Thanks very much. Um, thanks to Linda for putting together this really phenomenal meeting. I'm really enjoying myself, and thanks to the last speaker for setting some of my data up. I didn't know how integrated this would be, but it's uh, I'll get to the question of the mour pathway in the middle of the talk. Um, I've been asked to talk about IGF-1 and growth hormone action, muscle, and bone. And what I want to do is share with you some of the data we've put together over the last number of years, which really tried to dissect these two pathways using genetically altered mouse models. And by showing you these examples, I think it'll stimulate discussion, and I think they're revealing in certain ways with respect to what a growth hormone can do independently. This is a collaboration between Doug G. Ger Alo at Johns Hopkins and Karen Esser at the University of Kentucky, and it's supported by a grant from NIAMS, and that's the only disclosures I have. I was going to show um a slide to

justify the importance of the growth Harmon IGF-1 axis. But I think we'll stipulate, if the audience will permit me, that it's important in building both muscle and bone. What isn't known is, again, what I alluded to, and that is how these growth factors, these anabolic growth factors, how do they work together, or can they work independently through their own receptors. And that's exactly what we set out to do, and that's not a trivial distinction because growth hormone continues to be used in aging clinics and in off-label use to enhance performance. So, I think it also could have, if one could figure out the signaling pathways, there are clear differences in the way these two pathways act, and that's what I'll try to get at today. Now, the main problem, of course, is the interconnectedness. So, you have growth hormone made in the anterior pituitary acting in target tissues, the liver and elsewhere, through its receptor, JAT receptor, which can activate both PI3 kinase and STATs.

in particular, stat five, which is the stat that activates igf-1. So, igf1 is made in the liver and many other tissues in response to growth hormone. That ends up in the circulation or in these local sites. It activates its own tyrosine kinase receptor, which activates a common set of relatively common set of signal transduction pathways and produces very common effects in many tissues. So, the problem separating these two interconnected pathways, and that's why we've used genetically modified mouse models using prelock recombination to test both in vitro and in vivo the effects of these two growth factors. The two models that have been historically used to explain this interconnectedness are listed, are shown illustrated on this slide. The oldest one is called the somatomedin hypothesis, referring to igf-1 being a somatomedin, and that suggests that growth hormone increases igf1 production from the liver and from local tissues. And then the igf-1 actually is the main

somat ameden accounting for most of the effects of growth hormone arriving either from the circulation or locally. Uh, there is a second model, uh, called the Dual Affector Hypothesis, proposed by Green in a number of tissues, conver sites, and fat, and that suggests that growth hormone might have independent effects, uh, to increase the differentiation of initial progenitor cells, and IGF-1, again arriving from the liver or local sites, could expand that initially differentiated cell. And so, we use these two models as rough paradigms for what we tried to do, uh, using the genetic techniques. So, let's first deal with the bone effects or the skeletal effects of the two, uh, pathways to, uh, delete the growth hormone receptor gene in, uh, convites. We crossed mice carrying flock toal for the growth hormone receptor with mice expressing the cre combinat off a collagen 2 promoter, and we determined that these, uh, the deletion by using laser capture micro dissection. Uh, you can see the, uh, we, we removed most of the

growth hormone receptor but left intact here the IGF-1 receptor, which was important for the conclusions of the study. So, this just shows the remnants of the growth plates, which we laser dissected out. Surprisingly, the growth plates of the mice lacking growth hormone receptor were entirely normal. Mice grew normally, had absolutely no long bone phenotype, growth plate phenotype that we could detect at all. And other studies, which I don't have time to show you, suggest strongly that this is due to upregulation of local IGF-1. In addition, I remind you that IGF is available in the circulation. This probably rescued this phenotype. This contrasts markedly to what's seen if you delete the IGF-1 receptor using exactly the same genetic techniques. And I show you data from Dan Bley and Wan Chang, who are actually in the audience. On the top shows that the deletion using the collagen 2 promoter in a constitutive form ends up with mice even showing a growth retardation before birth and then after

birth, and many of these animals don't live. But if you use a tmox, an inducible collagen 2, you can show that the phenotype in converse sites in the growth plate that lack the igf-1 receptor, uh, there's a decrease in the proliferative capacity in the proliferative zone of the convites, ending up with a shorter hypertropic Zone. And data, again, I won't have time to show you from their group, suggests that there's an interaction with the igf1 receptor signaling pathway with pth, Indian Hedgehog signaling. So, suffice it to say, though, that igf1 receptor signaling is critical for growth plate elongation, whereas, surprisingly, growth hormone receptor is not and is dispensable. So, in terms of, uh, oh, so, so in terms of growth plate, it's the samatam Medan hypothesis that really abominates here. So, what about in bone and osteoblasts? And these are, uh, I'll just show you a few examples of work that

Doug Djalo did um basically uh we had made a uh osteoblast-specific uh knockout of the IGF1 receptor, a mouse that we made, and that uh model was used by Doug to test the effects whether growth hormone could actually rest resue any of these effects that we see, we saw in this MTH. One of the effects was is shown over here um and that is that if you if you delete the IGF-1 receptor from osteoblast, you end up at 3 weeks with a decrease in osteoblast number, that's shown here about a 50% reduction. So what Doug did would then come along with growth hormone and give fairly large doses of growth hormone to this animal that couldn't respond through the IGF-1 receptor, and he saw a marked increase in the control animals uh of growth hormon increasing osteoblast number, but absolutely no effect on osteoblast number in the in the mouth that lack the IGF-1 receptor. Again suggesting that the ability of growth

hormone to stimulate osteoblast proliferation, or the lack thereof. The r in this case was due to growth hormone stimulating at least igf-1 signaling. Let me just go back up a little bit if I can. So again, I won't have time to show you all the data, but he characterized the growth hormone receptor knockouts, and suffice it to say that they never really had any distinguishable phenotype different from the igf-1 receptor knockouts, any site that we could see. So in terms of growth hormone action in bone, the somatotrophic hypothesis holds in our view. We haven't looked at everything; I can't rule out other effects, but in the growth plate and in osteoblasts, the ability of growth hormone to stimulate anabolic effects is due to igf-1 signaling. We believe now I'm going to move the same experimental paradigm to muscle and spend most of the time talking about the same type of approach in

muscle, and I thank the previous speaker for introducing some of these concepts. We began our work with in vitro studies looking at the ability of both IGF-1 and growth hormone to stimulate in vitro cell proliferation of myoblasts isolated from mouse muscle, and then the fusion steps, which we'll get into. So, when cells are placed in culture, there's a... if you withdraw serum, proliferating myoblasts fuse in two stages. 24 hours, there basically... there you can show two stages of this. There's an early accumulation of myonuclei and then a second further accumulation. Mature myotubes form from that point on. These myoblasts obtain, you know, are phenotypically real muscle myoblasts. And here's some genes that are expressed, that they expressed all the components we're interested in, including the ability of, in this case, growth hormone to stimulate IGF-1 mRNA shown here over this time course. Now, so essentially, growth factors can stimulate the... the...

accumulate muscle in its final mature myotubes by either increasing the proliferative capacity of early myoblasts or increasing the Fusion steps so that more nuclei accumulate. We tested these two pathways in the in vitro system. The first was just to look at the ability of these two growth factors to stimulate proliferation, and we did this very acutely to make sure we weren't seeing changes in IGF-1. So in these cases, we're looking at BrdU incorporation at 6 hours in myoblasts, and growth hormone had absolutely no effect, whereas IGF-1 stimulated BrdU incorporation. I just remind you that this is before one would see an accumulation of protein. This is mRNA here, so before you would ever see accumulation of IGF-1 protein, growth hormone couldn't affect the proliferative capacity of these myoblasts in the same way. Then we looked at Fusion events, and we could separate these two into two separate phases if

growth factors including IGF-1 and IGF-1 were added for the entire 48-hour period after withdrawing serum. You get these preliminary Fusion steps and then an accumulation of myonuclei. Growth hormone added through the entire 48-hour period stimulated accumulation of myonuclei measured by the fusion index, which you saw previously from the previous speaker. However, if you added it during the first 24 hours and then withdrew the hormone and then measured Fusion index at 48 hours, there was basically no effect of growth hormone. However, if you added it during the second left it alone the first 24 hours and added it after withdrawing Serum at the after the second 48 hours, then growth hormone could actually stimulate Fusion, suggesting that growth hormone worked during the second phase of fusion. Then the question was which receptor was it? Was it through growth hormone receptor or through IGF-1 receptor? And that could be tested because we had both myoblasts knocked.

out with the IGF1 receptor and growth hormone receptor. Here's the control Fusion index. If you add growth hormone to cells, it lacks its receptor. Of course, you don't get anything. There's no receptor. IGF-1 can still activate Fusion or increase myonuclei accumulation. If you take cells now that MYBL that lack the IGF-1 receptor, there is no effect on Fusion again, suggesting that growth hormone stimulates Fusion in this case and expansion of the differentiated myotube through an IGF1-dependent pathway. We know that this effect of growth hormone IGF-1 actually involves NFATC2 and I4 accumulation. But seeing the previous speakers talk, it gave me a lot of interesting ideas about further dissecting that pathway. So, the conclusion from these in vitro experiments is that IGF1 stimulates MYBL proliferation, whereas growth hormone promotes through an I1-dependent pathway the fusion and differentiation of the mature myotube.

and both of these, uh, pathways probably intersect, at least in part, the nfat C2 pathway. So now we turned our attention to in vivo models, and again using clock recombination techniques, we made a mouse, analyzed two mice, one that lacked the growth hormone receptor. I'll show you that first, and then compared that to one that lacked the IGF-1 receptor in mice expressing the cre recombinase off the mef2 cre uh mouse that was made by Brian Black. And this is really a nice mouse because it really expresses uh cre only in uh muscle and skeletal muscle and not in the heart, as so many of the other earlier um muscle-specific promoters did. Uh importantly, again, if you measure growth hormone receptor ablation down here, you see that that's at six and 16 weeks. It's gone, whereas IGF-1 receptor, importantly for our studies, remained intact. So this promoter very worked very well for us. Histomorphometrically, mice that lack the growth hormone receptor had... This is basically histomorphometry.

standing with Myas and heavy chain, showing type one fiber specific type one fibers, and essentially the fibers had less SP nuclei, and there were less type 1 fibers in the mice that lacked the growth hormone receptor. There was also a decrease in the diameter of individual fibers. These are six-week-old animals, and the phenotype really remained out through 16 to 28 weeks. So, basically, the decrease in my nuclei and the narrowing would be consistent with our in vitro results, suggesting again that growth hormone, probably through IGF1 receptor, at least from the in vitro studies, had effects on fusion and accumulation of myonuclei. Functionally, this would be expected to produce mice that are weaker, and we tested this in very crude assays up to now, just grip strength in an apparatus like this. The growth hormone receptor knockout mice had decreased muscle strength, as shown by this, and then a rotor rod test, which is really a combination of balance, muscle contraction, and probably neuromuscular.

reactivity. Uh, also, they fell off these rotor rods at a greater rate than the control mice, again suggesting that functional impairment of this muscle that lacks the growth hormone receptor. Now, surprise came when we started to analyze older mice that the growth hormone receptor knockout in muscle and that they became fat. Here's a mouse that lacks the growth hormone receptor at about 10 to 12 weeks. They separated from the controls, and this really had everything to do with increased peripheral fat and really didn't change as far as we could tell lean mass. I might tell you that these animals had no change in bone density whatsoever despite changes in fat mass. Um, they basically, we immediately checked whether the food intake and energy expenditure was different, and they weren't. The animals ate about the same, had no differences in energy expenditure compared to controls, and this was done over a 4-day period using metabolic cages. They did however have high

serum glucoses and basically high triglyceride suggesting the uh the onset of insulin resistance. And this was actually confirmed using glucose tolerance tests and insulin tolerance tests. So these animals became severely resistant to the action of insulin. Getting to the question that was raised earlier, cells from these animals in culture also were resistant to insulin. This is just glucose uptake by 2-deoxy glucose in response to insulin. The control cells had a robust increase in uptake. Uh, it basically, if I'm sorry, this is just showing that insulin, uh, in the insulin-stimulated glucose uptake, and the control was, was, was, uh, was markedly increased. And if you basically had myiss that lacked the growth hormone receptor, this was severely blunted, uh, in the cells that lack the receptor. So Doug set out with a bunch of antibodies and a textbook and said, Where in the world is this insulin receptor pathway defect, defective or desensitized? And he basically found that every single step was downregulated from the top so

you basically the IR by IM immuno blotting was down, but this is the important probably thing, and it brings the point that somebody raised from the audience. There's a Serine on irs1, that's there's a number of serines that are phosphorylated, and when they're phosphorylated, they again through mtor signaling and S6 kinases, they phosphorylate irs1 and cause insulin receptors to be desensitized. And that clearly the serine 1101 was markedly increased in phosphorylation, and that led to decreased in downstream signaling events. So essentially, loss of growth hormone receptor resulted in this mouse that was decreased in terms of insulin sensitivity. So that was different. The question is what happens now when we contrast that to mice that lack the igf1 receptor using the identical technique. So again, kck's technique using the same cre. Again, this is just showing the confirmation that we decreased the deleted the igf-1 receptor had no effect on growth

hormon receptor, and this shows over here IGF-1 to some extent later on seems to go up, and perhaps this is some attempt to the cells to compensate. At any rate, they couldn't fully compensate for some of the phenotypes I'll show you now. The important thing to dismiss first of all, and I think the really interesting thing for discussion, is that these mice never became fat. So again, this is now where the with our first evidence for the separation of these two pathways. These mice, you know, in fact, if anything, they basically lost a little of their body weight over periods of time, but not remarkably so, and certainly never became insulin resistant or had any evidence of metabolic abnormalities. They did, however, have an absolutely identical histological phenotype in terms of decreased myonuclei over here. These are, I think, at 16 weeks. Decreased type one in type one fibers. A decrease over here in

myofiber cross-sectional area. And I'll just show you again this cont. This is almost phenocopying exactly what we saw in terms of the histological decreases in the growth hormone receptor knockout using that same Tech knockout technique. So, with respect to histology, these two mass models looked virtually identical, but they differed with respect to the metabolic effects. Now, we've had a chance in collaboration with Ken Campbell and Karen Esser at the University of Kentucky to start looking in a more detailed way at the functional aspects of some of these myofibers. And Ken Campbell has developed a technique where he can isolate individual myofibers from mice, culture them in a bath where he can change the calcium content, look at excitation-contraction coupling. And we gave him, he actually obtained some of our myofibers from the IGF1 receptor knockout mice, and we're doing now a series of studies comparing both receptor knockouts. But it was interesting in is in preliminary

experiments using this technique. The force generated in the IGF1 receptor knockout myofibers was less than the controls, indicating that there's a functional abnormality at the level of excitation-contraction coupling. Now, I should say that we don't know that could be a developmental phenotype. It says nothing about the acute loss of the IGF1 receptor or growth receptor, and those experiments are now underway again with Karen, who's developed a tamoxifen-inducible CRE, which we can come along and acutely disrupt these two receptors and look at this functional change, which we think is quite interesting, which again would support the idea there's some evidence, strong evidence in smooth muscle, that the IGF-1 pathway can impact contract apparatus and calcium contractility. So, with respect to muscle, there's a little bit of a difference in the growth hormone AON, seem to be able to function separately through the Rome receptors. Certainly, we think that the U, both of these growth hormones.

acting through stat five with the release of igf-1. Uh, we think that that really has, uh, is the main way that the differentiation pathway is controlled by growth hormone, is through the igf-1 receptor, and that'll, that is important for keeping, uh, uh, for the muscle fiber to develop and fuse normally. However, growth hormone through its receptor, not involving the igf-1 pathway, clearly has an important role in metabolism and insulin responsivity, because if you lack that growth hormone receptor, those muscles become severely insulin resistant. That's not what you see when you give growth hormone to somebody, so it's a little bit of an arbitrary model, but I think it tells us something very interesting about how these two anabolic factors can work together or separately in these two different tissues. Okay, that's basically it. I'll leave the questions for later, but uh, thanks very much. That was really nice. I've got a couple questions with you, growth hormone receptor knockout. I mean, you have um

osity in these animals, but you didn't have any change in food intake or energy expenditure. So, how do you explain that? Well, I mean, there is an insulin resistance then, you know, in if you if you block nutrient uptake in the muscle. Nutrients have to go somewhere else. Now, we didn't have, you know, we didn't measure any, and you're right, I don't, I can't explain it for now, but that presumably that nutrient excess would have to be deposited somewhere. I don't like to think of passive mechanisms, but you're right, we didn't see any food intake behavioral changes or that's correct. So, one thing we see this sometimes as well, and you set me up, no, no, no. I just think that what you probably need to do is maybe look at take veal samples and do bomb calorimetry experiments because in some of these models, you can get malabsorption, and that's where the energy may go. The other question I had is in the muscle. Did you

look at steatosis in the muscle because the impaired insulin action may be due to um di of glycerol or ceramide accumulation, which is known to inhibit insulin sensitivity. And given that your animals are already obese, that's likely. Did you look at that? We did, and they have it. Ah, okay, fine. Thanks, be San Francisco Tom. Your answer there actually sets up a question that I have for you that you alluded to just at the end of your talk, namely that when we give, I mean, kids who have, let's say, growth hormone issues, deficiencies, are fat, but when we give a lot of growth hormone, we have major problems with insulin intolerance. I mean, they get hypoglycemic. IGF, on the other hand, as you well know, when you give that, in fact, you run the risk of hypoglycemia. So we have these two very clearly big differences. Now, the question is, when you knock out growth hormone receptor, your growth hormone levels are going to be quite high. Uh, I would expect, I presume.

you've measured that, and you've seen that. So, the question is, when we give growth hormone to treat kids, um, are we, and are you in your mouse model, affecting places like the liver, uh, where you are, in fact, maybe contributing in part to, or the fat, the adipocytes, creating the insulin resistance in those tissues, and less so in the muscle? Or do you really think that what you're doing is altering muscle sensitivity to insulin, which I'm less confident of, even though you showed very nicely the IRS-1 searing phosphorylation. So, what do you think? I think it's mainly muscle. You know, we've looked at other tissues, and we can't really see insulin resistance peripherally in other tissues you might expect, and I can't, I really have no explanation there. There are ways to, I mean, I'm not, I don't think I haven't thought long enough. There are people that have spent their lives, you one or one of them, in terms of studying growth hormone action. It's a very different situation when

you're knocking something out, a receptor, for example, in a tissue, and you really don't have all the controls, as opposed to giving a hormone that can act on all these different tissues, including the liver, to make more IGF-1. So, I don't have a good answer for you, but I can tell you that the muscle becomes severely insulin resistant. We believe that that's a major de-phenotype that's probably accounting for the main accumulation of fat in muscle and probably in the periphery because I think that nutrients are really not being used in the same way in that tissue. Well, why doesn't the IGF-1 receptor do this? 'Cause it's working through the mTOR pathway. Yeah, I don't know. Don't know. So, there's something, I mean, obviously, and that's why it's exciting because we have the opportunity, I think, to really see new pathways which really hadn't been appreciated before, which you'd never find incidentally if you tried to do these experiments by giving growth.

hormon or taking it away. Yeah, it's fascinating. Tom, can you refresh my memory on whether Lon type dwarves have the same type of adiposity and insulin resistance that you're describing? I asked Steve Chernes like that, and he said there is evidence for, you know, people with growth hormone deficiency in receptors and so forth being small but actually having peripheral fat. But I don't want to make too much out of that because, again, there are other ways to explain, you know, or not explain that type of thing. But I was looking for the same, you know, kind of correlation in humans, and I asked Steve Chernis like who studies those people, and he says there is evidence for that. Well, okay.

Clemens: The Growth Hormone/IGF-1 Axis in Bone and Muscle

Scripts