Insulin Resistance and Atherosclerosis

The epidemic of obesity in the developed world over the last two decades is driving a large increase in type 2 diabetes and consequentially setting the scene for an impending wave of cardiovascular morbidity and mortality. It is only now being recognized that the major antecedent of type 2 diabetes, insulin resistance with its attendant syndrome, is the major underlying cause of the susceptibility to type 2 diabetes and cardiovascular disease.

In metabolic tissues, insulin signaling via the phosphatidylinositol-3-kinase pathway leads to glucose uptake so that in insulin resistance a state of hyperglycemia occurs; other factors such as dyslipidemia and hypertension also arise. In cardiovascular tissues there are two pathways of insulin receptor signaling, one that is predominant in metabolic tissues (mediated by phosphatidylinositol-3-kinase) and another being a growth factor-like pathway (mediated by MAPK); the down-regulation of the former and continued activity of the latter pathway leads to atherosclerosis.

This review addresses the metabolic consequences of the insulin resistance syndrome, its relationship with atherosclerosis, and the impact of insulin resistance on processes of atherosclerosis including insulin signaling in cells of the vasculature.

• I. Introduction

• II. Pathology of Atherosclerosis

• III. Biochemical and Cellular Mechanisms of Atherosclerosis

• IV. Metabolic Pathophysiology of Insulin Resistance Relevant to Atherosclerosis

• V. Inflammation, Insulin Resistance, and Atherosclerosis

• VI. Normal Insulin Receptor Signaling in Metabolic Tissues

• VII. Insulin Signaling and Cellular Responses in Vascular Tissues

  • A. Endothelium

  • B. Vascular smooth muscle cells

  • C. Monocyte/macrophages

  • D. T lymphocytes

• VIII. Abnormal Signaling in Insulin Resistance in Vascular Tissues

• IX. Conclusions, Implications, and Speculations

I. Introduction

DEVELOPING AN UNDERSTANDING of the molecular basis underlying the physiological phenomena of insulin resistance and its pathological sequelae of cardiovascular disease is one of the major imperatives facing human disease in the area of endocrinology and metabolism.

The discovery of insulin by Banting and Best (1) in the early 1920s was a milestone of medical research in the 20th century and resulted in the award of a Nobel Prize in medicine. The discovery and then availability of insulin coupled several decades later by the development of bioassays for insulin (2) allowed for two major findings. The first is that the hyperglycemia of diabetes had two distinct origins occurring in the absence and presence of circulating insulin, leading to the postulation of two forms of diabetes first proposed by Himsworth (3) and later confirmed by Bornstein and Lawrence (4). These two forms marked by nominally zero and high levels of insulin were earlier known as juvenile-onset and maturity-onset diabetes but are now termed type 1 and type 2 diabetes, respectively (5, 6). The second major finding was that the response or sensitivity of individuals to the glucose-lowering action of insulin could vary greatly (7, 8). This variability in insulin sensitivity led to the recognition of the existence of insulin resistance (part of the cause of type 2 diabetes). It was later established that insulin resistance clusters with a variety of risk factors for cardiovascular disease—dyslipidemia, hypertension, and hypercoagulability—which are physiological/metabolic disturbances that form a syndrome known as syndrome X, the metabolic syndrome, or the insulin-resistance syndrome (9, 10). Hu et al. (11) discuss the multiple definitions of insulin resistance and the insulin-resistance/metabolic syndrome. The insulin-resistance syndrome is tightly coupled to obesity, and the consequences of body composition were first noted by Vague in 1947 (cited in Ref. 12). Under the societal influences of the last century, the consequences of body composition may have remained a small phenomena of marginal interest. However, the altered lifestyles in Western and now in developing countries, which have led to an explosion in the proportion of the population with obesity, is driving an increase in the prevalence of insulin resistance, the insulin-resistance syndrome, and type 2 diabetes, and it is widely predicted that this will be followed by an increasing burden of premature cardiovascular disease. The underlying pathology of premature cardiovascular disease is atherosclerosis, known traditionally as macrovascular disease in the setting of diabetes.

A primary biochemical abnormality in most cases of type 2 diabetes appears to be insulin resistance, although its cause(s) and effects are far from well understood. The onset of insulin resistance commonly leads to relative insulin deficiency with a slow decline in the regulation of blood glucose with hyperinsulinemia and elevated circulating free fatty acid (FFA) levels. This is followed by a later decrease in the ability to control plasma glucose manifest as rising fasting plasma glucose or rising peak plasma glucose levels in response to an oral glucose challenge with intermittent and persistent hyperglycemia leading to a diagnosis of type 2 diabetes. From this natural history of disease perspective, in genetically susceptible people, type 2 diabetes can be seen as end-stage insulin resistance. Although the molecular mechanism(s) of insulin resistance is unknown, it undoubtedly occurs directly or indirectly in insulin-sensitive tissues most closely associated with glucose homeostasis including the liver, skeletal muscle, and fat. Whether or not insulin resistance occurs and how it might be manifest in the critical tissues of cardiovascular disease—the heart and vasculature—is controversial.

The proposition considered in this review is that insulin resistance is the primary biochemical defect underlying the current epidemic of obesity with cardiovascular disease and type 2 diabetes. If insulin resistance is the primary biochemical defect, then it will follow that lifestyle (13, 14) and pharmacological strategies that address insulin resistance, that is interventions that increase insulin sensitivity, may represent the most logical and ultimately effective therapeutic response. An older agent, the biguanide metformin, acts as an insulin sensitizer predominantly by reducing insulin resistance in the liver (15). Metformin therapy was associated with beneficial outcomes in the UK Prospective Diabetes Study (16). The thiazolidinediones or glitazones, which are peroxisome proliferator-activated receptor-γ ligands, have emerged in the last decade (17, 18). These agents ameliorate insulin resistance and have direct antiatherogenic vascular actions (19, 20). The potential role of thiazolidinediones in the prevention of cardiovascular disease has been reviewed elsewhere (21). The role of drugs improving insulin resistance is acknowledged but will not be the focus of this review. This review will focus on the metabolic and vascular pathology and biochemical mechanisms of insulin resistance and the association of these factors with the development of atherosclerosis and cardiovascular disease in susceptible subjects.

II. Pathology of Atherosclerosis

Vascular disease in diabetes is usually differentiated as microvascular and macrovascular disease, where the former is associated with diabetic retinopathy and renal disease and the latter with the consequences of medium-large vessel disease including ischemic coronary, cerebrovascular, and peripheral arterial disease. The metabolic syndrome, prediabetes, and type 2 diabetes, which all cosegregate with insulin resistance, clearly accelerate vascular disease and increase the consequential disease burden (22) (Fig. 1). The exact mechanisms for the increased susceptibility and progression of atherosclerosis in patients with diabetes are unknown. However, although patients with diabetes show more advanced lesion development (23), the macroscopic/microscopic appearances of atherosclerotic lesions are generally similar between patients with and without diabetes.

Fig. 1.

Natural history of insulin resistance and type 2 diabetes. The initial stages of insulin resistance are influenced by genetic factors and environmental factors with metabolic/physiological changes in insulin levels, lipids, and blood pressure. During the initial insulin-resistant period, or “Pre-diabetes,” the metabolic and physiological changes proceed in parallel with the commencement of silent atherosclerosis and cardiovascular complications (this is depicted by the white- light red area to the left of the complications section) in genetically susceptible people, before the onset of diabetes. The onset of diabetes involves the clinical presentation of hyperglycemia and hyperinsulinemia as well as the metabolic/physiological disturbances present in insulin resistance and the acceleration of macrovascular and microvascular complications, the latter being exacerbated by hyperglycemia. The microvascular and macrovascular complications are the major causes of morbidity and mortality in patients with diabetes. The diagram indicates that atherosclerosis and cardiovascular disease commence before the formal onset of diabetes—this period is thus termed pre-cardiovascular disease (“Pre-CVD”). HDL, High-density lipoprotein; sd LDL, small dense LDL; IFG, impaired fasting glucose; IGT, impaired glucose tolerance.

Macroscopically, the wall of the atherosclerotic artery is thickened and has reduced elasticity and the lumen diameter is narrowed due to the accumulation of lipids in macrophages and possibly a small component due to the proliferation of smooth muscle cells (SMCs) (24). The earliest deposit of lipid in the intima is called a fatty streak, and this appears as a yellow, slightly raised area on the luminal surface that enlarges and comes together to form irregular yellow streaks. The fatty streak contains accumulations of lipid droplets beneath the endothelium that are free and within macrophages. The fatty streak occurs in childhood and adolescence (25–27). The developing lesion contains gelatinous patches caused by edema and an increase in extracellular matrix molecules including proteoglycans. Intimal thickening at branching points of arteries appear as intimal cushions, which are composed of vascular SMCs (VSMCs), collagen, and other extracellular matrix proteins (28). The atherosclerotic plaque appears as a small, disc-like, slightly raised patch of intimal thickening with a smooth glistening surface, appearing yellow (lipid deposits) and white due to the superficial fibrous tissue. Areas of necrosis develop in the deeper part of the lesion, converting it to a structureless accumulation of extracellular lipid, cholesterol crystals, and tissue debris as well as neutrophils and other inflammatory cells (28). As the plaque thickens, the media becomes thin and weak in an attempt to structurally dilate and maintain patency (remodeling), and the lesion may erode into the media by disrupting the internal elastic lamina. A complicated plaque may be one that has cracked or ulcerated and a thrombus has been deposited on the surface or one that shows calcification that converts the lesion to a hard brittle plate (29, 30). The rupture of the plaque and acute thrombus formation may partially or totally occlude the lumen, leading to clinical events such as myocardial infarction, stroke, or peripheral vascular occlusion.

III. Biochemical and Cellular Mechanisms of Atherosclerosis

The study of the pathways and mechanisms of atherosclerosis in cell and animal models has given considerable insight into the underlying processes, but a considerable amount needs to be determined in relation to critical steps that may be subject to therapeutic interruption.

Over the last century, various hypotheses have been forwarded to explain the initiating events and factors that contribute to the development of atherosclerosis. There are three major hypotheses of atherogenesis, which have some common and distinct components, and they include the response-to-injury hypothesis, response-to-retention hypothesis, and the oxidation hypothesis.

Russell Ross (31, 32) is highly regarded for his enunciation of the response-to-injury hypothesis, which states that endothelial damage precedes SMC migration and proliferation, deposition of intracellular and extracellular lipid, and accumulation of extracellular matrix. Subsequently, the response-to-injury hypothesis was broadened to include endothelial dysfunction as a key event that initiates the inflammatory mechanisms associated with atherosclerosis (33). The endothelium is a dynamic monolayer lining the blood vessel wall and is responsible for maintaining vascular tone through the production of vasodilators [e.g., nitric oxide (NO)] and vasoconstrictors (e.g., endothelin-1). As well as its vasodilation properties, NO regulates SMC migration and growth and inhibits inflammation and platelet aggregation. The endothelium releases other factors, such as von Willebrand factor, plasminogen inhibitors, and prostacyclin, that regulate thrombosis by inhibiting platelet aggregation. It is the imbalance between the effects of vasodilators and vasoconstrictors and inhibition of the anticlotting mechanisms that contribute to and characterize endothelial dysfunction.

Beginning in 1995, the work of various groups was articulated by Kevin J. Williams and Ira Tabas (34, 35) as the response-to-retention hypothesis, which proposes that the central atherogenic process is the subendothelial retention and accumulation of lipoproteins by extracellular matrix molecules, such as proteoglycans. The importance of this hypothesis was strongly supported by the work of Jan Boren and colleagues (36), who showed that mice expressing genetically modified ApoB that had defective binding of low-density lipoprotein (LDL) to proteoglycans developed atherosclerosis at a rate that was appreciably less than those mice expressing normal ApoB and thus normal LDL-proteoglycan binding.

The oxidation hypothesis of atherosclerosis was affirmed by Joseph L. Witztum in 1994 (37). The central component of the oxidation hypothesis is the oxidative modification of LDL, which acts as an immunogenic stimulus for monocyte recruitment to the vessel wall and phagocytic uptake of oxidized LDL (oxLDL) by macrophages. The process of oxidation of cellular components and particles such as lipoproteins is thought to exacerbate atherosclerosis, although this is generally a feature of late-stage lesions (37, 38).

oxLDL dissociates from the proteoglycans and acts as a strong inflammatory stimulus in the vessel wall. The presence of oxLDL provokes a series of cellular reactions involving monocyte/macrophages, endothelial cells, SMCs, and (CD4⁺) immune cells. Selectins (P, E, and L) on the monocyte cell surface assist their rolling and tethering over the endothelium by binding to the P-selectin ligand on the endothelial cell surface (39). The movement of monocytes is stopped by intercellular adhesion molecule (ICAM)-1, vascular adhesion molecule (VCAM)-1, and integrins followed by monocyte attachment to hyaluronan cable structures through CD44 (40), which facilitates transendothelial migration and the presentation of monocyte/macrophages to the cells and lipid in the vessel wall. Adiponectin inhibits the production of integrins and P-selectin, a mechanism that may act to regulate the recruitment of inflammatory cells to the vessel wall. A reduction in adiponectin induced by oxidative stress may exacerbate monocyte adhesion to the vascular endothelium.

The atherosclerotic lesion propagates through a series of cell-cell and cell-lipid interactions, leading to inflammation, extracellular matrix alterations, and necrosis. Subsequently, the weakening and rupture of the cap of the atherosclerotic plaques, possibly by matrix metalloproteinases (MMPs) such as MMP-9, leads to plaque rupture, thrombus formation, lumen obstruction, and the acute clinical events of a heart attack or ischemic stroke. The major initiating factors for atherosclerosis are the modification of the extracellular matrix, particularly the glycosaminoglycan chains on proteoglycans that attract and bind atherogenic lipoproteins, and the inflammatory processes involving the binding of monocytes to the vessel wall, cell penetration, differentiation into macrophages, lipid engulfment, and foam cell formation. A schematic representation of the atherogenic cascade is shown in Fig. 2. The figure depicts the normal process of LDL transitioning through a nondiseased vascular media (Fig. 2A) and the binding and retention of modified atherogenic LDL and subsequent formation of monocyte/macrophage foam cells in a diseased vessel wall (Fig. 2B).

Fig. 2.

Hormonal and cellular mechanisms of atherosclerosis within the vessel wall. The diagrams are representations of transverse sections of a coronary blood vessel under physiological (A) and pathological (B) conditions. A, Physiological conditions. The cellular layers from the top down include the endothelium, media, and adventitia, with an internal elastic lamina between the endothelium and media layers. The media layer is comprised of well-oriented layers of SMCs. SMCs maintain tone and also synthesize matrix molecules including proteoglycans. Proteoglycans consist of a core protein (green) and one or more glycosaminoglycan (GAG) chains, which are heavily sulfated and thus negatively charged (represented as blue chains on the green core protein). The proteoglycans in the image have two GAG chains and thus represent biglycan, a major type of proteoglycan found in human vessels. Circulating lipoproteins in the blood such as normal LDL (>25 nm diameter, orange molecules) freely traverse the blood vessel wall a, Normal proteoglycans; b, normal LDL. B, Pathological conditions. Atherogenic stimuli associated with the metabolic syndrome (reactive oxygen species and inflammatory cytokines) arising from redox alterations in adipose tissues precipitate a metabolic abnormality in the vascular endothelium (endothelial dysfunction) and a cascade of events within the neointima of the vessel wall. The neointima (NI) is a less organized and morphologically distinct layer. Intimal SMCs secrete proteoglycans with one or more structural changes in the biochemical properties of the GAG chains (red extension of blue chain). Biochemically altered proteoglycans are more likely to bind and retain lipoprotein particles. In the case of patients with diabetes, the atherogenic lipid profile includes an increased number of small dense LDL (<25 nm diameter, pink molecules) particles that have a greater propensity to penetrate the vessel wall and a higher affinity binding for proteoglycans compared with normal LDL (178 ). The combination of altered proteoglycans and the abundance of small dense LDL leads to the entrapment of the positively charged LDL in the subendothelial space. Retained LDL particles are susceptible to oxidation and protein modification. oxLDL (yellow molecules) has a lower affinity for proteoglycans compared with normal LDL and is released from the proteoglycan GAG chains. The presence of oxLDL triggers an immunological response whereby circulating monocytes bind to the endothelium via adhesion molecules and possibly through the interaction of CD44 receptors with hyaluronan (green molecule extruding from the endothelium), a GAG without a core protein that acts as a CD44 ligand. Hyaluronan is synthesized on VSMCs and endothelial cells by hyaluronan synthase (represented by green oblongs on SMC). Monocytes traverse the endothelium and differentiate into macrophages that then entrap and attempt to remove the LDL. The presence of LDL overwhelms the macrophages, which are unable to phagocytose the LDL and the lipid-laden macrophages or foam cells (yellow cells on the right) are deposited in the subendothelial space. The atherosclerotic process continues over many years, and the complex lesion may weaken, rupture, and form a thrombus, which would then occlude the lumen and reduce blood flow to the heart causing a myocardial infarction. a, Endothelial dysfunction resulting from metabolic disturbances such as hypercholesterolemia and reactive oxygen species; b, proteoglycans; c, normal LDL; d, migrating and proliferating VSMCs; e, modified proteoglycan; f, small dense LDL bound to proteoglycans; g, oxLDL is released from the protoeglycans; h, monocytes bind to the endothelium via adhesion molecules and possibly interact with hyaluronan; i, lipid-laden foam cells.

The presence of T lymphocytes in atherosclerotic lesions is well established (41, 42); however, the factors that initiate their recruitment are still relatively unknown. Classical immunology shows that T lymphocytes are part of the acquired immune response by which the body’s exposure to various antigens stimulates T lymphocytes to directly destroy antigens. In atherosclerosis, the antigenic stimulus to T lymphocytes may be oxLDL, as antibodies to oxLDL have been identified in vascular lesions (43–45). Alternatively, the presence of T lymphocytes in atherosclerotic lesions may represent a secondary immune response induced by the release of chemoattractants from other cells (endothelial cells, VSMCs, macrophages) involved in atherogenesis. The implications of the impact of insulin resistance on individual cell types participating in atherosclerosis are considered in a later section of this review.

IV. Metabolic Pathophysiology of Insulin Resistance Relevant to Atherosclerosis

The spectrum of metabolic disturbances associated with insulin resistance extends beyond hyperglycemia and includes dyslipidemia, hypercoagulability, and inflammation. The importance of the insulin resistance syndrome is reflected in its recognition as the cornerstone of what is termed the insulin-resistance syndrome, syndrome X, or metabolic syndrome (10). Although the metabolic syndrome is becoming part of the obesity lexicon, it should be noted that there is controversy over the question of whether or not the metabolic or insulin-resistant syndrome is actually a syndrome if the use of syndrome implies a mechanistic link that draws the relevant parameters together (9, 46). The metabolic syndrome predicts higher levels of cardiovascular disease, but the question is whether or not the prediction has any stronger value than considering the aggregate effect of the individual risk factors. The major evidence for the existence of a mechanistically linked syndrome is that the cardiovascular risk factors occur together or cluster more often than predicted by chance (47). Further basic and clinical research will undoubtedly shed light on this intriguing and important question.

Obesity, and central obesity in particular, is a major underlying factor in insulin resistance. Insulin resistance in fat cells leads to increased lipolysis and the release of FFA. Some 40 yr ago, Randle et al. (48) proposed that insulin resistance occurs as a result of fatty acid oxidation, leading to inactivation of mitochondrial pyruvate dehydrogenase and ultimately decreased glucose uptake, and this was based on work in isolated rat heart. However, it would appear that glucose metabolism, including the role of insulin, can vary markedly between tissues. In skeletal muscle there is an apparently straightforward process of glucose uptake, mediated by glucose transporters, phosphorylation to glucose-6-phosphate by hexokinase, and further processing leading to storage as glycogen catalyzed by glycogen synthase. People with diabetes lose efficiency in passing glucose to glycogen such that the ability to store glycogen can be reduced by 60%. Both glucose transporters and hexokinase play a critical role in glucose uptake because it is the gradient of the parent chemical entity that determines the rate of influx. The phosphorylation of glucose by hexokinase greatly reduces the intracellular glucose concentration (to 100–200 μm), creating a large inward gradient for influx through the glucose transporter. Nevertheless, Cline et al. (49) have concluded that the glucose transporter GLUT4 is the rate-limiting step in insulin-stimulated glycogen synthesis. Increases in plasma fatty acid levels can inhibit glucose transport and hexokinase activity, implying that fatty acids can inhibit signaling through the insulin receptor.

The liver is the major site of glucose uptake and storage as well as insulin clearance. Insulin suppresses glucose production by the liver directly by binding to the insulin receptor on hepatocytes and inhibiting both glycogenolysis and gluconeogenesis. An indirect effect of insulin via suppressing lipolysis in adipose tissue is less FFA delivery to the liver with less glycogenolysis and less hepatic insulin resistance. Insulin is the major hormone responsible for maintaining energy homeostasis by coordinating the use of fat deposits not only in liver but also in muscle and adipose tissue. Adipose tissue stores FFAs as triglycerides and also releases FFAs and adipokines including leptin, adiponectin, resistin, plasminogen activator inhibitor (PAI)-1, and TNF-α. The acute response (90 min) of islets treated with FFAs is to amplify glucose-stimulated insulin secretion, a response that is mediated by the G protein-coupled receptor 40 (50). However, in human studies the insulin secretion rate is reduced by 30% in patients with type 2 diabetes infused with FFAs for 4 h compared with subjects without diabetes (51). The chronic treatment (3–7 d) of normal rat pancreatic islets (52) and rat insulinoma cell lines (53) with FFA in vitro decreases glucose-stimulated insulin secretion. Additionally, the 7-d treatment of fat-laden islets isolated from Zucker diabetic fatty rats with FFAs reduces glucose-stimulated insulin secretion (54). Leptin has been reported to have a role in insulin resistance via the adrenergic system and by acting directly on peripheral tissues (55). Mice deficient in leptin are hyperglycemic and hyperinsulinemic and leptin administration reverses these changes without altering weight, suggesting it may have an effect on insulin resistance independent of weight control (56). Adipose tissue TNF-α levels correlate with body fat and hyperinsulinemia in mice and have been shown to play a part in insulin resistance. The role of TNF-α in human insulin resistance is still to be clarified. In contrast, human plasma adiponectin levels are clearly negatively correlated with insulin resistance and have a stronger link to insulin resistance than body fat levels. Circulating levels of adiponectin, particularly the higher-molecular-weight glycosylated forms, may be a useful marker of insulin resistance and angiopathy (55).

V. Inflammation, Insulin Resistance, and Atherosclerosis

Insulin resistance, inflammation, and atherosclerosis appear to be linked via the metabolic syndrome, and the question is the nature of the association being either a common molecular pathology related to insulin receptor signaling or vascular consequences of metabolic abnormalities of insulin resistance. Some very important findings have emerged recently and provide biochemical linkages between obesity and insulin resistance and atherosclerosis. In mice, fat-derived cytokines activate the nuclear factor-κB signaling pathway in hepatocytes, and mimicking this activation by selectively expressing constitutively active IκB kinase-β in hepatocytes generates systemic insulin resistance most likely through the generation and actions of proinflammatory cytokines including IL-6 and TNF-α (57). In humans and mice, increasing obesity is associated with increased oxidative stress in the fat compartment and in circulating markers of oxidative stress (58). There are two potential mechanisms: the deregulated production of pro- and antiinflammatory cytokines and the elevated levels of systemic oxidative stress mediated through increased oxidative enzymes (nicotinamide adenine dinucleotide phosphate, reduced, oxidase) and decreased antioxidative enzymes (superoxide dismutase). Thus, obesity-derived proinflammatory cytokines and reactive oxygen species can generate peripheral insulin resistance but also directly impact on the endothelium to cause endothelial dysfunction and initiate the atherosclerotic cascade (see Fig. 2).

These studies provide a link between obesity and insulin resistance in metabolic tissues. Although the metabolic consequences are almost certainly proinflammatory for the vasculature, whether or not the changes have generated insulin resistance in cardiovascular tissues remains unresolved. Furthermore, the application and appreciation of the expression of deleterious inflammatory processes in the cardiovascular system is relatively new (33), and the cellular mechanisms linking cytokines with the role of the immune system in cardiovascular disease requires further consideration.

Inflammation is a term taken from other areas of medicine and applied recently to cardiovascular disease (33, 59, 60). It is poorly defined in its new context and confuses bloodborne and other factors from tissue sources that may be biomarkers or effectors of cardiovascular disease. We would take a narrow view of inflammation to be the binding, penetration, and proliferation of monocyte/macrophages in the vessel wall to destroy and remove deleterious stimuli, i.e., the process is vessel healing. However, when there is a secondary stimulus such as hyperlipidemia, the immunological response may become overloaded or chronic and the process of atherosclerosis is initiated. Many factors are then activated as a response to the initial insult. This insult may be insulin resistance, and insulin resistance is associated with elevated levels of inflammatory markers. The extensive range of inflammatory markers and effectors associated with insulin resistance and the impact of insulin resistance on their relevant parameters is documented in Table 1.

TABLE 1.

Inflammatory and other markers or effectors relevant to insulin resistance and cardiovascular disease (CVD) in humans

Inflammatory marker/effector of CVD	Changes in relationship to insulin resistance	Ref.
Acute-phase inflammatory proteins
C-reactive protein	Elevated	179
SPLA2-IIA	Elevated	180
Adipokines
Adiponectin	Elevated	55
TNF-α	Elevated	55
IL-6	Elevated	181, 182
IL-8	Elevated	183
Thrombotic/thrombolytic factors
Fibrinogen	Elevated	179, 184
Factor VII	Elevated	185
MMP-9	Elevated	182
von Willebrand factor	Elevated	185
PAI-1	Elevated, positive correlation	185–187
Tissue plasminogen activator	Elevated	185, 188
Thrombomodulin	Elevated	188
Soluble adhesion molecules
ICAM	Elevated	180
VCAM	Unchanged	180
Growth factors
TGF-β	Increased	189
PDGF	More potent effects	120
Monocyte/lymphocyte related factors
Monocyte chemoattractant peptide	Elevated	190
Resistin	Elevated	191
Toll-like receptor 4	Not affected	192
Glucose markers
Advanced glycation end products	Increased	193, 194
Glycated hemoglobin	Increased	195, 196
Lipid biomarkers
HDL-cholesterol	Decreased amount and particle size	197, 198
LDL-cholesterol	Decreased size and increased density	199
Tissue markers
Reactive oxygen species	Elevated	200
Other
Homocysteine	Elevated	201, 202

SPLA₂-IIA, Group IIA secretory phospholipase A₂.

Overall, these studies highlight the coexistence of systemic inflammation, inflammatory biomarkers/effectors implicated in cardiovascular disease, and insulin resistance but do not resolve whether or not the systemic inflammation initiates the inflammatory processes associated with cardiovascular disease and insulin resistance or vice versa.

VI. Normal Insulin Receptor Signaling in Metabolic Tissues

The insulin receptor plays a crucial role in mediating the effects of insulin including the rapid stimulation of glucose uptake (via the glucose transporter protein GLUT4) into its target metabolic tissues, muscle and fat. The insulin receptor is a transmembrane receptor tyrosine kinase (61) able to form homo- or heterodimers with the IGF receptor (IGFR) as disulfide-linked α₂β₂ tetramer proteins. Insulin binds with high affinity to the α-subunit of the insulin receptor, leading to subsequent autophosphorylation of the β-subunits on three intracellular tyrosine residues (62). Under physiological conditions, each receptor responds only to its own ligand (63). Under some circumstances, IGF-I appears to have metabolic effects similar to insulin (64).

The insulin receptor β-subunits are also subjected to intracellular Ser/Thr phosphorylation by protein kinase C (PKC) (65) and dephosphorylation by protein tyrosine phosphatases. Studies in mice deficient in protein tyrosine phosphatase-1B have shown that this enzyme plays a key role in attenuating the insulin receptor response (66). The insulin receptor phosphorylates at least nine intracellular signaling molecules including four intracellular insulin receptor substrates (IRS proteins IRS-1, -2, -3, -4). Studies with transgenic mice suggest that most insulin responses are mediated via IRS1 or IRS2 (67, 68). Phosphorylated tyrosine residues on each of these substrate proteins enables docking of a distinct subset of downstream signaling molecules containing Src-homology-2 domains. Docking proteins include the p85 regulatory subunit of phosphatidylinositol-3-kinase [PI (3)K]; adapter proteins Grb2, Nck, and Shc (69, 70); protein tyrosine phosphatase src homology protein tyrosine phosphatase 2; and the tyrosine kinase fyn (71). The diversity of these proteins enables the downstream signals to be amplified and diversified by serving as common substrates for a number of receptors and thus integrating signals from several signaling cascades (67). Insulin receptor docking proteins also serve as points of alteration in insulin receptor signaling efficiency.

In metabolic tissues, PI (3)K plays a central role in insulin-stimulated glucose uptake (see Fig. 3). Activation of PI (3)K generates phosphatidylinositol-3,4,5-triphosphate (72), which activates the serine/threonine phosphoinositide-dependent kinase kinase 1 (PDK1) and also activates an unidentified kinase, which together phosphorylate the highly conserved Thr³⁰⁸ and Ser⁴⁷³ residues, respectively, present in the serine/threonine kinase Akt/protein kinase B (73). Atypical PKCζ and PKCλ isoforms are also activated by PI (3)K and phosphoinositide-dependent kinase kinase 1 and have been shown to be required for insulin-stimulated glucose transport. The importance of Akt in glucose homeostasis has been demonstrated in Akt2^−/− mice (Akt2 predominates in insulin-responsive tissues) that exhibit insulin resistance and glucose intolerance due to elevated hepatic glucose output and reduced glucose uptake in skeletal muscle (74). Recently, a negative modulator of Akt has been identified and termed TRB3 (previously known as NIPK); it is induced in liver under fasting conditions and disrupts insulin signaling by binding directly to Akt and blocking activation of the kinase (75). Interference with Akt activation may contribute to insulin resistance in individuals with susceptibility to type 2 diabetes.

Fig. 3.

Insulin receptor signaling and the points of interaction causing atherosclerosis related to insulin resistance. Insulin resistance in metabolic tissue manifests as a down-regulation of the PI (3)K pathway leading to hyperglycemia and other components of the metabolic syndrome. Extrapolation from numerous in vitro studies leads to speculation that insulin resistance in vascular tissue causes down-regulation of the PI (3)K pathway, whereas the MAPK pathway proceeds unabated, with consequential proatherogenic effects. It is clear from the figure that insulin resistance in metabolic and vascular tissues leads to proatherogenic outcomes. EC, Endothelial cell.

In addition to the PI (3)K signaling pathway, insulin also activates the MAPK cascade (see Fig. 3). The insulin receptor phosphorylates the adapter protein Shc (or IRS-1 in skeletal muscle), which in turn binds to Grb2 and initiates a cascade of signaling events that lead to induction of genes involved in cell proliferation and differentiation. The MAPK cascade is not involved in insulin-stimulated glucose transport or glycogen synthesis and may relate to cell survival and proliferation, but it is likely to be associated with the processes of atherogenesis (Fig. 3).

VII. Insulin Signaling and Cellular Responses in Vascular Tissues

In most mammalian tissues, a significant fraction of both insulin receptor and IGFR occur as hybrids and form the majority of receptors in cardiac and skeletal muscle (76). In human endothelial cells, IGFR expression exceeds insulin receptor expression (64). Although insulin signaling is focused on metabolic pathways, IGF signals are directed to regulating cell proliferation, cell survival, and differentiation. In differentiated VSMCs, IGF-I activates the PI (3)K/Akt pathway and activation of SHP-2 blocks the MAPK cascade; however, in dedifferentiated VSMCs, e.g., in vascular injury, IGF-I activates ERK and MAPK via Grb2/Sos and Ras, resulting in proliferation and migration (77). A role for IGF in cell proliferation and differentiation has also been reported in hemopoietic cells and adipocytes (78, 79). The elements that control the balance between signaling through the PI (3)K and MAPK pathways may indicate trigger points in insulin resistance. IGF-I is increased in vessels after balloon-injury (80, 81). Thus, the vascular responses to both high concentrations of insulin (as occurs in insulin resistance) and IGF-I may be contributing factors to vascular disease. Of direct importance to cellular mechanisms of atherosclerosis is the relationship between circulating insulin levels and levels affecting the cells of the vasculature. Fasting plasma insulin levels in normal insulin-sensitive people are in the low picomolar range (50–150 pm) (82, 83). By definition, insulin levels are higher in subjects with insulin resistance and in type 2 diabetes and plasma levels in the nanomolar range (up to 1.5 nm) are observed (83). Although there are issues such as tissue concentration and penetration with respect to comparing circulating levels of agents with concentrations displaying actions in vitro, it is nevertheless potentially informative. By way of example, insulin has an antiatherogenic action (84) to inhibit the synthesis of the glycosaminoglycan hyaluronan (85) by human aortic SMCs with an efficacy of 60% and half-maximal effect at approximately 500 pm. Insulin (600 pm) treatment of intact rat aorta ex vivo enhances leptin-mediated relaxation as well as Akt phosphorylation and NO production (86). Thus, plasma insulin concentrations associated with insulin resistance are in the concentration range that affects vascular cell and tissue responses in vitro and ex vivo.

The major cells involved in atherosclerosis are endothelial cells, VSMCs, monocytes/macrophages, and T lymphocytes, and the following section considers insulin signaling pathways in each of these cell types.

A. Endothelium

Insulin receptors on human endothelial cells from the umbilical vein were initially identified and characterized by ¹²⁵I-insulin receptor binding studies almost 30 yr ago (87). The number of insulin receptors present on endothelial cells can differ depending on whether the cells are derived from human umbilical arteries, which have more insulin receptors compared with endothelial cells derived from human umbilical veins (88).

Stimulation of insulin receptors in endothelial cells activates the PI (3)K pathway, suggesting that these receptors are similar to those found in metabolic tissues. In contrast to skeletal muscle, insulin treatment of bovine aortic endothelial cells does not promote glucose conversion to glycogen, suggesting that these cells express little or no glycogen synthase (89). Endothelial cells from bovine aorta and human umbilical vein show no mitogenic response to insulin treatment (89).

In contrast to metabolic tissues, stimulation of the PI (3)K pathway in endothelial cells leads to an increase in the expression of the constitutively active endothelial NO synthase (eNOS) both in vivo and in cultured endothelial cells (90). A study using overexpression of the insulin receptor and an inactive form of the insulin receptor in cultured human umbilical vein endothelial cells has elucidated part of the insulin signaling cascade related to increased NO production (91). Insulin receptor activation leads to tyrosine phosphorylation of the insulin receptor and activation of Akt the serine/threonine kinase downstream of PI (3)K (91). Down-regulation of Ras, an upstream mediator of MAPK signaling has little effect on NO production by endothelial cells treated with insulin, suggesting that PI (3)K/Akt signaling is required for insulin receptor-mediated NO production in endothelial cells (91). Additionally, insulin-stimulated eNOS expression, protein, and activity are decreased in the presence of synthetic inhibitors of PI (3)K, wortmannin and LY294002 (90). Stimulation of eNOS by insulin treatment can be attenuated by the concomitant treatment with angiotensin II, which increases the phosphorylation of serine residues in preference to tyrosine residues on IRS-1, decreasing the interaction between IRS-1 and p85 [PI (3)K subunit] and reduces PI (3)K/Akt/eNOS signaling (92).

NO stimulates guanylate cyclase, the enzyme that converts GTP to cGMP. cGMP lowers intracellular free calcium, promotes vasodilatation, inhibits VSMC proliferation, and regulates angiogenesis (93). Thus, the production of NO by endothelial cells is viewed as a potentially vascular protective mechanism. Insulin and IGF-I increase the production of NO in endothelial cells by directly increasing eNOS, which converts l-arginine to NO and l-citrulline products (94). Insulin stimulates both eNOS expression and activity through the PI (3)K pathway (90, 91, 95).

Subjects with insulin resistance, type 2 diabetes, or other components of the metabolic syndrome demonstrate a phenomenon termed “endothelial dysfunction,” which is at least partly a result of decreased production or increased clearance of NO and thus impaired blood flow (96). Insulin-resistant patients display reduced vasodilatation response to cold pressor testing, indicating endothelial dysfunction (97). Family members of patients with type 2 diabetes also commonly have endothelial dysfunction (98).

Vascular endothelial cell insulin receptor knockout mice show normal vascular development; however, the expression of eNOS and endothelin-1 are significantly reduced in the heart and aorta (99). Mice with targeted disruption of the eNOS gene develop peripheral and metabolic insulin resistance (100). These two studies highlight the importance of maintaining a balance between NO and insulin for endothelial function. Conversely, others have recently found that insulin resistance does not alter the expression of eNOS in an animal model of the metabolic syndrome (101).

The production of NO by endothelial cells stimulated by insulin/IGF leads to a group of antiatherogenic effects including both antiinflammatory and antithrombotic mechanisms involved in atherogenesis. Antiinflammatory effects of NO include a decrease in expression of the adhesion molecules VCAM-1, ICAM-1, E-selectin, and a decrease in the secretion of proinflammatory cytokines monocyte chemoattractant protein-1 and TNF-α (102). Antithrombotic effects include a decrease in platelet adhesion (103) and increased prostacyclin production, which inhibits platelet aggregation (104).

Endothelial cell migration and proliferation is considered an antiatherogenic response to vascular injury such as hypertension, cigarette smoking, hypercholesterolemia, or hemodynamic stress, as an intact and active endothelium promotes vascular quiescence and vessel healing (105). Endothelial cell migration is induced by IGF-I and can be inhibited by the PI (3)K inhibitor LY294002 but not the MAPK inhibitor PD98059, suggesting that the PI (3)K pathway mediates IGF-I-stimulated endothelial cell migration (106). In an in vitro model of compensatory hyperinsulinemia, endothelial cells treated with insulin in the presence of a PI (3)K inhibitor wortmannin demonstrate activation of the MAPK pathway and downstream cellular effects such as increased DNA synthesis, increased expression of VCAM-1 and E-selectin, and increased interaction with monocytes (107). The interaction of endothelial cells with inflammatory cells may also be mediated by the extracellular matrix molecule hyaluronan, which is a ligand for CD44 receptor-positive inflammatory cells (108). Wistar rats fed a fructose diet to induce the metabolic disturbances of type 2 diabetes, i.e., insulin resistance and hyperinsulinemia, show an increase in hyaluronan deposition after balloon-injury (109). An interesting extension of this work would be to determine whether or not insulin mediates hyaluronan synthesis in cultured endothelial cells and thus alters monocyte/macrophage adhesion through the CD44 binding mechanism (Fig. 3).

Endothelin-1 is a 21-amino-acid peptide that is secreted by endothelial cells and induces proatherogenic effects such as vasoconstriction (110), increased vascular permeability (111), VSMC proliferation (112), increased production of IL-6 by endothelial cells and monocytes (113, 114), and increased proteoglycan synthesis by VSMCs (M. E. Ivey and P. J. Little, unpublished observations). In contrast, antiatherogenic actions of endothelin-1 include the stimulation of NO production (115). Sprague-Dawley rats with mild diabetes that were treated with high doses of insulin for 10 d showed an increase in plasma endothelin levels (116). Insulin-treated bovine aortic endothelial cells increase the expression and secretion of endothelin-1 in a concentration- and time-dependent manner (116). The stimulation of endothelin-1 secretion by insulin treatment of endothelial cells is not imitated by IGF-I treatment, and insulin-mediated endothelin-1 secretion is inhibited by genistein, a broad inhibitor of tyrosine kinases, suggesting that insulin mediates the increase in ET-1 through the insulin receptor (116). It is presently unclear whether or not the regulation of endothelin-1 expression by insulin is via the PI (3)K pathway or the MAPK pathway, and this remains an important and interesting pathway to elucidate.

B. Vascular smooth muscle cells

The presence of insulin receptors in human VSMCs was demonstrated in 1983 by ¹²⁵I-insulin binding studies (89). Both insulin and IGF-I receptors are present on rat SMCs, and these receptors are distinct from each other in terms of binding affinity of insulin and IGF-I (117). In rat SMCs, insulin has an affinity for the insulin receptor that is 1000 times that of IGF-I and IGF-I has an affinity for the IGF-I receptor that is 500 times that of insulin for the IGF-I receptor (117). Insulin and IGF-I receptors on VSMCs can be distinguished from each other using specific antibodies (118).

The insulin receptors on VSMCs are structurally and functionally similar to those in metabolic tissues (119). The number of insulin receptors is 10-fold lower in bovine aortic SMCs compared with bovine aortic endothelial cells (89). Vascular SMCs and metabolic cells show a difference in the predominant type of glucose transporter expressed (GLUT1 in vascular cells and GLUT4 in skeletal muscle). Insulin stimulation of glucose transport by GLUT1 in vascular cells appears to occur in a similar manner to the PI (3)K pathway-mediated glucose transport by GLUT4 in metabolic cells, except that GLUT1 is less dynamically translocated than GLUT4.

Insulin signaling mechanisms in VSMCs are similar to those in endothelial cells in that insulin binds to the insulin/IGF-I receptor, leading to tyrosine phosphorylation of the β-subunit on the insulin receptor. The insulin receptor then phosphorylates tyrosine residues on the insulin receptor substrates (IRS-1–IRS-4). IRS-1 and IRS-2 can activate PI (3)K/Akt or the RAS→RAF→MEKK→MAPK→c-fos cascade. Insulin signaling in VSMCs initiates proatherogenic cellular events such as proliferation and migration. Insulin-mediated bovine VSMC migration is inhibited by PD 98059 (MAPK inhibitor) but not wortmannin [PI (3)K inhibitor], implying that insulin-mediated activation of the MAPK pathway mediates VSMC migration (120). This is in contrast to IGF-I-mediated migration, which occurs mainly through the activation of PI (3)K (Fig. 3) (121).

Bovine VSMCs show a 3-fold increase in a mitogenic response to insulin compared with endothelial cells (89). Insulin-stimulated VSMC mitogenesis is mainly controlled through the MAPK pathway; however, the PI (3)K pathway also contributes to maintaining VSMC quiescence (120, 122). Similar to insulin, IGF-I also stimulates VSMC proliferation mainly through the MAPK pathway (121). Activation of MAPK leads to the proliferation and migration of VSMCs, which are somewhat simplistically referred to as proatherogenic cellular processes (105). Evidence suggesting that MAPK is a target for insulin-mediated proliferation and migration was shown in in vitro studies using PD 98059, an inhibitor of the MAPK pathway (122). Treatment of rat SMCs with PD 98059 inhibits the mitogenic response to insulin in a concentration-related manner (122). Additionally, insulin-treated bovine SMCs in the presence of wortmannin show reduced quiescence, suggesting that PI (3)K is involved in maintaining SMCs in the quiescent state and that blocking the PI (3)K pathway (as would occur in insulin resistance) allows insulin to activate the MAPK pathway to induce proliferation and migration responses (120).

Insulin acts synergistically with other atherogenic growth factors such as platelet-derived growth factor (PDGF) to promote the proliferation and migration of bovine aortic SMCs via a c-myc-dependent pathway (123). This may occur in parallel with the effect of insulin to induce VSMC quiescence through the activation of the PI (3)K pathway, which can inhibit PDGF-induced proliferation of VSMCs, but the PI (3)K pathway may be blunted in insulin resistance, and thus the PDGF and insulin combination would be proatherogenic (120). Insulin and angiotensin II have an additive effect to increase MAPK in mesangial cells, which have VSMC-like morphological characteristics (124). Similar to the effect of angiotensin II on endothelial cells, angiotensin II treatment of SMCs inhibits the association of p85 with PI (3)K and thus inhibits insulin-mediated PI (3)K signaling, representing a mechanism for insulin resistance after activation of the renin-angiotensin system (125).

Insulin at supraphysiological concentrations is more effective than IGF-I in inducing a mitogenic response in VSMCs (126); however, the growth effects of insulin and IGF-I are mediated through the IGF-I receptor, not the insulin receptor (118). There is no additive effect of insulin with IGF-I on VSMC mitogenesis (123). There are varied reports regarding whether or not hyperinsulinemia regulates aortic expression of the IGF-I receptor, with some showing that IGF-I receptors are increased in rats with diabetes given insulin (127) and others reporting that insulin-treated rats with diabetes have reduced aortic expression of IGF-I (128). Despite these different findings, it appears that insulin mediates proatherogenic effects in VSMCs, at least in part through the IGF-I receptor.

NO synthesis by VSMCs is regulated by inducible NO synthase (iNOS). Insulin treatment of rat SMCs increases iNOS protein, and insulin-mediated iNOS induction can be inhibited by both PI (3)K and MAPK inhibitors, wortmannin and PD98059, respectively. NO appears to both inhibit the migration of late-passage VSMCs (129) and promote the migration of primary cultured VSMCs (130). Similarly, NO has been shown to stimulate the proliferation of primary rat VSMCs (131, 132) and to inhibit DNA synthesis of later-passaged VSMCs (93, 133). The effects of NO on SMCs are mediated by cGMP-dependent protein kinase G (PKG). Estimation of NO effects in SMCs by PKG-dependent mechanisms is underestimated and compromised in in vitro studies because cGMP PKG is lost in cultured VSMCs after three passages, and this possibly accounts for the conflicting observations of primary vs. cultured VSMCs (134). Insulin and NO treatment of primary rat VSMCs has an additive effect on increasing cell migration, and cell movement is further enhanced when VSMCs are treated with insulin, NO, and PDGF (135). The effects of insulin on VSMC migration are attributed to the insulin receptor, not the IGF-I receptor (135). NO-induced VSMC migration occurs through the MAPK pathway (136), and insulin- plus NO-induced VSMC migration is dependent on the interaction of src homology protein tyrosine phosphatase 2 with the adapter protein Gab1, both of which are upstream mediators of MAPK (135). Insulin- and NO-induced VSMC migration can be inhibited by LY294002, the PI (3)K inhibitor, suggesting that both the MAPK and the PI (3)K pathways are involved in insulin- plus NO-mediated VSMC migration (135).

Targeted disruption of components of the insulin signaling cascade modulates the response to vascular injury in vivo. IRS-2 knockout mice show increased neointima formation after balloon-injury, indicating that IRS-2 has a protective role against neointimal damage, possibly because it is an upstream mediator of PI (3)K/Akt signaling (137).

Treatment of bovine VSMCs with insulin (10 μg/ml) increases glycogen synthesis by 250% (89). Insulin-treated VSMCs show an alteration in components of the extracellular matrix such as an increase in fibronectin (138) and a decrease in hyaluronan synthesis (85). Insulin increases the production of VSMC PAI-1, a protease inhibitor implicated in preventing fibrinolysis (139).

C. Monocyte/macrophages

Circulating monocytes are important inflammatory cells involved in the immune response. Vessel wall lipid retention stimulates monocytes to differentiate into macrophages in an attempt to phagocytose and remove retained lipid. Because macrophages cannot break down lipid and there is an imbalance between cholesterol influx/efflux, cholesterol remains in the macrophages and these cells are described as “foam cells.” Foam cells are deposited in the subendothelial space and the atherosclerotic lesion develops around the lipid-laden macrophages (105).

Insulin and IGF-I receptors are present on circulating monocytes/macrophages (140, 141). Insulin-treated macrophages show classical activation of the insulin signaling pathway. Insulin and IGF-I can activate tyrosine phosphorylation of the β-subunit of the respective receptors (141). The insulin receptor is tyrosine phosphorylated, leading to activation of IRS-2 and PI (3)K (142). Unlike the classical substrates required for insulin signaling in metabolic tissue, endothelial cells and VSMCs, IRS-1 is undetectable in monocytes/macrophages (141, 143).

Treatment of rat macrophages with insulin increases hexokinase activity, which phosphorylates glucose during glycolysis, and decreases glucose-6-phosphate dehydrogenase, a key enzyme of the pentose phosphate pathway, an alternative pathway to glycolysis (144). Of the glucose transporters, macrophages express GLUT1, GLUT3, and GLUT5, which are distinctly regulated during monocyte differentiation into macrophages (145). However, insulin treatment of mice macrophages has no significant effect on glucose uptake by macrophages (142). Treatment of macrophages with insulin does not increase phagocytosis (144).

Defective insulin signaling is implicated in macrophage foam cell formation (142). Reduced insulin receptor number, decreased insulin receptor/IRS-2 phosphorylation, and downstream signaling occurs in macrophages from obese, insulin-resistant mice (142). To support these findings, reduced insulin receptors on monocytes have also been documented in human monocytes from obese subjects (146). Insulin infusion by the clamp technique, while maintaining normal plasma glucose, reduces monocyte insulin receptors in a dose- and time-dependent manner (147). In addition, defective insulin signaling in terms of reduced receptor tyrosine kinase activity is observed in monocytes from nonobese, normoglycemic subjects with insulin resistance (148).

Macrophages from obese, insulin-resistant mice show an up-regulation of CD36, the macrophage receptor for oxLDL, compared with macrophages from wild-type mice (142). Insulin-treated rat macrophages from nonobese mice show an increase in CD36 receptor and increased binding to oxLDL (142). Macrophage CD36- and ApoE-deficient mice show reduced vascular disease compared with ApoE knockout controls, and thus CD36 is implicated in atherosclerosis (149). Macrophages lacking the insulin receptor obtained from insulin receptor knockout mice show an increase in CD36 protein when compared with macrophages from mice that express the insulin receptor (142). Treatment of normal mice macrophages with different PI (3)K inhibitors, wortmannin or LY294002, in the absence of insulin increases CD36 receptor expression and binding to oxLDL, indicating that defective insulin signaling, as would occur in insulin resistance, regulates macrophage CD36 expression (142). Monocytes isolated from patients with diabetes also show an increase in CD36 expression, and this may be due to defective insulin signaling (150).

Macrophages from mice with alloxan-induced diabetes produce more IL-6, TNF-α, and reactive oxygen species compared with macrophages from nondiabetic control mice (151). Macrophage apoptosis in atherosclerotic lesions may contribute to further monocyte recruitment by the release of cytokines and may thus aggravate the development of the vascular lesion and play a part in plaque rupture (152–154). Insulin treatment of human macrophages reduced apoptosis induced by incubating macrophages in serum-free medium, by increasing the expression of the antiapoptosis gene B-cell lymphoma-X (Bcl-X) (155). Insulin-induced expression of macrophage Bcl-X is inhibited by wortmannin a PI (3)K inhibitor but not by the MAPK inhibitor PD98059 (155). These studies suggest that under insulin resistance conditions the protective effect of insulin to reduce macrophage apoptosis may be lost because the PI (3)K pathway is blunted under these conditions.

D. T lymphocytes

Like monocytes, T lymphocytes adhere and infiltrate the vascular endothelium where they are immunologically active (156, 157). Both CD4+ (T-helper cells) and CD8+ (cytotoxic T cells) cells have been found in human atherosclerotic lesions (41, 42). T-helper 1 cells secrete cytokines such as interferon-γ, IL-2, and TNF-α and -β, which promote macrophage activation. T-helper 2 cells secrete IL-4, IL-5, and IL-10, which are also found in atherosclerotic plaques. Markers of atherosclerosis and insulin resistance such as increased serum FFA decrease normal T lymphocyte calcium signaling both in vivo and in vitro (158). C-peptide, a product of proinsulin that is increased in insulin-resistant subjects, colocalizes with CD4+ T lymphocytes in atherosclerotic lesions from patients with diabetes and acts as a chemotactic stimulus for T lymphocytes to adhere and penetrate the vessel wall (159).

Of all the cells involved in atherosclerosis, insulin signaling in T lymphocytes has received the least attention. Unlike monocytes, T lymphocytes do not have insulin receptors in the circulation; however, T lymphocytes have the unusual ability to express insulin receptors after presentation of an antigen in vivo and an antigen or mitogen in vitro (160). Once activated, T lymphocytes demonstrate classical insulin signaling such as increased phosphorylation of IRS-1 (161) and downstream metabolic effects such as increased glucose uptake mediated via PI (3)K (162). Stimulation of the insulin receptor on T lymphocytes increases the cytotoxic effects, allows for differentiation of the cells, and maintains the lymphocyte in the activated state after presented with a mitogen (lectin) or antigen. The insulin receptor allows the activated lymphocytes to accommodate extra energy requirements (162, 163).

Plasma insulin levels in vivo have an inverse relationship with the number of insulin receptors on T cells in vitro after the presentation of an antigen (164). The number of insulin receptors on T lymphocytes stimulated by an antigen in vitro is reduced after removal from patients with obesity, and there are even fewer antigen-stimulated insulin receptors when the T cells are from patients with type 2 diabetes (165). In subjects without diabetes, an acute increase in plasma insulin independent of plasma glucose decreases the number of binding sites for insulin on T lymphocytes after presentation of an antigen in vitro (160). Insulin treatment of activated T lymphocytes from normal subjects leads to an increase in the expression of IL-2 mRNA (161).

IGFRs are present on T lymphocytes in the circulation (166). IGF-I treatment of T lymphocytes decreases the expression of interferon-γ receptor 2 (167). IL-10 has antiinflammatory actions by decreasing T-helper 1-dependent inflammatory effects. IGF-I stimulates the production and mRNA expression of IL-10 in purified T lymphocytes (168). It is unclear which downstream insulin signaling pathways are involved in the effects of IGF-I on modulating the secretion of cytokines from T lymphocytes.

VIII. Abnormal Signaling in Insulin Resistance in Vascular Tissues

Insulin resistance may be due to defect(s) at some point before insulin binding to its receptor, an insulin receptor defect, and/or defects in downstream signaling components. Insulin receptor defects in specific tissues do not necessarily confer systemic insulin resistance. Pre-insulin receptor faults generally result from genetic mutations in the insulin receptor gene (rare) or alterations in delivery of insulin to its receptors giving rise to reduced insulin effectiveness. Defects in the insulin receptor that may contribute to insulin resistance include defects in receptor structure, number, binding affinity, and/or signaling capacity. It could be expected that an increased number of hybrid receptors (IGF-I/ insulin receptor) would reduce insulin sensitivity with these receptors preferentially binding IGF-I under physiological conditions; however, to date there is conflicting evidence regarding receptor structure, isoform expression, and binding affinity contributing to insulin resistance in type 2 diabetes in humans (62, 169).

Insulin resistance results in decreased FFA uptake and/or increased FFA levels in adipocytes and in the circulation. The basis for this change in fat cell metabolism leading to increased intracellular hydrolysis of triglycerides is not known. Insulin resistance often accompanies visceral obesity and elevated levels of the atherogenic factor PAI-1 (170). High circulating FFA levels enhance glucose output from the liver and reduce glucose disposal in skeletal muscle, thereby contributing to insulin resistance. To date there is no clear evidence to link any candidate regulators of fatty acid metabolism with insulin resistance. Increased FFAs stimulate the assembly and secretion of very low-density lipoprotein (VLDL) from the liver. Increased plasma levels of VLDL contribute to an increased accumulation of VLDL in the blood vessel wall in atherosclerosis. Indeed, the higher plasma levels of VLDL can result in higher levels of small, dense LDL and consequently deliver greater amounts of cholesterol to the vessel wall (171), where it becomes bound to proteoglycans within the extracellular matrix and can thereby contribute to plaque formation (172).

There has been considerable work on the direct actions of insulin on blood vessels and cells of the vasculature. The rationale for these studies varies from a view that these agents may be pro- or antiatherogenic. The concept of insulin resistance provides a particular perspective on these actions. If insulin resistance is associated with accelerated processes of atherosclerosis, then actions of insulin that are antiatherogenic are attenuated, insulin actions that are proatherogenic are exaggerated, or both. Thus, what are the direct actions of insulin in the vasculature, and how might insulin resistance be manifest?

Considerable evidence exists implicating vascular endothelium as a physiological target of insulin and, consequently, a potential link between insulin resistance and atherosclerosis. Endothelial dysfunction is one of the earliest detectable signs in insulin resistance, occurring even before the development of clinical manifestations. Insulin, a vasodilator, increases endothelial NO production and is regulated via PI (3)K-dependent/Akt insulin receptor signaling (91). Insulin normally stimulates the net production of NO via eNOS/tetrahydrobiopterin (BH4) activity. Impairment of NO production leads to insulin resistance such as that seen in eNOS knockout mice (100) and humans treated with an eNOS antagonist l-nitro monomethyl arginine (173), highlighting the involvement of eNOS and endothelial cells in insulin resistance and the potential to be proatherogenic if eNOS levels are reduced. Impairment of NO production has been demonstrated to be via a number of pathways including altered phosphorylation of eNOS, depleted BH4 availability, and impaired endothelin-1 signaling via ET_B (174). It is not evident whether a reduction in NO necessarily confers insulin resistance because not all disease states that have decreased endothelial NO have associated insulin resistance. Reduced NO production impacts not only on endothelial cells but also on VSMCs, the major constituent of the vessel wall responsible for vascular integrity and tone. In insulin resistance, VSMCs have an impaired responsiveness to NO, resulting in increased contraction that can result in an imbalance in vascular tone and may impact on insulin-mediated delivery of glucose to muscle in the insulin-resistant state (96).

Although the PI (3)K-dependent insulin signaling pathway is impaired in virtually all states of insulin resistance, the mitogenic MAPK pathway in endothelial and VSMCs remains intact and responds normally to insulin, and its associated cell effects may actually be enhanced rather than attenuated (107, 175, 176). Preferential signaling along this pathway may contribute to the progression of atherosclerosis with a concomitant increase in VSMC migration in the expression of cell adhesion molecules (VCAM-1, E-selectin) and cell interactions between vascular cells and macrophage/monocytes (107). VSMCs are also stimulated to increase production of PAI-1, resulting in reduced fibrinolysis and increased vascular occlusion and increased prenylation of GTPases Ras and Rho, leading to an enhanced mitogenic response of these cells to growth factors including PDGF (120). Macrophages also demonstrate alterations in their signaling pathways in insulin resistance. Macrophages from obese (ob/ob) mice are insulin resistant and display a posttranscriptional increase in CD36, a key molecule in the recognition of modified LDL, as a direct result of decreased insulin signaling (142). The elevated level of CD36 is a result of decreased catabolism, indicating a defect in receptor trafficking in macrophages, and results in increased binding, uptake, and degradation of LDL that may contribute to the development of atherosclerosis in the insulin-resistant state.

IX. Conclusions, Implications, and Speculations

It is clear from the above analysis that the exact molecular cause of insulin resistance is presently unknown. Several observations allow us to speculate on its origins. The key factors in insulin resistance are the presence of obesity and a genetic component. Increasing obesity is associated with increasing insulin resistance, but there is an overlay of genetic susceptibility. Thus, there exist lean people with severe insulin resistance and obese people without insulin resistance. The genetic component appears to have a subtle molecular manifestation and not an overt change in receptor structure or function. Numerous steps have been identified in the process of insulin signaling that convert insulin binding to increased glucose uptake in traditional insulin-sensitive tissues. Metabolic factors associated with obesity, such as FFA, can modulate insulin signaling pathways, so we speculate that there are molecular processes in the insulin signaling cascade that are susceptible to inhibition or activation by metabolic factors associated with obesity, and thus the spectrum of responses results from the interactions between these factors to yield a spectrum of insulin-resistant states.

Insulin resistance is associated with accelerated atherosclerosis, and a major question is whether or not insulin resistance occurs in cardiovascular tissues and whether this is the driving force for accelerated cardiovascular disease. From the viewpoint of antiatherosclerotic drug therapy, the recent discovery and introduction of the agents being chemically thiazolidinediones and pharmacologically peroxisome proliferator-activated receptor-γ ligands, which improve insulin resistance, along with increasing use of metformin, which also improves insulin sensitivity, provides a very useful opportunity to address the burden of disease associated with insulin resistance. In addition, the last two decades of vascular biology research have yielded much information on the biochemical and cell biology factors involved in the formation, progression, and rupture of atherosclerotic plaques, and it is hoped that this will lead to the discovery of agents that directly target these mechanisms in the vessel wall (177). Thus, greater understanding of the role of lifestyle interventions and their implementation and compliance coupled with new therapeutic strategies for targeting insulin resistance and the vessel wall should provide the pathways to alleviate the huge burden of cardiovascular disease that is confronting the world at present.

Acknowledgment

The authors thank Stephen Twigg, M.D., Ph.D. (University of Sydney, Department of Medicine) for constructively critiquing this review.

The laboratory has received financial support over the last several years from the National Health and Medical Research Council of Australia, Diabetes Australia Research Trust, Eli Lilly Endocrinology Research Awards, GlaxoSmithKline (Australia), Monash University, Department of Medicine, and the Baker Heart Research Institute, Division of Vascular Biology. Postgraduate students are funded by Australian Postgraduate Awards and the Biomedical (Dora Lush) Postgraduate Research Scholarship from the National Health and Medical Research Council of Australia.

Abbreviations

 
• eNOS,

  Endothelial NO synthase;

 
• FFA,

  free fatty acid;

 
• GAG,

  glycosaminoglycan;

 
• ICAM,

  intercellular adhesion molecule;

 
• IGFR,

  IGF receptor;

 
• iNOS,

  inducible NO synthase;

 
• IRS,

  insulin-receptor substrate;

 
• LDL,

  low-density lipoprotein;

 
• MMP,

  matrix metalloproteinase;

 
• NO,

  nitric oxide;

 
• oxLDL,

  oxidized LDL;

 
• PAI,

  plasminogen activator inhibitor;

 
• PDGF,

  platelet-derived growth factor;

 
• PI (3)K,

  phosphatidylinositol-3-kinase;

 
• PKC,

  protein kinase C;

 
• PKG,

  protein kinase G;

 
• SMC,

  smooth muscle cell;

 
• VCAM,

  vascular adhesion molecule;

 
• VLDL,

  very low-density lipoprotein;

 
• VSMC,

  vascular SMC.