 Good. Yep. Perfect. I'll start with that. Oh, here we go. Hang on a sec. There we go. So I have no conflicts of interest related to this presentation. Some of you know that I briefly worked for Nobel Nordisk, running their Type 2 Diabetes Research Institute in Oxford. And I've done some minor consulting and worked with a local small farmer on a totally unrelated project. And it's not a real financial conflict but I also helped start the Institute for Personalized Therapeutic Nutrition, which is a non-profit that seeks to translate some of this work I'll talk about today. So getting right into it and I will go a little bit quickly through the beginning part of this talk because a lot of it's published and I want to save some time for some unpublished data, which I'll discuss at the end of the talk. So we all know that there is an enormous number of people that have Type 2 Diabetes. Today's therapies for Type 2 Diabetes treat the symptoms of Type 2 Diabetes by sort of alleviating the work that the beta cell has to do or basically replacing it with exogenous insulin or bypassing its requirement altogether with SGLT2 inhibitors, et cetera. But none of these treatments that we have today have been shown to change the course, the natural history of the disease itself. And there's emerging evidence, which I might get to at the end of the talk, for the possibility of remission of Type 2 Diabetes with diet and bariatric surgery, but there are no drugs that we have out here that are disease modifying. If we could get to this larger pre-diabetes population, roughly twice the size of the diabetes population, perhaps we could protect the beta cells as they're on their decline here and before a significant hyperglycemia sets in. A lot of my lab is interested in regeneration and survival of beta cells, but actually today I won't talk about that work at all. I'm happy to answer any questions offline or on Twitter. Even earlier in the course of the disease, and there are probably 2 billion people on earth that are overweight or live with obesity, is a very complex time where you don't have hyperglycemia yet, but you have insulin resistance and hyperinsulinemia and adiposity. And if we could intervene at this early stage, potentially we could flatten these curves, if you will, and have an impact that could be disease modifying. The problem with this early phase is so many things are increased at once, it's actually quite difficult to determine what's cause and effect. A lot of your textbooks will tell you that insulin resistance and obesity are primary, and that hyperinsulinemia is a reactive consequence to the insulin resistance. I actually do believe this is possible. I know we know you can get hyperinsulinemia with drug-induced insulin resistance, and indeed with genetic insulin resistance such as folks who live with Donahue syndrome, insulin receptor mutations. But I want to talk today about whether or not it's also possible that the hyperinsulinemia in some cases could drive the obesity in insulin resistance, because the order in this causality has a big major implications for how you try to treat the disease early on. So the talk today will ask the question, is hyperinsulinemia, can it be causal? I'm not saying it's always causal, but are there conditions where it can be causal? So I'll start off with the human data and then move to our most models where we can address causality directly. I will also start off by saying that almost none of what I'll talk about today are my own ideas. We're truly building on the shoulders of giants, and so I'm recycling a fair number of ideas here. Some of these ideas haven't caught on I think as much as they should, but I'm thinking about Barbara Corky who in her banting lecture almost a decade ago asked hyperinsulinemia, cause or consequence, Walter Forreys who said diabetes, have we got it all wrong? And there's a number of researchers who advance the importance of understanding hyperinsulinemia and its causality. So what would it take for it to be causal? One thing is you'd have to see it before any of the other things. And these types of studies are correlative, so they have that issue. And I always have problems with correlative studies because it really depends on how accurately you can measure all of your different outcomes. But this is a quite a powerful study. These are hyperinsulinemic clamps on over a thousand individuals, this Trico paper and JCI, and a third of them had hyperinsulinemia that was in the absence of any insulin resistance. It also is correlative and predictive and this is work again by a true visionary in the field, Jesse Roth, who studied a long-term cohort and found that out of all the features that they examined, it was hyperinsulinemia that was the most predictive of which people would go on to get type 2 diabetes. I think there's other cohorts where it's seen in other directions, but this is an example of what can happen. Similarly, and I guess this shows that I don't have a massive pharma conflict of interest, but I think it's well known now that in all of the major trials of long-acting insulin, you do get, for example, weight gain. And there's some interesting work now coming out of Vanderbilt showing even in the context of type 1 diabetes that if you match for the glycemia, for example, of patients with Modi, that the exogenous insulin itself can drive some insulin resistance systemically. But the true test is whether something is necessary and that requires a loss of function experiment and this has been done in humans. It's been done quite a few times here. Here are some examples. One of the original examples is from a lesmede who used disoxide, which among a few things can directly inhibit insulin secretion from the beta cells and they saw weight loss in these hyperinsulinemic adults concurrent with what they saw in mice. Now, there are some doses and some studies where they don't see effects, but this has also been repeated more recently by loves at AL. I think this paper is only a couple of years old. But disoxide can have direct effects probably in the hypothalamus as well and can have effects that might be dissociated from its insulin effect. So what about insulin itself? Could we reduce insulin specifically? And that is an idea we had now about more than a decade ago and it led to most of the studies I'll talk about today. So the concept in these experiments and some of you have heard me go over this before, but so I'll go over it a little bit quickly. The concept here is to play with the insulin gene dosage is to control the amount of insulin that can be produced by the beta cells. And it turns out so mice and rats are somewhat unique. You have four alleles of insulin to play with, two from insulin one and two from insulin two. In this first set of experiments and the results are roughly similar regardless of which ones you study. This first set of experiments we've eliminated insulin two entirely to prevent compensation and we are comparing mice that either have one allele or two alleles of insulin one. So our control mice here have the normal complement of insulin one and you can see on the control diet here they gradually have slightly increased insulin with age. On the high fat diet here they become massively hyperinsulinemic and mice with reduced insulin gene dosage in the control diet at one year out. The mice with half the insulin gene dosage have half the circulating insulin and it's even more pronounced when we look at the high fat diet although they try to sort of mount an insulin hyperinsulinic response. By the time you're at one year of age the insulin effect is as if they were not even fed this diet. And what's truly remarkable is the mice can actually be reasonably happy physiologically with just one allele of insulin one and insulin one is the minor insulin. There's actually less insulin one made than insulin two. It's about 40% of the total. Aside from a time relatively early in life when they're sort of teenage mice and they're growing rapidly where you have glucose intolerance and this is exacerbated in SPF facilities actually you'll have more mice die in this phase. But these mice in this old conventional facility the vast majority of them made it through this phase and had basically normal glucose tolerance. What this means and this is an important thing here is that I'll define hyperinsulinemia for this purpose of this talk as the insulin levels that are in excess of those that are required to sustain glucose homeostasis. So when it comes to glucose this is truly extra insulin that I'm reducing at least throughout most of their life and this fits with what we know about the hormone the enzymes and the signaling pathways that control lipid homeostasis we know they actually respond to much smaller amounts of insulin than for example glute for trafficking does. And the bottom line of this first bit of research is that mice that were genetically incapable of hyperinsulinemia and this is in the absence of changes in glycemia were genetically incapable of diet induced waking. And the mechanism here is almost entirely a reduction in fat mass and fat pads and the other organs are basically similar size. You can see that instead of having swollen adipocytes they go back to being normal. The energy expenditure here is slightly increased and this is prior to weight differences and in the absence of any obvious changes in food intake. This is was very controversial at the time until some people that are more famous than me also got into understanding this but we actually purchased a we were relatively poor Canadian lab so we had to reuse our reagents so we got one Tacman mini array that we were going to use for all the different tissues including brown adipose tissue and it was to our great surprise that in the white adipose tissue we saw an up-regulation quite a strong up-regulation in ucp1, pgc1 alpha, p-par gamma, p-par alpha and obviously we now recognize this as the program for the browning of white adipose tissue and we showed this also at the protein level as well. The initial bit of work was the work of a talented graduate student Aria Maron who's now actually working for pharma company and it was followed up and I'll show you a lot of work from a really talented graduate student and now independent professor Nicole Templeman but she also later on noticed that there was more total amount of brown adipose tissue in these mice with reduced insulin. More recently a visiting fellow Diego Botticelli from Brazil looked at the fat pads at the much younger mice and saw an up-regulation of oxfoss enzymes in these so there's a few things going on that could contribute to increased energy expenditure. Regardless of that in the mice that have reduced insulin gene dosage and less fat you also see less lipid spillover into organs such as the liver and you see less inflammation so this puts hyperinsulinemia upstream of the inflammation as well so we see less markers of stress and ER stress and this is in white adipose tissue and in other other tissues. So the conclusion from this first part of the talk is that hyperinsulinemia is required for diet and reduced obesity in this model for inflammation and for the lipid spillover and this means that rather than simply being a consequence of insulin resistance and obesity here we can promote it up to being causal and I think that there are judging from the human data there are cohorts of humans and ages and groups that would that their pathway to diabetes would go through this as well and this means that theoretically if we catch it early enough and I was talking about flattening the curves before COVID-19 but now it sounds now it sounds even cooler but yeah theoretically you could actually do disease modification and end up with a situation where there's never any hyperglycemia. So the next sort of third of the talk I'm going to focus in on some work on insulin sensitivity and this will include some data that is brand new I saw it for the first time on last Wednesday. So what happens if we reduce insulin production to insulin sensitivity? In a lot of hormone situations we think of receptor desensitization a lot it's only in type 2 diabetes where we really don't think about that concept very often and the idea that the resistance happens before the high levels of the hormone as I said it's kind of unique to type 2 diabetes most other endocrinology systems and neurotransmitter systems we tend to think of it the other direction. So this is Nicole's really large cohort of mice this I'll spoil a little bit of the secret here but the reason these mice the cohorts were so large is this was part of our longevity study that we did it's also the reason you don't see things like clamps because those are terminal experiments but if you take these mice now insulin one is gone and they have two alleles or one allele of insulin two insulin two is actually the one that's conserved in humans for what it's worth and these are the females you can see that the reduction of fasting insulin is persistent it's quite a minor effect actually there's not a huge difference in fasting insulin and the very first hopefully everyone can see my cursor but the very first effect you see makes a lot of sense mice with less insulin have slightly higher glucose we wouldn't have seen this if the ends weren't 20 or 30 mice per group here but this this fits so a slight reduction in circulating insulin leads to leads to slight hyperglycemia but through most of their lives and this is the two diets diverging here this pink one here is slightly higher in fat it's about 60 percent 58 percent they're not matched diets so don't compare them directly and this one here is the breeder chow so it's got about 20 fat it's not a low fat diet at all but what you can see is that fasting glucose is basically perfectly normal in these mice they have that little bit more insulin on board so they never had the hyperglycemia in their teenage years like that other model did but if you wait long enough eventually out at 80 90 weeks you will start to see mice with less insulin have lower fasting glucose and the reason is that these mice have improved insulin sensitivity and this means that part of the reason for the age-induced insulin resistance is the hyperinsulinemia itself and this fits with the idea that insulin can down-regulate its own receptors so how does it do this more recent actually current senior graduate student in the lab Howard Sen took this on as his project he's been interested in understanding in non-islet cells and he focused his attention on muscle how does hyperinsulinemia affect the insulin receptor and actually these are not these are the types of experiments that were done 20 or 30 years ago as well and hopefully we're adding a little bit we've rediscovered them and we've added some unbiased mechanistic analysis but essentially this is also not a new concept so the way this experiment works if we've differentiated these skeletal muscle cells in culture these are a cell line and then we give them zero low or what is actually extremely high insulin levels we can take some measurements here we can also starve them for six hours and then do acute insulin stimulation and I won't show you all the data this has been uploaded to bioarchive and a very early version of this but we'll be doing a newer version soon but this is an insulin resistance model they do get resistance to the AKT pathway primarily but what's really remarkable is the insulin protein levels and we also know it's the insulin protein on the surface from biotinulation studies is dramatically reduced before starvation and after starvation and even if we look in vivo in some of the mice like I showed you before that have variations in their insulin circulating insulin levels you see this negative correlation with insulin receptor protein on skeletal muscle the mechanism for this or a substantial part of the mechanism for this appears to be down regulation at the mRNA level here you can see that the A and B splice isoforms are reduced they're the mechanism for that which had already been shown by some others and other cell types is at least partially due to the phosphorylation of Fox 01 at T24 that's an association but we think that that's important but what we wanted to do is find some additional novel regulators of hyperinsulinemia induced insulin receptor down regulation and so Howard did an RNA-seq experiment with a with a decent number of replicates we've got five replicates per group here and we have before serum and after serum and you can see that the the cells that are treated with the hyperinsulinemia are well separated and they they are separated from each other as well reassuringly the pathways that are altered in these in these treated cells are what you would hope for Fox host signaling insulin signaling map kinase signaling pf3 kinase signaling these are all the pathways in here and the individual genes you can scan through the interesting ones here but there's quite a few really interesting genes in here and we also used a tool uh Howard used a tool uh on networkanalyst.ca good old Canadian website and this tool takes your RNA-seq dataset and predicts which transcription factors are um are upstream of it so it helps it helps we were trying to understand the signaling between insulin signaling and all of these changed genes and so we got a list of predicted transcription factors these are the squares the big ones around the circle and then we cross-referenced it to things that were significantly changed in our own data set and that gave us 11 and then we did RNAi um knockdown in all 11 and uh we saw some regulators of irs2 uh several of them were regulators of irs2 but i'll just focus today on syn3a which we showed is a negative regulator of insulin receptor and so that leads us to this model here where probably both through foxo after the prolonged hyperinsulinemia and through syn3a and probably quite a few other factors will down regulate the insulin receptor and thereby reduce the amount of insulin receptors on the membrane and reduce insulin signaling so that's a mechanistic explanation for the insulin resistance uh so does this have any bearing to what is found in people so when covid struck i suggested to quite a few of the students that they learned some bioinformatics and uh you know it's really a wonderful time in science because so many people have done so many very powerful omex experiments and just put the data um onto the internet so we uh we're hunting through human skeletal muscle biopsy RNA-seq data sets to look uh both at the insulin receptor and other um genes that are in that uh similar uh network and and just to to look to see what correlates with uh circulating insulin levels and the two studies that we focused on so far this is a somewhat smaller study but very well matched very well phenotyped people from uh from Denmark from the Jensen group and these have controls so normal glucose as well as a drug treated they called them insulin resistant so these people are on insulin so some of the variation of insulin levels is exogenous and some is endogenous here and these are on oral anti-diabetics um and and a variety of those but when you look at them all together or even when you look at them uh individually you can see a really nice negative correlation between insulin and the insulin receptor mRNA and actually quite a few things including the IGF1 receptor and then there are some negative regulators of this pathway these genes over here which go the opposite direction we also got a hold of a data set which is the Finland UK uh United States rather fusion data set and um we uh it took us actually quite a bit of time to get all the data security stuff set up and uh Howard had to kick people out of the office because he's only allowed to use one computer for this but you see again this now this has hundreds of people but you you see again a um a correlation uh between insulin receptor gene expression and insulin also IGF1 receptor also IRS2 interestingly this is a stronger correlation in males than females in this larger data set so the conclusion from this uh second uh chunk of the talk is that hyperinsulinemia is sufficient to cause insulin resistance in part through insulin receptor gene downregulation which may include FOXO1 and SYN3A as molecular mechanisms so um I know there's a lot of expertise in lifespan analysis in Wisconsin so I wanted to talk about this and about uh cancer which is also obviously a major determinant of iceband both in humans and rodents but it struck me that these mice presented a unique opportunity to look at the insulin ligand and its effect on longevity so if you're familiar with the worm and fly literature you'll already know how the story goes again my ideas are not new it's been known for a couple decades now that decreasing uh insulin signaling or it or the insulin like peptides in in these invertebrate model systems can double lifespan and uh there's some evidence pointing a bit towards that in mammals but uh no one had looked at insulin specifically and a lot of folks had focused on IGF1 insulin is the ligand that we have the most control over it's acutely regulated by diet and a minute-to-minute level and uh and by fasting and and other features like this so I wanted to look at insulin itself so let's look at some of the um circumstantial evidence around the insulin signaling pathway that gave us uh some confidence here to look into it um there is evidence that IGF1 receptor uh can uh can extend lifespan in a heterozygous form shown here um all these studies are very difficult and they require a lot of money and a lot of mice and um you know sometimes you get different answers in different facilities but generally I'm trying to suggest there's some sort of some smoke around the area um there is also evidence in uh IRS1 um this is the replication paper withers and and and partridge again uh some sex specific differences and a slightly different effect on IRS2 um what's important here is that insulin and IGF1 could use the IRS's so it's not diagnostic for which is the ligand of relevance we also know that um uh caloric restriction or dietary restriction is associated uh with longevity this is seen in in in many many uh models I learned from Dudley today that there's even more nuances to this than than I'd appreciate it which is fantastic and then inhibitors of the joint insulin and IGF signaling pathway obviously Rob Mison as an example um here you can see it given late in life here there's a bit of a difference already and here um you can see some extension after the uh uh the treatment but what about insulin itself so we go back to these mice reminder that this is actually a pretty small manipulation they have only you know 25 reduction in uh in in insulin and for my money it's one of the smallest uh manipulations where folks have tried to look at lifespan uh we at both the time points we looked we could not see any difference in IGF1 levels or in any of the other hormones uh peptides we looked at uh reminder that these mice even when they're older are slightly leaner and you can see this in body mass and fat mass and protected from hepatic statuses um these are two-year-old mice roughly and here's the take-home message from here is that mice with reduced insulin gene dosage these are shown in the lighter colors have an increased um uh median lifespan and if you take the maximal lifespan of the top quintile it's also an increase in maximal lifespan. Every single mouse in this study went to a pathologist and it was blinded uh the pathologist wasn't blind but you know they didn't know which mice were which and they were trying to um assign a cause of death and this is obviously difficult because uh their mice and uh getting at cause of death retrospectively is not it's not that easy but they did their best and we we really couldn't find one specific thing that the mice were being protected from um there were hints there was you know a little bit of protection general protection uh from what was identified as renal degeneration it took a long time for the first mouse on this higher fat diet to get their first malignant cancer that was interesting to us um but generally the mice were kind of protected across the board by having less insulin so the bottom line here is that uh modestly reduced insulin can improve health and extend lifespan in mice so that cancer result and a few other ones which I didn't have a chance to show you um got us thinking about how do you test this directly obviously we were underpowered to study any particular cancer because the mice have a variety of types of cancer and the controversy and the interest was growing over what's called the insulin cancer hypothesis so it's well established that obesity and the early stages of type 2 diabetes are linked to multiple um cancer cell types but it's not really clear whether it's uh the hyperglycemia uh the inflammation that is associated with both of these a hyperinsulinemia a hyperlipidemia some it could be any of these factors um but we do know that hyperinsulinemia is an independent risk factor uh even as shown down here in non-obese uh participants uh so there was a whole bunch of human um epidemiological sort of circumstantial evidence uh suggesting a link and this is across quite a few cancers the link was particularly strong with pancreas cancer but um it's also uh the obviously the site of the insulin production and there's a sort of dual causality with pancreas cancer because both pancreas cancer and its treatments can cause diabetes as well so pulling apart the the causality uh there is a little bit tricky so this is again where we turned to our animal model these are now mice again with two alleles or one allele of the insulin one gene crossed on to um a mouse model of uh the early stages of pancreatic cancer so this is the pancreatic asinine specific uh Cree ptf1a uh driving the uh k12d mutant keras and this is exactly the same mutation that's responsible for 90 percent of human pancreatic cancer so it's really a um relevant model and this model was really spearheaded by uh janelle cop who we were fortunate enough to recruit um to our uh department a few years ago and she's been absolutely um instrumental in uh in these studies and teaching uh our joint student annie jong uh how to do all the pathology so as you can see here the mice again have a pretty small difference in um uh circulating incidence about half uh it's about half as much but you see a lot of very everyone knows there's a lot of variation in in insulin levels uh so much variation here in at one year it's not actually significantly different but what we know is for sure not different either significantly or numerically is the hyperglycemia so the female mice here there's absolutely no difference in um in glucose uh homeostasis and the mice again i'm not a pathologist and for those of us not used to doing cancer uh research when i first saw these uh pancreas sections i was like what it was the grossest thing i'd ever seen but yeah there's basically very little asinine left and this this pancreas is all filled with these uh pre-cancerous lesions called panions and panions come in uh grades one two uh three and and on and so the uh reducing insulin reduced the number of panions reduced their grade it reduced the um the inflammation shown here it reduced the um uh mucin um uh levels as well and this is quantified here in pancreas panion plus tumor area over a total pancreas area again mice with about 50% less insulin have about 50% less of the of the any of pancreatic cancer we're setting up to do this in xenografts and in models that have more advanced metastatic pancreatic cancer when you add in a p53 point mutant for example but the conclusion of this section here uh is that modestly reduced insulin uh can significantly lower the incidence of these pancreatic cancer precursor precursor lesions so i just want to spend uh some of the rest of my time and i want to leave lots of time open for questions to talk about um some of our unpublished data looking at the mechanisms so one of the things and we're we will have the data soon is uh we don't really know whether this is a direct effect on the tumor cells or not we think that that's part of it um but we're also doing the um insulin receptor knockout in the tumor cell uh population and um those mice are being analyzed or the sections from those mice are being analyzed as we speak but one of the things we wanted to do in our model was try to take a stab at which cell types and which mechanisms might be changed in the control uh hypersalamic condition versus the mice with reduced insulin levels and the first step of course just to simply ask what cell types have insulin receptors that are relevant and the answer is basically all of them so this is um from Annie's uh single cell RNA seek data set this is um uh from the KRAS mouse but you see the same thing if you roughly the same thing if you look in the human immune and the human pancreas data sets and you look for the insulin receptor um mRNA levels uh you know it's quite abundant on fibroblasts a variety of uh T and B cells macrophages um you know uh assener cells themselves doctal cells many cells will have insulin receptors and I also remind you that insulin receptors as with all membrane proteins you know you actually don't need a lot of mRNA uh to have um to have uh protein levels and that made me uh you know I drew this for a review that I'm supposed to have submitted already but yeah you can just imagine that the hyperinsulinemia coming from the eyelid cells could really act on the precursor to the cancer cells here the assener cells the doctal cells the blood vessels neurons any of these immune cells baroblast they all local adipocytes they're all potentially insulin target tissues so Annie uh did a single cell RNA sequencing experiment uh word of caution uh you know I think those of you who've already tried this know this uh but it is uh the pancreas might be the worst uh tissue to do single cell RNA seek uh because of course it produces enormous amount of RNAs and um and getting out intact cells um is is tricky uh so we did our best we sorted for live cells in all cases uh so we fact sorted them after isolating them and um so what we think here is that the gene expression is primarily uh from our our live cells and we also did some post-talk because exploding a post-talk sort of um gene garbage cleanup if you will because exploding cells even if you only take uh pure happy cells they're the surrounding media contains RNA from cells that didn't make it so there are some programs now where you can look for example if there is um uh one of the highly abundant um assener gene mRNAs and you start to see a little bit of it in every single cell you can informatically uh clean that up but either way we have well demarked cell types here um if anything the um the reduced insulin gene dosage made the cells uh maybe a little bit less um uh or sorry the hyperinsulinemia rather made the cells a little bit less mature in that they're the markers that they're famous for we're we're less abundant um we also see this interesting group of cells that are proliferating cells they sort of clustered amongst themselves and this is a group of proliferating cells which I don't have time to describe today but actually contains multiple cell types in here we're analyzing those ones as well but I'll just zip through some of this data because I think it's kind of interesting this is the type of data you get you get obviously RNA seek data it's not that deep so you're looking again mostly at the top sort of third express genes but you get some interesting insights you see um uh you know you can then take your individual genes and run them into the reactome pathway generator and you can get um significant pathways if you look across all the different cell types and I promise you I will blow this up in pieces so you won't have to try to read it but what you can see here is that uh for some of the pathways so these are the cells across the bottom and these are the the reactome pathways uh vertically here some pathways were changed in lots of lots of cell types so down here in this blue group you've got pathways that were changed in a lot of cell types and then you have up here at the top the opposite pathways that were only really dramatically changed in one or two different cell types so let's try to walk through from the bottom up some examples here so uh translation and mRNA processing were changed in a lot of different cell types more so in these proliferating cell population than in some of the other ones but you can see here the the uh reactome terms that that were changed signaling uh which was changed uh this you know these are a lot of a cluster actually of different pathways trying to make it a little bit more simple but here you can see map kinase and a variety of other signaling pathways you know as you would expect from um regulation of insulin and hyperinsulinemia uh this is just a grab bag of others um protein metabolism so here we have um assimilation, nedilation, uh deubiquenation, etc etc you can see there's some cell types you have changes in but not others general uh substrate metabolism here as well you can see differences the this proliferating cell group which also incidentally had the most pathways uh significantly different um seems to have an alteration in um in its metabolism a little bit here in the B cells and the dendritic cells as well there is obviously we have a lot of immune cells here so a lot of immune pathways came up matrix reorganization seemed to be relatively specific to fibroblasts DNA repair relatively specific to the proliferating cell uh group uh cellular response to stress senescence etc etc hypoxia was lightly seen sort of across the board cell cycle as expected in these proliferating cells but also in in B cells and and and not also in the asthma cells we saw a signal for more proliferation using uh Ki67 in our in our paper but it wasn't significant it was just a little bit under in all the different cell types so the conclusion from this section here is that modestly reduced insulin uh can significantly and differentially affect uh cell type specific gene expression so my overall conclusions here I'll just leave them up and um uh you know is that we right now we don't have a way of modifying the course of type 2 diabetes I didn't uh have a chance to show you but we and others are looking carefully at how uh different diets including uh low carbohydrate diets and that's what the IPTN is working on we have a a really nice paper just about to come out um uh building on papers from others but this is a randomized clinical trial which uh shows you can get people off of their medications essentially reverse or remit their type 2 diabetes um so you know I would never say you know insulin is good or bad obviously if you don't have enough insulin um you'll have diabetes but there's certainly evidence here that a lot of us are potentially carrying around more than we need and I hope I've provided some evidence for how this could cause insulin resistance and obesity and talk a little bit about the effects on cancer I want to spend a few moments to thank the really wonderful people who did this work most of what I showed today is the work of Nicole she is now a candidate research chair at University of Victoria she just started her lab um a month or so ago fantastic student I showed Aria's work Diego's work Annie's work Howard's work a lot of wonderful collaborators and funding agencies and I really want to take a moment to dedicate this lecture um to the late uh Susie Clee whose office is two doors down from mine um and uh unfortunately she passed the summer suddenly but her expertise in in genetics much of which she uh gained in Wisconsin uh with Professor Addie was really invaluable for for all of our studies and she made a major contribution I would like everyone to know that we'll be hosting a symposium in her honor in mid-late November and keep an eye on my twitter feed or email me if you want to know exactly when that is it might be the 18th of November anyway with that I will take any questions thank you Jim this was fascinating a great talk and I'm sure there will be many questions and the way that we will handle this questions is that you can either write your name to the chat or you can raise your hand and I will let you talk while we're waiting for the questions Jim is there a correlation between um or has anyone shown in the hyperinsulinemia and EMT epithelium the zenchemal transition oh my goodness uh there's a correlation with hyperinsulinemia and a lot of things but that um EMT is an area I don't know very much about so I I hesitate to answer but uh where there is hyperplasia and where there is cancer there's the correlation with hyperinsulinemia um so I really wouldn't be surprised um you know we're with collaborations you know there's the major um correlation with PCOS as well and and hypertension and we're always we we send the mice to lots of different people and collaborate so they're really a powerful tool to get at the specific effects of of insulin per se but I don't know the EMT question sorry we have questions from the audience Jason is going first Jason can you unmute yourself okay oh yeah hi Jim great talk um I had a question with regards to the mice um you showed the data that high fat diet really increases the insulin which is a little bit different than humans because dietary fat really has no effect on insulin production really what's what's the difference in mice because I see a lot of these conclusions where people talk about high fat diets therefore being bad but it's like but in rats that's a great question and in a longer format of my talk I really dig into that so that's a really astute point so um in adult humans the vast and we actually have a large series data set that that we're also soon to publish where we've done isolated islets dynamic analysis of insulin secretion by para fusion on now 120 different preps of human islets so we're kind of taking a survey and it is correct that the majority of humans have little to no response to free fatty acids as adults now actually there's probably some response pre-weening because the majority of you know breast milk is has a lot of lipids in it so very young human islets may actually be a little bit more like mice and when it comes to to mice you know I wouldn't I wouldn't take yeah I don't like taking the diet one to one I think what makes a mouse hyperslemic and what makes a human hyperslemic are actually different as as as you would expect them to be the the diets the natural diets of a mouse and the human are actually quite different um but but I believe that the the consequences of that hyperslenemia are conserved throughout mammals and then as I said all the way down to to flies and worms so um if you interestingly if you go really really really really high so if you go to a diet a mouse ketogenic diet which is basically giving the mice just essentially lard um you can't uh you can suppress insulin secretion so it's just that that shape of that bell curve is way uh way further in a mouse so um it it also depends which fat um but yeah so that you're you're 100 correct that the um uh the way that you can induce hyperslememia is different in mice adult mice and adult humans um but the probably the consequences of that hyperslememia are I think conserved pretty well across species okay next question is from Heiko Likert is joining from Germany I can allow him to talk if there's a technical glitch I will read his question from the chat if you can unmute yourself you can hi Heiko hey Jim can you hear me I can you sound fantastic yeah I saw it on Twitter obviously that you gave give a talk here in Wisconsin joined in uh 10 30 now night time but was really fun listen to you um Jim I was wondering the insulin resistance I think is super interesting upon the hyperslememia and Howard's data looked interesting in the muscle cells and so you know it it it was really fast that the insulin receptor also Egypt one receptor got downregulated like six hours but then in demo in your in your models it seems like the hyperslememia or blending blending the hyperslememia take much longer to have an effect on insulin receptors is that true it is true in that model we have another one which I didn't have time to show which we haven't published yet but um if you give the mice a high sucrose diet which also causes massive hyperinsulinemia um we can see this the separation in the so you give the mice with normal insulin uh gene dosage the sucrose on top of the diet I just showed you you see the hyperinsulinemia you can prevent that very quickly again but the insulin resistance that we see we see it very fast and we see the prevention very fast so in the sucrose model um which I apologize if I had five hours I would get to it but in the sucrose model we see it in weeks instead of almost two years so I I think there was something specific about those other diets that somehow um masked it you know they they're probably getting insulin resistant for other reasons then the high fat used in that other diet um is it kind of an old school diet I think it's a large so maybe they were getting some insulin resistance that is hyperinsulinemia independent I'm sure they were as well and that's that's thanks thanks Jim I mean I'm asking because of our opposite negative regulator of the insulin receptor and we wanted to know how quick how quick really the insulin resistance you know appears off I'll send you I'll send you that data set if you send me an email I'll just I'll just put it on I will thanks Jim was great thanks sure thank you um Andrea is the next one under it yes hi uh it was a great talk I'm curious to know what your opinion about um neutralizing anti body to correct hyperinsulinemia um that's an interesting question um you know I didn't get a lot of traction um when I worked for the world's largest maker of insulin for approaching um approaching as a therapeutic angle uh the idea of inhibiting insulin um but you know the biological rationale for this is as good as many other things you just have to get it at the right stage now uh I mean as an experiment I think I think it's cool as a as a drug candidate it actually sounds a little bit more expensive um for something that you could do with diet and exercise so this is everything I mentioned diet but uh exercise cessation is actually a also really robust stimulator of hyperinsulinemia as well so um you know exercise and diet are things that are potentially require less infusions but but theoretically it should work and actually it's the sort of thing that I I would love to see someone try in a in a clinical model obviously there are there are insulin receptor inhibitor drugs but it but those cause hyperinsulinemia as well so both the the all the urco knockouts including the insulin receptor knockout on the beta cell as as we just published last Friday um but a lot of these things will cause a hyperinsulinemia and and so then it's hard to tease apart insulin resistance from hyperinsulinia but your idea um of the neutralizing antibody I think that's something that should be at least tried you'd have to do it in the right first preclinical model and then uh clinically you need to pick the right population and we know from the geneticists that there's maybe five or six different kinds of type 2 diabetes there's probably the same number of you know genetically different obesity is at least uh so eventually we'd like to get to personalizing these things where you you can figure out which people are on are on a track that was initiated by hyperinsulinemia which people were on a track that was initiated by something else you wouldn't want to give the neutralizing antibodies at the wrong stage of the of the progression of type 2 diabetes or could then be kind of productive okay I I mean I wouldn't if I had to pick I would probably test it first in the cancer model yeah yeah so you know there are people doing um well we are setting up to do a clinical trial uh in pancreas cancer is tough because you know they about half the people get diabetes from the treatment and from the cancer uh but it's not even known whether you can control glucose on this basis in in those people and so our first trial um sort of a phase one tool just to see if I can be possible uh but yeah cancer would be a place to think about that you still have the IGF um and the same thing that insulin receptor blocker drugs have been tried um but then insulin the ligand can still act through um IGF1 receptors at including at physiologically relevant uh doses it's a smaller effect but it goes through uh uh goes through some of the same thing okay next question from Matt Flowers hi Matt can you unmute yourself hey Jim that was a great talk thanks um I was just wondering from a acute insulin secretion standpoint have any of the gene dosage manipulation models ever shown insulin secretion to be limiting enough to impair the time to return to uh eucalycemia especially after a very large stimulus like a mixed meal tolerance test or uh oral glucose tolerance test that's a great question so I can add and try to address it two ways one is it really depends on what insulin you you have left over so we actually know now that if you have one allele of the minor insulin insulin one especially again in our new spf facility a lot of these mice can't make it so that's just it's only in males but there's it's just not enough so clearly clearly that is that is an issue what is surprising though is is in the other models including when you have one a little of inch two left or the females of um the one a little of inch one left which females as we all know are tougher and more robust anyway um what what is interesting is that most the effect we see is on fasting insulin in in many of them and and they still they still have a response to glucose I mean there's maybe a little bit less we didn't induce a model that was paged out in class seven the first phase the peak was exactly the same and you could see a difference in the second phase by verification but the there's not a gene dose limitation on that first acute response to a square wave glucose but it's the sustained response and it's the basal that we think are more gene dosage correlated with but that's a great question great thank you okay um are there any other questions I see some hands are raised but the those already asked questions in the chat now is there good questions in the chat uh there is one but you already answered that one right and there's a much talk now about inhibition of glucagon from eve to all participants if you want to comment on that uh glucagon in in which in which participants sorry inhibition of glucagon in type one diabetes uh I'm not sure but maybe because that's that what if you want to type two yeah you know what I I think that we are we are we have been remiss we've been a bit insulin centric we understand that there there are other hormones we just tend not to talk about them very much uh but yeah clearly though insulin is the only non-redundant meal regulated metabolic hormone that I'm aware of so it's like there's insulin on one side and then there's glucagon and a few other things stacked up on the other side are there defects in glucagon um and in alpha cells in actually maybe both types of diabetes probably um I and therefore I think you know an ideal therapeutic would address them I don't think they're like the root cause of the disease though I don't think that um you know the alpha cells are still primarily spared uh at least in type one but yeah it's a bit of a hand wavy answer but I think that if we uh if you were to ideally replace the function of beta cells say with a transplant you would want to uh you would want to have beta cells alpha cells delta cells um you know vasculature neurons surrounding cells you know biology the eyelid has evolved as a as a micro organ and um you know I I still believe insulin is is is really primary but everything else is fine tuning and to me type two diabetes is a disease of the loss of fine tuning you know it's not um you know it's not catastrophic you know loss of the whole pancreas it's it it starts off probably as as fine tuning gone wrong uh so adaptation and again in another longer version of this talk I would talk about how hyperinsulinemia initially probably promotes an increase in in beta cell mass in response to you know to adapt and to um uh to give you that um adaptability in in youth and as an adolescent however as an adult if you're stuck with a certain floor a certain minimum number of beta cells which always leak out a minimum amount of insulin that's your hyperinsulinemia floor so if you're if you are um what might have been adaptive early in life may end up becoming maladaptive uh later in okay well uh I think if there are no other questions from non-training attendees then we will thank you again Jim this was a great talk I'm going to be applauding on behalf of the other attendees and then we can leave but you stay here with the trainees I see Chris Simra raising their hands so I think you will get a great session with our trainees okay and you've recorded this so um we can put it up in a drop box and people can look at it yes yes it's thanks for doing that I appreciate that thank you very much awesome bye to everyone except the trainees who I will talk with ma'am I'll talk to you soon bye