All right everybody can hear me OK so it's good all right so I start by thanking the Lalan in Chris for the invitation to speak to you today it's a real pleasure I want to begin with a short video and hopefully the audio will come off OK. OK. John. All right you get the idea if you were spending most of your time looking at the side of the video you can pat yourself on the back and consider yourself one of the typically developing individuals in our population that prefer to look at this biological motion capture that is sort of regularly oriented in that sort of tracks the sound track that you're listening to verses this side which is. And played in reversed version of that same video so this suggests that we have this kind of preference. Biological motion that facilitates our adaptive interactions with other individuals but it doesn't have to be this way turns out that toddlers that go on to be diagnosed with autism actually don't have that kind of preference early on they actually show this kind of equal viewing time for both sides of these videos suggesting then that there may be neural biological mechanisms for processing social and nonsocial information that might not be identical in our brains. And yet when we think about how the neuroscience of information processing has been studied over many decades for example to look at how the neural transfer me what kinds of neural transformations occur for cues to be havior what are mechanisms of learning in memory as well as reward in reinforcement almost all of these are studied in a nonsocial context raising the possibility that we don't actually have a deeper understanding of what goes on when individuals actually interact and pay attention to social cues so my lab has been really interested in the concept of social information processing trying to study these different areas all in more social contexts and the key then is to come up with experimental paradigms that use more natural social cues and behaviors so we can actually understand their underlying biology. So I came up many years ago as opposed to when I was at U.C.S.F. with what I called a kind of computational neuro eith a logical approach to studying the neural mechanisms of social behavior the general idea something that we as you know scientists often employ the fact that. Some stimulus comes in and it drives some kind of response in my lab we tend to focus on these contexts in which this stimulus response is more natural and generally social in context then you want to study the sort of the neural aspect of this computational nearly knowledge is to study kind of relevant neural circuits in the brain apply computational methods to look at how the neural code may represent either stimuli or responses. And then use constraints from natural stimulus statistics or what we know about the behavioral responses and how they change to constrain our understanding of this underlying code in the sort of modern day neuroscience Now we also can go a bit further and apply in rodents methods to test the causality of the kinds of mechanisms that we try to discover. So I'm going to tell you two stories today that use apply this kind of computational the logical approach the first is an area that my lab has spent over a decade studying since I've been here at Emory and even during my postdoc in mostly review findings from my lab and others on the auditory processing of communication calls in mice and the second story is a more recent collaboration with Georgia Tech B M E student in my lab who was jointly mentored by Larry Young studying social bonding in monogamous prairie falls. But first the archery process in community sounds in mice the big message that I want to convey is that learning a sounds meaning alters the feedforward coding. Of those sounds to improve their functional processing this is worth the actually spans many years beginning with a post-doc at Google in Delhi on in several graduate students two of which are Georgia Tech be amused students Franklin who's a bio you student in now most recently Kelly Chung who's in the lab who is a B M E student. All right so to introduce this topic I played back a Chinese poem which if you don't know Chinese made of just sounded like a bunch of words that were exactly the say but if you do know Chinese you may recognize those as the words of part of this poem that says line eating poet in the stone in its different intonations of the same basic sound to some degree. And the point is that people who recognise or understand that language actually very immediately understood how to parse those sounds and discriminate one from the other to create an understanding of the word so what is changing in the auditory system to support that immediate recognition of those communications here that's what my lab has been interested in actually more generally many labs are interested in this concept of neural plasticity what changes when your brain is able to acquire a new skill recognizing new stimulus etc and the basic idea here is that when a sound category is initially novel unfamiliar and occurring in some kind of behavioral context in the case of the auditory stimuli goes on in elicits responses throughout the auditorium you're axis all the way up to auditory cortex while this context will kewl cues may stimulate your other modalities and the. Together they combine to feed into poor sort of perception in action areas that drive your behavioral response to that sound in context now that behavior may have a consequence whether it's appetitive aversive or potentially just purely neutral no matter what this will occur with some Neuromodulation that bathes your brain Baze your neural circuits whether they're cool an urge to ignore nerd Jake or maybe even neuropeptide hormones like oxytocin or estrogen that are co-occurring and that Neuromodulation affects. Has affects widely throughout the brain but also within the auditory system itself the idea then is that you know this is something many people are actually interested in that that sort of dynamic modulation of neural activity by no modulators But I want to point out that these also have lasting effects because all of these you know modulators also affect neural plasticity so that when that sound category is encountered again in the future when it's more familiar and even when it's out of context its ability to drive perception in action is now improved so that's the underlying hypothesis that there is learning induces some long term changes that biases sounds feed forward processing. But exactly what are those long term correlates for sound memories that are choir to our experiences now a few years ago the sort of classic lore was that use it or lose it you have map expansion you create more territory in your brain to represent relevant stimuli and this is just an example from classic work by Greg reckons on in my first thinks lab. Where a monkey that is initially naive to a task has a representation of a frequency around two point five kilohertz the troll it really small but then you train that animal to discriminate sounds around two point five kilohertz and its representation of that frequency in increases or is enhanced so for many years that kind of map expansion is the map that when I say map I refer to the tone and topic map of sound frequencies within the auditory cortex that expansion was thought to be the sort of correlate of learning about a cell. A few years ago though Michael kill guards lad who who actually played a big role in our understanding of this kind of receptive field plasticity found that long term cortical map plasticity can actually DK without affecting behavior so actually map expansion is not a long term Coralie it's me actually be useful for learning but is not itself it's memory trace if you will so then what are the long term correlates of learning to functionally detect discriminate categorize or recognize in our case more natural social vocalizations or for that we have turned to a mouse maternal model that we hope to champion ever since I was a graduate student post-doc rather in Michael Merzenich an interest of trying you see a set. In this model we have mice that actually naturally emit a communication sound that's in the ultrasonic vocalisation range in particular mouse pups emit these U.S. elicit what's called the maternal retrieval which you'll see here in this. In this short video you'll be able to hear ultrasonic vocalizations through a bat detector in the. Corner here. OK So those are the calls you saw the mom responds go pick up the pup and bring him back to the nest it's a very robust behavior and if you look at the actual vocalizations elicit this response they actually form acoustic clusters in certain acoustic prendre spaces like duration and frequency the calls themselves are very simple single frequency whistles like what you show here and they're somewhat similar to actually calls made by adult males in the presence of females but there are also differences as you can see here these calls tend to be more frequency modulating and if you look at a large population of calls they actually form a distinct cluster in this acoustic space from what we see for puppets it's now this kind of acoustic categorization is reminiscent of what we see for human speech for example in human speech human bottles if you characterize sounds according to their form and frequencies they also form these kinds of acoustic clusters so then this gives us an opportunity to look at how neurons within the auditory system respond to specific acoustic features that defy sound categories. In this context the mouse model has further advantages which is that there seems to be a learning component to this recognition by maternal females first if you just take mothers and play back pup ultrasonic vocalizations or neutral sounds mothers a prefer to approach us fees about twice as often as compared to the neutral sound but naive virgins don't they go to both sides about will be offered. On the other hand if you give those virgins five days of experience caring for pups with the litter with the mother it too begins to show this preferential approach to pups suggesting that this preference can be acquired through experience. So that opens this great opportunity then to use this very natural behavior to understand what changes within the century system in order to support this recognition so now I'm going to review a bunch of findings from my lab and others that have led to some insights about this sort of underlying memory traces but first in a collaboration with actually one of Gareth's former students Nick Les occur we actually found that even before you look at mothers if you just look at naive animals there is already an intrinsic bias for ultrasonic sounds in mice in our case we use the C.B. a mice and that's reflected here we we saw this what we call a kind of central magnification of ultrasonic frequencies within the midbrain of the mouse and that's shown here where higher frequency sounds elicit a much greater response in the mid brain compared to the auditory nerve whereas the lower frequencies did not now Nick's lab further took. Micro electrode shanks and recorded systematically in the auditory nerve cochlear nucleus and in clay Killis and replicated these results using using unit recordings suggesting that perhaps there's a kind of acoustic phobia that emerges in the in the brain for these high frequency by the time you get to the to the auditory cortex. Mouse there is actually it's own field. Field that's dedicated to ultrasonic processing suggesting a further sort of elaboration of this frequency range. Now. Some of the earliest studies that we then went on to show first at U.C.S.F. and then continuing in my lab here at Emory was that as you compare naive animals to maternal animals there are long term forms of plasticity for those natural pople clues ations For example if you look at the ability of neurons to follow the natural five hertz rate at which mouse pups produce calls in mothers they're better able to do that compared to naïve animals if you look at the ability to convey information about individual calls for say detecting or discriminating them mothers show in Hants information compared to naive. But if you look at that classic phenomenon map plasticity and you ask whether or not mothers show an enlarged representation for calls compared to naive animals we actually did not see that occur so there was no tonnage topic map plasticity for ultrasounds whether we looked at frequencies vocalizations themselves or vocalisations themselves instead what we observed in both in anesthetize as well as awake recordings is what we call in enhanced lateral band suppression where Non's when you play back pup ultrasonic vocalizations say during the screen period here instead of being excited by the call. Actually being suppressed by the calls. And that suppression interesting Lee was stronger in maternal animals than in maternal animals but specifically in fields that were tuned to frequencies lower than the natural frequencies almost suggesting that as the animals are listening to these calls that are all in the sixty seven kilohertz range those areas that are tuned to lower frequencies are actually being more strongly suppressed and that was more recently confirmed in a completely different set of experiments using different animals different technique where we tracked how neurons responded to playback of calls from naive animals through motherhood to post weening and saw a progressive decrease in the firing rate specifically in these lateral band areas suggesting that this kind of contrast enhanced Mint may be a real effect for improving the ability to recognize these calls or maybe detecting them in background noise. So the picture then merges looks something like this is you go from a naive animal listening to pup calls through motherhood to post motherhood rather then this kind of picture where you have an enlarged representation of high frequencies shown here in red you actually have this kind of contrast enhanced where the lateral band areas are more toned down in their response. But I don't want to give you the sense that all of the changes are all about inhibition right there are changes in how neurons are excited by calls as well and so next I'll review some findings about how neurons are excited by vocalizations and first of all point out that even within the brain stem we actually found subcortical plastics. As you go from a naive animal to a mother responses to ultrasonic frequencies when you get up to the cortex we used a modeling approach to discover that. Neurons within our Dettori cortex can actually respond quite faithfully to the onset of these ultrasonic vocalizations if we model their responses based on how they respond to pure tones in for that kind of model we used basically a peripheral model of sound transduction in kind of conveyed it up to the cortex and what was interesting in that was that those neurons that were best predicted by our model turned out to be almost entirely putative Internet meaning that they are putative yabber urging Internet. Suggesting that the first response in cortex that's very reliably acoustically driven is actually an inhibitory response those neurons that were more likely to be putative excited Tory premeal neurons actually tended to be very poorly predicted by our model suggesting they were actually following the acoustics at the onset of the sounds so what's going on that's when we looked more closely at what changes during mother so then if you compare non maternal animals to maternal animals in how they respond to these pup calls or to acoustically matched adult calls that are matched in their onset properties not maternal animals actually respond about the same to both pup and adult calls but maternal animals actually preferentially. Respond as more strongly to pup calls over their acoustically measure calls feeding with this picture of this enhanced by. Is for pup vocalizations when they become behaviorally relevant to the mother how exactly are these neurons able to do that remember these calls are all matched in their onset properties and that led us to think about maybe these neurons are actually responding to the physical trajectory of these vocalisations across time how does the frequency change across time. And in fact we recently found that the offset responses in neurons actually in code this kind of trajectory and in change so to show that let me point out first that in port Tex in auditory cortex like many other parts of century cortex there's a huge heterogeneity in terms of how neurons respond to stimuli so for example here's a neuron that is we play back short ultrasonic vocalizations you get responses that are in this case. Quite strong right at the onset of the sound but linger even after the sound comes off and you see that more clearly here as we linked in the U.S. feeds in all of these cases there is a portion of the response that is there during this on portion of the sound as well as a portion of the response that appears after the offset of the sound so we call these types of units on in offset responding there are units like this that only respond during the on portion of the sound as well as units like this that respond just at the offset and only at the offset a very specific calls the different rows here are different calls with different acoustic properties now what is really cool is that if you look at the prevalence of these kinds of offset responding neurons that actually changes with motherhood. With an increase in the prevalence for offset only in maternal animals compared to Naama turtle animals and that's almost entirely due to this thick spiking population of putative pre-amble neurons rather than in this spiking population of inhibitory neurons further suggesting it's these premeal neurons that generally provide the output of cortical processing there are retuning to follow this sort of acoustic trajectory OK So that leads to this kind of model where in addition to this kind of contrast enhanced me we have individual neurons actually scattered throughout auditory cortex that are now responding better in mothers to pump calls over adult calls. So now how does this kind of transition occur what actually happens during motherhood to allow for this kind of plasticity to Iraq. And to understand that I'm going to briefly review some more recent work that's both preliminary in my lab and that was also recently published from other labs looking at biological mechanisms for learning in this context so my lab has long hypothesized that estrogen priming of century plasticity within Port Tex may be a mechanism for enhancing this and I won't show this today because of time but we now have preliminary data suggesting that estrogen. Along with social interaction with pups increases B.D.N.F. expression within auditory cortex for those of you who know about B.D.N.F. it's this neurotrophic factor that's strongly implicated in Century plasticity in centuries sensory systems. Other labs so for example his lab has shown enough. Other kind of mechanism that allows for this plasticity to emerge and he showed for example that puff odors a kind of multimodal interaction between pup odors and auditory processing can be found specifically in the auditory cortex of mothers but not virgins so when you present the odors of pups along with calls. The odor actually can modulator firing within auditory cortex. But again this doesn't happen in NY even. They've gone on his lab has gone on to show that there's a also a kind of disinhibiting mechanism during learning where pyramidal neurons when you present at the odor actually get more excited in response these vocalizations while interim your ons are actually being suppressed in the presence of those years so that suppression of inhibition is that then to allow excited Tory responses to increase potentially to allow for plasticity to it to occur and what's really cool is that Rob from his lab a couple years ago demonstrated a possible mechanism a neuropeptide oxytocin that may actually mediate this particular pup odors maybe releasing oxy Tosin within the brain this sort of social neural modulator and operating directly in our Tory cortex can lead to a kind of disinhibition like what are the Mitra he's led show where an presentation of oxytocin weakens inhibitory synaptic inputs while eventually enhanced being excited Tory synaptic inputs over time then this could cement a longer term change. In how these neurons respond to these ultrasonic calls so this is our kind of working model then for how. The auditory cortex learns about the importance of new sounds and reflects those changes in how to respond so we have this kind of model where a lot of neurons pyramidal neurons receive inputs from many other excited Torrie neurons but also from inhibitory neurons pup odors perhaps are acting through oxytocin release then leads to a disinhibition of pre-amble neurons by inhibiting. Putative Internet so that disinhibited or effect actually is something that other labs have found in other types of learning contexts such as in a fear learning context. In addition to this kind of disinhibited mechanism for learning our results also suggest that there is this kind of neural population contrast that is arising from a strengthened feedforward inhibition in specifically lateral band areas. So that's how we're thinking about this model now and there's lots to still to do. But if I could give a kind of summary of some important points I want to come back to later on one is that there's this kind of innate centuries circuit predisposition this acoustic phobia that we think we've discovered for processing evolved communication signals but then social experience also enhanced as the century coding of those vocalizations through neural plasticity mechanisms so this idea of predisposition in plasticity is something I want to come back to in this second system that we're going to I want to talk to you about which is. This model of social bonding in monogamous periods. Focusing particularly on this idea of social reward in reinforcement and this is work that essentially shows that dynamic cortical strategy activity can bias the formation of some of the social bond in these prairie bulls and this is work that was pioneered by another Georgia Tech B M E student lives in Ahmed a collaboration with Larry Young and this is being followed up by Jim Kwan who's an Emory neuro science graduate student so pair bonding and making a big transition here so I wanted to give you a little bit of an idea of why this is an important topic and for those of our generation this is probably a familiar movie to Titanic very strong pair bond right well that kind of pair bond happens in about five percent of mammals it's a reproductive strain is quite successful in about five percent of mammals. But exactly what happens in the brain that allows you to go from this kind of non bonded state to a bonded state one idea that's been out there is that there's kind of a over time an increasing reward value that's placed on your partner so that leads us to ask well how really is the Meza cortical limbic reward system activated in module A did during the formation of a bot. And to do that we're going to use this provable model of social bonding which has been pioneered by Larry Young's lab initially working with Tom Insel and now many other labs have sort of embraced this as kind of the canonical model of social bonding the general idea is that when you put a male and female prey able together and they cohabitate after enough time they form a kind of social bond which can be tested in the laboratory using what we call a partner preference test where you give the subject the chance to spend time huddling sort of motionlessly with its partner aka cuddling if you will or spending time huddling with a stranger and you find that in prayer Evil's unlike other types of closely related bulls prefer to spend more time huddling with their partner than their STRANGER And so this is. The laboratory signature of a pair bond formation. Interesting only when. Tom insulin and others went to look at the brains of prey voles and compared them to non-monogamous foals they found that oxytocin receptor expression was very different throughout key parts of the brain in particular prefrontal cortex and nucleus accumbens and when Larry when in fact in oxytocin antagonists site specifically into either the pre limbic part of prefrontal cortex or into the nucleus accumbens he could reverse. This kind of partner preference that was seen in control and it's suggesting that oxytocin acting specifically in those brain areas is important for bond formation. This is these two areas are part of a larger circuit involving dopa mean and oxytocin projections that we think is critically involved in social bonding but just knowing that those areas are involved where those sites are doesn't really tell us how those areas communicate with each other during social interactions leading to a bond formation So what exactly are those sites modes of action to study that we started to do neural recordings in prairie voles by implanting what we call a neural logger onto the brains of skulls of female probles as they co-habitate with a male prey ball. We videotaped them as we're recording so that we can synchronize neural and behavioral activity and then we have lots of undergraduates as well as graduate students watch those videos kind of porn if you will and score them. Or different kinds of behavior like me. Self grooming and of course huddling which is the expression of the box. As there as those behaviors occur throughout this six hour co-habitation time so here's. Where the male is mounting the female and maybe here's an epic where the females just self grooming by itself and then hears an epic where they are just sitting side by side. So once we had this behavioral records we could look at well when do these behaviors begin how long they last a tetra and one of the things that emerged quite quickly is that mating happens very rapidly when you put those together but huddling tends to happen much later not only that huddling is quite different from one individual to another and that's shown here in these huddling trajectories for individual voles what each line reflects one an animal begins to huddle and how long it's hardly So you see there's quite a lot of variability and we mark a particular time point to accumulate say five minutes of huddling time that gives us a kind of measure of the latency to begin huddling and if we look at that huddling latency in compared to how much time animals actually spend huddling by the end of the six hour co-habitation they are very strongly correlated which kind of makes sense right earlier animals begin to ho huddle the longer they're going to huddle but it turns out when you look at other behaviors that relationship between timing and duration doesn't necessarily hold so this was an interesting and important finding and suggests that we could use huddling latency as a kind of measure for how quickly a bond gets forked So what gives rise to that huddling latency. Well we looked at possible correlations. With a number of different behavioral measures including latencies and durations of tetra and none of these could explain the this kind of huddling latency so that led us next to take advantage of our neural recordings and look at potential relationships between this huddling latency and measures of neural activity within the two areas we placed electrodes medial prefrontal cortex and the nucleus accumbens and those recordings and example those recordings are shown here we recorded local field potentials and. Divided them into various epics that were synchronized with the behaviors of epics of mating self grooming and huddling for example and if you just look at what the power spectrum of those local field potential look like within the M.P. of C. or nucleus accumbens we see that there are there are oscillations in the local field potential and moreover those oscillations can be module aided by behaviors. In particular what lives in sort of locked on in noticed was that this very high low frequency peak at about five hertz which is kind of roughly a rhythm there was a very strong modulation of this. This peak across different behaviors and so that led us to look more closely at different measures of how these brain areas might be talking to each other that peak you can see is apparent in both M.P.'s C. and in nucleus accumbens So we wanted to know whether those might be actually related and whether activity in one brain area could actually effect the activity in the other and so we. Used a measure of what we call functional connectivity called The Cross frequency coupling the idea is that we take the signal from one brain area and filter it to. Represent say the low frequency component and then take the other signal and filter it to represent the high frequency component and if you can squint and see sort of that there's a phase relationship between when what phase the M.P. of c low frequency signal is at and the amplitude of the gamma oscillations in the nucleus accumbens that relationship turned out to be actually quite consistent there is what we call a cross frequency coupling where M.P.'s see low frequencies around five hertz modulator gamma activity in the nucleus accumbens at about eighty. And so this is an effective channel of communication between those two areas. So now with that channel we can just as say that cross frequency coupling across the whole cohabitation across all different kinds of behaviors and that shown here in the color coded fashion and you can see how do you square in make out that yeah there's kind of changes across time that mating tends to give very high. Cost frequency coupling but huddling tensely actually give you kind of low cross frequency coupling Now this is one animal's net modulation across time this is another animals and you see though that there's a lot of variability across animals and in fact if you look across all of these animals and look at their cross frequency coupling during different epochs of time whether animals are huddling whether animals are in this baseline period where they're just sitting by themselves or whether they're in this non huddling behavior which essentially we classified as as a more active. You're When ever animals are just not sitting next to each other and you see that there's a lot of variability but first of all huddling which is supposed to be this rewarding thing actually does not a listen activity within this reward pathway which is a little bit of a surprise to us but instead during these non handling activities there's a lot of variability across individual and moreover it correlates really strongly with this huddling latency suggesting that this brain activity this functional connectivity between these two areas was telling us something about how quickly animals will go on to to express their affiliation. Now this was a correlation that we observed when we took neural data across the entire cohabitation and compared it to the latency to huddle but it actually emerges across time. When we started taking data looking at the data in stored or ethics from the onset of cohabitation we found that even at about an hour into cohabitation there's already a nearly a point seven five correlation but even before that even at times zero which reflects a time point where the animals have not even met there's already a high correlation suggesting a kind of predisposition based on the neural activity in this pathway for animals to be kind of sociable with another individual but then that actually increases over time and if we look in a non hit circuit there's actually no such correlation so this is specific to this medial prefrontal cortex to nucleus accumbens circuit. This correlation seems to increase over this first hour which is the period during which meeting first begins so that led us to look more closely at what happens to this functional connectivity around needing And so this is mating right here before mating this is one animal where there's not much net modulation and after mating there's a slight bump in this this animal for a few minutes afterwards Here's another animal where you get a much bigger bump in their cross frequency coupling Here's another animal where you get almost no bump and when you look at all the animals together you kind of see that there's a color gradient that can be observed here and the coloring is actually based on how quickly animals huddle after they come out of meeting and in fact if you look at the change in that net modulation and correlated with that huddling latency from coming out of meeting there's actually pretty strong correlation suggesting that what how these animal how these brain areas are communicating as they come out of their first meeting event is saying something about how quickly they want to come together and express their behavior. And that is significant for about a couple minutes after they come out of meeting and even if you subtract out that baseline correlation I mentioned earlier there is still a significant correlation between these two suggesting this is a separate effect than any baseline predisposition to be sociable with each other on the other hand if you look at our control behavior self grooming there was no such enhanced moment that was correlated with with their huddling leans. OK So that suggests mating is this experience that can in change how the brain responds to the partner which then correlates with how quickly they're going to be affiliated can we now actually see if it's causally involved so for this we used the. Channel road option and opted genetics approach to try to bias the formation of a bond in Prairie Evil's that would not normally form a bond and so this just shows expression of Datsun within the nucleus accumbens we injected our virus into the medial prefrontal cortex and put a fiber over the nucleus accumbens to. Excite the projections into new Cummins and then we allowed the animals to cohabitate only for a limited time three hours in a way in which normally bonds would not be formed the male was actually jailed sitting under a cup unable to physically mate with the female but every time the female walked into the social chamber it got a five hundred stimulation object stimulation of these imports from him P.F.C. to new clues come. And then so if we look at how these animals are stimulated compared to control animals that just got a flora for virus that does not actually gate activity within within the. Transfer active neurons there's no difference in terms of how much optical stimulation they got or in terms of how much time they spent in the different compartments of this cohabitation chamber. But then when we took them to the partner preference test when we looked at our control animals most animals actually showed a preference for the stranger over the part or consistent with the idea that I. Our our cohabitation paradigm does not lead to a bond formation. On the other hand those animals that receive ten or a Dobson showed a preference for the partner over the stranger and when we looked at the difference in time for huddling with the partner versus the stranger there was a significant effect with a preference in stimulated animals for spending more time with the partner so this they to frequency oscillatory drive can into nucleus accumbens can causally by the emergence of this affiliate behavior. So this puts together a kind of working model then of what's going on in the middle prefrontal cortex at low frequency phase modulator the nucleus accumbens in the Gamma frequency range and that strain the modulation actually varies across subjects and correlates with the huddling latency but first mating actually can improve that brain affiliated behavior correlation and. Showing cause ality directly in Hansing that. Activity can actually shift the preference towards a part. All right so now stepping back in and trying to put this together with the auditory story told you about the big picture insights that we're gaining from this model are that again we see this kind of. Individual predisposition this time in the cortical limbic circuit for sociability but that. Social interactions can dynamically inhance the ability to switch how an individual responds behaviorally to another individual who is presumably a source of social world ward or for that individual. And that brings me. To a recent paper. Specifically this idea that it keeps reoccurring a predisposition it was published by Warren Jones an army klyn from the Marcus Autism Center down the road from and Emory earlier this year which found that there are individual predispositions for seeking out social information and they did this by looking at human infants and looking at how they view videos of social interactions with caregivers with other infants and amazingly when they compared monozygotic twins to die as I got it twins or other age. Kids in Manas I got it twins there is a strong correlation between one twin and the other in terms of the macroscopic level how it is bending its time viewing those videos specifically how it's spending time viewing social information like it's time spent looking at the eyes and also the mouth there's a strong correlation there from one twin to another that's present in monozygotic twins but not in dire twins or others this is at the macroscopic level more amazingly when they look at the individual traces of sick Codd's visuals a cause how they actually viewed videos monozygotic twins demonstrate a greater power probability relative to dyes I got it twins of moving their eyes at the same times for each the cowlick I movement by twin one within three hundred fifty milliseconds there was an eighteen point six percent increase in twin two's probability of also making a by suggesting that there's an innate. Genetically potentially driven salience for social information that's there from birth and actually is preserved up to I think they look at the. One months there is still a strong genetic drive for how they view these videos so that predisposition thing something we often ignore in neuroscience actually seems like it could be a key for understanding individual differences and eventually making the jump to potentially mental health but so then stepping back I hope I've been able to demonstrate how this kind of computational neural youth a logical approach to social information processing has given us some new new insights new lessons first in this case of arbitrary process in communication calls it suggests that actually instead of map expansion we should be thinking about different ways of enhancing at a population level the neural contrasts for how sounds that are communication sounds or behavior the relevant are represented in the brain in the social bonding case that we should look at focus on this cortical Strike Eagle circuit that may be hoping to switch animals towards more affiliate of behavior which kind of fits in the language of how people study this political struggle circuit which is often in the context of drug addiction and also habit formation Maybe love is like forming a habit for your partner for example all right and with that I'll end with acknowledgments especially funding from an ID CD for auditory work and I image for social bonding work collaboration's with Larry Young and and also. Frank Cady and Kelly who were so important for a lot of electrophysiology in the caves in Louisiana today here in Jim Kwan who were involved in this social bonding system that will. Thank. You. Well. I don't know the full answer yet but I can give you a couple of anecdotes one is that in a few animals lives in actually tried stimulation in twenty herds those animals actually developed a strong stranger preference so it may actually be that frequency does matter. But we haven't followed up on that there was only a couple of them. So. Now. Very. So first of all in rodents odor in maternal behavior are is critical very important for maternal behavior. In our model how I think of it is odor is kind of the innate. Response innate signal that these are pups if you're a mom be maternal but the auditory system and these vocalisations are dishes cues that hope a mother locate pups and respond to them and so I think of the auditory system as kind of learning from the all factory system about the cues of the pups and that's where the odor modulation of auditory cortical activity there may be mediated through oxytocin is a really exciting possibility as far as individual vocalizations from different pups in my case there's no clear evidence that individual pups matter sex of pups may matter so in rats males versus females mothers seem to pay more like to pay more attention to the males than the females there was some studies done by. Peg McCarthy's lab in University of Maryland that suggests that male pup vocalizations are particularly interesting to the mothers but in terms of individual animals when pups are really young. There's a lot of individual variability in their vocalisations But as they mature after about P thirteen which is when their years actually open. Thirteen their vocalizations become more similar to each other in certain characteristics and just you know ecologically it may be you know these mothers have a lot of pups all at once they will actually Co care for pups of of their sisters pretty easily they'll harem raise pups so it may not be so important for them to invest time into individual pup the way it would be for humans for example. And. So adult male mice can eventually show a preference for pup calls but they take longer so. They have to have more than one litter of experience before they'll show that recognition. They can show it but it just longer so and that makes it logistically harder to do experiments on them so we haven't really investigated their neural responses much. Thank you.