Yong Xu - Biochemistry Colloquium - Fall 2026

All right, I guess people are moving in. We can just get started. Well, welcome everyone to our um biochemistry seminar series. My name is Wayey. I'm assistant professor in the biochemistry department. It is my great pleasure to introduce our first speaker of this semester, Dr. Yongi from University of South Florida. So Yong is a physician scientist who completed his medical training in China from from Tongji Medical University and then he went to Canada and obtained his his PhD at the University of Alberta. Then he decided to pursue a postto training with Dr. Joel Mquist at UT Southwestern in Dallas and in 2010 he started his independent laboratory at Baylor College of Medicine where he was promoted to tenure full-time professor and director of the obesity research center there and in 2025 which was last year the University of South South Florida was very luckily able to recruit Yong to set up a brand new center that is dedicated for for molecular neuroscience and psychiatry. So Jung has made phenomenal phenomen scientific contributions to the understanding of molecular and cellular pathways control or how the central nervous system controls whole body metabolism and physiology. And I'm not going to brag about his science because you'll learn from yourself in a few minutes. But I just want to emphasize that he's not only a great scientist but also a great great mentor. He has trainees setting up their labs around the world and I've had the opportunity to become friends with many of them and they all unanimately told me that they they think that the support and guidance received from Yong has been greatly greatly um helping them moving along in their scientific careers. And another objective evidence is that this year alone three postos in his lab have received NIHK99 independence to pathway um awards. just three year three postoc in one year. This is just how remarkable his teaching and his mentorship and his general support to the young scientists are. So I think Yong is going to tell us about some nervous system mechanisms about whole body energy balance and Yong please take it away. Thank you. Uh first of all really appreciate that you invite me to come to this beautiful campus during this beautiful time and I'm very impressed with uh everything you guys have here in the department. Very impressed. Actually I'm going to while I go home I'm going to tell my uh eligible postto fellows to try to apply for faculty position here because this is a great place for them to start and also thank you for the for the really kind introduction. I think it's very hard for me to really live on with that. Uh anyway, so uh I'll share with you uh two stories that are recently emerging to the lab and hopefully this will entertain you for the next one hour or so. Um first let me ask uh have you guys done this the ice bucket challenge which I know was initially started to promote the awareness and donation for ALS which is a motor neuron disorders and I have to admit I haven't done this but you know the uh ice bucket challenge when I think about it it actually itself can promote benefits for metabolic disease. For example, when we were exposed to extreme cold, in order to survive, we need to defend our body temperature and this requires more energy to be spent. In order to do that, we need to refill an energy by eating more food, right? And by the same time, there is cumulating evidence from both animal and the human studies to indicate that repeated code exposure can improve our incident sensitivity in the glucose homeostasis. So if we put all this together, we have a very interesting scenario where we can improve our metabolic benefit, metabolic health without necessarily cutting down what we eat. We can eat as much or whatever we want, but we can still maintain metabolic benefits if this is whole all true, right? However, of course, we know that the code exposure may not be practical for perhaps most of the humans especially in a matter of repeated code exposure or chronic manner, right? So that promoted us to start to asking ask a question. Can we figure out a mechanism by which the code exposure produce the metabolic benefit and then can we use it to develop a code matic if you will to produce the metabolic benefits without actual code exposures. With that question in mind, we performed this first experiment which is quite simple. We put our wild time mice into the sixderee environment for um six degrees. was cold for two hours and then we perform imunino staining of this protein called C force. So C force is a early immediate gene. It can be rapidly in the transit turned on whenever the neuron is activated by all kinds of stimula. So by looking at where the C force is expressed we can find out the neurons that were activated by this code exposure for two hours. And then by doing this we found of course many different brain regions that have C force expression. But very quickly we focus our interest to this particular region. The name is dorsal perventricular hypothalamus posterior part and then the short form is DPVP or dorsal PVP. It actually doesn't matter what is it called. What it matters matters is number one it really is robustly activated by code exposure and two is when we look into the literature we very surprising to us that we found very little is known about this region. actually there's no known function that's been associated with this brain region. So this got us interested and started to look into this region what they were doing in the context of code exposure. In order to do this type of experiment we use a technology called trap which stands for targeted re combination in active population. I know it doesn't make much more sense right so let me just explain this again. So I think trap this abbreviation is much better because literally with this technology we're trapping whatever two proteins in a group of neurons that were activated. That's basically what it is and how it works. Imagine that we have a mouse that have two transgenic alo. One of them is this Rosa 26 driven expression of t tomato which is a reporter which is preceded by tulock p size flanked stop sequence. I'm sure many of you guys use similar leo like this right? Normally the cell wouldn't express the reporter because there's a stop sequence. However, if this cell has a crebinase cre is going to recognize this two locks p size and then remove the stop sequence. Now the expression of Tito tomato will be constantly driven by this ubiquitous promoter loss 26 right and then this mouse also have another transgenic alo which is our trap 2 al this is a force promoter driven expression of cre er again I told you force is a early immediate gene its expression can be rapidly and transiently turned on whenever the neuron is activated so imagine this mouse carrying these two alals were exposed to code and then some of the neurons will be activated. This activated neurons will express force and the force promoter then will drive expression of CRE. But we know CRE is not going to really make any difference because CRE do not go into the nucleus to perform DNA recombination. However, if we timely pair a injection of a tmoxifen with our code exposure, tmoxifen is going to drag the cre into the nucleus of that activated neurons in the neurons in this in nucleus of this neurons cre is going to recognize the locks p and we move the stop sequence and then drive the expression of t tomato. Now these mouse have all their code activated neurons labeled by t tomato. Right? A few things I want to emphasize about this trap uh technology. Number one is once trapped always trapped. For this particular neuron that is trapped because the expression of the reporter now is driven by ubiquitous promoter. So this neuron unless this neuron is dead this neuron will permanently be labeled by t tomato. Right? Number two trapping can only take place during that narrow time window of injection. This is very important as you may imagine after this trapping protocol the mouse will go back to his home cage and continue to live his happy life. But a lot of things is going to happen in the rest of his life which can always turn on expression of force in certain neurons. You know somebody come to the room turn on the light grab the cages put in the hood open the cage and grab the mouse and maybe do an injection. Every single one of these procedures can cause activation of certain group of neurons. However, as long as you don't give another tomaxine injection, you're not going to trap anything else in these animals. So, this give us the specificity of this trapping event timely associated with our code exposure. Number three, we can essentially trap any two proteins of our desire into this group of activate neurons as long as we have a credependent alo either a mouse alo or maybe a vector that is c-ependent. Right? This will become important in later departmental. So the first experiment we did with trap is exactly what I just told you. We trapped tea tomato in the neurons that were activated by code and then just show you example. Again we see tomato expression in this dorsal PVP vision a lot. And then with this tool we can also perform something fancier. We use light shim microscopy to scan the tomato signal in the 3D mouse brain that's been cleared as a transparent. So shown here is actually from Rossio to quad and where we see the PVP or doso PVP vision where the Tmato label cells but more important question is what are these neurons are doing in the context of code to answer that question we use another technology called Dread which stands for designers receptor exclusively activated by designer drug. How many of you guys ever heard of it? Oh yeah. Okay. I I see some hands but this explanation doesn't really give you much more information again right so I'll try it again so imagine that you have a g- protein couple receptor that has been mutated so this is our designers receptor the receptor has been designed in a way that this receptor no longer respond to any of the endogenous liant instead only respond to a synthesized chemical called a CNO this is our designer's drug notice that this version of the designer's receptor is coupled to GQ. So when single bind to this GQ couple designer receptor, this is going to need to the excitation of the cell that express this designer receptor. Now we can package this designer receptor into a AV vector that were made in a creep dependent manner. Remember what we can do with trap? we can combine this AV vector with our trap animal and now we can deliver the expression of this uh DRA receptor or designer receptor in activate group of neurons that were activated by cold that's what we did here so we made we had a virus AV virus that's credependent to express this designer receptor couple to GQ and then we can inject this virus only into the dorsal PVP vision of the trap 2 mouse and then we can expose this mouse with cold for two hours. At the same time, we give a tomax injection. Now, this receptor will only be expressed in the cold activated group of neurons in this brain region. If this animals recover from the surgery and the code exposure, now we give the C no injection, we're going to selectively activate this group of neurons. And our question is when we mimic the activation of this neuron as they were exposed to code, what happened to these animals? what phenotype we may observe. The first phenotype we observe is this dramatic increase in feeding. They're eating like crazy compared to other control groups. So this drive animals to eat. Also, we noticed that this increased their body temperature. In the third experiment, we put these animals in this thermal gradient box where the animals were allowed to freely explore this environment with a gradient temperature from 15 to 46°. And then they can just choose whatever environment with the temperature they would like to stay. Can you guess where they going to stay the most? If you look at the black curve, the control animals like to stay in this zone of between 28 to 30. This is considered to be a thermal neutrality zone for mice. They want to stay there. They stay there most of time because they don't have to spend extra energy to defend their body temperature. But if we give the animals a single injection which is this orange curve here, this mimic the activation of the dorsal PVP neurons in response to cold except that we didn't give them code. They actually they have choice to go to anywhere they want in terms of ABN temperature. Now they like to spend more time in a warmer area. They are seeking warmth. We can use the same mouse model and give them repeat CNO injections on daily basis. So this can mimic a repeat cold exposure. What happened to this animals? They constantly eat more. Every day they eat more and every day they have increased body temperature. This combination resulting a no change in body weight. But what is intriguing is they have improved glucose homeostasis. They have better glucose tolerance and a better insulin sensitivity. To summarize what I show you so far, the activation of this group of neurons in the dorsal PVP which were activated by code basically can recaptulate many of the cold responses. Eating more, increase body temperature, seek warmth, but this is associated with better glucose homeostasis. Right? Okay. In this field of body temperature regulation, the dogma is that changes in room temperature can be first sensed by our skin and this is transduced through nerve terminals back to the brain which eventually trigger all kinds of behavior and metabolic changes in order to deal with this changes in room temperature which I think it is true. However, as a neuroscientist, our field has always been wondering the same a question that is whether there are neurons in the brain that can directly sense changes in temperature in the local area, whether that can also contribute to this defense of body temperature. And a few years ago, actually three years ago, this paper came out from neuron from my friend's lab, which I think provided pretty important evidence to suggest that this may also be the case. What they did is they simultaneously measure the body temperature and the brain temperature of the same mouse when they manipulate the room temperature. The room temperature is this black curve here goes from 25 to 35 or even higher 38 and then drop back to 25. And the red curve here is the measurement of a body temperature. So you may imagine the mouse body temperature goes up and down when the room temperature changes. What is really striking is when they measure the brain temperature which is this blue light blue curve here. This follows exactly as the body temperature they also fraate when the ABN or room temperature is changing. What that indicates is the neurons living in our brain are also exposed to changes in temperature in their local little environment. So this prompt us to ask whether the neurons that we identified in the dorsal PVP they can directly sense the changes in their local environment. To do that again we use our trap protocol and then we can perform the patch clamp recording of the brain slices from the mice the trap two mice that have been trapped cold trapped so that the code activate neurons are t tomato positive. So we can record from them and then this chamber this spring brain slices in this little chamber which has a peruse to go through them and then we can manipulate the temperature of peruse to change the environmental temperature for this brain cell. In other words for this neurons and then we can record from this tomato positive neurons. When the temperature goes down we see this neurons become more activated as demonstrated by this increased frequency of this vertical spikes. Each one of them represent an action potential. So if you see more of them during the same period of time means they have higher firing frequency indicating the neuron is activated right then when we uh increase the temperature back to 34 degree and then this firing frequency recover this is a quantification it's a perfect almost perfect negative control experiment in the same slides we can also record the neurons that are not tomato positive so these neurons are surrounding neurons but they are not activated by cold and then when we locally change the environmental temperature yes they are activated by code. So this data indicates that our code activated neurons here they can directly sense the code exposure and then the next question is how and then there's a number of ion channels that have been known as temperature sensor. So based on this we look into the publishable published single cell RNA seek data to ask a very single a simple question which of this ion channel as a temperature sensor are highly expressed in our neural population and quickly we identify this channel called KCK2 which is basically a potassium channel as demonstrated here. So KCK2 is again a potassium channel. What is unique about this channel is at the room temperature or in a baseline the channel constantly opens. This allow the potassium which is a positive ion go from inside the cell to the outside the cell because this is positive going outside the cell. This leads to inhibition of the neuron. Right? When the neuron was or the cell exposed to cold temperature the channel close and then this prevent the leakage of the potassium ion to going outside the cell. Therefore, this needs to in activation of the cell and then we can record from those tomato positive neurons and then we can change the temperature again. And then here we use a voltage glam uh protocol to record the overall the total potassium current going out of the cell. And what we found is the lower temperature can reduce this potassium current and then this recover when the temperature goes back up. Of course, this recording itself does not tell us this changes in in the uh potassium current is mediated through KC and K2 because we're recording overall potassium currents. However, in this condition, when we pretreat the slices with this compound called spatting, which is a selective inhibitor of K and K2 channel, you know, when we do this, we found this fluctuation of the currents was totally abolished in response to the temperature change. So combine this data this suggests that the obs the these observed changes in potassium current in response to code are likely mediated through KNK2 channel and then we test the effect of this KN2 inhibitor sparting directly in the animals. We directly inject this compound into the dorsal PVP region of live animals and then just see what happen to these animals. They eat more. They increase their body temperature. They seek warmth. Again this recaptulate the code responses. Then we did the same experiment in another group of animals where we overexpress a mutated case in K2. This mutation basically make this channel always open no matter whether there are phase 2 code or there are phase spatting. So you just this is sort of like a gain function mutation but they block the effect of spatting. And then at a physiological level yes we do block effect of spatting. there's no changes in whatsoever by spotting spanning treatment and then we use gas uh SGNA for K2 combined with a cerependent castine AV virus and then we inject into this dorsal PVP region of trap 2 mice and these mice as you may imagine go through the code trapping protocol so we can specifically knock out KK2 only in code activate neurons in the dorsal PVP and then we have proper controls of And then this knockout basically leads to increase in food intake, increase in body temperature. The animals have no changes in body weight but they have improved insulin sensitivity. So again this fully recapulate the cold responses. Right? Because we basically identified the first function of this underststudied brain region the dorsal PVP. Our reviewers asked us to do a bit more characterization of this brain region. That's why we performed this single cell RNA sig analysis of our own and then this time we expose animals to code and then we cavity tissue do the single actually single nuke RNA sik and that's why we can detect force expression so these are the code activate neurons and then we found this highly colloicalized with this transcription factor called the left one and then we perform imunino crystal chemistry just to confirm left one is highly expressed in this region this is in code exposure We found a lot of left one neurons were activated by code as demonstrated by C force expression. So this confirm left one is a good genetic marker to enable this brain region that were activated by code and there is a left one cre mouse model. So we combine that with our dread approach remember the DQ. So we can inject our virus predependent AV virus into this region of left one cre mouse. This way allow us to specifically activate this group of neurons. Again when we activate this group of neurons the animals eat more have increased body temperature they see warm they recaptulate the code responses and then because the left one is a transcription factor and actually I started Korea by looking at ya alpha as a nuclear receptor transcription factor. This made me just can't stop thinking maybe left one as a transcription factor has something to do with our key expression of our key uh code sensing machinery KC K2. So we just look into the case to promoter region and we define a number of potential binding sites for left one and then we perform cheap assay to show that the left one protein can bind to this potential binding sites of KC and K2 promoter and then in the cells we use luciferous assay to show that overexpression of left one can robustly increase KC and K2 expression and then we can largely abolish that by making mutations of this potential binding sites. So this in suggests that the left one can directly transcribe KC and K2 channel and then there is a left one flux animals and then we inject AV cre into this region of this flock animals. This allow us to knock out left one in this region and then one of the things we notice that expression of KC and K2 is reduced. So again this confirm that left one is a positive regulator as a transcription factor for KC and K2 transcription. At the physiological level, we notice there is increase in body temperature and also the animals seek warm. Again, this recaptulate the uh cold responses. So to summarize this part of my talk, so left one neurons in the uh dorsal PVP region can directly sense cold to induce hypopasia, induce body temperature, increase body temperature and energy expenditure and trigger warm seeking behavior. This overall result a improved glucose balance and then we found that Kin K2 channel is the code sensor for those neurons and the left one directly regulate the KN2 transcription. More importantly in my opinion is activation of this group of neurons such as by KN2 inhibitor I think can be a good strategy to improve glucose balance or metabolic health despite the chronic hypopasia. So that is a a perfect scenario that I would hope for personal I can eat as much as I can whatever I want but I can maintain my metabolic health right for the second part of my talk I'm going to share with you our journey with this in my opinion very magical molecule or gene called tripy5 so uh trip 5 uh is one of the family members of this trip family trip stands for transient receptor potassium uh caden channel uh trip 5 is one of family members is form basically form a calcium channel. So when it opens the calcium can go from outside the neuron to inside the neuron. This will lead to excitation of the neuron. So our interest on trip 5 stem from our collaborator Dr. Salaf Ruki from University of Cambridge UK. So Salaf is pretty well known in the field of obesity because she um contribute to our understanding of human genetic basis for obesity. So he she collect a large cohort of u early onset severe obesity kids and then perform axon sequencing those kids and identify genetic variance that may be associated with uh this their obesity. Now one of the genes she identified is tripzy5 and this actually is um two boys that carry the uh trip 5 gene mutation. Um you see their body weight curve this is from birth to about 18 years. I want to draw attention to this area first. This is our what we usually consider as normal from 5 percentile to 95 percentile in terms of body weight. Even you are in the normal range but at 95 percentile that is very big boy right but look at these two boys they're totally out of the chart very very obese. And then in her cohort she was able to identify seven different point mutations in trip 5 protein that were associated with early onset severe obesity. So this is where we start collaborating with her and then we took all the seven mutations we characterize in the cells to make sure that all of them are loss of function mutations for trip 5 ion channel and then we picked the one this one K34 D because this is the most severe loss of function and then we made this into a mouse. So we have a knocking mouse that mimic that carry the same mutation that we observe in humans that have obesity and then this mouse recapsulate the human phenotype. They eat more and they develop obesity. So the combination of this human genetic data and the mouse genetic data in my opinion is perhaps the most compelling evidence that we could hope for to make a statement that trip 5 loss of function mutation can cause obesity in humans. And because of this work, this gene has now been added to the the uh genetic screen for human obesity. Also in the same paper which is uh a paper we are very very happy about and we also found that this mouse model we captured a number of other human phenotypes seen in this uh mutant carriers including autism anxiety and the sleep disorders indicating that trip 5 also play very important role in this different aspect. I'm not going to talk about any of this today, but instead I'm going to tell you a very interesting story that develop along this uh trip 5 journey. So this is the two boys that I just show you for the body weight curve. And then this is their siblings. They also carry the same mutation. As you can see, they're pretty heavy as well. This is their uh BMI. This is another one. And then all these boys inherited their mutation from their moms who are also very obese. And then Salava is a physician. She interviewed these two moms and very quickly she realized that these two moms suffered very severe postpartum depression during every single one of their pregnancies. And then she told me through a zoom meeting at the time this that's during the covid pandemic. So she told me that I can clinically diagnose this very easily. However scientifically I can't even make association because I only have two patients. And I said don't worry we have the mouse we have the mouse model carry the same mutation and then we can use the mouse model to characterize whether this mutant mice also develop postpartum depression. I have to admit by the time that I said that a promise to saddaf I never really even thought about that my my lab is going to study this disease postpartum depression. So both myself and my first author the postto fellow um start to research or learn about this human condition. So apparently about 10% to 30% of mothers will suffer postpartum depression which is a very significant problem. Believe it or not about 10% of fathers claim that they have postpartum depression. Okay I saw some people shaking heads right I don't I didn't believe this I will come back to this later then. Okay so the symptoms of postpartum depression of course have all this depression behavior because it's called depression right? I'm not going to read this but I just want to point out the most severe consequence will be the suicidal ideation. So it's pretty serious uh stuff. However, I want to point out that postpartum depression and the depression you probably talk about on common basis are two different things. If you say certain professor have RO1 very well scored but didn't get money that professor is depressed that's different from postpartum depression. That is called major depressive disorder. It's a different disease compared to postpartum depression. The ideology is different, treatment is different and the mechanism is likely different as well. Okay. And then one of the important uh thing to distinguish this two is the timing. The postpartum depression has to take place during the postpartum period. Understandable, right? Another important characteristic of the postpartum depression is the mom the moms do not like to take care of their babies. That's why when we use mouse model to characterize this human disease, not only we need to look at depression like phenotype but also we need to look at look look for their uh maternal behavior. So shown here is a movie where we perform this classic assay in the field. It's called a pup retrieval assay. So this is a wild time mom. This is a mutant mom on the right. So the assay is goes like this. So, we put their pups at the opposite corner of the net and then we in reintroduce the mom back to the cage. As you can see, this wild time mom right away start retrieving the pups one by one back to the net. She's so smooth. She's so efficient. She's so great, right? But look at this one. This is our mutant mom. By the time that our wild time is done, you see where she is, the mutant mom. So, she's definitely slow. And also another thing I hope you already realize is another phenotype. Our mutant mom does not care about net at all. They already destroy the net. Whereas our wild time mom maintain the net in very good shape. Both these two signs indicate that our mutant female mutant mom have impaired paternal care. Right? And then we also need to characterize their depression like phenotype because we're trying to model this for what? postpartum depression, right? But unlike human, you can ask them whether you are depressed or not, right? For mice, you cannot do that. There's a number of assets that we can do to try to estimate this. One of the assay is called a full swimming test. So, basically, we have a big jar with very deep water and then we put animals into the water. Mice do not like to be put in the water. Mice do not like to be put in the water. I said it twice because it is going to be very important. Okay. Okay. Mice do not like to be put in the water. So once they're in the water, initially they are going to struggle. They're going to try to get out. But you set up the experiment in a way that the water is so deep and also it is so far away from the edge. It is just impossible for this mouse to get out. So for a while the mouse will realize that it will become hopeless will give up. Once the animals give up the mouse will spend time like this. We call that immobile. So you give the animals enough time which usually six minute six six minutes is enough. You just count how much time this mouse spend as immobile. The longer it the time is the more depressed this mouse is likely to be. Got it right. So what we see is our mutant moms spend more time as immobile indicating that they are more depressed compared to a wild type moms. Another assay we use to estimate depression is called sucrossse preference test. It's very easy. You give animals a free choice of two bottle. One is water, the other one is sucrossse. Just like humans, mice like sweet stuff. So if they have free choice, they're going to drink a lot of their liquid as the sweet stuff. As shown here, the wild time mom drink about 80 to 90% of the liquid as a sucrose. But our mutant mom have significantly reduced that preference to sucrossse. This is something we called anhidonia. This is another sign for being depressed. So both of this indicating that our mutant mom are depressed by this time. You may be wondering, okay, what about you look at this uh females when they're virgin, when they're not moms, right? So, we look at that same exact assay, but this time our mutants are wild virgins, so is their wild type controls. There's no phenotype whatsoever. So, this highly specific and then really we capturate almost a full spectrum of postpartum depression. So this basically also indicate trip 5 play a very important role to maintain normal mental health in the postpartum period and then the loss of function mutation of this this gene can cause postpartum depression. Now you still remember that 10% of fathers who claim that they have postpartum depression, right? How many fathers do we have here? Raise your hands. Don't be shy. Have you ever experienced postpartum depression? Okay. How many mothers do we have here? Okay. Do you have partner or your husband ever told you that they don't want to take out babies? They feel like they're depressed during when you are nursing. Okay. No. So when I read this, I don't believe it as well. But I'm not saying it just from uh this is from scientific papers. People did the epidemiology analysis and find that 10% of fathers have postpartum depression. But I doubt it because you can imagine how they diagnose it, right? So you give you hand over this man a a a a list of questions and then they can answer it and then you use the answers to evaluate whether they're depressed or not. But I bet everybody in this audience, every man in this audience can manipulate your answers in a way that you get whatever diagnosis you want it to be. Right? So that's why when I first saw that 10% number, I thought this 10% of men are just lazy. They don't want to change diapers. They don't want to wash dishes. Right? And then okay, I'm depressed so I can be excused for this. Right? So I'm going to show you one man that belongs to the other 90%. >> Yes, this is myself. This is myself about four years ago when we accidentally had our third child. I can explain why it's accident. But uh but you know, I hope you appreciate the fact that first of all, I'm not depressed. I'm pretty happy, right? And also I hope you admit that I'm pretty skilled at what I'm doing. Right? I don't know about all the men, but I know for C-57 black mice as a virgin, it's not something everybody's born with. There's a famous experiment in the field. If you took a cohort of C-57 black male mouse, virgins, and then each one of them, you give a few pups, of course, nonbiological pups, right? Because they're virgin. They couldn't have their own pups. Then you just observe how they deal with the pups. Randomly this cohort of male virgins is going to show one out of three different behaviors. Many of them is going to attack pups. Many of them is are going to ignore. Only a few of them will retrieve the pop just like the mon. But what is really magical is if you allow this male virgins to mate with female and they become fathers. Now you give them same pups again just nonbiological pups just to be fair, right? All of a sudden they all become super daddies. they re every single one of them will retrieve the pups. So this tells us this paternal behavior or fatherhood behavior is highly dependent on experience. So that's why when we characterize the paternal behavior in our mutant mice we need to look at it in two different stage when they're virgin and then when they become fathers. So shown here is first of all wonder version again our male version on the left wild type and a mutant male version on the right and I'm playing at 10x right so I hope you can appreciate that both of them are equally bad compared to our moms that we just observed in the last movie right because I'm playing at 10x I'm not showing you when they are attacking because that's not ethical so I'm just showing you the ones that they're not attacking right all right so Now, I'm going to um jump to the second half of this little movie and I'm going to stop here because I want to draw everybody's attention to this side of the screen for the next few seconds while I resume because if you miss those few seconds, you miss the whole thing. Right. There we go. Our wild type fathers on the left. He is done. And then look at our mutant father still working on it. So I hope you are convinced that this same mutation in males can also impair the fatherhood behavior right okay what about depression like phenotype so we look at force women test again when they're a virgin no phenotype but when they become father they are more depressed now I start to believe at least some of those 10% of fathers they are telling us the truth they may really experience postpartum depression because apparently we identify a genetic basis or mod molecular basis for such a thing that could happen. Right? All right. So we keep going because one of the things that really struck me is this dramatic behavior change from the virgin male to you know a father male right so what happened in their brain that can mediate such a dramatic change they don't take care of the babies at all now they become super daddies super daddies right we know trip 5 definitely is in the equation because trip 5 lost fun mutation can impair that behavior so we did a simple experiment we just did imunostating of trip 5 throughout the brain in two conditions One is when they're aversion. The other one is they become fathers. And very quickly we found this brain region called MA pre-optic area. It doesn't matter again what is called MA has a lot more expression of tripz5 when they become fathers compared when they're virgins. So this is just a demonstration and MOA is famous for knowing uh their function for maternal behavior in females and one of the important lure population that are for maternal behavior is those that express estrogen receptor alpha again because that's why that is a focus when I started my career. So just sort of by habit let's look at whether trip 5 and the yafa collocicalize in this part of the male brain and then what we found is uh when there are virgin there's about 60% of yafa neurons in this region co-express low level of trip 5 protein but when they become fathers almost 100% of yafa neurons in this region express very high level of trip 5 protein and then next question is whether this play a role in the the paternal behavior behavior that we observe. In order to do that, we have the trip flocks animal and then we cross with yafa cre flippo. So this cross itself does not need to any knockout because flippo doesn't work around lock. However, if we inject a flippo dependent cre into the ma region this will allow us to specifically lock out trip 5 only in ya positive cells only in this specific brain region. Right? And then we have proper controls. I'm not going to show you this but more importantly is we look at the paternal behavior. So this case we allow our controls and our knockout animals both become fathers by mating with females right then we look at how to deal with the pups again the wild type fathers as I showed you earlier every single one of them is super daddies they all retrieve pups but look at this when we knock out trip 5 in the mo yava neurons remember they're still fathers but we largely impair this behavior this is just quantification of you know how long how time they need to finish the retrieving jobs like mab is like a piece of cake finish in no time but our knockout takes forever. So they have impaired paternal behavior. Then we we asked at the time we thought maybe a crazy question. So if trip 5 here is required because our knocking out in fathers damage their behavior is also sufficient can we increase trip 5 in virgin males and then can we make this virgin male to show fatherlike fatherhood behavior. So that's what we did here. We made a a vector that express trip 5 in a c-ependent manner and then we inject this into the ma of yafa cremise. So this will overexpress 5 only in yava neurons in this region and then we have proper controls and this time we maintain both group as virgins. They're not touching any other females. Then we give them pups as I show you earlier. If it is wild type virgin males this is their behavior. Most of them either attack or ignore the pups. Only few of them will attribute the pups. But now look at this. Simply by over expression trip 5 in this small group of neurons in MOA in virgin male they're still virgins we basically made a molecular father they take care of their babies much better than the wild type again this is a quantification okay then the next question so we know okay trip 5 undergo this dramatic increase in expression when the male go from virgin to father and then we now show that this increase in trip 5 expression is physiologically important not only required but also sufficient to drive this behavior. Then the question becomes what is trigger trip 5 expression increase. With that question we look into the hormones into the circulation at different time points. virgin few days after mating or you know uh after winning um and then we look here I'll just show you two hormones prolactin we didn't see significant changes but testosterone this is the major male hormone in the circulation we found increase starting from 10 days after mating maintain to be high and also pick at 3 days after birth and then maintain to be high even after you know a few weeks after after birth right of the pops not not the males focus, right? So then we wonder, okay, maybe this activation of testosterone is important. Then to test that, we did a pretty cruel cruel experiment. So here we have our virgin wild type males and then we put them together with the females. So they're getting mate. Then once we confirm the female is pregnant, we cast castrated the males. So we remove their testes which is where the testosterone will be produced mostly right then it will still still stay with the the the mom and the pops and that will perform the behavior. Okay. Then this is the sham surgery. So this is still normal males or testies are still there. They still have balls and the testosterone. You see the the the the father is staying in the net and taking care of the pups at this particular time. But you look at castrated males and this is the retrieval assay. It take much more time to ret finish retrieving and some of them can't even not finish it. So testosterone seems to be required for this behavior. And then we know uh testosterone and the enzyme activity of this enzyme called raises can be converted into estradile which is the female home but you know males also have it if they have aromatus right and then we know that in moa there is a lot of expression of aromatus. So it is possible that the tachosone in this region can be converted to estra die. So then we ask a question whether this process is important. So we took the uh testosterone flux animal. This is the gene name for not not testo I'm sorry aromatase. This is the gene name for the testosterone. We took the aromatus flux animal and then we inject a cre into this moa region. So we prevent the conversion of testosterone into extra duct. So similar experiment uh setting and then we show this can impaired the potental behavior dramatically and then now okay so we have toning in the circulation it can be converted locally to produce estradiol and then we found this is all important so if estradile is important it has to act on estrogen receptor right so we know that this neurons express estrogen receptor alpha because that's where we started so then the next thing is did knock out estrogen receptor alpha in trip 5 uh positive cells in this region. So again we have yafa flops animal cross with trip 5 flippo animals that we made and then we can inject flipo dependent cre into this region. So this will result in a specific knockout of ya alpha in trip 5 neurons in this brain region and then we have proper controls and again this division impair the paternal behavior. All right. So then the next question okay so yeah alpha remember that's a nuclear receptor it can function as a transcription factor. So can YA alpha increase the trip 5 expression? That's where we started, right? Trip C5 expression increase dramatically. We look into the trip 5 promoter. Unfortunately, we found no clear yeah alpha binding motive. Instead, we found quite a lot of binding motif for other transcription factor called SP1. And it's well known in the field that SP1 can form a complex with YAFA protein. And then in response to ashra signal this can also stimulate gene expression. So to confirm this possibility we use yafa antibbody to perform chip assay and then we did find that this antibody can pull down the binding sites for sp1. suggest that yeah alpha does interact with this promoter region likely through SP1 interaction and in the cells we overexpress SP1 to look at trip 5 luciferous uh promoter activity u and we found no significant changes but when we overexpress yeah alpha you can stimulate a lot of uh trip 5 expression so put everything together I think the the model is like this so about trip 5 I hope you can remember three things number one is what makes a good father. So at the end of the day I realized that the initial trigger is still the male hormone testosterone. We need to have this elevation of testosterone in the circulation. This will get into the brain and through aromatase to be converted into estradi. Astradial can act on this receptor alpha which function as a transcription factor to interact with SP1 to stimulate trip 5 expression and trip 5 as a potassium channel can increase the excitability of this neurons and this mediate the potennal behavior. Number two, I hope you remember especially and I'm making voice for our men for our gentlemen is you know if any of this pathways is somehow damaged for example mutation of chipsy5 or maybe other things that damage this pathway this will impair the paternal behavior and then perhaps cause postpartum depression. So postpartum depression in male is a real thing. So that's number two. Number three, I haven't showed you yet. I'm about to show you now. So remember, we start all this inquiries about because we're curious about this postpartum depression, the depression, right? So for all the models that we had, we dump them in the water, the force swimming test in order to test whether they're depressed. Now, I'm about to show you perhaps the most crazy the craziest po uh force swimming test that you're about to see in your life. So this is the model that we overexpress 5 in yava neurons in this region. Remember this mice, although they are virgin males, they show fatherhood behavior. They can take care of pups, right? So, pretty amazing mouse. Now, we dump them in the water in this full swimming test. And the control on the left, the overexpressor on the right. Remember the things that I said twice. Mice do not like to be put in the water, right? But look at our overexpressor. What they are doing? He is diving. He's diving spontaneously. Have you ever seen diving mouse? I work with M for more than almost 20 years now. Other than this model, no other models that I work with ever show this spontaneous dive behavior. I can guarantee you that. And then you look into the literature, you probably won't find anything. I did it. So this really struck us. So we decided, okay, let's seriously look at this phenomenon. So the post made a big tank large water area very deep 28 centimeters which is about three times of the body length of the mouse and look at overexpressors you know I can look at this whole day without without getting bored but at the end of the day the question still remain I don't know what this mile is trying to tell me at biological level I don't know what it means because in the literature I can't find any relevance uh reference talking about mouse diving so I don't even know what to do so one of the task here is I show this to everybody and then if you can do force warming forcewing test in mice if you have that in your I cook protocol next time when you have animal pro have a mouse model please test this if you found similar phenotype email me we can work together for for interesting maybe a small but interesting paper. Okay. All right. Enough fun, right? All right. So, at the end of it, just want to share with you, you know, what is the big picture here? What what what I'm trying to do. Um, about 16 years ago, when I start my own lab at Baylor, my focus was how the neurons regulate feeding behavior, which is, you know, fundamentally essential for the animals to secure sufficient nutrient in order to survive. But over the years over and over again the molecules the genes the neurons the neurosircuits that we study for feeding regulation led us to make discoveries about other behaviors that are also important for survival. Like today I tell you trip 5 as example. We found that trip 5 can regulate feeding but also parenting behavior. Both of this are essential for the survival of the individual as well as the survival of the species. Actually, we have many other projects in the lab where we discovered overlapping mechanism for wide range of behaviors and all of this could be very important for survival such as aggression, mating, socialization, escaping behaviors. So over the years I have been convinced that our mother nature has prepared us a very small set of genes which I called survival code. The survival code act on a complex and overlapping neur network to co-regulate all the essential behaviors for the animals to maximize their chance of survival. And then this survival code is highly conserved through the evolution. can also regulate behaviors in humans. Indeed, many of the human diseases actually result from the deficits in this small set of the genes or the deficits in the neur network that were regulated by this small set of genes. And again, trip 5 is example loss of function mutation of trip 5 in humans associated with obesity, anxiety, autism and postpartum depression as I told you today. And then we can fully recapture all this using a mouse model carrying the same trip mutation. Because of this, we have a perfect entry point to first of all establish the causality of this gene with the human disease and more importantly we can use mouse models to review the detailed mechanism for the development of this disease and hopefully identify potential targets for intervention. So the big picture here or I should say my vision is to continue to tackle this survival quote with a goal to leverage what we learn from this fundamental survival behavior from animals to make a difference in the diagnosis, treatment or prevention of human disease. Okay. Finally uh just to acknowledge the people who did this. Um actually I moved my lab um uh from Baylor to USF last year almost exactly one year ago. I was very fortunate that the majority of my lab about 20 people decided to go with me and then when we were get there and then we we build the center we build this group to to have the the same size again and the work that I presented today is led by two really talented fellows now junior faculty in the lab uh Hano and Yun Jang and both of them are K K9 awardes and then I'm definitely going to try to push them out and then maybe in your department and then this work is supported by end and also a European research council synergy grant as well as NH grant and this is our collaborator Saddaf Ruki uh for the trip work and with that I thank you for your attention I'll take your questions if you have any thank you very much for that wonderful talk I am interested in that first part uh of the talk on the accord exposure >> and uh they increased insulin sensitivity >> and the animals were you said were hyperphagic >> and they were eating you know any was it any any diet they were they were hypopagic on any diet what was the composition of of of the diet is it >> with both no uh low-fat diet and highfat diet, it's the same. >> They will eat more. They will eat more. We test two different diets, >> right? >> The 60% highfat diet and also the boring child diet we call lowfat diet. >> Okay. So more so, so what do you think is the ins, you know, the signaling uh from that you know the brain area to enhance this insulin sensitivity? M and there perhaps many things that the brain signals will come to a number of peripheral metabolic uh tissues uh such as fat tissue, liver and muscle and I mean this has been uh well studied by other groups. there's direct innovations. Uh for example, what I can remember is the brain can uh innovate brown fat for example uh indirectly through multiple synapses and then influence uh the brown fat function even secretion from brown fat that can uh ultimately increase insulin sensitivity. Uh so our work does not really provide detailed uh pathways in the peripheral part but we build upon what is known that you know code exposure itself if you do this repeatedly in mice or in humans you do have certain benefits in incident sensitivity and know we did observe that yeah >> I was really fascinated by uh the data you showed at the beginning where cold um exposure to the neurons themselves in the brain uh recapitulated the whole body cold exposure uh phenotype. >> So what does that mean in terms of um practical application? I mean I wonder does has have people done experiments where they just cool somebody's head like go out in the in Wisconsin in the winter put on a hat? >> Yes. >> Is that going to affect >> um body metabolism even though the rest of your body is warm? I I don't know whether people have done that. Um but I'm hoping that people can at least experiment that but my from my perspective is I don't really think that is practical for humans especially in repeated manner. Uh that's why I think uh once we identify this group of neurons we continue to try to look for what is the sensing mechanism there. We hope to always identify a molecule that can be handled managed by pharmacological ways. For example, the case in K2, we identify SA but has channel and we show that this selective inhibitor does recapture a code response to improve insulin sensitivity. So, of course, you know, Spartan perhaps is protected by pattern. We can we can do much about it. But, uh, if I have the capability is we're going to use case 2 as a target to identify other uh inhibitors that could, you know, be developed into a medicine. Yeah, there there's a question in the back. The lady in the back. >> Thanks. I thought it was really interesting how you proposed that there was a small set of genes that regulate a bunch of behaviors essential for survival. But, uh, looking at your data, I thought it was really interesting how if you have a mutation in one of those genes, that causes problems with like so many different systems. So, do you have an idea about why evolution would favor something like that? >> Um, I wouldn't say that evolution favors something like that. It's just uh this gene is just the the gene is so essential that you just can't have you just can't survive better without it without the proper function of the gene. That's my interpretation. is unfortunately a very small portion of the individuals somehow they they have this mutation and then they start to develop all this really uh dramatic phenotypes which are not good. Uh so yeah I hope that sort of answer your question and instead I think the most of our this this normal function is um sort of conserved in order for the animals as well as humans to have the best chance of survival. >> Thank you for that. That was very interesting. I thought it was actually fairly well established that oxytocin is in charge of all the pup retrieval and nesting. But you and actually that oxytocin is induced in a sympathetic way in males and is kind of responsible for this behavior. But you didn't mention oxytocin. Where is it in your brief? >> It's in that paper. And so in females we characterize because the same mutation trips mutation in causes this postpartum depression like phenotype. And then in the paper we identify the mechanism is through the trip 5 expression in oxytocin neurons in the parent trigger nucleus hypothe. Um so that is reported this all in females that is reported in the paper. What we did there is we knock out chips5 in oxytocin neurons. We recaptulate the phenotype and then we did the opposite from the trip 5 mutant background. We overexpress a wild type ch5 in oxytocin neurons alone. We can rescue all this phenotype. So this in that cell paper that I I mentioned but in males this is the following study in males the first thing we look at is oxytocin we knock out trip 5 in the PVH and then we found the depression like phenotype but we do not see the paternal impair pur impair so that got us interesting this so we start looking for the trip 5 expression level in virgin versus father and then we found ma so right now we're still trying to wrap up this to to to to fill all the holes that we have in this paper. But I think this model emerge in a way that it's the same gene again survival code but in male or females they engage different or overlapping but not entirely the same neurosircuits to regulate this crucial behaviors. So that's my my my model for this. Uh I have two questions. The first is like how do you say any like uh why the KN2 confirmation change like uh is that directly activated by the temperature change or like it's indirect? >> Yeah. So this is getting out of my comfortable zone. I'm not a structure biologist and even not a biochemist. uh so and but that has been reported by others uh in early work so it's casing to was reported as a code sensor and I I can get back to you after I read the more details and how I guess at the functional level they can confirm that right you expose uh you post to the to the cold temperature you found the current changes right but whether they really show there is confirmational changes in iron channel structure I I don't remember >> yeah okay yeah got it thank you and the second one is uh I noticed that when you drop the mice into water uh it seems the father wild type mice become hyperier than the warden uh male mice like they the the immobile time become shorter compared to >> you're very good observer and you know I wouldn't be surprised I mean being a father is a happy thing I'm I'm serious I'm serious. >> So I mean it is it is the social bonding you know with the opposite sex or with your your blood it is general a very happy thing. So the postpartum depression um if they do have it I think something fundamentally is wrong either at the genetic level or neuroscircuit level. So I think they have they this patient do have biological basis for that. It's not they are lazy. >> Okay. God just thank you like wonderful. >> Yeah. Thank you. >> All right. If no more questions, let's thank you for very exciting seminar and thank you everyone.