Dr. Willa Hsueh is Professor of Medicine and Chief of the Division of Endocrinology, Diabetes, and Hypertension at the UCLA School of Medicine, Los Angeles. She will discuss cardioprotective effects of insulin-sensitizing interventions.

What I'm going to focus on is to continue Dr. Haffner's discussion about inflammation. You heard about the issues with C-reactive protein (CRP), an important marker of inflammation. I would like to extend this to the vascular wall and talk about implications of PPAR-gamma ligands in the vascular wall.

I will focus on some of the issues with regard to inflammation. This slide highlights what we know about the beginning of the atherosclerotic process in the vascular wall.

LDL cholesterol is clearly very important for atherosclerosis. Clinical trials have shown us that as we lower the burden of LDL cholesterol, we decrease cardiovascular mortality. As we lower the burden of LDL cholesterol, though, we lower that amount that gets trapped into the vessel wall, and the trapped LDL becomes oxidized through different processes, different steps. This process of oxidation, however, is increased in insulin resistance, leading to an increase in oxidized LDL in the vessel wall. This is measured by small, dense LDL cholesterol as a marker of that moiety of cholesterol that is more easily oxidized because it has larger levels of apolipoprotein B.

The macrophage, however, becomes important because it gobbles up the oxidized LDL cholesterol, and rather than taking cholesterol out of the vessel wall, it breaks down into foam cells. These foam cells are so lipid-laden that they stick into the vessel wall, and as they accumulate, begin to form a fatty streak. That is the first important change in the vessel wall, the first marker, lesion of an atherosclerotic plaque. Then when smooth muscle cells start to migrate into the area and form a fibrous cap, and these foam cells often die away, leaving a necrotic lipid core, we develop a more advanced atherosclerotic plaque.

Now what about the role of the macrophage? As the endothelium is damaged, monocytes stick to this vessel wall, and they stick because there's an increased expression of adhesion molecules, intracellular adhesion molecule 1 (ICAM), vascular cell adhesion molecule (VCAM). As the process of vascular injury proceeds, the chemokine, macrophage chemoattractant protein-1 (MCP-1), comes now, is increased, secreted from the endothelial cells, and the vascular smooth muscle cells. The monocytes become differentiated and activated, and now become active macrophages to take up oxidized LDL cholesterol.

This process turns out to be quite important. How do we know this? Because if we take a mouse who is at high risk for atherosclerosis and genetically remove MCP-1 or macrophage colony stimulating factor (MCSF), we can actually rescue this atherosclerosis-prone mouse from foam cell formation.

A major review in one of the important scientific journals, in the United States anyway, the Wall Street Journal, indicated that coronary heart disease, Luces' identified prime suspect, inflammation of the artery wall. It turns out that is a very interesting review because they put in lay terms how do you knock out MCP-1, how do you knock out MCSF, and what the result is. So there is a lot of focus now in atherosclerosis to try to inhibit this inflammatory process.

Dr. Goldstein already alluded to this issue about insulin resistance, and why do we have this proatherosclerotic milieu in terms of the, is it insulin or is it insulin resistance. I think it is a combination of hyperinsulinemia in the setting of insulin resistance.

This is a diagram looking at some of the signaling mechanisms by which insulin acts. We know that there are two major signaling mechanisms, one that is clearly important for glucose transport. This is the phosphoinositol-3 (PI-3) kinase pathway. When insulin binds to its receptor, it turns on insulin receptor substrates now defined by Ron Cahn and Morris White. There are four of them, which become phosphorylated, ultimately turning on this PI-3 kinase pathway, which has clearly been shown to be important for glucose transport.

In addition, insulin is a vasodilator and stimulates nitric oxide in endothelial cells. This pathway has been shown by Michael Kwan and Allan Baron to be very important in also stimulating insulin-mediated nitric oxide production. When we have an impairment in this pathway, either through genetic causes or through signaling because of substances from fat that turn off insulin action in skeletal muscle, we have an impairment both in glucose transport and in endothelial cell function. Allan Baron clearly showed several years ago that patients who are insulin resistant, even before they have diabetes, have defects in nitric oxide production, in resistance arteries. And we have been showing similar data in coronary arteries as well.

Insulin is also a well-known growth factor, mediating the RAS, RAF, MEK, and mitogen-activated protein (MAP) kinase pathway, which is not only important for growth of vascular tissue, but this pathway also mediates migration of cells including vascular smooth muscle cells, monocytes and macrophages as well as endothelial cells. This pathway has been shown in the vasculature, heart and kidney to stimulate or regulate production of plasminogen activator inhibitor type 1 (PAI-1), which you already heard is an important prothrombotic substance, as well as emerging data that suggest it is an important profibrotic substance. In addition, this pathway may also mediate endothelin production. We have dual effects of insulin, one on the vasculature to stimulate the protective white knight, nitric oxide, and the other to stimulate a pathway that is proatherosclerotic.

What happens when a subject is insulin resistant? We have already defined it as a defect in this pathway with impairment in glucose transport by definition, and an impairment in nitric oxide production. George King showed in animal models, and Larry Mandarino in humans, that while this pathway is defective in insulin resistance, this pathway operates normally. Here emerges an imbalance between what we define as insulin resistance, defects in glucose transport, endothelial cell dysfunction, but normal proatherosclerotic pathway.

This insulin resistance is certainly related to inflammation. I think a question that we are addressing is what are the common links between insulin resistance and atherosclerosis. We already know that free fatty acids are increased in insulin resistance. There is increasing evidence that free fatty acids aggravate arterial injury. Tumor necrosis factor (TNF)-alpha is known to be elevated in insulin resistance. There is increasing evidence that TNF-alpha is increased in injured vessel wall. We have known now over the last several years that CRP predicts cardiovascular events in patients with or without diabetes. It is actually a predictor of diabetes.

Oxidized LDL cholesterol is an important inflammatory substance in the vessel wall. In addition, we now know that insulin resistance subjects have more of this oxidized LDL cholesterol.

There also is increasing evidence that angiotensin II is a major culprit in the vessel wall. The HOPE Trial that was recently published told us that it is important to inhibit the renin angiotensin system to protect the vasculature. There is some increasing evidence that perhaps insulin resistant subjects, with or without diabetes, may have increased sensitivity to angiotensin II.

Here we have angiotensin II as an important mediator of multiple mechanisms that are involved in insulin resistance. Our well known risk factors--hypercholesterolemia, hypertension, diabetes, smoking--all add up to causing endothelial cell dysfunction. Decreased nitric oxide and increased tissue expression and activity of the angiotensin-converting enzyme (ACE) lead to increased local vascular tissue levels of angiotensin II.

Angiotensin II stimulates the prothrombotic factor PAI-1, to promote thrombosis. Angiotensin II stimulates VCAM expression and ICAM expression on the endothelium to promote inflammation. Angiotensin II stimulates endothelin to promote vasoconstriction. It is like insulin. It is a growth factor and causes vascular remodeling. It also stimulates matrix metalloproteinases to make existing plaques unstable to promote plaque rupture. Therefore, angiotensin II is an important culprit.

This slide looks at the vascular tissue in terms of signaling. Insulin and the MAP kinase pathway have already been reviewed. Angiotensin II is also well known to stimulate MAP kinase in a variety of tissues, specifically the heart and vessel wall. Other growth factors that are implicated in vascular disease, such as platelet-derived growth factor and insulin-like growth factor 1 (IGF1), are known to increase MAP kinase activity. This leads us to discuss hyperglycemia and its effects. Hyperglycemia also stimulates the MAP kinase activity. Therefore, there are multiple growth factors, vascular injury factors, and hyperglycemia that all stimulate this common signaling pathway.

One of the questions we ask in the future, in terms of approaches to cardiovascular disease and insulin resistance, is: how do we target this MAP kinase pathway to interfere with this pathway and potentially protect the cardiovascular system? The peroxisome proliferator-activated receptors (PPAR)-gamma exist in vascular wall. If PPAR-gamma ligands are added to cultured vascular cells such as smooth muscle cells, monocytes, and macrophages, MAP kinase activity is not attenuated in the cytosol, however, activation of important transcription factors is prevented. Two of them are in the Ets family of transcription factors, Elk-1 or Ets-1. If we prevent the phosphorylation of Elk-1, we inhibit growth. If we inhibit Elk-1 phosphorylation, we also inhibit the turning on of another important transcription factor, Erg-1, which regulates a whole cascade of inflammatory genes in the vessel wall.

In addition, if we turn off another transcription factor in this family, Ets-1, we inhibit the production of matrix metalloproteinases, and inhibit vascular migration as well as the factor, which is implicated to cause plaque instability. Some very important processes are inhibited, not only growth and migration, but inflammation and plaque instability in the vessel wall. An important question is whether this whole system, the MAP kinase system, is an important therapeutic target.

What happens when you add angiotensin II to animal models? Allan Dougherty, in Kentucky, and Bob Taylor, at Emory, and our group have taken models of atherosclerosis in mice. If you remove the LDL cholesterol receptor, the mouse cannot metabolize cholesterol. On a high-fat diet it starts after 2 months to develop fatty streaks, which are shown by this oil red O staining in the aorta.

If you infuse angiotensin II, however, you get this massive atherosclerotic change in the vessel wall in those same 2 months with the high-fat diet. This shows a dose of angiotensin II delivered by osmotic mini-pump in levels that also elevate blood pressure. However, patients with diabetes and insulin resistance are often hypertensive. So this is now a model of advanced atherosclerosis with necrotic lipid cores covered by fibrous caps. These advanced lesions are not only found in the aorta, but throughout the carotid, throughout the coronary arteries, and even smaller resistance vessels in our model. Angiotensin II accelerates lesions by 40-fold in this animal model.

What happens if we give PPAR-gamma ligand, such as rosiglitazone? Here angiotensin II accelerates blood pressure, but the rosiglitazone does not substantially reduce angiotensin II-accelerated hypertension.

This model, also after 3 months, but not after 2 months, actually develops hyperglycemia and hyperinsulinemia. Rosiglitazone does not attenuate glucose and has a modest effect to lower insulin in this model. It also does not decrease the hypercholesterolemia or the hypertriglyceridemia seen in this model.

However, if we look at lesions, we see what we believe are direct effects of rosiglitazone on the vessel wall. This slide shows the angiotensin II-accelerated lesion. Again, showing massive changes with advanced atherosclerotic plaques and attenuation by rosiglitazone.

When we quantitate this, we see the 40-fold acceleration of lesions with angiotensin II. Rosiglitazone attenuates this very advanced process beyond the fatty streak by 60%, which is very substantial attenuation.

This slide depicts the reverse experiment showing the LDL receptor knockout mouse that has its bone marrow irradiated. Then you give it stem cells, which either express PPAR-gamma or you give it stem cells, which are null for PPAR-gamma or have no PPAR-gamma in them. What you see is the LDL receptor knockout mouse with PPAR-gamma, and a 30% increase in atherosclerotic lesion area if you give it back donor cells, which have no PPAR-gamma. What I showed you was that PPAR-gamma ligands attenuate modest and even very advanced atherosclerosis, while if you take out PPAR-gamma from the monocyte macrophage, you get an acceleration of lesions, just the opposite. This type of data tells us that PPAR-gamma, at least in the monocyte macrophage, is very important to prevent the atherosclerotic process.

There are a number of ways in which PPAR-gamma, at least in the monocyte and macrophage, has been implicated in atherosclerosis. There have been data over the last several years that turning on PPAR-gamma in monocytes and macrophages inhibits the inflammatory process. Although these data are not shown, every agent ever used inhibits migration of monocytes and macrophages. In addition, there are some data that show that many of these agents, at least in very high doses, inhibit inflammatory actions, including production of TNF-alpha, the interleukins, and inducible nitric oxide synthase (INOS) protein in monocytes and macrophages.

We have already reviewed the fact that they inhibit matrix metalloproteinase production in macrophages and vascular smooth muscle cells. Our first observation with these agents in the vessel wall was that they inhibit vascular smooth muscle cell growth and migration. Recently, there has been some exciting data demonstrating that these agents inhibit the angiogenic process. Judah Folkman, about a year ago, published data suggesting there is angiogenesis that occurs in atherosclerotic plaques and that perhaps angiogenesis was important to maintain atherosclerotic plaque growth. If he used angiogenesis inhibitors, he could inhibit atherosclerotic plaque growth, especially those plaques that were over 250 microns in size.

There has been a lot of interest in new vessel growth in plaques. In fact, several years prior to that, Peter Libby and colleagues demonstrated that human atherosclerotic plaques do contain small vessels. These small vessels rupture and leak, and this causes more inflammation in the plaque and more plaque instability.

In addition, there have been a number of different studies which showed troglitazone, and now rosiglitazone, improve endothelial cell function. The next few slides will review the story of angiogenesis a little bit farther. There is increasing evidence that inflammation is an important link between atherosclerosis and insulin resistance.

These are some unpublished data by Dr. Dandona from Buffalo, New York, who looked at markers of inflammation in patients who were insulin resistant but without diabetes, and the effect of rosiglitazone. Shown in this slide are decreases in generation of reactive oxygen species, decreases in macrophage P47 expression, decreases in CRP, and also decreases in circulating chemokine MCP-1. In human studies we are seeing decreases in inflammation in response to rosiglitazone.

In addition, there is early evidence that the first PPAR-gamma ligand we had, troglitazone, could inhibit a subclinical marker of atherosclerosis, carotid intimal medial wall thickness (IMT), and could reduce actually thickness in patients with type 2 diabetes, while placebo controlled patients had a progression of carotid IMT.

Another study looked at women who had a history of gestational diabetes who have a high probability of developing diabetes. These women received troglitazone over 3- to 3.5-year period and not only saw that troglitazone impaired the development of diabetes but also that there was slowing of progression of carotid IMT.

Early evidence suggest that these agents can, in humans, attenuate more definitive markers of atherosclerosis, such as carotid IMT. Studies with rosiglitazone to address that issue are in progress.

Now what about atherosclerosis and angiogenesis? An important area where angiogenesis may contribute to pathology is neovascularization in the eye.

We took retinal endothelial cells in collaboration with David Hinton, in ophthalmology at USC, and added vascular endothelial cell growth factor (VEGF) and induced growth and migration of vascular endothelial cells as well as tube formation.

VEGF is very important in diabetic proliferative retinopathy. The vitreous of patients with proliferative retinopathy contains much higher levels of VEGF compared to vitreous from patients who do not have proliferative retinopathyIn this graph, white is rosiglitazone and black is troglitazone. There is a dose-dependent decrease in VEGF-induced growth, migration, and tube formation.

In fact, if you want to look at the tubes, you add VEGF and see capillary tube formations in vitro. If you add troglitazone or rosiglitazone, as seen on the last slide, you attenuate that process.

If you take a rat, however, and subject it to hypoxia for several hours a day, a hypoxic response develops similar to the response that patients with diabetes who develop in their retina, and that is increased production of VEGF. This slide shows staining of the retina for VEGF either in the animals who were untreated, or animals who are treated with a vitreous injection, which is an injection of rosiglitazone or troglitazone directly into the eye. Both groups develop an increase in VEGF expression.

However, if you look at neovascular formation in response to that VEGF in vivo, here in the animals that are just injected with placebo, you start to see capillary tube formation. This is the retinal membrane. You see neovascularization. If you, inject rosiglitazone or troglitazone directly into the vitreous, you see now an attenuation of this neovascular formation.

In fact, if you now count the number of nuclei on this retinal membrane, you see a significant attenuation by either rosiglitazone or troglitazone.

There is increasing evidence that the PPAR-gamma ligand is certainly on endothelial cells, and one of its functions is to inhibit growth and migration, at least VEGF-induced. This has potential implications for the role of angiogenesis in atherosclerotic plaque growth and for neovascularization.