uh thank you to my very dear colleague Greg Jeff for nominating me and to the colleagues who wrote in support of um of of the nomination um and thank you all of you for coming tonight and also for the people who have uh joined online um I'm going to continue and start by thanking people uh I've I've never done science by myself I've I've done all science part of the team and so I'd like to acknowledge all of uh members of my lab past and present who contributed all of contributed to the body of
work work that uh this award is for um I thought i' also list list uh all of the people who I uh published papers over the years just to drive point the home of how International and collaborative Sciences there's more than 130 people listed here uh from 25 different nationalities uh and I'm going to single out one person which is doing Z which was my scientific adviser in Portugal who took me into his lab when I was a second year medical student a rather useless one actually um and and he just
let me have fun in the lab um he taught me how to plan experiments think about experiments read papers think about problems and then he also really pushed me to come to the UK to join the welcome trust PG program at UCL that every that will run for many years which was a really important step in my career now this lecture is about neurons synapses and the Instinct of survival as John said and um and I'm going to spend a little bit of time introducing to you uh instinctive behaviors which are beh
aviors that animals do without prior experience they are behaviors that animal doesn't have to learn uh and therefore they constitute an expression of innate knowledge the animals are born knowing how to do these behaviors and um and they can be very simple a simple reflex or they can be complex series of actions but they all have in common the fact that they have um evolved to promote survival there are many things that animals have to do in order to survive and evolution has hardcoded some of
these things in the DNA to to give animals a head start when they are born um and to illustrate some of uh these instinctive behaviors I'm going to show you um a video a sequence of videos from uh that the BBC Natural History unit has very kindly um shared with us um and the video should be coming up in a little bit and we have the video please uh there we go so this video uh and shows follows uh one of the most uh remarkable instinctive behaviors uh in the animal kingdom which is the the salmon
run where every year there are millions of Salmons that leave the ocean to go and find the river that where they were born to spawn and to eventually die and uh the particular summon run we're going to follow here uh is uh happens in the northwest coast of America uh that it starts in the Pacific Ocean and then salmon will migrate about 5,000 kilometers to the rivers of British Columbia and not only they have to navigate to the coast they uh when they get there they have to navigate these Maze
of rivers in order to find the exact one where they were born and which they can do with really remarkable special precision and so this illustrates two key instincts the Instinct of navig or the Instinct of migration where animals migrate over long uh distances to find uh places with better resources here the resources are the shallow Waters of the rivers which are safe place to with eggs and also the Instinct of navigation animals are very good at knowing where they have to go and and getting
there now this journey takes um many months and during this journey they will face a lot of challenges uh and the first challenge that they face are coastal predators and the most deadly one is this is the baldheaded eagle which uh captures and predates on Fish with really remarkable a remarkable degree of skill and precision and this illustrates one of the most important instincts in animal World which is pration which shapes all of other behaviors the top priority of all animals is not getting
killed you might go out to find food and if you don't find food that's fine you can come back the other day but if you fail to avoid a predator you not you're probably not going to come back the next day and this is the topic of today's lecture uh avoiding Predators now for the fish that make it and make it past the Predators and there's there's a lot of fish that make it past the Predators uh the challenge is not really over yet because this is what they face right it's the proverbial uphill b
attle uh where fish will have to swim upstream uh jumping over Rapids and uh rocks uh but which they can do with really remarkable degree of physical press um just like they if they have been practicing this all their lives but they haven't this is the first time they're swimming Upstream uh yet they do it and they do it because it's the result of millions of years of of evolution now there's another problem however is that there's some bears are waking up and they're really hungry um and so gri
zzly bears have been hibernating here at the top of these Snowy Mountains they have given birth and now they are going to start a journey on of their own that will last many months um to go and meet the salmon run and and get some food um and they do this and this illustrates first again this this really important Instinct of navigation where animals are really good at knowing where have to go and and getting down there but also another one which is you know these bear cubs is the first time tha
t they've been out of there then and yet here they are uh going down a really Steep Hill you know perhaps not as uh elegantly as um as their parents certainly much more elegant than me going down a ski slope um and but they do it right and and they do it uh and this is a very scary descent but they do it um because they have to in order to survive right [Music] uh and this so this journey will take them a lot of months and then they will eventually go to The Rivers where uh the salmon are and wh
at they have to do well the first thing they find is that uh there's a bunch of other bears that had the same idea and they'll have to fight for position here they are um uh but then they have to catch the fish right and catching a flying fish very slippery fish U that's flying past you is not easy uh and this illustrates another very important aspect of instinctive behaviors which is uh these animals might have been born with a drive to catch fish even maybe a set of basic actions to catch thes
e fish but then they have to refine this skill over years through experience um bear cups for example are really bad at this they can't do it and this illustrates you know this is one this is true for most instinctive behaviors they provide a basic blueprint that sets a basic set of AC action that is then refined through experience to give animals the best possible chance of adapting to to to the version of the world that uh that they were born in um now a lot of fish do make it pass it and then
eventually they reach the shallow Waters here uh of the rivers that they were born and they'll engage in a more peaceful set of behaviors uh courtship nesting uh and eventually spawning and um and the the start of of a new generation which will uh grow up here for about a year a year and a half go back down and then four years later come the all the way back up all right so now the these behaviors um are some of these behaviors are very species specific right not all species travel 5,000 kilome
ters in order to breed but there's one that I mentioned that is really common to all species which is avoiding Predators uh and and this is because uh from the moment I was life uh there was uh there were always species predating on each other and a very strong evolutionary pressure to develop behaviors or evolve behaviors to avoid predators and one of the most universal um behaviors for avoiding Predators is just to run away to escape which is a behavior that has been reinvented time and time a
gain over evolu um during Evolution and at its most basic level Escape is a locomotive action that moves an animal away from a thre ideally toward safety but because the enormous diversity of uh species of their habitats of the Predators uh Escape can be very varied right you can escape by flying by jumping by running by swimming but despite all of this diversity Escape can be decomposed into a set of of common elements if we think about it all animals the first thing you have to do in order to
escape is they have to detect a threat they have to use their senses to figure out whether there's a threat there or not if there is a threat they have to choose or decide what to do next they might decide to escape um or they might decide to do another defensive action such as uh stay very still to avoid being detected or even engage in the fight if the Predator is too close if they do choose to escape then they will exe they will have to execute this Escape as fast and as accurately as possibl
e they should really know where they're going not getting cornered for example uh and then they have to terminate the Escape ideally as soon as possible as well why you escape is very costly um not only in terms of energy but also be cost of missed opportunities when you're escaping you're not doing many other things that uh you also need to do in order to survive and as John mentioned um all of these processes are modulated by experience uh for example you might learn that something that you th
ought was really scary is not scary anymore or you might find that there's a new safe place in the environment that you should get to that you can get to more fast for example or faster that you should use and the goal of our research is to to understand how the brain implements all of these steps and we think that the answer to this might be interesting because not only we learn the biological bases of of instinctive behaviors of behaviors of survival but also because many of these steps are co
mmon to to many other behaviors for example the decision to escape uh is a decision but we make decisions uh uh every day we we choose actions every day we execute actions every day so maybe um by understanding how the brain implements this process in a simple Behavior we can learn something about how the brain solves wider problems in general now ideally we'd like to understand the human brain uh because it's the brain that we care about the most it's the most powerful thing that Evolution has
produced and it would be great to be able to fix it uh when uh when it goes wrong our research however uh is done at the very cellular level we want to understand the biological processes behind each computation uh and this means that we need to get quite messy with the brain we need to do invasive recordings uh that give us the the prec the spatial and temporal resolution that we need um and we want to take things out of the brain put artificial things into the brain and clearly this is somethi
ng that we're not going to do in a living human so we go down the the phenetic tree and we stop at the mouse uh where we can be a little bit more invasive while making sure that the animals are happy we go to great great extense to ensure that all animals are very happy but where we can have this level of resolution and precision that we need now the mouse brain is is smaller than the human brain uh has about a thousand less times the number of neurons uh but if we if if I slice these brains now
like this you're seeing the brains from the Top If I slice them and then flip them and I blow up the mouse brain you can see that the anatomy is different but actually many of the regions that exist in the human brain can can be Iden can be identified uh in in the mouse brain and for most of the stalk Al going to be focusing on a part of the brain called the midbrain which has a very clear equivalent between the human and the mouse and if we look at the neurons actually of human if you compare
human neurons with mouse neurons they're actually even more similar so their physiology the way they work is very similar their Anatomy is very similar the main difference is that the mouse neuron is is about a third of the size of the human neuron that these ones here are drawn to scale all right so for you to be able to follow this lecture I need to explain to you how a single neuron Works which is something that I think we understand really well uh after more than 100 Years of research uh and
which I'm going to try and summarize for you in about four minutes um so this is a cartoon of vura the first thing you can notice is that a very asymmetric cell type is very non-uniform uh it has that the Center cell body the main role of the cell body is keeping the the neuron alive um and then at the top uh has this protrusions that form a a a tree like structure that John already mentioned to you that are called dendrites and dendrites the main job of dendrites is receiving input um and the
input arrives into neurons uh via synapses synapses are um specialized Junctions between neurons and so the dendrites have a bunch of synapses to receive input from other neurons at the other end we have this long thin protusion that is called an axin which is responsible for generating output and passing information to the next set of neurons and at the end of this axon there's also some copses so a neuron has copses at the dandr to receive input and cabs at the axent to transmit uh um to trans
mit output uh to the next set of neurons now the most important property of neurons is that there are electrical devices there's a membrane potential uh between mem potential difference between the outside and the inside of the mirror and so if we look if we record this potential when the when nothing much is happening um the nerve in it's sitting is kind of bobbing around what we call the the resting potential is just kind of sitting there now if the nerve starts receiving a lot of input uh the
potential changes and at sometimes at some point is going to hit a little threshold that is indicated here by this dotted line uh and generate what we call an action potential which is a very large and brief change in the memory potential which is responsible for passing down information so these action potentials are going to activate the synapses um H where is this oh this is tricky wow okay uh yeah anyway the action potential goes like this and then activates the copses um so what happens at
the syapse uh this is a blown up of the cups you can see the aent you can see the dandr and the main property of the syapse is that it has these vesicles that are filled with molecules uh of neurotransmitter uh which are typical amino acids for example glutamate and when an action potential arrives at the axin the um it causes this membrane this vesicle to fuse with the membrane release neurotransmitter and that and it's going to then act on receptors on the side of the dendr and cause a potent
ial change at the dendr so the result of this is get you get a potential down the axent you get this little magic of synaptic vesicles and then the potential the potential propagates to the next axin now one important very important aspect about copsis that I'm going to need you to remember for the duration of this tol is that most of the times an action potential comes down the accent nothing happens because of biological constraints the probability of actually releasing releasing a vesicle is
very low for example uh in the cortex the most advanced part of the brain you need about five Action potentials to release one vesicle so the probability of release is about 20% and during my PhD working working with yuk Goa and Kevin staras at UCL we developed some methods to to measure this release probability directly uh using a bunch of different techniques including electromicroscopy and this what I'm showing you here is a reconstructor of a cups in the hippocampus where you can see the dan
dr and the axent and the synaptic vesicles inside the axin and what we did was to develop methods where we could deliver a fixed number of action potentials to this to the signups to a single signups and then turn the vesicles that have been released black literally black and then we can go into the electric microscope and just count the number of vesicles we knew the number of action potentials so we can compute directly the probability of release and this using this and other methods what we f
ound is that the release probability can be is a very Dynamic Property of the neuron that can be changed up or down uh depending on what's going on in the neuron if the neuron is really active speaking really loudly release probability goes down if the neuron is really quiet release probability goes up to make the neuron louder and so it keeps it in this dynamic range where it can integrate and pass down information um so what happens when a vesicle is released so I told you that uh vesicle acts
on receptors causes a Chang in membrane potential but the catch is here that the change in membrane potential is very small and so in order for the neuron to actually generate any kind of output you need to add up a bunch of these uh syap vesicles together and um this defines the way neuron integrates this information defines the input output function of the neuron so what is the neuron do with the input it receives and here for example I'm illustrating you a straight line which is a linear inp
ut output function meaning that the neuron just adding up Sy upses at a constant rate and during my post talk working with Michael Hower and Beverly Clark one of the things that we set out to do is measure this input output functions in a bunch of different neurons and what we found is actually if if you look at the synapses uh here we when we shifted copses we shifted we decided to activate a bunch of copses at the very tip of the D right all clustered together what we got was that an input out
put function that was really super super linear so uh we we started getting much more activity than you'd expect if everything was linear suating and this was exciting because we did a bunch of more experiments to show that this kind of property allowed neurons to do very powerful computations such as um discriminating between temporal sequences and we were able to actually pin down the exact mechanism to the activation of a specific iron Channel called an nmda receptor and this together with th
e work of many other colleagues um kind of painted a picture of neurons as really powerful computation devices that can generate a variety of input output functions that means performing a bunch of different computations just depending on the type of uh uh the type of ion channels that they Express and also the spatial temporal pattern of activation of their copses so this is this is how a neuron works at least I think so um so now that you know this how does these neurons how do these neurons g
et together and work together to compute escape and this is what John is alluding was alluding to in his introduction and it's it's it's really spot on we know a lot about how neurons work but what we need to know now is how are these properties actually used to compute anything useful in the living brain right how do they serve behavior um and so I'm going to walk you through this and I'm going to tell you two stories one story about the decision to escape how do mice decide to escape depending
on what's going on in the world and then a second one on mechanisms of how mice navigate to shelter and the first this first project was led by uh Dominic Evans and Vanessa Temple and for all the experiments we do that I'm going to show you today and for most of the experiments that go on in the lab we take mice that we U put in an arena which is about it's kind of a round table about about this High uh has to be high enough so that they don't jump out um and then these might have never been in
this Arena and the arena has a little Hut at one end that mice find through the natural exploration they get in there and actually quite like it and from there they start making kind of out ings to explore the rest of the environment and in some of these outings we present them with threats and John already mentioned one threat that we like to use is a kind of a a shadow that that comes from above that mimics may be a predator maybe an object in collision course but that mice and actually all a
nimals including humans innately re react to so these animals I've never seen the stimuli but they're going to react to it and this is an example this mouse is going to see one of these luming stimuli and right and then turns around goes to the shelter that's the behavior you want to understand it's pretty simple uh this is U the behavior rendered from above so the blob so this blob there is the mouse and then it goes all right so the the first thing we do so we want to understand how animals co
me to the decision of escaping or not so the first thing we did was a behavioral experiment where we varied the contrast of this of these threat stimulus and this is kind of the equivalent of a mouse has having to figure out if there's some bird of prey coming at it in when the sky is really clear and it's sunny versus when it's cloudy it's rainy and they can't really see properly right so we're just going to vary the saleny of the threat and to see what happens to the decision uh so I'm going t
o show you this is the same Mouse exposed to to three different contrasts and the videos are aligned to when the stimulus is presented okay so my mouse is coming out of the shelter and when it gets to the end here he going to see the stimulus this see what happens right so you can see that there's a very clear modulation of of the Behavior Uh when the contast very high Mouse turned around ran straight to the shelter when the contrast is very low it really took its time and it had to see the thin
g five times in order to be convinced that yeah I should get out of here um and so if we plot the this data so the probability of escaping goes um up as the contrast increases W this is really funky it's extremely nonlinear as well W uh and the re and the reaction time goes down so when a contrast is really high uh the U they react really quickly and so the next thing we do did was to to model this to create can we come up with a simple model that describes this and what we did was to actually t
ake models that have been uh developed for human decision- making to model human decision making and apply it to this to this um to this behavior and and this type of model is called the drift diffusion model or a diffusion to bound where the gist of it is that there's a threshold for you to make the decision and you keep integrating evidence towards or against it right and so you're sitting there thinking about okay shall I go or shall I stay and at some point if there's enough evidence then yo
u cost the threshold you go if you never cost the threshold you never go and so what what the way we model is that we thought okay the animal is integrating some sort of threat and the threshold is a threshold for escaping so when the contrast is really high uh this variable uh the threat variable Rises very quickly and hits the threshold very quickly and so you always escape and you escape very very early on if the contrast is low it'll take some time kind of to drift up so increases and then i
ncreases a bit and then increases and if you if you see enough eventually you you hit thres old uh and and sometimes you don't get it and sometimes you do when when you do get it you you hit it late so that's why the time to react is is uh is longer so then we went into the brain to try and uh understand how the mouse brain might be implemented this type of model and we started in a place called The superus as I show you uh as I mentioned before is in the midbrain it's at the back of the brain i
t's a region that's very well studied it receives direct input from the eye which is important for us because we're showing them visual stimulus uh and this input arrives to the superior ccul via these cells called retina gangler cells that make synapses onto to some of the neurons in the culus uh and work from many colleagues has shown that there are neurons in the superi culus that are sensitive to looming stimuli so they react very strongly to looming stimuli so we thought this is a pretty go
od candidate for processing uh this type of threat the second region we focused on is just below it still in the midbrain uh which is called the peral gray I I might call it P sometimes and superal I might refer to as SC sometimes uh and this perative FL gray sends accents to the motor centers so it activates motor commands and it has long been known to be important for defensive Behavior so if you electrically stimulate it you you get you get a mouse to jump and actually interestingly in humans
um humans um if you put an um a human in the scanner and you kind of uh you scare it the scare here is uh the the threat of receiving a little electric shock that is kind of painful so you don't want to have it and if the threat is really imminent the peral gray gets activated also if during a neurosurgery you activate the perial gray the patients report a sense of dread of being chased like a panic attack and also patients with post-traumatic stress disorder have an active overactivation of th
e pactical gray so this is clear a reason that is very important for defensive behavior and it has a clear homology between humans and and mice so we're going to see what happen what goes on in these two regions during escape and to do this we um we use a very nice technique developed by col the Stanford uh where you can mount tiny microscopes on the head of the mice so the microscopes are about the size of a coin and we're going to take advantage of the fact that when an action potential happen
s so that large potential difference uh that sends information out calcium gets into the cells and so what we can do is we put a a molecule um that turns fluorescent when you see Cene okay so we can just sit there and whenever a cell lights up that means that the cell is active so we point our microscope to the super cicus we point the microscope in other animals to the partical gray and we use a virus to genetically modify the cells and express a calcium indicator uh and then we see what happen
s and so this is the activity profile in the superc so when a threat comes the activity starts increasing and then animal escapes and uh activity in the culus Peaks more or less at uh during Escape now when we look at what happening at the per in the peral gray the profile is quite different so during the part of the threat nothing happens and then you have a jump in activity just before the animal starts to uh starts to run when we look at trials or situations where the animal didn't Escape for
whatever reason for example when the contrast was low we see that the super is actually still activated and the P to Tool gray is dead silent nothing happens so if we think back about to our drift diffusion Model A decision- making model what he suggested to us is that the activity in the super culus might be representing the level of threat that is being thresholded and the activity in the partic ray represents the result of this thresholding operation this tells us whether you should go or no
t so to to further test this and to establish a causal link between these activity profiles and the behavior of the mouse uh we used another technique called optogenetics where we can play the same tricks where we use a virus to genetically modify the cells but now we're going to express a new molecule that makes the the cells sensitive to light so that whenever we shine blue light we can activate the cells um and the experiment we're going to do is so we express this optogenetic tool in both of
these regions and then we're going to activate um a small number of cells a bigger number of cells more and more and more and see what happens to the behavior and we can do this just by changing the intensity of the laser that we apply to the brain um and so what I'm going to be plotting here is the speed of the mouse these lines are kind of different Trials of of the speed of the mouse during thre presentation uh and if the speed is above this line here it means that animal has escaped and if
it's not above this line it means that animal is just kind of walking around so we activate some cells in the ccul nothing happens we activate more cells you got one Escape we activate more we got more escapes and as you keep cranking up the stimulus we get progressively more and more and more escapes and so if we plot this the probability of escaping kind of smoothly goes up as the the laser intent as we activate more and more cells in the ccul and the smooth curve is really similar to the curv
e that you get when you actually lever the the the real threat stimulus now let's do this for the periodical gray we activate some cells nothing happens we activate more cells nothing happens we activate the next level and boom you always get escape and then again always escape and always Escape so we B is we get a really steep curve meaning that you either don't get Escape or you get Escape all the times and this kind of activ this activity first first we learn two things learn one thing which
is the activity in these cells is really closely related to is sufficient to cause the animal to Escape right these animals are not seeing any threat which just hacking into the system to make the animal run but it also tells us cement our view that the activity in the super ccul might be the the the variable that being thresholded and that the the the activity in the PG is represents the command to initiate escape and because of it we thought okay maybe the thresholding that line actually sits
between the superal and the paral grade so the first thing we then wanted to know was whether the two uh these two these two populations of cells are connected to each other and to do this the first thing we did is use a really wonderful technique developed by several colleagues including maret rep part who's here uh called where we can use a virus to infect a group of neurons uh here we've infected a group of neurons in the paral Gray we've turned them blue and then what this virus is going to
do is going to travel back up one syapse and one cups only and turn the other neurons uh in a different color right and so um we're going to ask whether we see we get neurons there um are in the superior fcul and indeed we do uh this is one example and we can see that all the magenta neurons are the ones that give input to the blue Nance and this tells us that the superior cus gives a lot of input to the parac gray so the next thing we did then was to record from from these individual neurons so
um we're going to record what's going on inside the cell um and we're going to express our oogenetic tool in the superior ccul so we can activate the ccul while we're recording from the perod lorian era so this is Illustrated here we're going to shine light we're going activate the pink cell and we're going to recall what's happening in the blue cell which is the p and this is what happens the first thing that this tells us is that okay yes they are connected but this is a tiny really crappy re
sponse and the reason why this response is very small is because the cops have a very low release probability so remember what I told you most of the times when action potential comes you you don't get any release and these cses these sys have a really really low release probability now when you activate this um the sign up connection a ton of times you can overcome just by uh because you have a lot of them you can activate uh you can generate a lot of activity in the downstream neuron and and w
e this is we think this is pretty cool because um this is the equivalent of um what Happ this is seeing what happens in in these neurons the equivalent of low contrast being the single input so if if the contrast is really low nothing much happening because the release probability is really low and in order to actually uh get a lot of input and get animal to escape you have to have a ton of a ton of uh activity in the super ccul that you only get when you have really high contrast stimulus and t
his is basically a biophysical implementation of this thresholding operation that is done at the synoptic level right and so the final thing we wanted to do here for this this project was to test whether actually this pathway is really necessary for the behavior well because that's our model right and so the prediction is if we block the CTIC connection the animal should not Escape so in order to do this we use yet another trick uh where we this time instead of using light to change the activity
of the neurons we use the tool um where we can it's called The chemogenetic Tool where we express another molecule in the pink neurons in the spheric neurons so such that makes neurons sensitive to a drug so when they see the drug the neurons get silenced and so um here we're just going to check if this tool works so we're recording activity in the culus we see these these Action potentials oops Action potentials and then this is what happens um in the peral gray so the cells are connected and
then we're going to deliver the drug we're going to block the synoptic connection and then we can see that the response is gone so what happens to the behavior this is a mouse where we doing this block now comes the stimulus and this is what happens so the first thing you should notice and the most important thing is this animal is not escaping right um and so this tells us that uh this sign pration is really necessary for escaping but there's another interesting aspect here is that the animal a
ctually reacted to the stimulus that reacted with a different type of defensive response this animal is actually freezing which is interesting because it suggests to us that actually Coptic connection is uh you know it's necessary for evoking Escape but the threat information is actually being rooted to another circuit that that can initiate defensive freezing so that when we block a scape uh it goes by another pathway that generates freezing and so the summary here um is that we have um we thin
k that that's one the core of a circuit for computing Escape decisions is we have this layer of superior ccul neurons that integrate thre information uh we have the PG neurons that initiate or command initiation of the scape and then we have this threshold that is implemented by a synaptic connection and the key here is a very low release probability that implements the this um this threshold now once you decided uh um to escape you should really know where you're going um and in in mice I've al
ready shown uh several times that what do is they escape to a shelter it's a very sophisticated thing to do because shelters are by definition places where the attacker can't get to it can get to you and so they offer long-term survival uh and so we decided to investigate this process and actually understand the mechanism by which my navigate to shelter and this work that was started by huin Val the first thing that hubin did um was to note that I've showed you this already uh you scare a mouse
and it runs in a straight line to the shelter which which is a trivial finding um in many ways but so regardless of whether you use a looming stimulus a loud sound you put the looming steam in different places the animals always run in the straight line to the shelter and this will become more important now what we wanted to understand next is well how do mice get get to the shelter what strategies they use to navigate to the shelter and um mice can you can mice and people can navigate to uh to
places using a lot of different strategies and perhaps one of the simple things that this animals to do was to look to see where the shelter is and just run there right so we did a lot of experiments to test whether this is the case but I'm going to show you the result of one which I think is very informative which um we let the mouse know where the shelter is it goes there escapes there and then when one of the times when it comes out we sneak up on the mouse and we move the shelter to a differ
ent place right and now we're going to see what happens so the shelter is now moved the mouse could just look at it but we scare it and and this is what it does it goes to where the shelter used to be and kind of stays there hanging around searching um and it um and it does not ever go to the new one so this is uh an extremely robust effect that all when we do this rotation experiment all mice will escape to where the shelter used to be I not where it is we've done lots of other experiments like
doing this in the dark for example where animals can do perfectly fine uh and the overall conclusion from this type of simple experiments is that we think that are forming a memory of the where the spal location is of the shelter and use this memory to to navigate there now the the spatial world is not as simple as this uh it has often lots of complications in order to get where you want to go and one of the complications we we thought would introduce is a very simple one but it's just a simple
obstacle between The Mouse and the shelter and then asked uh this is worked by then by Philip shamas who then ask well how do the mice behaving under these conditions and the first thing we found is that with a little bit of experience mice mice are pretty good at this and instead of now turning and running to the shelter what they do is they turn to the edge of the obstacle and um and get to the shelter which is the most efficient way of getting to the shelter so you target the you target Edge
you turn and then you go to the goal so the idea is that they might be using this this Edge as a sub goal they know that they need to get there in order to reach the final goal so how what strategies they use to navigate in this situation so again we did several EXP expent but we did one which is kind of conceptually similar to the one where we we move the shelter which is let animals learn that there's a barrier there and then we're going to lower the barrier and then we're going to see what t
he animal does and the idea is that if the animal could see that the barrier is not there they should just run straight right you can probably guess what happens come on Mouse there you go um the thread's going to come and the mouse goes around as if the barrier was there but it's not there anymore all right and um again this is a very robust finding uh so here we're plotting just the the the trajectories of the mouse the blue lines is when the mouse targets the obstacle Edge there's lots of blu
e here uh and when the obstacle goes down there's still quite a lot of blue here a lot of them still choose the obstacle Edge and this suggested to us again that they're using a memory based process to know where the edge of the obstacle is and so overall if we look carefully at at both of these behaviors um we what we what what happens is that the very first thing the animals do is they Orient where they have to go so when there's a shelter the first thing that the mouse will do is Orient to th
e shelter and then run and this is why the trajectory Escape projectors are are are all in straight lines because when they start running they're already facing where they have to go right and this is a good strategy right because if you started running while you're facing you kind of end up going kind of around about so you turn in place and then you go um and this is true also for when there's an obstacle when there's when there's an obstacle they target The Edge and then they go all right so
so the basic blueprint for escaping is turn and then go and so this suggested to us that mice are probably keeping constant track of of of a vector of where they have to go and the vector is defined by by the angle between the heading direction of The Mouse and the shelter so how much do I have to turn to in order to get to the shelter and the distance how much do I have to cover and then important observation here is that we never told these mice that they were going to get scared right and whe
n we do it uh they immediately go to the shelter so during exploration uh it means that there's a very strong drive for mice to know where safety is an innate Drive they they found it as a safe place and they they will always keep track of that in case they get scared and this also makes the prediction that if we're going to look inside the mouse brain we should see a representation of where the mouse thinks the shelter is um during the during during exploration um so we decided to to see if tha
t's the case and we know we know we know a lot about special representations in the brain J Keef got the Nobel Prize for finding that there are cells in the brain that represent where you are at any given time point they're called play cells uh and we could have started there and try to understand how these systems might be representing where shelter is um but to to tee joint a little bit um we we ignored BL cells for the time being um and and actually instead we decided to uh to start from the
action side and our reasoning was all right the mouse first thing the mouse does is turn the head to where it has to go so let's look at the place of the brain that makes the head turn and then ask what makes the head turn to the right place all right now in mammals uh so I've told you a lot about the superus and everything that I told told you about this kind of uh in this middle part middle part of of the culus but the super culus is uh better known actually for its role in orienting the head
and orienting the eyes and even um orienting the body sometimes and this is when the head moves in this direction uh it's um carried out by this more lateral part of the of the culus and there's a lot of work showing that so we and this is true for all vertebrates from lamp R to to humans and so we we thought okay the super is probably doing the head turn so what tells the ccul where to turn the head to now we know that it can't be sensory input because we I've already shown you that this is a m
emory process so it can't be input from the eye for example and so we looked for places that could convey some sort of spatial memory into the ccul and based on Anatomy we focus on this part called the r spinal cortex that I might refer to as RSP that is known to be very rich in special representations and since accents Str straight onto the super cus and so this is work led by hop and Val and Di comper and so we went then to look for spatial representations of where the shelter direction is whi
le the mouse is exploring so this time we really wanted to have very high special Precision so instead of using the head mounted miniator microscopes we used a silicon probe that we can also uh stick down um stick down the brain and record uh the voltage directly with very high tempo Precision um this is what it looks like if you kind of slice the mouse brain coronally like this uh and this is um uh where the probe was and with one signal probe we can record activity in the Ros spinal cortex whi
ch is right above the cicus and in the cus itself and so we're going to ask whether there are neurons that have information about the shelter angle which is the the Ang angular distance between uh where that Mouse is heading and uh and the shelter and we Define here zero as the mouse is heading at the shelter and so indeed we find these cells uh what I'm showing you here is uh um the activity of a single neuron um as a function of where it's facing and and each dot here represents one action pot
ential the point that you want you to take from this is that there's a lot of yellow dots there's a lot of spikes or action potentials when the mouse is facing the shelter and this is represented here so this neuron cares the most about tell this neuron tells the mouse when it's actually facing the shelter this is another way of representing this uh when this Arrow just tells you where uh which direction uh does does this neuron care the most about now to really make sure that these neurons care
about the shelter and not something else like some something else behind the shelter what we do is we now going to rotate the shelter and this time we're actually going to give time for the animal to learn where the shelter is so after a while they actually figure out okay the shelter has moved I should start going there which we can confirm by letting seeing the mouse Escape there and then we're going to see what happens to the cells and what we find is that this this the this firing profile k
ind of rotates in space but the neurons are still firing when the mouse is facing the shelter except now instead of facing uh East uh they have to face North right so these neurons care about not Northwest Northwest or east they care about their relative position to where the shelter is all right now uh our model here uh that we're testing is we think that the cortex RSP tells the culus where shelter is all right so we're going to do an experiment where we're going to inactivate these cells thes
e neurons that projected the culus and see what happens to the information there we use the same type of chemogenetic strategy where we can use a drug to inactivate inurance and so we do this in the retrograde manner which means that we we're going to only be inactivating the neurons in the cortex this blue neuron here blue neurons here that s accents to the superior ccul all right um this is an example of a superior ccul cell uh that is tuned again to pretty much where the shelter is now we're
going to inactivate the blue neurons and this is what happens this this cell loses its tuning right so and this is a very again very robust effect about 60 or 70% of the neurons lose their tuning and this tells us that the superus really does need to hear from the rpal cortex where um where the shelter direction is now again similar to the previous project now this does this matter for the behavior because um uh so to test this we did again the same type of experiment this time we're actually in
stead of activating the whole neurons we're going to deliver the drug right on top of the copsis we can do using a little canula so we specifically in activate the copsis and we're going to see what happens to the mouse behavior so this is a a little guy that has had the sign uptic Connection in activated and what hopefully you saw from this movie is that he reacted when the threat came but instead of turning around and going to the shelter it just started running in a pretty random direction ri
ght uh and the reason why they do this is because they perform the wrong orientation movement but they Orient instead of orienting to the shelter they Orient to some random place and start running and then turn around start running and eventually they actually stop in the middle of the Arena without run reaching the shelter and and you know if this was a mouse being chased by a predator this would be the last mistake that this mouse does right so this tells us that this Coptic connection is real
ly fundamental for the animal to perform this correct orienting movement and Escape successfully now the final thing we wanted to do uh is to to know how how um how information about the shelter direction is trans is is is moved from the Ros spinal cortex to the superus so again we we looked at the connectivity first and actually I haven't told you about inhibitory NS but the brain has both excitatory and inhibitory neurons uh excitor neurons excite inhibitor neurons inhibit um and we wanted to
know whether the cortex is is connected to either excitatory or inhibitory urance in the sub culus or perhaps both so we use the same type of viral strategy to to to look at this uh first in excitatory Nance so we infect with a virus of excitatory neurs here and then we see where the virus goes and it goes to a bunch of places including the rosenal cortex we do the same thing for the inhibitory inurance and we see that the profile is really very very similar right so this tells us that the corte
x sends um its accents and makes s uptic connections with both excitatory and inhibitory inance we then went and again recorded from these individual neurons uh so we're going to record from the superior cicular neurons while acting activating input from the rosenal cortex um this is what happens when we record from exat neurons we confirm they're connected and the connections are pretty small we do the same for inhibitory neurons and I actually see these connections are actually much stronger a
nd so it's much easier to to actually drive the excit the inhibitory neurons in the SC to to to far Action potentials than excitatory ones and then we several more experiments we actually were uh found out that not only the retos spinal cortex send accents to the SC to the excitatory and inhibitory neurons the inhibitory neurons inhibit the excitatory ones so that is what we call the fit for kind of Lal inhibition the reason why this is kind of I'm bringing this up is because what this means act
ually when you activate the Ros spinal cortex what you most the most what you do is actually inhibit the whole SC right because it's much easier to to activate the superus neurons and the inhibitory neurons and the inhibitory neurons shut down excitatory ones and this is a little bit puzzling because you know we're thinking that the Ros spal cortex is actively putting information in the superc but what we're finding here is actually what it's doing is shutting down a whole thing right so how doe
s this work and to try and shut some light onto this uh we we did a bunch of modeling and we took all of our data our experimented data and we bu mod models of um of of this of this circuit and trying to understand what models fits the fit the data and the upshot of this is that what we think happens is that we have a layer of neurons in the Ral spinal cortex that it's tuned to different directions and it probably computes uh the tuning uh the directional tuning from things like play cells for e
xample and other types of cells and then these neurons send their accents to the exodor neurons in the Subic ccul which just inherit uh um their tuning and so when a mouse is facing in a certain direction you have this particular uh set of neurons activated and when it turns ahead you have another one so you can go back and forth and um the superus kind of always knows where it's facing because the Ral spinal cortex is kind of saying it so but so what does inhibition do and what we think happens
is that the inhibitory neurons are uh have an inverse connectivity pattern and we have some evidence for this such that when the mouse is facing in a particular direction what happens is that the inhibitor iner inhibit all of the excitatory iner except the one that's pointing in the right direction for the shelter all right and the result of this is that you completely clean the representation of shelter Direction in the super cul so that there is no um uh doubt of where you have to go in case
you have to all right um so I'm going to finish here um what have we learned uh well we learned that U instincts are innate behaviors there are encoded in a genome that set out a blueprint of actions and drives that is critical for survival and these actions and drives can be modified by experience to give the animal the best possible chance of adapting to the world they were born in um we've learned that neurons are very powerful computational units because of their properties and the propertie
s of the synapses and that together we can put these neurons together to form very simple computational modules such as the first one I talked about where you have with only two layers uh you can build a little module for compute for computing instinctive decisions and this type of model can be completely hardwired you can just encode this in the DNA make these neurons connect them in this way and you've built a mouse that can make this type of Escape decisions but I've also shown you that you c
an do this in more make these more complicated networks uh that in this case in the example I gave you are important for keeping track of a particular location in space uh and therefore they kind of represent an interface between learn and inate knowledge I'm going to finish by thanking again all of the members of my lab um they're really amazing group of people I'm inspired by them every day they um they're smart they're creative and they have the most important um attribute that the science ne
eds scientist needs which is a good sense of humor uh so we have we have a lot of fun um I'm very grateful to the S Welcome Center for um the support and everybody inside that Institute uh for making it a really amazing place to work uh science is not cheap um and so I'm very grateful for um the support that we receive from the Welcome trust uh who have been supporting me since my PhD uh the gas P ala Foundation have also been very long-term supporters uh the European research Council and the MR
C Laboratory um for MC biology where I started my lab and finally my family for uh all the 11 support and I think my my daughter is watching online with her cousins in CB Bridge Hi H I'm on YouTube don't eat too much pizza before bedtime um and thank you very much uh for coming tonight thank [Applause] you so thank you very much Thiago and um congratulations to the committee for making a very good choice um Thiago has agreed to answer some questions we have about oh little under 10 minutes so I'
ll take uh questions first from the audience that's here and then uh we'll see if there's any uh questions on on on the chat so are there any questions always way in the back yeah uh their microphones and if I see any of the hands y first off pempa from brain cares uh it's a beautiful talk uh I was wondering if you had any clues on uh what might be the molecular cues responsible for the structural stuff that you've uh in the structural circuits you've you've shown the mo for setting up the circu
its I do not have a clue um I do not know no so that's a very important and very interesting question actually uh it's it's it's actually spot on right so if you're going to build a system that that works in this way you need to make sure that everything is wired properly right and go to the right places to the right neurons with the right properties uh we don't know um there might be developmental uh neurobiologist colleagues that will know but we don't U but that's knowing that is really the k
ey for understanding how you can wire a system up to to develop this yes thank you I guess the concept of concept of synaptic tagging I guess who you think would be the tag maybe I mean synaptic tagging normally is is a mechanism that that we think is important for for changing the the weight of copses which which may may well be important in this particular case for um learning where a shelter is for example but to establish the the system I guess uh you need to root the accents to the right pl
aces and make sure they're connected to the right IDs with the right properties and that's a very hard I mean that's a really interesting problem and also very hard one because you need to know you also want to know the function of the neurons afterwards right so let me take Chairman's privilege and ask you a question um you've concentrated on um the whole escape from threat circuit and you very nicely laid out the components of it um do you think this circuit can be used for other things like a
pproach to rewards and if not the whole circuit and I suppose one would wouldn't want to think of the p as being involved in the apprach re Bo but can you think of I mean there some parts of it can be repurposed for other other uh yeah so so depends I guess it's the question the answer so what we think so we did an experiment where we we trained a mouse to go to a reward right so we play a sound and the mouse has to go and collect a reward but we gave this animal a long 10 seconds to get there o
kay right when we do this and we get rid of this pathway the mouse is perfectly fine but when we look at what the mouse does instead of doing the snap orienting and running it kind of starts running where it's going and kind of goes like this right and takes its time there and for that this pathway is not important now what we think this pathway might be important for is maybe not necessarily just for escape when you have to get get to the goal really short right and so you know the way I think
about it is you know you have the hippocampus and the H and cells doing these really complex computations they put it there in rosenal cortex that puts this information in the superal which will have limited capacity right and depending on what on what your context is you keep track of of the most relevant goal so if you need to get there you use the system to get there really quickly because you don't have to recompute the whole thing from scratch right and so that I think can be repurposed for
many other behaviors that need to be implemented really quickly okay right that's how I think about it good thank you other questions right here in the front yeah SC y I'll oh you need a microphone yes yeah there are so many variables possible with this VI if you had two shelters MH and they were equidistant from where the animal is but one time the animal had to go around the barrier to get to it and the other time it could go straight to it the distance is the same but maybe the effort requir
ed to turn is that's a variant of that question that I've never heard very good uh the the answer is they would prefer the so the time is what would matter what we think we've never done that exact experiment it's a very good experiment if we if the if the if the shelters at equid distance uh they uh what matters a lot is what where they're facing right they might be far away from they might have one close to the back but they're facing this way and there's one there depending on the distance th
ey would prefer there I think what matters is a mixture of of effort and time right and that's we're actually doing exactly that experiment trying to come up with a model of what are the key variables that are important for the mouse to choose where it has to go right and I think you you get it right exactly that that's a effort or time that it takes to get there exactly hi yeah kind of yeah can you hear me y um yeah that was amazing I just wondering I was quite fascinated by the fact that when
you turn off the circuit um they no longer Escape but they start freezing and so there are kind of alternative strategies and and clearly in your scenario there's really only one obvious strategy but have you explored indecision or where conflicting options and is it a binary switch or is there you know do you have to decide one and then stick to it or can you kind of look at what the circuits are that are choosing that so yeah that's a very good point so um so when animals are fac through threa
t there's actually a very well established relationship between the imminence of the threat and the accent they choose right if the threat is very imminent uh they will escape but uh if the threat is a perceived uh longer distance they they will actually freeze instead um actually when we do these experiments we low contrast the first thing that animals do is they freeze and they freeze for a while and then when another one comes eventually they run so I think we have that decision there that in
decision that that you're referring to already in in this in this assay and so then the question is how does it work um that's something that we've been trying to nail for a while y Yar is here actually been work have been working on this for uh for for a long time the the the model kind of the standard model would be that the the freezing we know quite a lot about freezing and it's supposed to be carried out by actually the more vental parts of the p and so one model for to solve this that we w
ould like to show or test and we like it to be true uh is that the superior ccul sends accents to both the dorsal part that does escape and the vental part that is freezing and that the threshold for activating the the the the freezing circuit is lower than for activating the one you know maybe you could just have a higher release probability there and that would actually Implement that behavior it's going to be horrendously more complicated than that but this this would be a way of doing it yea
h yeah congratulations thank you for allowing somebody like me to understand thank you so what happened to the Circuit when innate behavior uh is ignored for example when an animal decide not to follow suit yeah so that's very good question and actually um so uh work from from Troy margary actually that we collaborated with has shown that you can actually you know these animals figure out that actually nothing bad is going to happen to them if you keep looming they they you know if they take out
the loom and they tape nothing bad happens they do it again nothing bad they learn very quickly that uh to actually suppress and that Troy has shown that actually they can suppress this for a very very long time for up to months so the question is how does it work we don't know uh but so I think what we know is that what we're looking at is and Steve linday is here is doing the these experiments together with Marcus ston Jones um we we think that whatever happens as a learning process that that
's going to change the activity in the super liquid right and it's going to make the super lius and that part of the superus that cares about threat much less responsive to Threat by by inhibiting it right and that's that's what what we think is one of the key targets for modulating this instinctive behavior is just by acting on the threat detect of course you can act at many different levels right you could act at the signups you can act at Escape initiation but there seems to be a pathway ther
e that's kind of optimally poised to to shut down and to decrease the sensitivity to threat yeah and the final question ah oh thank you a privilege this is for my brother by the way um so you give the idea that circuits are quite deeply conserved but are there presumably there's um animals that don't care about this Behavior as much perhaps if you go up the food chain uh is that the case and so how do you expect the circus to change in those species yeah that's a very good point so so as you go
up the fot chain you get there's less and less Predators um humans are still scared especially baby humans are still scared by this looming stimulus suggesting that this this this structure is still there but I think what happens there is that you know I I don't know maybe this the circuit has um you know shrunk to and so that is not important but what these higher species have is a really big cortex uh that has the ability to to modulate all of the instinctive behaviors and probably act to shut
down the fact that for example an infant baby will react to this and and it takes much more for a human for an adult to actually react might have to do with the fact that you just learn to to parse the sensory reward more uh in a more efficient way and that will come from the cortex that can just shut this down right I think that's one thing that definitely happens but it's still possible that the circuit has um um evolved you know to to be to be redone yeah well very good thank you very much a
nd it only remains for me to read the citation and to present Thiago with his um prize all right the it's an official I have to read this Dr Thiago Branco has awarded the Francis criek medal and lecture for 2023 for making fundamental advances in the molecular cellular and circuit basis of neuronal computation and for successfully linking these to animal decision Behavior thank you very much Thiago and I've got something all right got a metal supposed to open [Applause] this okay as a metal all
right and thank you very much and good night thank you
Comments