Vascular biology was revolutionized by the simple experiment by Furchgott that demonstrated dose-dependent relaxation in isolated coronary artery rings in the presence of endothelium. This discovery led to many new questions and experiments.

Miller and Vanhoutte discovered endothelium-dependent relaxation in bony fish, showing that endothelium-dependent control of vascular tone is an ancestral form of local regulation of vasomotor tone. Vanhoutte and colleagues confirmed the presence of endothelium-dependent control in human arteries. This is now an accepted principle.

What is the mechanism for endothelial-dependent vasorelaxation?  One possibility is cell-to-cell contact and transmission of an electrical signal. This most likely occurs in very small blood vessels, where the contact between endothelial cells and smooth muscle cells (SMC) are very close. In large blood vessels, the endothelial cells produce endothelium-derived relaxing factor (EDRF), which has a very short half-life of a few seconds.

EDRF is nitric oxide (NO) formed from L-arginine by the endothelial cells, which contain the enzyme NO synthase (NOS). NOS transforms L-arginine in nitric oxide. NO diffuses to vascular smooth muscle and activates soluble guanylate cyclase (sGC) leading to increased levels of cyclic GMP, causing relaxation.

Competitive inhibitors of the enzymes were developed, including L-NMMA, to analyze the physiological role of NO. Using these competitive inhibitors in intact organisms showed that NO is involved from memory to erection. NO has become one of the most important mediators in biology.

EDRF is not the only vasoactive factor. De Mey and Vanhoutte, in 1982, concluded from a series of complex experiments in human coronary arteries that functional coupling between endothelial cells and the arterial media involves at least three different pathways, two of which depend on the metabolism of the fatty acid.

The presence of NO produced by NOS is known. But, certain endothelial cells in certain blood vessels, particularly human blood vessels, generate sufficient prostacyclin to contribute to endothelium-dependent responses. In the third pathway, endothelial cells cause the release of endothelium-derived hyperpolarizing factor (EDHF), causing relaxation by hyperpolarizing vascular smooth muscle by opening a potassium conductor.

Some of the most important ways to explain endothelium-dependent hyperpolarization includes gap junctions, direct contact between cells, and production of potassium. A very important hypothesis developed by Shimokawa and colleagues, has very convincing evidence, particularly in humans, that hydrogen peroxide can act as a very important hyperpolarizing factor. Presently this is probably one of the most exciting aspects of EDHF research.

Nonselective agents are still being used to investigate the action of EDHF in humans, thus little is known about its physiological importance. However, Vanhoutte maintains that EDHF is important, particularly in small blood vessels, and that EDHF becomes very important when NO is insufficient. EDHF as a backup system for failing NO production is an area of research pursued by Vanhoutte.

The only likely source of acetylcholine (ACh) is cholinergic nerves, which enervate very few blood vessels. Nerves enervated by cholinergic nerves are found in the adventitial medial border. Any ACh liberated from such nerves must escape the acetylcholine esterases around the nerve, traverse the VSMC and find a hole in the elastic membrane to reach the endothelial cells. Because there are few data to support this process, Vanhoutte and colleagues have looked for other stimuli that would have more physiological relevance as releasers of EDRF, NO, and EDHF.

The shear force exerted by the flowing blood is one important stimulus. Vanhoutte and colleagues showed in the experimental setting that increased shear stress on endothelial cells releases NO and relaxing factors. This explains the well-known phenomenon of flow-dependent vasodilatation. Measuring flow-dependent vasodilatation remains perhaps the best way to test endothelial function, particularly in humans.

Bradykinin is a potent endothelium-dependent dilator, as shown in work by Vanhoutte and colleagues in canine basilar arteries. They also showed that bradykinin acts on B2 kinin receptors, causing the release of NO and EDHF. Medically this is important because bradykinin cannot be dissociated from converting enzyme, which transforms the inactive peptide angiotensin I into angiotensin II, which is a very strong vasoconstrictor and growth promoter. The converting enzyme also is the most important destruction pathway for bradykinin; it is the enzyme responsible for the breakdown of most kinins in the body.

The endothelium-dependent effect of bradykinin is important for the pharmacology of ACE inhibitors. This group showed in coronary arteries with and without endothelium that the ACE inhibitor perindopril markedly augments the EDRF-releasing properties of bradykinin in the presence of endothelium. This is thought to be due mainly to ACE inhibitors protecting bradykinin from breakdown, so that it can act longer on its receptors, causing more NO and more EDHF to be released.

This group has also demonstrated that most large arteries, when they contain endothelium, also contain precursors, kininogen and kallikrein, in sufficient amounts to cause local vascular production of kinins in sufficient amounts to stimulate B2 receptors on endothelial cells and cause the release of EDHF and NO.This group showed that a B2 kinin antagonist annuls the effect of perindopril in isolated canine coronary arteries with endothelium under shear stress, and that the antagonist blocks the shear stress-induced relaxation. They concluded that somehow shear stress was upregulating the production of NO that involved the local generation of bradykinin.  Drexler and colleagues showed in humans that the local injection of a B2 antagonist inhibited flow-dependent vasodilation. Vanhoutte maintains that shear stress turns on the production of endogenous bradykinin, which acts on B2 kinin receptors, causing augmented release of NO and EDHF. And, this is responsible in part for the protective action of ACE inhibitors. EDHF is responsible in part for vasodilatation and the decrease in blood pressure. Their experiments also make it easier to understand why ACE inhibitors lower blood pressure, even in patients with normal or low levels of renin and angiotensin II.

Thrombin is almost an ubiquitous releaser of EDRF.  In coronary arteries, thrombin causes impressive relaxation in the presence of endothelium, but not in the absence of endothelium. Thus, blood vessels with healthy endothelium will begin to generate NO if thrombin is formed in the vessel.

In coronary artery rings with and without endothelium, injection of human platelet suspension causes vigorous contraction of the smooth muscle in the absence of endothelium. But this is not seen in the presence of endothelial cells. Platelets cause a massive release of NO.

Serotonin, one of the most important platelet-derived products, causes vasorelaxation in porcine coronary arteries with endothelium, as shown by Shimokawa and Vanhoutte. Pertussis toxin, a Gi-protein inhibitor, inhibits most of the response to serotonin, but not the response to bradykinin. From this and other experiments in his lab, Vanhoutte concluded that two families of G proteins, Gi and Gq, in the endothelial cells cause release of NO as a result of platelet aggregation.

Physiologically, the response to aggregating platelets and thrombin is extremely important.
The initiation of platelet aggregation in a blood vessel with normal endothelium will cause the release of 5HD and ADP and set in motion the local coagulation cascade for the production of thrombin. Those substances act on the endothelial cells to cause large amounts of NO to be fused to the smooth muscle and relax the blood vessel. Simultaneously, NO, particularly if prostacyclin is produced simultaneously, will exert a formidable feedback on the aggregation process at the interface between the blood and the endothelial cells. In the absence of endothelium, the platelet products serotonin and thromboxane fuse to the smooth muscle and cause vasocontraction. The blood vessel shuts down, causing what is called the vascular phase of hemostasis, which permits proper clot formation. So, the presence or absence of endothelium can contribute greatly to survival.

This response is not the same in all blood vessels. Veins are poor donors of nitric oxide. Experiments in canine blood vessels and humans vessels have shown that arteries relax in response to increasing concentrations of ACh, whereas veins do not. This helps to explain why veins are not as good as arteries long-term for coronary bypass grafts.

Gender can affect the release of NO. Miller and Vanhoutte showed in ovariectomized rabbits that estrogen, in the form of 17 beta-estradiol, chronically augmented endothelium-dependent responses to ACh in femoral arteries. Work from other groups has confirmed that estrogens upregulate the expression of NO. This may be responsible for the protection against coronary disease women are afforded pre-menopause.

Exercise is healthy for the endothelium, because regular exposure of the endothelium to increased flow in the heart and in the limbs will cause upregulation of NO. Dietary intake is of fundamental importance. Shimokawa and Vanhoutte showed in porcine arteries with endothelium that chronic intake of cod liver oil markedly potentiated platelet aggregation, due to the intake of omega-3 in saturated fatty acids. Polyphenolic extract from Bordeaux wine augments the release of NO, as demonstrated in the experimental setting in rat aorta.

Age also downregulates the expression and activity of NO synthase. The coronary response to ACh, as in index of NO release, decreases with age, as shown in the experimental setting. Smoking has been shown in several studies to downregulate NO synthase. Also, in smokers the ability of the endothelium to defend the organism by releasing NO is curtailed.

Obesity can lead to endothelial dysfunction, based on very convincing data. In one study, by Taylor et al., in rat mesenteric arteries, ACh caused an increased response in the diabetic animals compared to normal glycemic animals. The direct effect of NO as evaluated by giving sodium nitroprusside was not affected.

Hypertension also leads to endothelial dysfunction.  In Dahl rat aortas of normotensive rats and hypertensive animals, salt-dependent hypertension caused a clear cut blunting of the response to ACh, illustrating endothelial dysfunction, in work in Vanhoutte’s lab.
In the setting of diabetes and hypertension, other factors, such as free radicals, also interfere with NO.

Is endothelial dysfunction just a consequence of disease, or could it play a primary role in defining the conditions that will lead to atherosclerosis? The same risk factors for cardiovascular and coronary disease are also involved in the development of endothelial dysfunction.

Endothelial cells grow during maturation. They do not grow after they touch each other, and they form a monolayer. They remain in place for years without changing. Yet, in the human body, endothelial cells are apoptotically programmed and will begin to die 30 or 40 years after maturation, and are immediately replaced by cells growing from the rim. But this regenerated endothelium that begins to invade the vasculature does not function as well as the original endothelium, as shown in work by Shimokawa and Vanhoutte.

In denuded porcine coronary artery in vivo, they showed that the endothelium was regrown one month later. In this regenerated endothelium, platelet injection caused an increased response, compared to native endothelium, because the response to serotonin was dysfunctional. Six months later, this problem remained. This is selective for serotonin. The same experiements showed no response to bradykinin or ADP. There is a selective dysfunction of the signaling cascade. It is believed that the receptors for serotonin are still present.

The Gi proteins are present in the regenerated endothelial cells, as shown by work by Vanhoutte and colleagues examining Gi family immunostaining in endothelial cells from native coronary endothelium and endothelial cells regenerated for 28 days. However, their function in the regenerated cells is downregulated.

In primary cultures of native and regenerated endothelium derived from pigs, they showed that the native endothelial cells are similar in size and contain once nucleus. But, in the primary culture from regenerated endothelium, the cells are very large and contain more than one nucleus, similar to what is seen with very old senescent endothelial cells.

The same phenomenon is seen in human coronary endothelium in a normal coronary artery. The native cells are similar in size with one nucleus, but in the endothelium regenerated to cover an atherosclerotic plaque, the endothelial cells are very large, again demonstrating premature senescence of endothelial cells.

In cell cultures of primary native and regenerated endothelium, they found that the regenerated cells had a 2-fold higher uptake of modified acetylated LDL than native cells.
In denuded endothelium in intact animals, they also found a 2-fold higher intake of modified LDL in regenerated endothelial cells. So, even when exposed to normal levels of lipids, the regenerated cells would accumulate much more LDL.

Increased modified LDL is the only factor known to selectively remove the response of serotonin in endothelial cells. Cox and Cohen showed that an increase in oxidized LDL blunted the endothelium-dependent response in porcine arteries. The blunted response is very similar to that seen with pertussis toxin and in regenerated endothelium. Vanhoutte and colleagues found a significant increase in the generation of oxidized LDL in the regenerated endothelial cells in the cell cultures.

This intracellular increase in oxidized LDL, as long as it remains moderate, selectively removes the physiologically important serotonin signaling cascade in the endothelial cells.
As a consequence, in the areas covered with regenerated endothelium, there is a chronic shortage of NO. NO inhibits contraction of smooth muscle, platelet aggregation, proliferation of smooth muscle, oxidation of LDL, expression of adhesion molecules, and the adhesion of platelets and white blood cells. NO also is a superb inhibitor of endothelin.

Shimokawa and Vanhoutte also showed obvious signs of myocardial infarction in the denuded porcine coronary artery after serotonin injection. NO inhibits the expression of adhesion molecules. Under chronic reperfusion in another model, Vanhoutte showed that blood cells did not attach to native endothelium, but that platelets, white blood cells, and red blood cells adhered to the regenerated endothelium. Hemodynamically, this amount of material will not impede blood flow, but it does increase the presence of growth factor, while NO, a good inhibitor of cellular growth, is decreased. Vanhoutte and colleagues have shown in native vascular smooth muscle cells that increased endogenous NO inhibited cellular growth, and in contrast, in regenerated cells there was less NO production and increased cellular growth.

In a pig with denuded left anterior descending (LAD) coronary artery fed a cholesterol-rich diet, at 2 months there was a remarkable form of accelerated atherosclerosis, only in the denuded areas, compared to a pig with a denuded LAD fed a Mediterranean diet in work by Vanhoutte and colleagues. Clearly, endothelial dysfunction is causal in the genesis of atherosclerosis.

To study atherosclerosis in cardiac transplantation, Vanhoutte and Perrault planted a second heart in the abdomen of pigs. The native heart serves as the in-built control and the implanted heart represents cardiac transplantation. At day 30 after transplantation, vasorelaxation in the porcine artery with endothelium is blunted in response to serotonin, compared to bradykinin. This is thought to be beginning of atherosclerosis.  In the artery in the pig heart, microscopy reveals that endothelial dysfunction has permitted cell adhesion, which will then initiate the other aspects of the atherosclerotic and inflammatory process.

Dysfunctional endothelial cells is truly a key event that sets in motion the atherosclerotic process, because it sets the stage for the growth of the atherosclerotic material.