them to the butterflies. So performing quite poorly. And but of course the gain is that they have colonized all the worlds very successfully and they can torment us in any kind of any time of the year while butterflies are much more sensitive, maybe. And yeah, if an animal can avoid an interneuron then this of course also comes at the price of flexibility but of course our findings do not exclude the existence of interneurs in the medalla. Of course there will be inter-o-material opponents weighing from different types ofomatidia, different spectral sets in adjacentomatidia, then interneurons can compare the signals from them, of course at the price of the spatial resolution because then the precision, single-matidium precision is lost. But that's not something very new because we also do it. We also compare spatially different spectral receptors and our resolution of color vision is also inferior from the acromatic vision. So the last question or I should say questions appearing are from Tom. I will start with the first part. How well do the spectral zero crossings tend to match any difference in a respective spectra in cone-specific wings? We have not systematically measured the spectra of the wings and I can immediately tell that those points where the spectra crossed zero are not fixed because those receptors tend to have different moods of opponents. You can have a sleeping beauty where there is no opponents in the butterfly and it's very difficult to get any opponent signals in the receptors. They can be eventually enhanced with receptor, depolarization with heart injection or you can have an opponent with very strong flavor and so what I'm suggesting here is that the opponent may be state dependent. So the next thing we are going to test is of course the effects of neuromodulators on this opponent and this will then affect also the crossing zone. So it's not totally fixed. But the wing spectra are well measurable and this is of course an attempting thing to measure because we all speculate that the butterfly vision is sophisticated also to see their cone specifics. So yeah. The continuation of this question again from Thomas, do you think that this is a major driving force in selection or is it possibly more about wider statistics across the scene that kind of thing as well? Yeah, I mean we have seen highly conserved sensitivity across species. So you can see that this is quite a generalistic visual system not very particularly tuned to the wing spectra. Of course it would be probably good to measure exactly these crossing points and correlate them but yeah, a blue butterfly with the red receptor apparently uses the red receptor to maybe evaluate the host plant. Maybe that's a major driving force, the quality of the greenery where the position takes place because the spectrum of the non-quality leaves is very much changed in the red part or the far red part. So I think lots of this is anyway devoted to the generalistic vision not so specifically tuned to the wings. But of course then you have this fancy broadband and violet receptor simpapelio for example which are in any color vision model excluded because they do not seem to contribute to generalistic color vision. So the violet and broadband receptors in papelio but they can of course be used for other tasks like wing recognition or broadband receptors could measure the general illumination, et cetera. So there are exceptions to it of course. All right, thank you very much. So as there are no more questions appearing in the chat I would like to thank you once again Gregor and I would like to remind the audience that they can join us in the post-talk informal get-together even as bystanders if they want to continue following this discussion because soon I will be terminating the live broadcast. So you can either follow the link that you will find posted there or we will see you in one of our next talks. And given that there are no questions appearing for now like I have a couple of them that I would like to ask you Gregor. So one of them is like you saw it towards the end of your presentation that the retinal mosaic changes like based on the adaptation and it appeared that there was like a dose of dental gradient. Is this like something consistent or it was like just one? This has been noted basically in all butterflies with measured eyeshine. This is very predictive, predictable. So the top part is enriched in shortwave interceptors and the bottom part is enriched in long-wagel interceptors. So this is very usual. Okay and the second maybe a little bit crazy. Is there any circadian adaptation of the mosaic that we know of? Whoa, this is cool. I mean, as I said, this may be all stay dependent but the mosaic as such is fixed and it cannot change but of course the opponent connections may change somehow but what we have seen is actually only intra-o-material connections. So no inter-o-material stuff going on. Thank you. And one last question that is of general nature. So you mentioned that in some butterflies we don't see these opponents with that green opponents. Do we know if these butterflies like displayed different behavioral or ethological components that separates them from the rest or this is not unknown? I would like to be able to answer this. There's been theories put forward in the lab that maybe those with simplified mosaics, simplified visual system to chromatic are more towards generalists being present earlier around migratory species like Vanessa but the monarch is also migratory and has beautiful red receptors. So there is no clear answer to this of course. And even in the coloration as I said, morpho with the red receptor is like very counterintuitive. Yeah, I guess it's quite a complex niche to dissect and like find the cause and effect relation between behavior and expression patterns. Okay, so at this stage I would conclude I think the live broadcast given that it's also the first talk that the baby is attending and that you go into like an interview mode and ask you how you feel Gregor, how you feel baby and so on. But yeah, I would like to thank everyone for tuning in. And again, as I already said, if you want to tag along, please do make sure you click on the link that is available in the chat. Thank you very much. Thank you folks for listening and thank you for inviting me and to see you just around the corner. Okay, and we are officially off air.