Editor’s Note: This story, first published in 2007, has been updated.
It may start with a fat-rich diet, physical inactivity, smoking or an unfortunate inheritance.
Whatever the cause or causes, the delicate inner lining of an artery is injured—scratched, if you will—revealing the underlying collagen infrastructure. Platelets rush to site to repair the injury and prevent bleeding. They attach themselves to the exposed collagen, and recruit other platelets to join them.
Along comes thrombin to further sound the alarm and “set” the clot. Thrombin, an enzyme, chops up a protein called fibrinogen into short, sturdy strands of fibrin that trap platelets and other blood cells like fish in a net.
Just enough clot can save a life. Too much clot, however, can take it, by blocking an artery and triggering a heart attack or stroke.
This may be what’s behind the high rates of heart disease among people with type 2 diabetes and obesity.
Currently about 66 million Americans—nearly a third of the adult population—are obese. Most will develop heart disease. “There is no greater threat to American’s cardiovascular health,” warns Douglas Vaughan, M.D., former chief of Cardiovascular Medicine at Vanderbilt University Medical Center.
The prevalence of type 2 diabetes also is burgeoning, and is expected to double—to nearly 40 million people—within a decade. At least 85 percent of these people will die from blood clots that stop their hearts or their brains. Women seem to be more vulnerable than men.
“Unless the underlying mechanisms responsible for these events can be identified, there will be an unprecedented number of diabetic patients suffering thrombotic episodes in the next 10 years,” predicts Stephen Davis, M.D., Ph.D., chief of the Vanderbilt Division of Diabetes, Endocrinology and Metabolism. (See “How to tell if you’re at risk”).
In 2006, Davis and Vaughan joined forces with their colleagues at Vanderbilt to tackle one of the most urgent mysteries of modern medicine—why exactly do obesity and diabetes increase the risk of life-threatening blood clots?
Vaughan has since left Vanderbilt to chair the Department of Medicine at Northwestern University Feinberg School of Medicine. But his former colleagues are continuing their inquiry through a Specialized Center of Clinically Oriented Research (SCCOR) supported by the National Heart, Lung, and Blood Institute. Their goal: to investigate bleeding and clotting disorders, and to rapidly translate their discoveries into clinical practice
Colorized image of a retinal scan reveals a tiny clot (white spot at the intersection of two blood vessels). This occasionally is the first sign of heart disease in patients being screened for something else—in this case, diabetic retinopathy. "The canary in the coal mine" analogy is particularly apt," says Lawrence M. Merin, who directs the Vanderbilt Ophthalmic Imaging Center, "as the vast majority of these patients have no visual symptoms. The clinical significance of these lesions is most important; if left untreated, there is a 50 percent mortality rate within five years of identification of these lesions."
Certainly diet and exercise can help people avoid the serious health consequences of obesity and diabetes. But lifestyle changes are easier said than done. And it is becoming increasingly clear that genetic predispositions influence risk—and the effectiveness of risk-lowering interventions.
“If you have a set of inherited risk factors…, even modest hypertension (and) modest elevations in lipids… may have a more deleterious effect on you than they do on another individual with the same numbers but a lower set of genetic risk factors,” says Samuel Santoro, M.D., Ph.D., chair of Pathology at Vanderbilt and a SCCOR participant.
Better understanding of the molecular and genetic contributions to clotting disorders could lead to new therapies that are more effective and have fewer side effects. It also could help doctors identify and manage patients at risk of serious complications during and after heart procedures such as angioplasty.
Another goal of the SCCOR is to help open the door to individualized, preventive medicine. If genetic variations or polymorphisms that increase the risk for thrombosis can be identified, it may be possible to screen patients in advance, and make adjustments in the drugs they are given, the procedures they undergo and the lifestyle changes they make so that they never experience that first heart attack or stroke.
Powerful platelets
In the early 1960s, the landmark epidemiological study conducted in Framingham, Mass., pinpointed high blood pressure, smoking and high cholesterol as major risk factors for cardiovascular disease. Within a decade, coronary care units, angioplasty and open heart surgery had become the standard of care, and drugs had been developed to lower blood pressure and cholesterol levels.
But while blood pressure drugs and cholesterol-lowering statins remain a pharmaceutical cornerstone in the effort to prevent heart attacks and strokes, “non-traditional” treatments may be equally as important among people with diabetes and obesity. Vaughan believes these people may have “a special environment in their arterial vasculature that promotes cardiovascular complications.”
One of the key players is the platelet—the disc-shaped element produced by the bone marrow that promotes blood clotting. While it does not have a nucleus, and therefore is not a cell, it is bristling with receptors, enzymes and other factors that allow it respond to—and powerfully influence—its environment.
Among them:
— Collagen receptors, notably alpha 2, beta 1 integrin and glycoprotein VI, which enable platelets to attach to exposed collagen at the site of a tear in the blood vessel lining; and
— Protease activated receptors (PARs), which, when activated by thrombin, trigger the aggregation or clumping of other platelets into the clot.
“When platelets encounter collagen, they don’t just stick to it; they are stimulated to form aggregates,” says Santoro, whose lab at Washington University in St. Louis discovered the alpha 2, beta 1 integrin collagen receptor in 1990. “It is the first component of blood clotting.”
The platelets of patients with type 2 diabetes have been found to “over-express” the integrin receptor. As a result, they may bind more readily to exposed collagen, thereby increasing the clotting risk.
This idea is supported by a mouse model engineered by Santoro, Mary Zutter, M.D., and their colleagues in 2002, the year before they moved their lab to Vanderbilt. The animals, which lack the integrin gene, exhibit a “profound deficit” in platelet adhesion to collagen.
Zutter, who directs the Vanderbilt Division of Hematopathology, is continuing her efforts to model the mechanism of collagen binding, while Santoro and others in his lab are conducting studies in humans to better understand the genetics of receptor over-expression.
Meanwhile, Heidi Hamm, Ph.D., chair of Pharmacology at Vanderbilt, is trying to understand why the platelets of people with type 2 diabetes seem to be “hypersensitive” to other clotting stimuli, including thrombin. She is focusing on the thrombin receptors, G protein-coupled PAR1 and PAR4.
G-proteins are intracellular molecular “switches” that transmit signals into the cell by attaching to G protein-coupled receptors in the cell membrane. While at the University of Illinois at Chicago in the early 1990s, Hamm and her colleagues solved the G-protein’s three-dimensional structure.
They also showed that a peptide, a piece of protein, could prevent the G-protein from hopping on and off its receptor each time a signal came to the cell. This suggested that the switch for a single pathway, such as the one activated by thrombin, could be disabled by blocking a specific G-protein.
“Maybe PAR signaling is different” in diabetes, Hamm says. “If we find it’s so, could we direct a therapy that could be able to quiet down the platelets?”
Through their collaboration, Hamm and Santoro also have found that the collagen and thrombin receptors influence each other’s activity. That, says Santoro, “suggests some very interesting pharmacological interventions.”
Clot busters
Over-active clotting is not the only problem faced by people with diabetes and obesity. They also seem to have a faulty blood thinning system.
Normally, there’s a whole series of natural anti-coagulants, including protein C and antithrombin III, which circulate through the bloodstream and inhibit clot formation.
Another set of short-lived anti-coagulants generated by the blood vessel lining include nitric oxide and prostacyclin. They inhibit platelet aggregation and dilate blood vessels, thereby improving blood flow.
“Mother Nature must have been very concerned about the possibility that we would form clots inside our blood vessels because we have several different mechanisms that prevent this,” Vaughan says. If the natural anti-coagulants fail to do their job, there’s a “back-up,” called the fibrinolytic (clot-dissolving) system.
The king of the clot-dissolvers is plasmin, an enzyme that hacks apart the fibrin mesh.
Produced in an inactive form, plasminogen, in the liver, it is “liberated” to do its work by other enzymes, called plasminogen activators. These enzymes, in turn, are inhibited by proteins called plasminogen activator inhibitors.
Plasminogen activators like t-PA and their inhibitors, notably PAI-1, thus maintain a balance between too much clotting and not enough. It’s a balance that’s all too easily tipped—in favor of thrombosis.
For example, too much fat in the bloodstream (a characteristic of obesity) and too much glucose (the hallmark of diabetes) can increase levels of PAI-1. So can activation of the renin-angiotensin-aldosterone system (RAAS), which regulates blood pressure.
“For a protein that’s involved in regulating the clot-dissolving system, it’s rather puzzling to see how it’s influenced by a variety of different factors that you would think have nothing to do with clotting or protection from clotting,” says Vaughan, who is internationally known for his research on the fibrinolytic system. “All these things that get messed up in obesity and insulin resistance end up driving PAI-1 production.”
In the 1990s, Vaughan, Nancy Brown, M.D., and their colleagues at Vanderbilt helped identify the critical relationship between the blood-pressure regulating and clot-dissolving systems.
Diabetes and obesity pull the dial to the left (toward clotting) by making platelets "stickier" and more sensitive to clot-forming thrombin, and by increasing production of PAl-1, which inhibits t-PA and thus clot-dissolving plasmin. Similarly, too much fat in the blood can increase fatty acid oxidation, which in turn, accentuates the platelet-clumping actions of thromboxane over the blood-thinning actions of prostacyclin.
Key players in blood pressure regulation include the angiotensin converting enzyme (ACE) and the hormone aldosterone.
ACE converts angiotensin I into a small protein called angiotensin II, which causes vasoconstriction (narrowing of the blood vessels) and raises PAI-1 levels. Aldosterone, which regulates blood volume, also increases PAI-1 production.
Drugs that block the ACE enzyme, called ACE inhibitors, are widely used to lower blood pressure. By squelching excess production of PAI-1, they also can reduce the risk of a clot-induced heart attack. So can a class of diuretics that block the hormone aldosterone.
In 2002, Vaughan and his colleagues led by then post-doctoral fellow Mesut Eren (now a research assistant professor in Medicine) engineered a strain of mice that overexpresses a long-lasting form of human PAI-1.
As they age, about half the mice spontaneously form clots in their coronary arteries, without evidence of hypertension, atherosclerosis, or high lipid levels. This is evidence, Vaughan says, that high PAI-1 levels, in themselves, can precipitate clotting and a resulting heart attack.
The Vanderbilt researchers are continuing their work to determine what factors increase PAI-1 levels both in animals and in patients.
In the meantime, drugs that specifically inhibit PAI-1 are being developed. Vaughan predicts that patients at risk of heart attack in the future may have their PAI-1 levels checked, just as they are currently screened for high levels of cholesterol.
The peroxide connection
The anti-coagulant system also is a delicate balance that can easily get out of whack.
Prostacyclin’s ability to prevent platelet clumping, for example, is matched by thromboxane, which does exactly the opposite—it stimulates platelet aggregation and constricts blood vessels, thereby limiting blood flow.
Prostacyclin and thromboxane are members of a family of lipid molecules that include the prostaglandins, and which are involved in everything from inflammation to smooth muscle constriction and blood pressure regulation. They are generated by the cyclooxygenase (COX) enzymes.
Thromboxane is produced in platelets by the COX-1 enzyme, while prostacyclin is a product of the COX-2 enzyme acting in the blood vessel lining. Both COX enzymes are inhibited by aspirin. During the past 30 years, John Oates, M.D., founding director of Vanderbilt’s Division of Clinical Pharmacology, has helped define the role of these critical compounds in high blood pressure, cancer and other disorders.
In the mid-1980s, for example, Oates and Garret FitzGerald, M.D., currently chair of Pharmacology at the University of Pennsylvania, found that low doses of aspirin inhibited thromboxane production without unduly lowering levels of prostacyclin. Their work helped form the basis for the use of low-dose aspirin to reduce clotting risk in heart patients.
People with diabetes, however, seem to be “resistant” to aspirin’s protective effect. Oates, who is leading a fourth major project in the SCCOR, believes this may have something to do with peroxide.
The COX enzymes actually have two binding sites: one for arachidonic acid, the fatty acid precursor to prostacyclin, thromboxane and the like; and the other for peroxide.
Peroxide is a ubiquitous cellular messenger that in this case activates the COX enzymes like the starter of a car. It is produced when fats in the body are exposed to oxygen (lipid peroxidation)—the same thing that causes butter to go rancid.
Too much peroxide, however, will overcome aspirin’s ability to block COX activity. “We’ve found that when the enzyme is exposed to peroxide, aspirin binds less readily,” Oates says.
In 1990, Jason Morrow, M.D., Jackson Roberts, M.D., and their colleagues at Vanderbilt discovered a series of bioactive prostaglandin-like compounds they called isoprostanes.
Produced by lipid peroxidation, isoprostanes are now recognized as the gold standard to measure oxidative stress. Smokers have high levels of isoprostanes. So do people who are obese or who have high levels of low-density lipoprotein, the “bad” form of cholesterol, in their bloodstream.
Another unhealthy consequence of too much fat thus may be aspirin resistance.
Genetics also may play a role. “It’s conceivable there may be a polymorphism (mutation) in the COX 1 gene related to aspirin resistance,” Oates says. “Once we identify groups of people who are susceptible to aspirin resistance we can devise all kinds of strategies” to reverse it.
Paradoxical effects
People with high levels of fats and glucose in their bloodstream are not the only ones at increased risk of dying from a clot—so are people with hypoglycemia—not enough glucose to satisfy the body’s energy needs.
Hypoglycemia is the complication most feared by patients with diabetes who must inject insulin periodically to lower their blood glucose levels, Davis says. Unless recognized and treated immediately, severe hypoglycemia can lead to convulsions, unconsciousness and, occasionally, death.
Several reports have linked severe hypoglycemia to serious cardiac and cerebral thrombotic events—heart attacks and strokes. Corticosteroids—hormones that regulate carbohydrate and fat metabolism—may be involved.
Davis and his colleagues helped pioneer methods for teasing out the complex interactions between insulin, corticosteroids and other factors that regulate the supply and delivery of glucose and other fuels required by body tissues.
Recently they found that experimentally induced hypoglycemia resulted in significantly increased levels of PAI-1 in normal subjects. So did administration of drugs that mimic the actions of corticosteroids.
One of those hormones, cortisol, is secreted by the adrenal glands in response to psychological or physical stress, including hypoglycemia. Acting through corticosteroid receptors, it stimulates release of glucose by the liver. But it may also increase the risk of clotting.
Davis and his colleagues will try to find out whether activation of a specific corticosteroid receptor is responsible for hypoglycemia’s pro-thrombotic effect.
“This information is urgently needed,” he says, “bearing in mind… that increasing numbers of (people with) type 2 diabetes are experiencing hypoglycemia due to the move towards intensive metabolic control.”
While the Vanderbilt researchers say they are not out to discover new drugs, their work could lead to new avenues for drug discovery.
For example, Hamm’s research could lead to a more effective way to block thrombin.
There is already a thrombin blocker on the market, bivalirudin, which binds directly to the circulating enzyme. The drug is given intravenously to patients undergoing coronary procedures such as angioplasty. The goal is to prevent clotting without thinning the blood too much.
Bivalirudin is superior to the classic blood thinner heparin in preventing bleeding complications during coronary procedures, but it is no better in preventing subsequent heart attacks or deaths. That may be because the drug also inhibits the paradoxical anti-coagulant properties that thrombin exhibits at low concentrations.
A drug that blocked the thrombin receptor on the platelet could avoid this problem, Hamm says.
More drugs, however, are not enough.
“There are big issues that haven’t been answered,” she says. Among them: why deaths from cardiovascular disease among women with diabetes remain stubbornly high compared to their male counterparts.
Is the disparity due to an increase in smoking among women? Are women less likely to receive appropriate treatment? Or do they have a different biology that isn’t picked up in studies of new drugs conducted largely in men?
That’s why basic research is so critical. “There are cures that are out there,” Hamm says. “We just can’t predict where (they) are going to occur.”