 Hello, good morning everybody. Thank you very much, Mary, for having me here. Thank you very much to INCF to give me this opportunity and to I really look forward actually to those days to to discuss how we can interface experiments and modeling and and that analysis and I guess the the subject on which I've been working for the the last 15 years or so is particularly amenable to this type of of interaction and what you just actually described so What we are going to go through is a series of a of Datasets that we obtained in rather recently in the last few years and our attempts at actually modeling this dataset to understand what's the interplay between receptor organization at excitatory synapses and the synapse function and Really, I have actually although I'm really not a Very serious modeler. I'm just doing that slightly on the side but certainly over the older years I came to to realize that really Modeling our our data has been extremely helpful in actually understanding what what what they mean And that's kind of the message. I'm gonna try to convey to you So bridging super resolution imaging Single-molecule tracking of Empire receptor dynamic organization and synaptic function So our object of interest are excitatory Determinaturgic synapses. I'm sure you all know That those mostly occur on the small Protrusions that are called spines. They are pretty small and that's the main reason for using super resolution imaging Methods to look at their organization those synapses are really at the base of brain function And of course understanding how they work is extremely important to understand human behavior memory learning and also many Your generative disease and and brain pathologies and We're gonna concentrate on the post synaptic side of these Synapses trying to understand. What's the organization of these receptors and How the organization of these receptors and this dynamic organization actually has an impact on on synapse function So these glutamatergic synapses They harbor a whole variety of molecules and in particular a whole variety of receptors We're gonna concentrate just on one subtype of receptor which are called the amper receptors Which are cationic ion channels Opened by binding of glutamate and those are really the receptors that mediate all excitatory basal synaptic transmission That's kind of another schematic view of that post synaptic Synapse whole variety of molecules actually thousands of of molecules and Most of them are known I would say by now at least from their their molecular sequence What really we don't know is how they actually Organized together although from biochemical data We have a lot of information on how they actually all interact with one the other Actually how they are actually really organized at the nanoscale is really something that's just beginning To be understood and one of the message I really want to Convey to you today is that actually knowing the molecule is just not enough We really have to understand how they are localized and how they are dynamically interacting over over time and for example one of the reasons for that is that if you look at the way Glutamate is released from the presynaptic side What we are gonna see is that actually When a vesicle is released in the and releases its content in the synaptic cleft Actually, there's only a subset of the area of the post synaptic membrane That's really sensitive to that glutamate because there's the number of molecules of glutamate releases actually pretty low and so knowing whether receptors are here or here or here actually does make a difference and One of the hallmark actually of this the organization of this of these synapses, which is also important is that although there's lots of different molecules their actual numbers is pretty low and Sometimes it's it can be as low as just a few of them. For example, if you look at the NMDA receptors that we won't talk about today Some synapses only have one or two of these receptors and most cases It's only around 10 receptors So getting numbers and how they organize is really important so That's of what's called the the scaffold of these of these synapse The other aspect that's very important is that this overall organization is extremely dynamic And we actually know that synapses are dynamics since a very long time Over 40 years ago In fact, it was found by in these seminal work from this and Lomo that actually the efficacy of synaptic transmission is extremely plastic And you I'm sure you all know these seminal experiments whereby recording Synapse efficacy in the in the epicampus those others found that when you do high frequency stimulation of the affluent pathway You can change very rapidly and for a very long time the efficacy of synaptic transmission either Potentiating it or depressing it depending on the way you make the the stimulation and very early on the authors proposed that this could be The cellular basis of memory and after lots of bloodshed and the wars in the field. It was actually We are coming to an age. I would say that's rather accepted now that indeed plasticity of synaptic transmission is one of the core mechanism of Learning and and memory and so reconciling this complex organization of the of the old synapse and the plasticity of this function is really a very challenging question that That needs concerted efforts for many many fields including Neuro informatics So in terms of this plasticity A turning point Occurred about ten years ago. I would say for when a number of labs actually found out that one of the Basic principles one of the basic mechanism of these changes in the synaptic efficacy Occurred through change in the number of receptors in the post-synaptic membrane It's not the only mechanism But it's it really it's coming apparent that it's one of the major mechanism of plasticity of these excitatory synapses at least in the hippocampus And so the idea is that if you increase the number of receptors in front of the release site to potentiate synaptic transmission If you decrease the number of receptors you Decrease synaptic transmission and so really became Pretty important conendrum to understand how you could at the same time have this very complex Organization of the receptors under the same time being able to regulate very rapidly Their numbers and maintain that over time And that's one of the really the major question. We've been we've been asking over there over the years So getting into this organization of the receptors and how it can be dynamically Organized one has to realize that our view of the organization of these receptors has been changing Dramatically over the years If you look at the very first images of those ampereceptor localization by em back In the night in the mean 90s at that time using a post-timing Immunogold only a few gold particles were found in synapses and and those type of data actually set Something that was very wrong a dogma that there was only a few ampereceptors per Synapse and they were all concentrated in the post-synaptic membrane and as new data become available that view actually changed quite extensively Up to the current view which was initially developed, I guess from freeze fracture replicas Finding over hundreds of receptors in the synapse But also lots of receptors in the extra synaptic membrane and as we'll see there are actually much more extra synaptic receptors than synaptic receptors and then recently development of super resolution optical methods actually allowed to use optics To visualize receptor organization from this first work from the lab of java is on using storm to look at organization of receptors and then Only last year actually three papers came out nearly simultaneously one from our group I'll describe and then two others Using a variety of super resolution methods together with electron microscopy Finding that actually ampereceptors are not diffusely distributed in the post-synaptic membrane But seem to be organized in very small nanoclusters and I'll start by Describing those clusters of receptors and then go on to discuss what could be their function and how Then we'll go into how is the din receptors dynamically organized into that So in terms of dynamics, that's first slide introducing that Beyond the localization and organization of receptors in the post-synaptic density as I told you we now know There's a lot of receptors also in the extra synaptic membrane and something I think quite important. We found More than ten years ago now is that receptors can actually exchange all all the time between Synaptic locations and extra synaptic location through surface diffusion and we found that by doing single molecule tracking tagging little flow force to the receptors and doing videomicroscopy Finding that receptors can diffuse Constantly between those different sites at pretty high rates and I'll show you our recent developments of that receptors also actually traffic between the surface and the intracellular Compartments of the of the neuron by recycling events. Those are extremely important. Also, they occur on a slightly slower time scale while those diffusion events occur in the millisecond to second time scale Those recycling events occur more in the tens of seconds to minutes timescale and just for matter of choice I won't talk at all about these these events today, but we can discuss them later on what we have got to get into is really understanding How are those receptors organized as I said in the PSD? How they do exchange between the PSD and the extra synaptic membrane and how this fast exchange by surface diffusion actually impact fast synaptic transmission So some of the open questions and I won't have time to go through all of them. They're just to In echo to you to what you were saying in your introduction Some of the things that I think are important to look at from a neuroinformatic perspective Are the one the first question that we are going to tackle it is what's the dynamic of these nanoscale organization of receptors? How does it impact sign-ups function? What's the function of this fast amper receptor diffusion? What's the balance between synaptic and extra synaptic receptors? What's the dynamic interplay between surface diffusion and recycling? What are the sites of amper receptor insertion of removal? How does regulation of this traffic contribute to sign-ups plasticity and really it's really my deep feeling that Understanding all these different questions really really requires not only new experiments But also modeling data to actually be able to put all this in understandable framework So I've divided my talk in mostly three parts first. We're going to look at static data looking at how receptors are organized at the nanoscale and then Show you what are the evidence we have that amper receptors are mobile not only between synaptic and extra synaptic sites, but also Actually inside the PSD and then and and to look what could be this the function of this quick amper receptor mobility in fast synaptic transmission So first I'll show you our recently published data that amper receptor are organized in nanoclusters for that we've been using both Electron microscopy, but mostly single molecule based super resolution methods So I'm not going to talk extensively about those methods at all just to give you for those of you who don't know The the hallmark of what I think is a very important Development in the field of imaging that occurred a few years ago in terms of obtaining nanoscale resolution Images with optics of a molecule organization. And so the basic principle Is to use a dye it could be a genetically encoded protein or an organic dye You can photo switch what whatever the method is not so important and you photo switch that dye between One color and another color or between a dark state and an open and light state And you're gonna image the dyes one by one in order to go from a very blurry image to a very high resolution image and The idea is that you're gonna image those dyes one by one or a few at a time So that you can resolve each of these of the single organic dyes of each of the single dye sorry and You can actually localize each of these single dyes with a very high precision which can go down to a few nanometer Because you perfectly know what's the point threats function of these dyes and the difference in aspect of the spot and then you do that over and over again until you can actually reconstruct a super resolved image That has gained more than an order of magnitude in terms of resolution a variation of this method that you can use that I'll show you lots of data with is called Dynamic palm or single particle tracking palm where at the same time as you look at the localization of the molecules You also look at their movement. So you get you get at the same time both The super resolved image, but also the dynamics of all the different molecules So some data here obtained on these receptors tagged with one of these photo switchable proteins That's called deos. That's normally in a green form and then can be switched to a red form and You acquire stacks of images With literally hundreds of thousands of data points given all the location of all the different single molecules They all are activated and then the photo bleach one by one Getting a single image takes several minutes but you can actually accelerate that now with new cameras and Better dyes and go down to a few seconds to to obtain a whole stack of of images I won't talk at all about it Here, but just to let you know that of course gathering all these data and analyzing all these single molecule data very Efficiently actually requires a lot of informatics not your informatics specifically but a lot of informatics and There's not a week in nature methods where there's not a new algorithms and new methods to actually Analyze all these single molecule tracks, but I won't go into that at all today So anyway from these single molecule data You can actually reconstruct a super resolved image and there we found really our first big surprise in this project is that going from And an image of a spine where you have the those ampereceptors which look very diffuse when you look with this enhanced resolution you see this Nanoclusters of receptors that seemed very concentrated with the In that spine this is another image you're looking now at endogenous receptors with a variation of the single Molecule methods which is called storm. This is the classical diffraction limited whitefield image of ampereceptors Distribution, this is the super resolved storm image You see it's extremely spotty and if you look at the individual spines here in the boxes You see that in in all cases the receptors are very concentrated in those small clusters These are 3d renderings of these of these images those cultures here Those hippocampal cultures we are using are very flat So 3d is not so impressive, but you can see in any case that these are really individual clusters So what are the properties of these clusters? They are very homogeneous in their size With this is the distribution of their size obtained from Gaussian fit with a size in the order of 80 90 nanometer Diameter what's extremely valuable is the actual number of nanodomains per spine and actually so looking at those properties with various Superresolution methods including this the This other stead method You see that the the number of nanodomains per spine is very viable on average It's around one and two nanodomains per spine, but it can be none or Up to five six nanodomain per spine and I'm not showing you the data here But what what we found is the actual number of nanodomain per spine is Extremely proportional to the actual overall size of the of the spine when we try to submit that that data we actually had some resistance from a variety of reviewers and Because this was this was very different from what had been seen previously By electron microscopy data and so it was thought that these this type of organization We are seeing might be an artifact from the technique the optical technique We were using and so we are pressed pretty hard to to Use EM to look at this type of data And as I showed you previously in fact most of the previous EM data Hadn't seen this type of organization of the receptors which were which was a bit surprising But we actually found the the solution for for this It was just a method a matter of the labeling method that was that had been used and so we went on and used pre-embedding it's actually life-sale labeling of surface amper receptors using a new antibody very efficient for the extracea domain of These amper receptors for the grade 2 subunit and there we found immediately that indeed Those receptors at the M level were localized in these in these clusters Here, it's not all synapses. You see some synapses exhibit some Exhibit diffuse labeling but in fact most of the spines do exhibit this type of of organization and if you look at the size of these clusters They are more about the same side. We found with Supervision optical microscopy Some more data on these clusters something where we've been pretty interesting interested by is actually understanding And determining what's the number of receptors that are present in each of these clusters? And so that's one attempt we've made at understanding that So this is a cluster here and in those images the storm images you see also these individual spots here and We thought that this could actually represent individual receptors so what we did is Take those Smaller spots look at their distribution of fluorescent Intensity we found that they were actually perfectly fit by two gossians and this could be Coming from either from the fact that they're actually in these receptors two subunits of this grade two We are labeling or the fact that the the antibody we are using have two fluorophores We haven't quite deciphered the reason why Anyway, what was important is that actually all these single spots here We thought could correspond to individual entities and we just then took the overall intensity of individual Nanoclusters and divided them by the median intensity of these single single spots and we found on average that each of these Nanocluster Had about something like 25 of these individual nano objects And so it they was suggesting that each of these clusters had on average about 20 to 30 receptors what was pretty interesting is then if you compare That number to the number you get from what's called miniature excitatory post-synaptic current that means the the number of receptors that are activated by a single vesicle You find on average in these cultures For these mini currents an amplitude of about 20 pico amps that corresponds More or less to the same type of receptors So there's a good correlation between the amplitude of the number of receptors activated by a single vesicle released and The number of receptors you could have in each of these single nanoclusters So that's about the amount of data we have so far on on the organization of these of these receptors And the question we wanted to ask then is what's the actual? What's the impact on synaptic function of this organization of the of the receptors and so we thought that actually modeling Could actually pretty be pretty informative to try to to understand what could be the impact of this receptor organization on synaptic transmission So we actually took a model that we had developed Based on some public published data or some published models already a while ago Taking a pretty classical Kinetic scheme of amperiseptors As you all know so amperiseptors are those pretty pretty complex tetrameric structures They are activated by the sequential binding of two glutamates that bind here in the ligand binding domain What's very important is that they can either open go to this open state here Or go to a series of desensitized states and that's actually there's a very recent paper Just came out a few weeks ago on some new crystal data, which actually can assign The structure of these extra sero domains in the desensitized states to various States kinetic states here. Basically. What's happening is once glutamate has bound here There's a big dissociation occurring in the end terminal domain here That really changes very dramatically the overall organization of the end terminus and brings it in various Desensitized states some of them being pretty stable and we'll come back to them so we use that kinetic model based on the rate constants that have been published in the past including including by us and started to look at The properties of the synaptic current Evoked by release of a single release of molecules we released Around three thousand glutamate molecules and we took a metric of matrix of receptors with a nano cluster here comprising various numbers of receptors as most of the time a five by five here and mimicked Recorded miniature EPSC Adapting the parameters here so that we get a good fit of the by simulation of the normal excitatory post synaptic currents and the first thing we found Which was actually already Published by Lisbon and Raj Ghaffari Ten years ago is that because those receptors have a pretty low affinity for glutamate which is in the order of several hundreds of micro molar The overall area over which receptors are activated is actually pretty small And you see the the full width have maximum of the Gaussian where receptors are activated is around something like hundred nanometer and We think that's very important because then it really means that depending on whether you release your glutamate Physical right on top of a cluster or next to a cluster is going to make a very important difference in the post-synaptic In the post-synaptic response, so we've been exploring that with having some interesting things first of all modifying the where you release the vesicle as compared to the location of the of the Nano cluster of course as you move away From the now cluster You're going to decrease massively the amplitude of the response because the extra cluster Density of receptors is much lower. It's not so surprising What what we've been a little surprised by is that actually? You see that you can actually do your release up to 15 nearly hundred Nanometer away from the cluster before having a very dramatic decrease in the amplitude of your current So probably and that's already telling us that there might not be to have to be a perfect match between the site of release and the location of the cluster to get an important response Then we've also been Looking at synapses with with several clusters and Looking at the impact of the inter cluster this Distance and of course you see that as clusters are more separated You have less smaller and smaller responses But probably the most interesting result we got from these simulations is that if we get that if we Take an inter cluster difference, which is the average of what we see in spines, which is in the order of 300 nanometer here And then you release your vesicle you see that indeed you get a pretty flat response Because as you move your release site in between the clusters as you lose and the response from one cluster You start to get response from the other cluster and so you see a very flat amplitude actually Saying that indeed if you have several clusters in a given PSD The actual location of the release site might not be so important with respect to the to the cluster location So that's basically where we are now we are doing extensive Experimental work to try to visualize the site of release with respect to the site of the cluster But with so far without success. So only modeling is left to us up to now to understand this this relation between site of release and cluster location So let's move on now To the more dynamic aspect of this This organization of the receptors and I'm going to start by showing you some of the data We have that actually demonstrates that those amper receptors are highly mobile and Then we'll go on to understand what's the impact of this receptor mobility on Fast synaptic transmission. So first of all these receptors are very mobile We found that by a variety of means starting by tracking receptors by very big latex particles more than 15 years ago now More recently using either Q dots or I would say more recently using small smaller tags But this is kind of the hallmark of what actually convinced the community that these receptors are indeed mobile Tagging those receptors by those small quantum dots, which are a few nanometer organic semiconductors Which are extremely photo stable doing video microscopy. This has been actually very Developed now and it's pretty easy to do you can see a whole zoology of movement of receptors Some receptors being moving very quickly some others being stuck in spines and Some receptors actually exchanging between a synaptic and an extra synaptic site and this has been this type of movies Have been very important to actually convince the community which was initially very resistant I would say to this idea that receptors are not all stable in synapses But in fact most of them are actually pretty mobile So these type of movies are very amenable to tracking and to to analysis because of this high accuracy you get from single molecule tracking and Probably the most important finding we made analyzing this type of data Is that so not only extra synaptic receptors seem to be mobile as you as you see on this track of extra synaptic Movements, but actually when we looked at the movement of particles which were co-localized with synaptic stains We also found a pretty high proportion about half of the of the tracks Which displayed pretty strong movements also and that was very intriguing to to us then by comparing the what's called a mean square displacement, which is on average The surface explored by the receptors Comparing the surface explored by receptors in the extra synaptic membrane and in the synaptic membrane for these mobile receptors We found a very big difference Not so much in the instantaneous movement, which is this Incident in this diffusion coefficient, which is not so different There's only a factor of three or four here between the move the instantaneous speed of movement of those synaptic versus extra synaptic receptors What's very different is the curvature of these plots, which indicates to you that those synaptic Movements here are extremely confined. So it seemed from this data that Synaptic receptors were mobile But we are moving and moving at pretty high rates, but we are moving in a confined environment and Actually, we thought even even us and more reviewers That this could be probably an artifact of tracking those receptors with those Particles here and in fact as I show you we are very much convinced now that in fact, this is actually true So this is the distribution of diffusion coefficients for synaptic receptors This is very systematic of what very representative of what we find About half of the receptors here are immobile. You see this is a log scale Very low mobility virtually immobile receptors But half of them here are very mobile and we are going to really focus extensively on these mobile synaptic receptors to cut a long story short In fact using smaller dies either palm or single organic die tracking We always find the same type of Distribution finding half of the receptors being mobile half of the receptors being immobile This is some of the data we've been obtaining using palm now Dynamic palm on life cells where you can track also the movement of the receptors You can find the whole diffusion map of the movement of the receptors In these neurites and if you look at on spine heads where the post-synaptic density is You see that we find the same results as we had with the q dots that the diffusion coefficient in the spine Here is slower indeed, but not so much slower just a factor of three or four slower that that in in the exosynaptic membrane and then finally By doing time-lapse palm we could actually combine the local is the identification of the localization of the clusters of receptors together with their mobility and what we found I think it's Very very interesting is that not surprisingly if you look at the clusters of receptors Receptors are immobile on these clusters not so surprising. They are compactly. They are compact and there But most interesting in between clusters receptors are extremely mobile and if you plot the mean square displacement If you look at clustered receptors, they are virtually immobile So that represents about half of the receptors, but the other half that's in between the clusters They have a mean square displacement that shows that they are extremely mobile moving nearly freely although With some more confinement than the exosynaptic membrane so altogether We are pretty convinced now that indeed you should see the PSD as a place where amperis receptors are either highly clustered very immobile or Extremely mobile in between the clusters and we start to have some evidence that they do exchange in between those clusters and outside of the clusters But I won't show you show you that today And then in terms of proportions on average, it's about half of the receptors that are mobile and half of them that are immobile So What's the function of this movement of receptors? So first of all the global Global function of these movements It's something I would say pretty classical. So I'm gonna go pretty quickly. It's a classical diffusion trapping model that you can compute whereby receptors are either trapped by post-synaptic molecules or freely moving when they are untrapped and This type of model actually recapitulate the overall data pretty well You can see receptors exchange in between a trapped state and a freely diffusing state And so using this type of model we can we could actually recapitulate most of the data we had First of all looking at the mid-square displacement at the variability You can get very similar data using these small or classical experiments You can model the impact of neural development with the increase in the synapse number in the dendritic shaft mimicking perfectly with this diffusion trapping model the decrease in mobility that you see over time and during the during development You can of course adjust the different parameters by comparing the experiments and the data finding a k-on and a k-off for this trapping of the of the receptors and then something which is probably most of most interest is that you can recapitulate some of the excitatory synaptic current fluctuations Due to this mobility and that's something that's been completely overlooked So as I'm sure some of you know you have variability in synaptic responses and this Variability depends on the amplitude of the response getting more variability on smaller responses than larger Responses and you can actually find the same thing Looking at the mobility of receptors and so we think that actually mobility of receptors in and out Synapses actually has a strong impact on this variability, which was initially only Attributed to viability in transmitter release So we think that actually viability in receptor numbers fast viability in receptors numbers due to fast diffusion actually has an impact on On the coefficient of variation of synaptic responses and that's I would say an experiment I like a theoretical experiment I like so looking at the viability on the number of receptors by simulation on these post synaptic areas and then mimicking an experiment We are going to go in detail right after by immobilizing receptors by cross-linking them with antibodies You see you suppress the viability In the number of receptors hands in the in the amplitude of the post synaptic response And that's also something we had found experimentally a while ago Looking at the viability of excitatory post synaptic currents when you immobilize receptors you get less viability So that was telling us and this is confirmed by those simulation experiments That's actually receptor mobility is fast enough to actually impact fast synaptic transmission and that's something that was that came as a complete surprise to us and Also to the community and I would say that's not completely accepted yet And so I would like to really develop a little more on onto that in the next few minutes To see how could actually fast receptor diffusion impact fast synaptic transmission Coming then to this last part What's the actual function of fast and per receptor movement in fast synaptic transmission and probably the most interesting finding we had on that aspect came In that paper we put we published now six years ago where By immobilizing and per receptors by cross-linking them with antibodies We found a very strong impact on short-term plasticity Basically short-term plasticity is this process whereby when you do a sequential High-frequency stimulation of the cell you see changes in the amplitude of the response here in those in that cell The the amplitude was similar in between the two poles and after cross-linking You see a big depression of the second pulse and basically the first pulse is not affected You only affect the second pulse And you can recapitulate that by a ion-trophoretic application of glutamate and so the question was really how could Receptor mobility impact this short-term plasticity And so to understand that we have to go back a little to the scheme of amper receptor activation that I've Kind of showed you in a more complex way just before so As you remember receptors are initially in an inactive state binding of glutamate They get open in a few milliseconds and then they get desensitized and that desensitized state is relatively stable And it takes tens to hundreds of millisecond depending on the receptor composition for receptors to recover from this desensitized state and this Desensitization actually contributes to what's called per pulse depression short synaptic depression and So the idea is that when you do this high frequency stimulation You see a depression of the post-synaptic response and that depression is due to the fact that Either you lose Transmitter or you desensitize the the receptor in fact in most cases in most synapses Synaptic depression has been attributed to loss of pre-synaptic transmitter and the the role of post-synaptic Desensitization has been a bit probably under looked for for a reason. I'm going to tell you in a second So this post-synaptic depression as you've understood I'm sure is the idea that when you release transmitter Receptors get desensitized very quickly and because it takes them a while to recover from desensitization If you have glutamate release on the same location On the a few tens of milliseconds after Receptors are still desensitized. So you get a smaller response and as I said This implication of receptor desensitization in per pulse depression is a bit controversial And we think one of the reasons for that is that what has not been taken into account in this type of rezoning is actually receptor movements So the idea we have is that indeed in this recovery from per pulse depression You have not only recovery from desensitization, but you also have recovery by exchange of receptors and the idea is that When you release glutamate you activate receptors they get desensitized but then those desensitized receptors can actually exchange by diffusion and be exchanged by Exchange for my naive receptors So that you get a faster recovery if receptors are mobile than if receptors are not mobile And I must say when we initially submitted this This data there was very strong reaction against it saying basically that it's absolutely impossible that receptors move fast enough to exchange in inside the PSD and Modeling was really very helpful at that time to actually show that it's a in fact it is possible You can have fast enough exchange of receptors To explain this type of recovery and the basic reason for that is dual first of all Receptors move much faster than when we thought and second the area over which receptors are activated is pretty small So I guess What we really need to understand and that's where modeling is going to be very helpful It's what's the impact between receptor organization and fast synaptic transmission and rates of exchange Trying to understand what's the role of the area over which receptor activated how how they exchange How how this is related to their kinetic states? So just to show you a few of our attempts to go into that to finish this this talk Comparing the organization of receptors And their impact in this short term plasticity. So back to the scheme Doing now sequential glutamate release at various interstimulus intervals looking at the recovery From synaptic depression either when receptors are immobile on where receptors are mobile first of all Computing the extra fraction of exchange receptors depending on the zone over which receptors are activated There is a very strong impact of course if you assume that receptors are activated over the whole PSD Like over 400 500 nanometer you you don't have exchange because diffusion although it's fast It's still a bit. It's not that fast. And so you you really have a cut off around 100 nanometer if you want to have a sizable exchange of receptors and so that's where actually the affinity of Empire receptors is going to be very important because Depending on their affinity the area over which they are going to be activated is going to be very different And then of course over time This depends the rate of exchange the fraction of exchange receptors depends over time Of course if you have a very big sign a big area of activation of receptors It takes forever before receptors are activated But you see if you have a hundred nanometer area of activated receptors You see that within ten millisecond you have nearly half of the receptors have had time to exchange between One state and the other then Looking at the impact of this diffusion on the recovery from purpose depression This is the model. This is the actual experiments You see they are kind of similar although not perfectly parallel yet you see When you cross link receptors or when when you bottle diffusion of of zero You have a slower recovery from depression One difference between the the model and the data you see is that in the experiments They do depress a little more But that's actually it's just a matter of adjusting the the parameters of the of the kinetic model of the receptors So I would say to finish We really think now that this exchange of receptors between the site of release and The rest of the PSD or the exosynaptic membrane is going to be having a very strong impact on this recovery And we think that's very important for a number of physiological processes I'm not showing you at that at all because I'm my time is over But basically we've been able to make a variety of different experiments That show that by regulating the mobility of the receptors you can switch from one state to the other for example, camcain is activation regulation of Empire receptor binding to PSD 95 really immobilized receptors and induce strong per pulse depression Reciprocally if you remove the extracellular matrix, you can actually accelerate receptor movement and get faster recovery so All together, I think we should really change a little our view of the the organization of the receptors and the impact of this Mobility on their function We think the the location of these ampereceptor clusters with respect to the release site is an important thing to understand We have no clue now whether actually receptor release occurs on the clusters or randomly anywhere What we do know is that receptors are very mobile in between those clusters and this mobility has a strong impact on this recovery from shortened depression I think that's going to be a pretty important venue to to explore in the future And I'm really convinced that modeling this whole as this whole aspect of signups function is interesting Just to finish I should really mention two very important people here for what I've told you about today Martin Heiner which is now a group leader in Magdeburg has really been the discoverer of the impact of receptor mobility on Short-term synaptic plasticity when you want to postdoc in my group and a lot of the modeling I've been showing you is done in collaboration with Olivier to mean which is in the Institute. I thank you very much for your attention Was with those elements also part of the model of the model of the miniature EPSC sizes or where you're actually trying to match it strictly to The distribution of nano of the domain sizes. Yeah in that in that work. It's strictly matched to the Variants of the receptor size actually It's tricky Strictly Looking at the the variance in the receptors So the distance from Thomas is not taken into account at all. This is a very crude and what's going on in the MPA receptors Is there anything like this been done for the NMDA receptors and if so How do you see the relationship? That's that's that's of course a very important question because NMDA receptors are extremely important for synaptic plasticity The problem is NMDA receptors are much more difficult to work with and that's although we are desperately trying to do that We are much less advanced into that What I can tell you is I mean the the preliminary data we have now Is well first of all there are much less NMDA receptors than Empire receptors nearly 10 times less Yeah, so there's nearly 10 times less receptors and we don't find them Organizing clusters so so far the data we have actually show NMDA receptors to be Rather randomly distributed over the the whole PSD And it kind of makes sense because NMDA receptors have a much higher affinity for glutamate than an Empire receptor So probably they are local their location is a bit less Important with with that respect in terms of that of their dynamics NMDA receptors are very interesting in the sense that they display a very broad Dynamics depending on their composition. So what we do know and that's actually published is that NR2A receptors Are extremely mobile. They are very very stuck to the to the PSD They barely move and are to be containing receptors are highly mobile not quite as mobile as Empire receptors But they do still exchange a lot and in fact, there is a paper published by my colleague rock Just recently in Embo just a few weeks or months ago showing that actually Changing the mobility of these and are to be containing receptors has a very strong impact on the LTP induction Because because of modifying that so so there are things to be done there sorry the interplay between the number of vesicles being released and the mobility and number of the receptors because it's really the sum of those two that should or the sum the Interaction of those two that should give you the dynamics and the plasticity And it would be interesting to know how that I mean, I know I'm asking a lot Yeah, no, no, no, I fully agree. In fact, I fully agree. I think that's one of the biggest point that we have not addressed yet in fact in this model and in the actual to understand those data that Mobility has an impact on short term plasticity one of the biggest challenge is the the frequency at which you you have Simulta you have really not to say you have sequential release on the same sign ups So we've we are starting to put that into the into the game But what really we what we really want to have is actual experimental data to look at Single release sites and that that's actually pretty difficult and I'm open to any any suggestion for that Because really I mean the idea there's a fierce battle in the field and I was just a couple weeks ago at another meeting where you have Completely opposite views from different people some of them saying that you know You cannot have More than one Hertz release at a given release site some other people saying but you can have up to a hundred Hertz release At a given release site and that that's gonna make a very big difference in the the actual data Yeah, yeah, absolutely I'm not sure the paper is out but Martin Heiner has actually a paper in press about that comparing intern neurons and So shaft synapses, I would say and the spine synapses which is also a thing and so It's slightly disappointing in the sense that we thought there would be a very big difference There is a difference. So in fact the Receptor exchange seems to be a bit faster in the in the intern neurons, but it's not as big as we thought initially I My view now is that it can be nearly completely explained by the shape of the Of the spine by the spine shape which which does induce some confinement with respect to the shaft synapses Yeah, that's actually a big big part of my group is actually working on that So we do know a lot of the molecules which are involved in stabilizing the receptors piez 95 Oxylery subunits and all that What we don't know and it's actually coming a bit as a surprise. We don't know what's holding the receptors together in these nanoclusters We thought it would be very easy. I mean we thought it would be tarps and piez 95 and it doesn't seem so easy So so I in fact, I don't know one idea I have is that there might be some lateral interactions In between the the end terminus of the door of the receptors that may help because those ampariseptors are really a tower I mean they are very different from the image we had of them Some years ago before the crystal structure. They're really a big tower And so if they're in those clusters, they are very compactly. They are very packed And so certainly Interaction in their end terminus domain could play a role in that the one thing we are very surprised that by is the Extreme homogeneity of the size of the clusters, you know, they are they are much more homogenous that what theory could Would if you just model the diffusion trapping Organization you would have a much high much broader Variability in the size of the clusters so there has to be some type of additional interaction that holds them together And we haven't found the the mechanism yet