Beta Cell ERKonomics. Functions of ERK1/2 in Pancreatic Beta Cells

By: Nihvcast

Good afternoon, everyone. Welcome to you in the masur auditorium and all of those listening by videoconference and i have already assured our speaker that the number of people who will soak up the science that she presents is considerably greater than the people she can see in front of us. Thank you for those of you who turned out in person because that's always a little bit more interactive that way. I'm delighted to introduce our speaker, melanie cobb who is the professor of pharmacology at ut southwestern medical center and holds the jane and bill browning, jr. Endowed chair in medical science. If you're watching the calendar, she was scheduled to speak to us on april 4, and if you weren't paying attention to what else happened that day, you might wonder what happened to that plan. Well, it was not exactly in her plan not to get on a plane that day and it wasn't in our plan to not haveler arrive.

It had something to do with a tornado that plowed through dales and disrupted the dallas-fort worth airport for quite a few days. So we were very pleased that despite that particular unfortunate occurrence, melanie was willing to reschedule and come back and speak to us today, may 30. We arranged a thunderstorm last night, which delayed her plane from landing for two hours. So, yes, the tradition continues. Hopefully, the flight home will be uneventful. Dr. COBB OBTAINED HER UNDERGRADUATE DEGREE AT THE UNIVERSITY OF CHICAGO, Ph.D.

In biological chemistry at wash u after postdoctoral training at mount sinai and a three-year period as a research associate at albert einstein. She landed a faculty position at ut southwestern and essentially been there since that time for most of the remarkable career she has carried out, studying a variety of issues, particularly well-known for her work on map kinases at ut southwestern. For a time she was the dean of the southwestern graduate school and now is professor and holder of endowed chair. For all of us who have struggled trying to understand map kinases and map kinases and map kinases and kinases and i don't know how many ks we are up to, it's a relief to have somebody explain this stuff. Not only doe we have a distinguished speaker member of the national academy to lead us through this, but she made seminole contributions to this whole field, particularly in the original identification of erk1 and erk2 and many other steps that followed now being applied as she will tell us about today in terms of their function in the pancreatic beta cell, most obvious implications for diabetes, a disease which is a virtual epidemic proportions right now in our country and increasingly so in the rest of the world. So it is a great pleasure to have her here today. Please join me in welcoming professor melanie cobb. [ applause ] thank you very much for the introduction and thank you for inviting me twice.

I'm glad i actually made it one of those times. Okay, so just briefly today i'm going to tell you all stories about pancreatic beta cells after a little bit of background on erk1 and 2. So, i think it's become increasingly clear that erk1 and 2 are involved in all kinds of processes in cells and in fact, i can think of few events that don't have some input from these particular map kinases. They are key in proliferation and cell lineage commitment and maintenance of differentiated functions and that's what i'm most interested in and why we looked a lot at pancreatic beta cells. They are misregulated in all kinds of disease syndromes, particularly cancer due to activation of upstream activators of the paths way by mutation and a number of cancers. There are a range of developmental syndromes that end up impacting the activity state of erk1 and 2. Some of these are among the most common developmental syndromes in people, and there are also negative regulators, tyrosine kinase signaling that are involved in causing erk impactive and developmental issues.

Beta Cell ERKonomics. Functions of ERK1/2 in Pancreatic Beta Cells

So, a number of years ago through collaborations with betsy smith and one of her postdocs and students who might be here as a fellow, we solved the crystal structures of erk2 and the unphosphorylated state here and in the phosphorylated state. IT LOOKS LIKE A TYPICAL PROTEIN KINASE, ONE OF THE 9s THINGS IS IT IS A RELATIVELY SMALL PROTEIN SO WE COULD SEE THE WHOLE PROTEIN. It's 41 killadaltons, it has a smaller end terminal domain and a larger c terminal domain and atp binds at the interior of the interface between the domains. It's activated by two phosphorylations and in the case of erk2 that activation is really profound causing about a 50,000-fold increase in p cat and the two phosphorylations are in this activation loop region here and they cause structural rearrangements in the protein to cause this activity increase.

A couple of points i'll come back to later a little bit about the map kinase insert. These purple regions are regions that are not contained in the core kinase domain. They are inserts and map kinases in particular. This insert here and winding up the back of the protein, we have a c terminal region that has an important 310 helix here and helix at the top of the molecule. In terms of the detection of activity while we use for many years in the complex kinase assay, now just about everybody uses antibodies to the phosphorylated forms of erk1 and 2.

A comb of reasons those are so good, the two forfoes -- phosphorylations sites are separated by a single residue and the second is, these are abundant proteins as we think of signaling proteins sometimes as perhaps somewhat rare but these may be on the order of a couple of 100 nana molor or micromolar in certain cells. So they are easy to measure and detect. And so, i'm going to talk about this pathway and i just for the sake of full disclosure want to point out that erk2 fits here in this example of the signaling network and it's very easy to focus on one little pathway and lose site of the big picture, which is all of these interconnections and so, more and more we are trying to acknowledge these interconnections and figure out how they really contribute to what all happened when this pathway is turned on. Okay.

So much of what i'm going say will focus on specificity in different ways and the first thing i want to point sought although this cascade can be activated by all kinds of ligands, almost every single agent that effects cell function will change the activity of erk1 and 2. There is one step where incredible specificity is generated in map kinase path expwais i have 3 here. This is the map kinase step and just to clarify, this is the map two kinase step or map kk. AND IT IS AT THIS STEP THAT WE HAVE INCREDIBLE SPECIFICITY AND THIS JUST SHOWS HERE PHOSPHORYLATION AND ACTIVATION OF IN THIS CASE, PHOSPHORYLATION OF ERK2 BY THE MEK AND ITS PATHWAY COMPARED TO THE MAP 2Ks AND THE MAP 38 PATHWAY. There is more or less absolute specificity difference. While lots of things turn on this pathway, they have to work through this particular cascade to do it.

So i mention the that many agents control this pathway and can change its activity. Almost a dizzying number and it's sometimes hard to get a grip on how different activating mechanisms can lead to different outputs. And what can you see here is old data b-20 years fold a whole bunch of different laboratories we worked on this problem. Activation of erk1 and 2 and pc12 cells by two different ligands. Egf is a rapid increase in activity that is transient and ngf, nerve growth factor that can be detected the next day. So the differences in ligands activation mechanism has a big impact on the output of the pathway so it's this pathway that is associated primarily with proliferation of pc12 and ngf with nurrite extension.

So phosphatases have been implicated as one mechanism for controlling this type of ligand-dependent kinetic specificity. And this will come up again when we get to beta cells in just a minute. So another important factor in determining what these enzymes do, they can be activated by a number of ligands but they don't always localize in active form at the same place. And i'll show a couple more slides about that but first i want to point out there are large number of scaffolding proteins that detect these enzymes and localize them in particular places depending on which ligands have activated them. So for example, ones that were particularly interested in are these nuclear poor proteins and 3 or 4 of these are known to be very tight binders for map kinases. There are some proteins like eea15, which keeps erk2 out of the nucleus.

There are scaffolds that localize to the cytoskeleton and various organelles, for example, and ones that have very distinct cell type dependent functions such as k -- ksr proteins. Scaffolding has a lot to do with it. This maybe one of the least-understood aspects of how these proteins are controlled. About location.

We started looking at nuclear localization along a number of years ago, quite a long time ago. And what we found was that if we used recombinant erk2 and perm brides cells, we could see that the protein localized in the nucleus whether or not we added an energy source or cytosolic transport factors. And here say study with gfp erk2 and also localizes to the nucleus. In this case, we compared to nuclear localization in constitution studies to nlsvsa import substrate used frequently to look at beta one dependent import. And you can see here that import of nlsbsa requires transport factors and requires an energy source and gets into the nucleus equally pretty much regardless of additional factors added. And import carriers compete with erk2 for entry because erk2 binds to a similar location on nuclear poor proteins as the import factors. Here, for example, is a schematic model of nuclear poor proteins on the nuclear poor and some of these proteins contain a whole string of fx-fg type of motifs.

These were shown by a lab, in particular erk1 and 2 and select activity and so erk can bind to these fx and fg motifs common on these nuclear poor proteins and you can see here just a pull-down assay showing binding of erk2 to the c terminal domain of nuclear pore 163. So 358 nuclear pore protein 358 and ppr have also been shown to bind and i believe there are two or three others that have also been shown to bind. So erk2 probably spends a lot of time around the nuclear pore.

Interestingly, most of the characterization of the localization of erk2 was done in transformed cells, the standard cell lines that people use in the laboratory like rack 52 cells. If we use erk2, you can see in the unstimulated state, there might be a signal from one cell. But if you look at cells stimulated with egf orester for 10 minutes, that much of the protein is localized in the nucleus although you can see some localization around the cell periphery in the case of egf. On the other hand, if you look at untransformed cells, cells that were instead immortalized but not transformed, like human foreskin fibroblasts, and a lot of other cells that have been immotor a.alides like this, you see something somewhat different. Stimulation with egf doesn't cause that pro mountained localization whereas stimulation withester does.

So in nontransformed cells which haven't been the models we use very much but by-and-large in non-transformed cells, the localization of these enzymes can be quite different depending on activating ligands. So, that's all the background. The rest of what i want to tell you focuses on the role of these enzymes eone and two and pancreatic beta cells -- erk1 and 2 -- one of the reasons we were interested in their actions and beta cells is that we found early on that they were important for regulating insulin gene transcription and i'll have a good bit more to say about that. One of the fascinating things about beta cells that we have learned is that erk1 and 2 are activated by all of the signals that stimulate insulin secretion and inhibited by agents that prevent secretion. So they are really wonderful markers stimulation beta cells to productive insulin release. I don't think they are directly involved in secretion or if they are, it might be a more minor role than some of the other roles, nevertheless, they are wonderful as a read out of stimulatory agents on beta cells.

So we used it and we learned a lot about how beta cells work using this model. So just to show some pictures of beta cells on one of my points here is that this is a colored-in model with alpha cells in red and beta cells in green. And you can see here as the mouse islet from a normal mouse and here is a human islet from a relatively normal human. And one of the things that you see is the ratios of beta cells are quite different and the whyed that all of the alpha cells are located on the outside of the islet doesn't really hold up when you look at normal human islets so much. And so, while we use human islets for all the studies we can to look at signaling, we try to reproduce everything in human islets, we have used beta cell lines a good bit. I think the point here is that mice might not be the perfect model to study control of pancreatic beta cell function. It's just architecture issue alone is important. So, i'll show this slide twice.

So the things i want to really point out is that all of these agents that activate insulin secretion, amino acids, glucose, ach, glc1, all activate erk1 and 2 and do it through calcium influx and they all require calcium dependent fasfoe taze to be able to activate erk1 and 2. Agents such as epinephrine which acts through the alpha 2 add nergic receptor and turns on gi and gi blocks insulin secretion and blocks erk1 two activation by all of these agents. So all of these nutrient sensing events depend on calcium neuron. Agents that activate erk1 and 2 through more conventional tyrosine kinase ligand-type mechanisms, they don't require it. They can activate erk1 and 2 whether calcium is in its presence or inhibited. Somehow there is insulation between the glucose nutrient sensing pathways and the typical pathways that we think of as regulators of erk1 and 2 and most other cell types. So, first some evidence for what i'm telling you about this nutrient sensing insulin demand select activity. So here is the comparison of erk1 and 2 activation.

Here is total amount of erk in the lie pays of these cells and then a measure of activity using abilities to phosphoerk1 and 2. You can see 11 or 3 mill i molar glucose which is low. 11 is pretty high, glucose concentration. Erk1 and 2 shows activity although it is increased at 11 relative to 3. If we add a cyclic amp increaser, a more or less mimic of glp1, we get a potentiation of activation. When we looked at this concentration in more detail, we see erk1 and 2 are actsivated maximally by about 8 or 9 mill i molar glucose. And resting concentration, we have now been using about 4 or 4 1/2 mill molar glucose instead of the standard for a normal glucose tollerrance concentration -- you can see here, we still see activation at 3.

So i guess the important point is that eone and two are activated in beta cells under normal conditions. The amount of activity changes but it goes up and down over a fairly small range. One reason why we can't do all of the experiments we have done in islets is the following. Here is what we see if we look at alpha cells and both are cultured cell lines. If we look at alpha cells, you see erk1 and 2 are low at high glucose and high at low glucose.

So, just the opposite of what happens in beta cells. So alpha cells work an opposite way. So if we look at whole islets, you can imagine that the signal that we would see from beta cells will be mitigated somewhat reduced a little bit by the or maybe a lot, by the contribution from alpha cells.

And here you can see an activation very nice activation by insulin in alpha cells. Okay. So alpha and beta cells don't work the say same way and if we want to understand the signaling, we are somewhat restricted to using beta cell preparations or isolated beta cells from impact islets which we have done on numerous occasions. So, the issue of nutrients and secreted activating erk1 two by mechanisms different from growth factors. Here is evidence to support that. And it's important to point out that sq506 is an inhibitor of calcium neurons and basically this is the molecule that will prevent erk activation and this will dampen calls newspaper activity. Here is glucose activation. If we add calls neurin activity to that, we see a significant see expression of activation by glucose.

If you look at glp1, you see something similar, stimulation of erk1 and 2 with glp1 and a blockade with calls in neurin inhibition. And this is very clear. This is true with all the agents that enhance insulin secretion, all the nutrients and hormones. Okay.

Growth factors that activate erkgon two and any cell you look at. Here you can see forma lester blocking. No requirement for calls neurin to get the pathway turned on and here is egf. Nice activation of erk1 and 2 but no blockade by calls neurin. So nutrients in growth factors activate by mechanisms even though we presume they use all the same kinase cascade. Everyone thinks about controlling the pathway. What about secretion inhibitors? i mentioned before that epinephrine blocks insulin secretion from beta cells and that happens through alpha 2 ab nergic receptors and here is activation of erk1 and 2 by glucose. If we include an alpha 2 selective agonist, we block activation of erk1 and 2 and if we use epinephrine itself, we also block activation of erk1 and 2.

If we look at a helix cell where people have characterized how these pathways work, you can see that control conditions, there is no activity deductible. If we add epinephrine, we can activate erk1 and 2 and turn it on. If we add the uk, selective alpha two add nergic agonist, we can also activate erk1 and 2. So turning on gi in an average cell that is used for this kind of experiment will activate erk1 and 2 and beta cells that completely turns it off. And again, the growth factor insulation component of this argument, you can see here that glucose again is blocked bite alpha two abner gick agonist but there is no effect on that drug and activation of the pathway. So just putting that all together, we have all these nutrient sensing mechanisms, all of these mechanisms telling the cell where insulin needs to be secreted and they all require calls neurin in order to turn on this pathway. And this will inhibit insulin secretion, blocks erk1 two activation. So, i want -- one of the questions that really puzzled us for a long period of time was how the phosphotaze calls neurines is required to turn on this kinase cascade.

What is it doing? i'll show you evidence that it is working by dephosphorylating the map kkk or map 3k up stream of this cascade. And all of this work was done by a former graduate student,lingling. And the first thing she had to do was figure out what the map 3k was up stream of the pathway. There are multiple rav isoforms. Rav is considered to be the mab 3k upstream of eone and two. She assayed all three of these.

She couldn't measure any activity from ara. And here is what she got with craf. You can see basically no activity detected.

Just to show you she could really do the assay, here is a pma stimulated hela cell with the same type of cock police department assay phosphorylating the kinase substrate here -- on the other hand, immunoprecipitated b rav, activation by glucose and reduction in activity by calls immune. One ofone of the -- calls i neuron. This is the expectation was that craf would be the activator. It's the most commonly considered activator in a typical cell for the erk1 and two pathway and so whatlingling did was to immunopreimsipitate craf and show she could bring down the protein by blotting it and then also blot with an antibody to phosphosenior 338, a key activating phosphorylation site on the protein.

So phosphorylation is the site and would indicate it had been activated but if anything, the content decreased following expose tower glucose. So, another experiment here she did a time course of stimulation with glucose, brafip kinase again and treated with calls neuron -- cyclosporin a. She could reduce that braf kinase activity.

So that is the increased straightforward and then she did another experiment based on a lot of work from a number of laboratories initially from joe's lab but a huge number of groups have really presented incredibly valuable information that shows how complicated this kinase pathway can be. And that involved the idea that these isoforms can dimer ice, heterodimer ice and the partners in the heterodimer combinations both positively and negatively influence each other's activities. SO JUST TOASTER IF THAT WAS GOING TO IN BETA CELLS, WHAT SHE DID WAS IPd CRA.

And blotted for braf. Here is her craf blot. It was present in untreated cells or cell stimulated with glucose and here is her braf blot. It did not coprecipitation with -- coprecipitate but it did in the stimulated cells. So we have the issue of combinations of these molecules being present in some of the immunoprecipitation and obviously effecting each other's activity. To answer that question what she did was to use rnai in beta cells and this is not such an easy prospect in beta cells because just about the best knock down that is we have achieved in these cells was about 50%.

Here what she did was knocked down bra. And you can see erk activation is attenuated. This is her braf blot. This is the reduction from this to this. And here is just a plot of the activity. Here is cra.

And this one was a surprise. She stimulated and when she knocked it down, she actually got an increase in erk1 two activity. So short-term what that suggests is that consistent at least with other data in the literature that craf could suppress the activity of braf in a relatively rapid assay. Interestingly, when she looked at a couple of hours of stimulation with glucose, she found the opposite. In this case, craf was activating and forcing more activity through braf. So, a complex time dependent response of these kinases to activation by nutrients. So, how was calls neurnin this smith what was it doing? she figured out that calls neurin essentially dephosphorylates a negative feedback site on bra. And that site needs to be dephosphorylated in order to allow continued activation of braf.

If the site phosphorylated and that site isn't reversed basically, braf will be turned off and erk can't be activated any further. So a couple of different experiments to put that idea together some of which depended on determination of braf phosphorylation sites by debbie morson's lab a couple of years ago. So, here you can see she stimulated with glucose and here she is assaying braf activity using phosphorylation of kinase meth one again and there is the increase in activity. And she blots calls i neurin and the activity is decreased. We know braf is sensitive step.

If she blocked erk activity, she does not get the -- she still can measure braf activity, and if she inhibits calls neurin and blocks erk activity, with the mec inhibitor, she doesn't suppress braf activity any longer. By blocking erk she is preventing the feedback inhibition. And here you can read the red lines if you couldn't follow what i said because it's fairly confusing. So, she identified a site that cemented to carry most of the inhibitory activity and that was 401. So she then expressed braf in cells and compared the wildtype protein to the alanin mutation and so here you see stimulation by glucose and depression of activity by blocking calls neurin with p401braf, she found a stimulation of activity higher basal activity to begin with, and there was no suppression by inhibiting calls neurin presumably because this site can no longer be phosphorylated stow doesn't need to be dephosphorylated.

So all together that gives us a model which is calcium influx enhances calls neurin activity and it's important function is to dephosphorylate braf so it can continue to be lead to activation of erk1 and 2 by nutrients. So, next, a few slides on what we have done to look at erk1 and 2 and regulation of regulation of gene transcription and we worked on this for a long time but a lot of the work was done by a former graduate student and then much of it has been done by a former postdoc. And this was the key experiment that michael did that really made it important, every detail of how we did these experiments became important and so, what he showed here, it was well-known that glucose could suppress gene transcription but we didn't know anything about the time course. So what he did was to show that glucose activated the insulin gene promotor on the short-term. This is two and 6 hours. By 12 hours, he found a suppression of that stimulation and by 24 hours, and thereafter, he found suppression below basil transcription left -- level.

So that wasn't two surprising. The surprising part is we blocked erk1 and 2 activity, we reduced the gene transcription in response to glucose so the -- so we were basically close to basal levels at this point. But if we blocked erk1 and 2, he restored the suppression caused by high glucose at long time. So we are looking at a biphasic effect of glucose a multiphasic effect of erk phosphorylation. We block our activity early and prevent a stimulatory effect on the promotor, we block erk activity late, we prevent the inhibitory effect of glucose by the insulin gene promoter. So this became really a key to how we did many of these experiments that come next. So we are really basically comparing rapid or somewhat acute stimulation with glucose to a more chronic long-term stimulation.

And this just shows the first experiment that we did in collaboration with mike german and steve griffin at ucsf. If the cat reporter as a n-this case it was in freshly isolated mouse islets and here you could see relatively low promotor activity whether erk is added and transfected. What he stimulated was 16 mill molar glucose. You can see increase in activity and when he blocked erk1 and 22 with a mec inhibitor and including kinase, pretty much potently blocked activity from this promotor in the islet. And so, over a number of years, we and many other people identified substrates that are transcription factors that work on the insulin gene promotor and the important factors that have been identified are neurod1, pdx and mop a. Interestingly, two of these, particularly neurod1 and to some extent pdx one, are important points of substrates and then in fact, neurod1 requires a phosphorylation in beta cells to function well. Erk also phosphorylates p300 that's been shown in the literature.

It phosphorylates the common general ebox dimer partner for normal d1, e47 and phosphorylates cebt beta which is an inhibitory regulator of insulin gene transcription and that has been shown for a number of years for example by work from habner's lab and it can phosphorylate the c gene. We tried to figure out how all of these factors inner plate -- interplay with erk signaling to regulate the promotor. One of the important things that is a nonissue if you're working on islets from animals but it's a significant issue if you're looking at cultured islet cell lines, the maintenance conditions for these cells generally the cells are maintained at very high glucose concentrations. We wondered why that was and so michael lowered the concentration to what would be considered a somewhat more normal glucose concentration and what he found was after a couple of days, he lost expression of mop a and he kept the cells in very high glucose. Inhibiting erk activity with the mec inhibitor doesn't do anything to this. So the glucose concentration itself has a big impact even on a relatively short-term of expression of mafa. We know that gene expression goes up with low glucose. You can see here and again, by habner and a number of other groups have shown that cp beta is induced when cells are exposed to very high glucose for a long period of time.

So you can see here 48 hours on high glucose is enough to induce the bait a so it's important to treat these cells on a particular way so you know what the state of the relevant factors is that will be controlling promotor activity. Okay. So in this case, what we did was to look at what happened to factors that we know are important, mafa, neurod1 and pdx1 and cpd beta. These are preimsipitation assays so you can see bands. And these are cells treated for 30 minutes with glucose and compared to the control, you can see increases in binding of mafa and neurod and pdx to the neurogene promotor.

If we use the mec inhibitor or calls neurin inhibitor we block thosish actions. It's just a substantial reduction. Rap mice in, inhibitor of mtor c1 doesn't do anything. If we depolarize with potassium, we see binding of mafa and both are blow blocked by inhibiting erk 1 and 2 activity. If we look at those incubated for a couple days with high glucose without removal of that glucose, what we see here is that dropping the glucose for 30 minutes and then adding 11 milli molar glucose, no binding of mafa, neurod1 or pdx.

But we see binding of cep beta concentration, and it's higher in 11 mill i molar glucose. Interestingly, if we treat these same cells with potassium depolarization, we can still see binding of maf and this is just that the beta is preventing these other factors. So, one of the observations that we made that gotten us particularly interested in studying this in more detail is what happens with the signaling molecules themselves here we did chromatin immunoprecipitation on human islets with antibodies to erk1 and 2, antibodies to downstream substrate risk two which can phosphorylate histone tail and calls neurin and the phosphotaze. You can see we can see irk one and 2 are associated with the insulin gene promotor under stimulatory continues and we can see here with glp1. If we add fk interaction is reduced. We can see that under a number of different conditions and with other promotors so we are now exploring the density of interactions of erk1 and 2 with chromatin and beta cells.

Just for comparison, again my point about ligands, if i haven't made that strongly enough yet, here again we have cells comparing glucose eg. And ngf as activators of these cells and looking at activated erk1 and 2 by immunoblotting. If we block mef, we can lip limb nat the signals and reduce them substantially. Here we are looking at just chromatin immunoprecipitation and you can see that we are comparing the insulin and the phos promotors and glucose and egf don't activate very much and beta cells but ngf does and you can see a big difference in interactions and if we add to a 126 which blocks activity, you can see we can dampen glucose dependent interactions consistent with the chips i was showing before. We don't really inhibit egf signaling in these cells and you can still see a activity and you can see reduction in both types of interactions in response to ngf. So different ligands do different things and that includes different dna associations with activated molecules downstream of those ligands.

So one of the things that we noticed was erk1 and 2 had an impact on histone acetylation and beta cells and this is to show you not so much about the histone modifications themselves but about the glucose dependence of these interactions. So here if we look under normal conditions, you can see stimulatory glucose and increase in acetylated h3 and h4 and increase in binding of p300 to the insulin gene promotor. If we block erk erk1 two activity, all of these are reduced.

If we look at cells that have been exposed to high glucose for a couple of days, the pattern is quite different. You can see here that under low glucose conditions we can see association of p300 and acetylation of histones and blocking erk1 and 2 at high glucose enhances their association relative to glucose alone. And this is basically a comparable type of experiment. Here we are looking at cytokine action and i'll get back to that in a few minutes. But the combination of high glucose il and beta cytokine, causes basically an inhibition of acetylation of histones and binding of p300 and binding of polymerase two to the insulin gene promotor. But if we block erk1 and 2 under these conditions, so this is a exacerbated chronic glucose situation, exacerbated with the cytokine, then we no longer lose the interaction by blocking erk1 and 2 activity. And this is more of the same.

This is in human islets. Again, preincubated in a resting glucose concentration we see pretty normal behavior with stimulation of histone acetylation as we add glucose, but if we incubate these islets in high glucose for 3 days, we get the same kind of reversal i was showing you before. Instead of increased acetylation as we increased glucose concentration, we see decreased acetylation and so basically, what we are seeing here is that the condition of the beta cell, not the activation of the kinase pathway, is a major determinant what have is on the promotor and how it can be regulated by the signaling pathway. And this might be the most obvious set of experiments showing that. So here we looked in more detail at what happens when cells are treated with either resting or stimulatory glucose under chronic conditions along with il1 beta. And so erk1 and 2 is activated by glucose but it's also activated by the cytokine. Perhaps a little bit differently under each condition but basically, all of these conditions will cause erk activation. The stress associated map kinases p38 and junk, these are activated by the cytokine.

So you can see here addition of the cytokine independent of the glucose concentration, activates these protein kinases although erk is activated under all of these conditions. So, now what happens? when we activate all of these pathways? these are time courses chromatin, immunoprecipitation and what you can see here is that the erk1 two bind right away and still associated even under this condition which i'll show you in a moment. Has turned off this promotor completely. P38 likewise jnk likewise and risk likewise and all the signaling molecules bind away when they are stimulated in assay and they stay there. What about factors that are important for insulin gene transcription? you can see mafa binds but in the presence of the cytokine with high glucose, it's gone after an hour. Neurod does stay on. Pdx one also drops off.

P300 and polymerase are also gone. So one of the things that happens, it takes about a day of high glucose to induce the bp beta but if we add il1 beta, we see increase in this molecule by a couple of hours. Here is the 4 hour time point. So it's a much faster induction of this inhibitory transcription factor. If we did nuclear runon at 4 and 24 hours just to see what was actually happening to transcription in the insulin gene promotor and at 4 hours in the presence of the cytokine but at moderate amount of glucose, you can see we still have transcription initiation ongoing so we could see nuclear run on. But if we looked at 16 mill i molar glucose at 4 hour time point which corresponds to this last column, you can see a suppression of transcription.

There wasn't any ongoing initiation so there was no nuclear run on. But we could measure. So, i guess the main point here is that the kinase may be bound with the promotor and they are regulated the same way regardless of the conditions in the beta cell. So if you lower glucose erk1 and 2 activity will go down. If you increase glucose, the activity will go up. But if the cells are treated under circumstances that promote their resistence, their glucose innocence activity, then what erk does in the signaling standpoint will not be the same because the protein compliment in these cells is different. Okay. So, a point of great interest to me but not necessarily immediately relevant to beta cell function is what is the molecular basis for erk erk1 and 2 chromatin association? we are working hard to,000 better.

First erk1 and 2 binds all kinds of dna associated factors. They bind a lot of hlh proteins, complex factors, p300 and many other binding proteins. They also bind others. Fsk, another histone kinase so it was reported at 3 years ago now that unphosphorylated erk2 can bind to dna. We find that phosphorylated erk2 is the better dna binder but we also see direct binding to a number of different regions of dna.

So we are exploring what that might mean. And one idea is that it may occur through basic residues in the map kinase insert suggested by this paper but also before that by a paper from maria schumacher who determined the crystal structure of a bacterial kinase has a eukaryotic kinase fold associated with bacterial dna. And what she found is this is the part of the kinase, the structure of the kinase looks just like this.

This is the part of the kinase that was touching the dna directly. And this is the map kinase insert which is unique to this particular branch of kinase. So far we have done a lot of sequencing and we have identified a number of interactions in beta cells. There are a number of interactions that overlap with p300 binding sites. This may account for some interactions we can detect. So overall, erk1 and 2 and transcription. Map k activities reflect in coming signals. They are on if they have been stimulated to be on and they are not if they are not.

They are present in transcription complexes. Responses depends on what is there. Changing glucose concentration and beta cells alters the profile of the transcription factors. So erkon and two regulate contacts histone modifications and other kinds of transcriptional responses and reduction in beta cells.

And briefly for the last couple of minutes, i'd like to tell but studies that have been carried out by a postdoctoral fellow in the lab. And the results of these experiments were quite shocking to me, very gratifying and i hope we can turn them into something more significant even then. They might appear here. So this is a collaboration with somebody in cardiology and professor j snider.

The chemical strene about 10 years ago almost for molecules that enhance cardiac differentiations and reporter assay and he found a group of compounds that started looking at how some of them work. He noticed the compounds could increase expression of neurod cells. So we thought it would be a good idea to look since neurod really is beta two. So we tried these compounds on islets and we get as many islets, human islets as we can handle and when we can't use them all acutely for experiments, we put them in the incubator and when he got this compound to work on, we actually had islets that had been in the incubator for a year. Human islets sitting there in the inch beirut. A this point they don't make a lot of insulin -- incubator -- they might secrete just a little. He started looking at them and he found -- this happens to be a 6-month-old set of islets.

HE FOUND THAT WITHIN TWO DAYS INSULIN mRNA WENT UP BY MORE THAN 60 FOLD AND THIS EXPERIMENT BY TREATING WITH THIS COMPOUND AND THAT WAS PRETTY SHOCKING. Interestingly, glucogon went down several fold. He did or looked at insulin secretion and you can see here the compound increased the amount of protein secretion quite substantially.

And so, a question is really this is an increase but is it significant? how does to compare to what a fresh islet would secrete? so he compared islets that had been in incubators this time for three months with ones that we had fresh from the transplant facility at uab. And he either treated with the drug in the light bars or not. And then he compared insulin content. PROTEIN NOT mRNA. And you can see here this is the amount in the fresh islets and after treating the old islets with the drugs for two days. You can see that it was in the right order of magnitude for insulin production from fresh islets and the compound increased insulin content of the fresh islets after only two days. So, this was a really amazing result to me. So he analyzed the actions of this compound in a number of ways and he looked at a lot of changes in gene expression and we start out with the three genes that are essential for insulin gene transcription, neurod/beta 2.

Pdx one and mafa. They are all in creased as a function of time after exposure to the compound. That was really good. But more surprising still was after again a time course of treatment with the compound he looked at glukokinaise and glucose sensing apparatus basically beta cells and they are increased a lot. Really huge increase in glukokinaise expression. Then he stood back and looked at the genes involved and the beta cell differentiation pathway and they are listed here and more or less in sequence of their expression during beta cell differentiation and they looked at everything on here and some of them are shown here. A lot of them show up in multiple states of course.

And you can see here, huge increases nkx6.1 and 6.2. Huge increases in neurojen path 4. These two factors are not highly expressed in mature islets as some of the other others and they went up very tremendously right at the beginning and then began to drop slightly following further time of treatment with a compound. So, everything he looked at basically increased with time of treatment with this drug.

And so he did look a little bit at histone acetylation some signaling in these cells following drug treatment. And this is a time course. And i want to point out that this time course looks quite different from what i showed you with glucose or protein coupled receptors and beta cells where we saw acute responses within the first few minutes. You can see a little bit of activation of erk2 here, 15 minutes. But tremendous activation at about 4 hours that is somewhat persistent although dropped somewhat. This looks like ngf-type of response although it's not very pronounced early. So it's quite delayed. Doesn't look like any obvious signaling to me.

You can see a time dependent increase in the acetylation of h3 and h4 and here not such a big change in h3k27 trimethylation. And so this is men 6 mouse beta cells and here we have human islets and you can see to the extent that he looked, these changes are quite similar. Interestingly, with when i deal with this compound it would effect histone transferase activity but an actual fact there was no effect on -- sorry. One idea that everybody expected was that it would effect hdac activity. It didn't. But it did increase hat activity, treatment way drug compared to tsa. And you can see big increase in hat activity and nuclear extracts from the cell. And when he reconstituted p300 and looked at a effects of erk on activity as well as, you can see drug increased activity relative to untreated and that erk1 and 2 did somewhat suppress activity compared to the drug treated without erk inhibition.

So, we have compounds that seem to really enhance beta cell function. The improved islet function in culture and transplant ef seas and diabetic mice. We haven't done enough animals yet, treated them to have statistical significants but it seems very likely based on our current data that the drug can be used to treat animals and increase their glucose tollerrence. What we don't know that the point is what are the mechanisms of action. We are looking at this very intensely and we'd also like to know how long the effects are stable. If we take the drug away, do the effects stop? once we remove the drugs, how long will beta cell function remain improved? and finally we are looking at similar compounds that seem to work at lower concentrations. We were looking at 10 and 20 micromolar compounds. We would like to be looking at nanomolar concentrations so similar compounds to see if they improve type ii diabetes or perhaps rejuvenate islets for transplantation or stabilize them for transplantation.

So, just all of this came from the interest in erk signaling and beta cells and we are hoping it will go further than that. So people who did the work, here they and are a lot of people in the lab have contributed in particular, kathy who is actually maintained islets in culture for us for about 10 years. Thank you very much. [ applause ] thank you very much. We have time for questions.

Please use the microphones in the aisles so individuals watching by video can also hear. This is very interesting story. It certainly raises the question of what is the earliest response that happens when you have the compounds? you showed us a variety of transcripts that seem to be going up fairly dramatically. Have you basically done a complete global analysis? to see what happens in the very earliest moments after expose and you are see whether you can identify what might be driving this in terms of the primary offense. no. We haven't done that. We started -- we have done some initial microaway experiments so we haven't done anything global or quantitative.

What we tried to do is identify direct targets by affinity chromatography and we have some candidates but i'm skeptical so far because i'd like compounds that bind a lot better than more efficacious before we -- more pelt ent before we really get serious about -- potent -- what potential targets might be. But i think that is really our goal at this point. Initially, they are to show efficacy in animals and try to identify direct binders. thank you for complete coverage of the kinases involved.

So we understand about the signaling pathways. Why are there so many drugs for type ii diabetes? besides metformin? i'm interested. why are there so few drugs for type type ii diabetes? and cardiovascular complications. and complications. I'll tell you one reason why i think there are so few drug is because we don't really understand much about beta cells. I have preconceived notion busy how these cells would be regulated and they are incredibly fascinating but they didn't exactly match my preconceived notions in a number of ways. So, they don't have properties that exactly match anything else that has been well characterized. And there aren't exactly a lot of them.

Other than looking at these cultured cell models in rodents, it's quite difficult to look at beta cell regulations, particularly because the other cells in the islets don't work the same way. So i think it's, we just plain old don't know enough. The more we know the better off we will be. But understanding more about receptors on beta cells is very important and learning how to manipulate those receptors so we don't desensitize the film. One of the things that is quite striking about beta cells is you treat a beta cell for 10 minutes with something and the effect can be very good to treat for an hour and it will start to be bad.

Not good. So, they get desensitized or rewired or whatever or how far you like to think of it, very quickly. Which means prolonged therapies may be difficult to achieve. It's not a drug company concept to look for short assay agonists. and regarding the kinase, and specificity. Do you have any effect? [ indiscernible ] i continued understand that. on the kinase -- inhibitors -- okay. About kinase inhibitors.

I think one of the major things i have learned is that kinase inhibitors probably aren't a good idea here because the signaling is very responsive to what is outside but beta cells change depending upon what is outside. So you can still grow beta cells in 25 mill i molar glee cows, take it away for 15 minutes -- glucose and add glucose back and erk will be turned on but the the beta cell will have negative regulatory factors and won't have any of the things -- some of the things they need in order to control what they do properly. So under those conditions, only bad stuff will happen. Inhibiting the kinase may or may not make a difference. So i think what i learned is that the glucose regulation of what is in this cell and the kinase regulation are distinct. Erk1 and 2 regulate insulin gene transcription but they don't control the compliment of factors expressed in the beta cell that can regulate any of these transcriptional events. that was related to my question.

So you saw that with long term glucose treatment, you see the change in transcription factor profile. So if you treat with your small molecule isx, do you see an inhibition of that change that you see with long term glucose? because it seemed to be the opposite effect with your molecule. so i will say that we didn't look at cebt beta, at least i can't recall the data if we did. We probably did but i don't remember it. But all of the factors that are important to stimulate or stimulatory transcription, they are all initudes.

And what happens? these cells -- induced -- the cells and islets are maintained in high glucose. So adding the drugs still has positive effects even though the cells are maintained in high newicose. okay. One follow-up.

Is the small molecule the effects you see with that, are they sensitive to calls neurin inhibition? let us thank our speak are for an elegant presentation. [ applause ].


Dacarbazine is an antineoplastic chemotherapy drug used in the treatment of various cancers, among them malignant melanoma, Hodgkin lymphoma, sarcoma, and islet cell carcinoma of the…

Views: 372 By: Audiopedia
[Biophramaceutics] Therapeutic Hormones Part 1

1. Hello, everyone. My name is Minsoo Choi.I’m in a laboratory for Systems Biology and Bio-inspired Engineering. I will give you the lecture of front part of chapter 11: therapeutic…

Views: 153 By: Biology and beyond
Endocrinology – Hypoglycemia: By Robyn Houlden M.D.

Hypoglycemia is a fact of life for most people with diabetes on glucose lowering medication. However, it is an uncommon clinical problem in other patients. There is no plasma glucose…

Views: 1 334 By:

Allotransplants is the transplantation of cells, tissues, or organs, to a recipient from a genetically non-identical donor of the same species. The transplant is called an allograft,…

Views: 142 By: Audiopedia
Meet Darrell Gray, II, MD, MPH, Ohio State Gastroenterologist

My name is Darrell Gray II. I'm a gastroenterologist at The Ohio State University Wexner Medical Center. I had the blessed opportunity to work with a lot of gastroenterologists…

Views: 1 447 By: Ohio State Wexner Medical Center
Endoscopy at Ohio State: Why and What to Expect

The gastroenterology department does endoscopy. Endoscopy is actually inserting a flexible a tube into the gastrointestinal tract. This allows us to visualize the lining of the gastrointestinal…

Views: 115 By: Ohio State Wexner Medical Center