 And I think we are officially live. Great. Hello, everybody, and welcome to yet another session of the Sussex Vision Seminar Series within the Worldwide Neuroinitiative. I'm George Caffetzis, a former master's student in Thomas Euler's lab, and newly arrived PhD student with Tom Badden. As your host for today, I would like to first and foremost thank Tim Vogels and Panos Bozellos for putting forward this very initiative towards a greener and much more accessible seminar world. And having said that, please allow me to get back to the reason we all gathered here for today and introduce our guest from ARHUS in Denmark, Associate Professor Keisuke Yonehara. His research interest in motion sensitive neurons actually dates back to his PhD and first postdoc. During this time with Masaharu Noda in the National Institute for Basic Biology in Okazaki, he focused on genetically identifying different motion sensitive cell types in the retina and of course tracing their projections to higher visual areas. In 2009, he relocated to Switzerland and the lab of Botondroska where he systematically expanded his focus to include the development, the physiological function and the disease of these specially asymmetric motion sensitive circuits. In his own lab located in ARHUS since 2015, they investigate the aforementioned circuits and their contribution to computations in higher areas by exploiting a wide arsenal of molecular techniques, optical or electrical recordings of neural activity and behavioral analysis. Today, we have the pleasure to be hearing about their latest findings which deservedly caught a lot of attention in his talk entitled Sign-up Specific Direction Selectivity in Retinal Bipolar Cell Axon Terminals. Without any further ado, please all welcome Professor Yonehara. The stage is all yours. Okay, so let me share the screen. Can you see it? Yes, everything is visible and your laser pointer as well. Okay, great. Okay, so thank you George for kind introduction and also Tom for inviting me today for this exciting worldwide neural seminar series. So today I'm talking about our new findings about retinal direction selectivity, particularly direction selective axon terminals in the mass retina. So the first steps of vision occurs in the retina which is accessible part of the brain. The photorecept, there are two types in photorecept that roads and the cones which receive a light incoming from the environment and generate a neural signals which are transmitted to downstream except the bipolar cells which has more than 15, about 15 types, distinct functional molecular types. And then these signals are passed to downstream ganglion cells which are the output neuron of the retina. The bipolar cells release glutamate onto the dendroids of ganglion cells. So these three cell types form main excitatory visual pathway in the retina. And these excitatory pathways are modulated by inhibitory neuron such as horizontal cells and the macroin cells which is very important for the sophisticated computations in the retina. The task of retina is to extract different visual features such as colors, edges and the motion from the scene. And those extracted features are encoded in different ganglion cell types and transmitted to brain areas. And strikingly at least eight types among 40 types of ganglion cells are specialized for detecting motion direction. And these cells are called direction selective ganglion cells which was discovered by Horace Barlow who passed very recently and he found these cells in rabbit retina models around 60 years ago. And as this anime shows, this neuron shows spiking activity when light stimulus move across its receptive field. And this response is direction selective. The response is larger for certain direction which is called preferred direction and the opposite direction is called null direction which evokes a minimum response. There are two classic types in direction selective cells in the mammalian retina. One is called on-off DS cells which send axons to the lateral geniculate nucleus and superior colliculus and considered to send signals to visual cortex. And these cells have a broad speed tuning and four subtypes and each subtype responds to one of the four cardinal directions either superior or inferior temporal or nasal directions. Another type is called on-direction selective cells which have a three subtypes responding to either superior temporal or inferior directions on the retina, but recently sub by 12 found the fourth subtype which is also tuned to nasal direction but with a small directional tuning. And these cells send axons to the nuclei of accessory optic system which is composed of nuclei such as the medial terminal nucleus, nucleus of optic tract, those are the terminal nucleus. And this on-DS cell pathway mediates eye movements for gauge stabilization. It's called optokinetic reflex. Then sub by 12 found that the preferred directions of on-off direction selective cells are aligned with the optic flow axis, particularly a translation induced optic flow axis. So as animal move through the space forward, it creates the movement to create visual flow which is called optic flow. As the optic flow directions are indicated by these blue arrows. And this is a projected optic flow patterns on the retina and this expands from certain point in temporal retina and expands into nasal directions. And they found surprisingly that the different genetically labeled on-off DS cells preferred directions are aligned with this optic flow axis on the retina either translational axis or this gravitational axis for vertically tuned DS cells. So this work indicated that the on-off direction selective cells may be a specialized detector for self-motion induced optic flow. Recently our lab tried to test this hypothesis by cortical imaging. So what we have done is to head fix a wake mouse and did the imaging, calcium imaging from different higher areas as well as the primary visual cortex and showed binocular stimulus which simulate translational or rotational optic flow. Then we found many cells which is sensitive to either translational like, this is like this cell response to rotational stimulus and this neuron response to translation. And we mapped the cells in different areas and I don't talk details about here but what we found is that there are many translation or rotation selective cells in higher visual areas particularly area called rostrolateral or anterior area. Then these optic flow sensitive responses were significantly reduced in mice in which retinal direction selectivity is disrupted by knocking out FRMD7G. So these experiments kind of supported an idea that retinal on-off DS cells detect the detector of optic flow patterns. And somehow there's a signal from left and the right are integrated in different cortical areas to create different tunings to translation or rotation. If you are interested in this work more please have a read this bio archive paper. Then now I come back to retinal introduction. Then the starbuster macrin cell is an interneural in the retina which is a key player in retinal direction selectivity. So I will I tell short that these starbuster macrin cell has a cell body and processes radiating away. Then these processes have a centrifugal direction selectivity. So they release cold release gather under such a recording efficiently in response to motion moving from cell body to the tip of processes. And also another work by a Japanese group showed that retina in which starbuster is ablated genetically roots direction selectivity in the retina. So the key circuit mechanism that create retinal direction selectivity has been considered as a there are two key mechanisms. So one needs that centrifugal direction selectivity in the starbuster processes. And the second mechanism is specially asymmetric inhibition from starburst cells to DS cells. So this is purple wines DS cells and this magenta wine it's starburst cells in this figure. The very interestingly starburst cell processes that point in the null direction of DS cells makes synapses and transmit GABA together with a suture recording. But there are much less synapses from processes pointing in the preferred direction of DS cells. But this process but transmit a suture recording non-synaptically. So as it's callinagic transmission is more symmetric whereas GABA inhibition is asymmetric. So when this mechanism is combined with the centrifugal selectivity but starburst cells this DS cells can achieve direction selectivity. So for motion in preferred direction basically this light can evoke spiking activity in DS cells. But for motion in narrow direction first these processes are excited and GABA is released so that these dendrites are suppressed. This is the basic mechanism. What the mechanism that has been supported by different experiments. And DS cells or ganglion cells in general and the macrin cells receive glutamate glutamaturgic excitation from bipolar cells. And it's been under big debate whether these bipolar cells are directionally tuned or not. So many works have recorded IPSC and EPSC from cell body with DS cells and found that inhibitory input is direction selective and also excitatory input is also direction selective. Such a way that so the IPSC amplitude is larger for narrow direction and smaller for preferred direction. Whereas EPSC amplitude is larger for preferred direction but smaller for narrow direction. So it's been believed that direction selectivity may be created by directionally selective invasion and the directionally selective excitation. But they have been debate on this. So as a diamond group suggested that imperfect space clamp could explain apparent direction selectivity in EPSC. And also but another lab showed that this space clamp imperfect space clamp may not fully explain apparent EPSC in DS cells. So the question we addressed in this work is bipolar cells directionally selective and addressing this question should be a key for our understanding of how motion is extracted in the retina. But there is conceptual problem if we assume direction selectivity in bipolar cells. So because motion is computed in four cardinal directions there should be either like four different bipolar cell types each sensitive to one of the cardinal directions or individual bipolar cells should perform parallel processing of four different directions. Or bipolar cells may not be tuned at all. But if model A is the case the retina may need four different bipolar cells for covering all directions, which may be, but it's a bit unlikely that the retina has so many different bipolar cells dedicated to motion computation. But model B looks like very complicated and bit difficult to imagine how this kind of complex processing may be enabled. Then several labs have addressed this question by imaging analysis including myself. So we like seven years ago what I did is to do rabies tracing from DS cell on DS cells with the rabies virus expressing a calcium indicator. Then we were able to label type five bipolar cells that connected to on DS cells. And did the imaging from calcium imaging from type five bipolar cell terminals but we couldn't find the direction selectivity. We also did a glutamate imaging from on DS cell dendroids but we couldn't find the convincing selectivity. And Park et al performed glutamate imaging from genetically labeled on DS cell dendroids. And they also didn't find the directionally selective glutamate releases. And Chang et al performed calcium imaging from type five and seven bipolar cells but they also did not find the directionally selective bipolar buttons. So these works together suggested that bipolar cells may not be tuned for motion direction. And my lab find a new mechanism which can explain a directionally tuned EPSC observed in cell body on DS cells which does not require tuned bipolar cells. So what we did is to do glutamate imaging along dendroids on DS cells. And we found that fast transient bipolar inputs are biased to another side of the dendroids and slow sustained inputs are more biased to preferred side dendroids. And our computational simulation suggested that this kind of a space-time wiring could create directionally tuned EPSC. When light stimulus move in certain speed around 200 microns per second, which is quite slow, then these different input could occur near simultaneously resulting in high efficiently estimated EPSC amplitude. But for narrow direction inputs are deco-related in time. So the amplitude of input may be lower. So we suggested this as a mechanism to detect for on DS cells to detect the slow speed for mediating optokinetic reflex. But it is important to remember that there is diversity of bipolar cell types in mouse retina. There are 14 or 15 distinct molecularly, functionally distinct bipolar cells in mouse retina. And no one has really performed a cell type specific glutamate imaging from individual types. As many people may be aware that axonal calcium levels may have a non-linearity. So axonal calcium may not be a good indicator of transmitter releases. And so we decided to do cell type specific glutamate imaging based on some motivations. So one is that this whole at all reported that alpha seven nicotinic acetylchlorine receptors are expressed specifically in type two and type seven bipolar cells in mouse retina. So these bipolar cell outputs could be modulated by the citricoline from stupas and macroline cells. And the second motivation is our work in collaboration with the Gautama Avatura Money Lab. So what we did is to express citricoline sensor in DS cell dendrites and showed the motion. And we found that dendrites are locally tuned. Some rods show are directionally tuned as cholinergic input. But just next rods show very different tunings. So we found that this kind of micro segments with the tuned cholinergic inputs. But as a globally, we didn't find any bias to any directions. So it's locally tuned, but was not really a globally tuned. And this tuning was degraded by blocking acetylchlorine esterase. So probably in normal condition, acetylchlorine esterase degraded acetylxylacetylchlorine and preventing them to spread too far. So this work suggested that by stupas cell butons could transmit directionally tuned cholinergic input to neighboring structures. So based on these results, we thought this type two and type seven bipolar cell terminals which express alpha seven cholinergic receptor could be modulated in a direction selective manner. Then from here, I will talk about this bio-archive reprint which is entitled Synapse Specific Direction Selectivity in Rectinal Bipolar Cellulaxion Terminals. And all of the physiological experiments were performed by a very talented postdoc Akihiro Matsumoto in my lab. And we collaborated with Shai Sabah at Hebrew University Israel who did a 3D electron microscopy connect mix analysis. So we wanted to do imaging from type two and type seven bipolar cells. Then we tried several different techniques but they didn't work. And one day I wrote to Shindwan at UC San Francisco and he wrote me that he found that AAV with the CAG promoter can express genes only in type seven and type two and the rod bipolar cells. So we made this AAV with the CAG promoter expressing this SF-Igol sonifer, 184S new sensitive glutamate sensor and injected into Maasai sub-retinally. And we found that these type two seven rod bipolar cells are nicely sparsely labeled with the igol sonifer in retina. And we injected this AAV in Maasai in which starburst cells are labeled with the TD tomato so that we can confirm the depth of imaging. So these type seven axon terminals are a little bit lower than this on-chat band and type two is a little bit above the off-chat band. So we were able to make sure that we image from this specific depth. Then we showed motion stimulus moving in eight different directions. And these are individual axon terminals of type seven bipolar cells. So these are color coded based on the correlation, noise correlation. So we assumed that butons with the higher noise correlations belong to the same cell. So probably this is butons from the same bipolar cell. This is from next cell. And it was exciting to see that we show that we found that some butons have a high direction selectivity like in case like around 0.5 or four. And the buttons preferred directions were heterogeneous even among the same cell. And we performed, I'd like to thank Philip Berens for giving us feedback to do this bootstrapping after we deposited the preprint. And we did the bootstrapping and we found that the 38% of type seven bipolar cells. are directionally tuned. Then we made polar plot and this is density plot of this polar plot. And it shows four distinct peaks in those are ventral and the nasal and temporal directions suggesting that this is tuned, these are tuned in cardinal directions similar to all of the cells. And we did a statistical analysis that they did and we found that these butons are indeed tuned along four cardinal directions like four peaks. Then next we wanted to understand the regulations of the direction selective glutamate releases. For this we performed pharmacology. So this is reconstructed axon terminal of a single type seven bipolar cells. And we performed glutamate imaging from these different butons. And we showed motion stimulus and in control, for example, this buton one show direction selective glutamate signaling. Then we are locked alpha seven as it's called in receptor. Then general trend was that release in preferred direction was reduced. Amplitude of the Eigl-Sniper signal was reduced in preferred direction. And we also found TTX reduced direction selective index in tunings. So several works prior works including work by Weiwei lab showed that wide field and McLean cells are connected to type seven bipolar cell terminals. We thought these may be candidate to provide inhibition to type seven bipolar cells. And we use that's why we tested the TTX. So basically TTX should block the activity of GABA agic wide field and McLean cells. And this TTX also had effect reduced to the SI in some butons. For example, these butons blocking cholinergic transmission reduce releases in preferred direction, but TTX increase release in null direction. And we marked these are different pharmacological sensitivity across axon terminals. And we found heterogeneity in regulation mechanisms. 24th percent of the axonal butons were sensitive to both acetylcholine and voltage sensitive sodium channel blocker. And 6% of the butons were only sensitive to cholinergic blocking. And 36% was only sensitive to TTX. And 34% was not sensitive to any need of blockers. And we found that the butons that are sensitive to cholinergic blocking and TTX have a higher direction selectivity in control condition compared to other butons such as sensitive to only cholinergic or TTX. And we found that TTX effect was occluded by blocking GABA transmission. So this supports that TTX effect is on wide field McLean cells rather than TTX sensitive bipolar cells. So we wanted to see if really the starburst the McLean cell is the key cell type that create the direction selectivity in bipolar cell terminals. For this, we used diphtheria toxin receptor transgenic mice crossed with the chat cream mice to express diphtheria toxin receptor selectively in starburst cells. And injected diphtheria toxin ligands intraocularly. Then this ablated most of the starburst McLean cells. And we performed the glutamate imaging from type seven type two bipolar cell terminals in this ablated retinas. And we found that selectivity is significantly reduced. This is histogram of direction selective index. So in control, there are some butons with higher selectivity but in ablated retina the population is shifted to lower selectivity index. And in control, around 40% butons were tuned but in starburst ablated retina only 2% was tuned. And we couldn't see any more effective pharmacology in starburst ablated retinas. So we think not only cholinergic transmission to external butons, but also a geoburgic inhibition from a wide field of McLean cells also relies on starburst McLean cells. Then next, we wanted to see the microcircuits converging onto the axon terminal of type seven bipolar cells. We collaborated with Shai Sabah to perform connectome analysis. So he traced sample. So first what he did is to find the on-off DSLs and the found the connected type seven bipolar cell. And this is the type seven bipolar cell which was traced from on-off DSLs via this synapse. Then he traced all the McLean cell inputs to different terminals, butons. So this is an example image of synapse from wide field cell to bipolar cell. And these are reconstructed wide field cells connecting to type seven bipolar cell which is black here. And we also found many known wide field McLean cells connecting to type seven bipolar cell. And we found that 20% of the butons receive inputs from both wide field and the known wide field McLean cells. 12% receive input only from known wide field and 20% receive input only from wide field cell. So similar to pharmacological experiments we here again we found heterogeneity of a potential regulation from different bipolar cells. This is kind of similar to pharmacology results. Then, okay, so next question was, so the pharmacological experiments suggested that the GABA inhibition from a wide field the McLean cells may be directionally tuned. This is better not yet we are sure but we explored how wide field McLean cells could be tuned. So then because a starburst ablation abolished all direction selectivity, we traced from wide field cells to starburst cells. Then actually we found synapse seeds wrap around synapse from starburst cell to wide field McLean cells. But also, but somehow these synapse from starburst to wide field cell this synapse was not close to bipolar cells. But this based on this process angle of starburst cells we can infer the preferred direction of wide field cell over axons. And these estimated preferred direction was kind of random which matches with our glutamate imaging. So this is the summary of the first part. So this is the schematic of potential circuit mechanisms that establish direction selective glutamate release from type seven bipolar buttoon. So this is a preferred direction. This is another direction during preferred direction stimulus. So this is a situation calling it's non-synaptically transmitted from starburst to bipolar cell buttoons and this enhanced release. I didn't mention that we couldn't find the synapse from starburst to bipolar cell axon which is in agreement with the past results. So probably this chained colonnage transmission is non-synaptically mediate. And GABA release from wide field to the McLean cell it's kind of suppressed by probably from starburst to McLean cell via GABA. And it's already well established that DSL dendrites also receive this recording in preferred direction. In contrast, during narrow direction motion this GABA release from wide field to McLean cell is not suppressed. So GABA is suppressing terminal and the acetylcholine is not transmitted. So the glutamate release is diminished during narrow direction. And as well established this DS cells further receive GABA's input null direction GABA's inhibition from starburst cells. So it is very surprising that different serially connected cells have a kind of coordinated tunings which all based on starburst cells. So this kind of add a new picture into the circuit mechanism of retinal direction selectivity. So the actually cardinal direction selectivity is already established in type 200, type seven bipolar cell timers. And it is enhanced refined further at the dendrites of DS cells. And somehow starburst cells kind of coordinating the aligned preferred direction of the pre and post. Oh, sorry, this in from next slide I will show how this tuned glutamate inputs it's transmitted to different direction selectivity cells. But this is more about how tuning is established. So next question, so from until here we I showed you that how tuning in bipolar cell terminal is maybe established. So next question was whether DS cells receive this tuned input from type seven bipolar cells. For this we did glutamate imaging from the dendrites of all of DS cells. We use the, I think cut clear mice targeted with the Eiglesunifer and injected one of the cell with Alexa for 594 to do targeted imaging from the single on of DS cell. And we showed this trap stimulus which was developed by Burden and the Euler lab. We also showed the motion stimulus. And we found that some region of internet show a chained glutamate inputs. And based on this response to trap we performed clustering unsupervised clustering which is based on sparse PCA and Gaussian mixture model. And we found the six distinct glutamate input types. Yeah, so each type show distinct response to trap stimulus. And here it's a fraction of individual types. The next we wanted to know which group may correspond to type seven bipolar cell. For this we used pharmacology. So we know that type seven input is sensitive to blocking the type alpha seven nicotinic acetylcholine receptor blocking. So we added this alpha bungalow toxin that did the imaging. And we found that the group three it's selectively blocked by this bungalow toxin. So we reasoned that the group three corresponds to type seven bipolar cell. Then we, here is a histogram of the direction selectivity of the boutons. And we found that the group three contains many, like around 40% of the boutons are direction selective. Whereas other groups contains less selective boutons. But interestingly, group one also contained some, level over churned, which boutons, which is 15%. And these boutons were not sensitive to this bungalow toxin or TTX. So we don't know what this is, but this could be one of the type five. We need to continue our study. Then we mapped preferred direction of the individual group three inputs along dendrites of the form of DSLs. And this was striking that most of the inputs were tuned to the same direction. And this was, this is a tuning of individual glutamate inputs. And this is a tuning of a firing as assisted by CELA attached recording. So, and we determined the direction selective index of each group in aligned with the firing preferred direction or randomized direction. And group three showed a high direction selective index aligned to preferred, firing preferred direction, but not to the randomized direction. Group one was also showed selective churnings to firing preferred direction. Yeah, sorry. So this work suggests that somehow all of DSLs receive tuned input collected selectively among heterogeneous retuned boutons. So bipolar cells have a heterogeneous retuned terminals, but somehow these DSLs can collect only preference matched inputs. This was striking result. And next we asked if tuned glutamate release from type seven may affect direction selectivity in DSLs. Here we recorded the EPSC and try to block type seven activity by adding so in control condition EPSC amplitude is larger for preferred direction compared to narrow direction. So alpha bungalow toxin reduced DSI a little bit and the additional TTX further reduced DSI as indicated in this quantification. So this suggests that this type seven input likely contribute to direction selectivity in EPSC recorded from DSL. And here we are showed that effect of TTX can be occluded by blocking of GABA transmission advance supporting idea that the TTX effect originate from a wide field of the McLean cell. Okay. And we also found an interviewing observation that blocking alpha seven colonnagic receptor also blocked direction selectivity in IPSE recorded in DSL. So the amplitude of inhibition is reduced for narrow direction. So this is probably because type seven is known to make synapse with the starburst process. So probably blocking alpha seven colonnagic receptor impaired GABA release from starburst cell processes. So these suggested that type seven bipolar cells have a diverse functions for direction selectivity. So one mechanism is to send the two directionally tuned glutamate to DSL. So the second is to drive GABA release from starburst cells. So by at least two distinct mechanisms type seven bipolar cell may contribute to retinal direction selectivity. So this is a summary slide showing a schematic of multiplexed external direction selectivity for target specific transmission. So as I have shown you, the single bipolar cell perform parallel processing of motions in different for cardinal directions. And these selective signals are transmitted in a target specific way. So that, for example, this upward tuned ones are selectively transmitted to upward tuned DSLs. Leftward ones, selectively transmitted to leftward tuned DSLs. So by such organization you don't, retinal does not need four different bipolar cell types and they can use single bipolar cell type efficiently for coding all motion directions and transmit to different pathways. And these are probably are external tunings are weak and these are kind of amplified but at the dendrites with DSLs. So DS direction selectivity is first generated in axon terminals and they are amplified further in the post synaptic DSLs. And okay, but so this looks like very, the connection is extremely specific and the people may wonder how that's possible. I thought about it and I want to talk about one potential mechanism for creating kind of tuning in bipolar cell terminals. So this paper, Cetlamangian and Matsumotetal bio archive with Gautama Outramani, we showed that dendrite of DSL has locally tuned colonnagic signaling along dendrite. So these are micro domains with different colonnagic tuning and this null direction tuned patch should receive a GABAGIC inhibition from the star bus cell. And other colonnagic signals should be meditated by non-synaptic transmission. Then if it is likely that there is a micro volume, I would call in which a colonnagic transmission is tuned. If bipolar cell which express a Cetlamangian receptor makes synapse with this volume, then automatically this tuning will be aligned to preferred direction. So, but if to test if this is the correct or not, we have to first test if tuning of type seven bipolar terminal and this post-synaptic dendritic micro segments are tuned to the same direction. And such specific wiring could be established by bipolar cells avoiding GABAGIC synapse from star bus to DSL. So we are also interested in developmental mechanisms. What is first established? Maybe colonnagic micro domains are first established and the bipolar cells found the right domains or bipolar cells make synapse first and somehow colonnagic system find this right way, right location to set a selecting. Then I think this is the last slide. So of course, as many of you are aware, we had a discrepancy about tuning of bipolar cells. So I want to discuss a little bit why there is a discrepancy from prior results. So it is important that we found that glutamate direction selectivity in glutamate release is the population of the tuned population is not really a major as a whole in bipolar cell population. This is the result of all glutamate inputs on DSL. So only like around 10% was tuned. This could be easily missed if one performed population imaging. But if imaging was targeted to type seven specifically, then 38% was tuned. So I think this may be the one reason why this selectivity was missed in previous studies. And also in this our work, our raw size was very small compared to prior works. In this work, we did a noise correlation to determine the size of the rods, which looked like worked very well. If because neighboring rods are tuned in different directions, larger rods could easily blend up different tunings. And lastly, there may be potential non-linearities in action processing imaging for estimating releases. Like a work by a firm guess for showed that relationship, there is a non-linearity in calcium mediated current and the post-synaptic glutamate mediated currents are not really linear. And also any sensor has a non-linearity like G-camp also. We are not sure about this could be one possibilities. And finally, I'd like to thank people involved in this study, Akihiro Matsumoto, who is really talented, fast and organized postdoc who did all these works, physical experiments. And Shai Sabat Hebrew University performed the connector analysis. And I'd like to thank other lab members in my lab in office, university, Dandleight. And I'd like to thank David Barson who provided us with all of these scanned electron microscopy volume. Then I'd like to thank Lonebeck Foundation, which provides our major funding and also ERC starting grant and the other funding like NovoNordisk and the Velux Foundation, CarSpark and the DFF Danish Research Council. I'm very good reaction. Thank you very much for your attention. So thank you very much. Kisuke for this wonderful presentation of your recent efforts, holistic efforts, genetic lines, pharmacology, recordings of activity and anatomical EM investigations. We already have some questions in the chat. The first one is from Marla Feller. Are some preferred directions more impacted than others by bungalow toxin? Right. We couldn't find that directional bias in bungalow toxin effect. Yeah, we think that was kind of same for all directions. Okay, thank you very much for this. I just want to remind the audience that you can post your questions on the chat right now. I will soon also post the Zoom room link that we are currently sitting in that you can either join directly or later on for the informal discussion. I continue with a question from Greg Schwarz. Have you thought about the biophysical mechanisms of functional heterogeneity among bipolar cell terminals? How are they electrically isolated? Sorry, I was a bit distracted, sorry. Yeah, no worries. So I mean, it's an excellent question, so I can repeat it. So the question is, have you thought about the biophysical mechanisms of functional heterogeneity among bipolar cell terminals? How are they electrically isolated? I see, yeah, that's a good point. Yeah, sorry, I haven't really investigated that part really. I know that some bipolar terminals maybe have gap junctions, but I don't know about type 7 and biophysical rights. So at least type 7 and type 2 are not expressing a voltage-dependent sodium channel, this we know. But otherwise, I'm not really fully aware of what other voltage-dependent calcium channels are differentially expressed between different types. That should be investigated, yes. Yeah, sorry for interrupting you. I don't know. So what we found is that type 2 and type 7 are almost the same. We could call them like sister bipolar cell types. In terms of directional modulation, we couldn't find the difference between type 2 and type 7. So they may express very similar channels. Yeah, I mean, it's one thing to talk about multiplexing and how to actually establish it within the same cellular space. The next question, sorry, yeah. Maybe it was also about how these different butons are coupled or not coupled, right, in the same bipolar cell. I think it was most, I mean, and Greg can correct me if I'm wrong, mostly how do you manage in the same cytoplasm to have two different functions to have this functional heterogeneity, like to separate one compartment? Yeah, so we didn't really measure electrical activity this time. Just we used the glutamate sensor, so it's a bit difficult to answer. But I think voltage imaging or something could be interesting to look how they behave differently. Next question is from Ana Vlasic. What overall proportion of the bipolar cell inputs measured on the direction selective ganglion cell dendrites were group 3? Do they represent enough of the inputs to explain the tuning in the excitatory current? Right, that's a really good point and I'm also a bit puzzled because this 3D EM reconstruction showed that only like 4-5% of bipolar cell was type 7. But I think more type 2. So it's not really the major cell type, bipolar cell type. So my working hypothesis is that somehow transmission efficiency or used receptor may be not identical among bipolar cell types. So type 7 may have some mechanism that can contribute more to the DSL excitability. If I am allowed to jump in, I remember like from your recent work with Gautam and also like from a paper from Greg Swartz like last year that they checked the signatures of different cell types. Like we have the expression of nicotinic acetylcholine receptors on the dendrites of the direction selective ganglion cell as well, right? So I'm a bit puzzled like how do you bridge these two together? That the multi-site rapid cholinergic transmission, how does it complete the story? Right. So I think I should end this one of the last slides that am I still sharing? No, I stopped sharing but I can. Can I share? Yeah, try sharing it. So this one. So Gautam and our study showed that this dendrite DSL received locally tuned cholinergic signaling. And so we think that our hypothesis is that tuning the site where this type 7 bipolar cell makes synapse with DSL may be also tuned to the same preferred direction. So if this is the case, the tuned cholinergic input and the tuned bipolar input could synergistically excite local dendritic site. Yeah, and we look like as a confirmatory step that this is the actual direction of motion. But my question is mostly like because we are talking about signups less communication. Oh, yeah. Unless you have really specifically expressed there the receptor, then you cannot account for this, right? Because the direction selective ganglion cell will be seeing different directions from a certain whole line. Maybe I'm not phrasing it correctly. We can discuss it later on because like I already see a lot of people are. Yeah, sorry. Sorry. I will proceed with the questions from the audience. The next one is from Serena. Have you thought to remove T2 and T7 bipolar cells contribution to see whether direction selective ganglion cells keep their preferred direction? Yes. So we are now we have a cream ice which labels type 2 bipolar cell, which is net one cell reported by Joseph St. Slap. And we are making the type 7 cream ice now. So our plans to update or ideally kind of a temporary silence releases and to recording some DSLs. Because we found that the pharmacology is a bit difficult to interpret. It affects multiple mechanisms. So yeah, that's what we want to do. Going back to Bungarotoxin effects. Alan Polsky asks, would Bungarotoxin affect direct input from starbust amacrine cells to ganglion cells? Would it affect the estimation of bipolar contribution to DS in the ganglion cells? Right. So that's important points. So we tried to stay out of a seven out. So I think we had a Bungarotoxin Alexa die. So we tried to stay in retina with this toxin die to see if DSLs express binds to Bungarotoxin. But somehow staining didn't work well. So we have to improve the experiment. But yeah, but our yeah, but yeah, I agree that it could be a big issue in interpreting our data. Yes. Next one up is Gautam Awadramani. Do you number, do you numbers of type seven input to ganglion cells match up with the anatomy from a health data that we have? I think like, yeah, I'm interpreting us. Do numbers of type seven input to ganglion cells match up with the anatomy information that we have from the health data set? It's not a major discrepancy. It's kind of, so he's talking about our EM data and Gautam imaging. I assume so. So our EM data suggested that like 4% of on bipolar input was type seven. And Gautam imaging was like, I think what was exact, but less than 10. But it's important to remind that there are four DSL types that we traced only one of the type with EM. And Gautam imaging was done from four types. So it's possible that there is a difference between DSL subtype. So we have to look into this potential type related difference. Yes, I see your point. And the next one again from Gautam is why not measure different signals in DS in SAC dendrites. I could send a signal to DS dendrite. In star mastama dendrites, yes. Oh. Yeah, I got the point. Yeah, actually we wanted to do then I think we are doing now. Okay. Yeah, there are many interesting questions about Gautam input from type seven to star mast as well. Yes. Next question just posted from Theromalini Vaithyanathan. And I'm really sorry if I mispronounced it. Do you expect similar multiplexing demonstrated in bipolar cell terminals in each ribbon? In each ribbon. Yeah. Yeah, I'm not sure about each buttons is supposed to be each ribbon synapse. So yeah, I'm not entirely sure how to interpret this question myself either. Maybe similar degree of multiplexing, but to be honest, I'm not entirely sure. I follow myself. Yeah, I think it's a bit difficult. I think one needed to do simultaneous Gautam images and use the exactly the same sample for EM study. To look into really fine. More space spatial resolution or something. Yeah. We just treated the buttons as a single. Right. So. Okay, great. This was the last question that appeared on the live chat. I posted again the link for this zoom room that we are sitting at this very moment. Some people are already here. Soon I will stop the live transmission. So in case you want to follow up the informal discussion about the zoom room, in case you get latest latest findings, please make sure to join ASAP. This room room. Thank you very much. Thank you. So people are joining. Needless to say, I'm not a moderator anymore. It's an informal follow up of your talk. So people feel free to proceed with your own questions. Great job, guys. Okay. Really interesting stuff. Thank you, Greg. I just logged on to see everybody, but I really.