Oxidation, Inflammation, and Hypertension

Research related to reactive oxygen species (ROS) and vascular disease performed in Harrison’s laboratory and by other investigators in recent years was reviewed in this special lecture, as well as their role in the genesis of hypertension and thus how they could serve as a target for treatment.

Virtually every aspect of atherosclerosis lesion formation can be contributed to ROS, and their production is increased by numerous risk factors, including high cholesterol, diabetes, and hypertension. ROS contribute to pathophysiologic phenomena, such as adhesion molecule expression, vascular smooth muscle growth, apoptosis, lipid oxidation, remodeling of the vessel through activation of matrix metalloproteinases (MMPs), and altered vasomotion. 

As a consequence of ROS, hypercholesterolemia, heart failure, diabetes, hypertension, cigarette smoking, aging, and nitrate tolerance stimulate an increase in the production of  superoxide by the endothelium, which reacts very rapidly with nitric oxide (NO), leading to a loss of NO in the formation of peroxinitrite. The rate constant of this reaction is extraordinarily rapid and is nearly diffusion-limited.

A major source of ROS in vessels is the enzyme NADPH oxidase, which is composed of two membrane subunits, gp91^phox and a docking subunit, p22^phox, plus cytosolic subunits including p47^phox, p67^phox, and small G-protein rac.

In recent years, the presence of NADPH oxidase in vascular cells and the endothelium, vascular smooth muscle cell (VSMC), and adventitial cells has been recognized. In this setting, the gp91^phox is replaced by an alternate homologue called Nox proteins, which plays an important role in vascular biology.

NADPH oxidase in VSMC and in the endothelium is activated by common pathophysiologic stimuli. An important stimulus is angiotensin, which acts on the AT-1 receptor, leading to the activation of phospholipase D and protein kinase C, which phosphorylates P47^phox and leads to translocation of P47^phox to the membrane. A parallel pathway involves activation of the tyrosine kinase C Src, transactivation of the endothelial growth factor (EGF) receptor, and ultimately activation of the small G-protein rac. ACE inhibitors, angiotensin receptor blockers, and statins have been shown to inhibit activation of NADPH oxidase to some degree, at least in cultured cells, if not in vivo.

Activated NADPH oxidase that is generating ROS can act on other enzyme systems, thereby causing those systems to generate ROS. The interplay between these various systems is very important, where activation of one can lead to activation of others. For example, ROS such as peroxinitrite can oxidize tetrahydrobiopterin in NO synthase (NOS), leading to uncoupling of NOS. ROS can promote the conversion of xanthene dehydrogenase to xanthine oxidase, which can then generate ROS. Further, free radicals can cause mitochondrial DNA damage and alter electron transport through the mitochondria, which can cause the mitochondria to generate ROS.

Work in Harrison’s laboratory in human endothelial cells showed that mitochondria treated with angiotensin for several hours produce very large amounts of hydrogen peroxide, whereas untreated mitochondria produce very little hydrogen peroxide. A catalase or pretreatment of mitochonrdria with apocynin, an inhibitor of NADPH oxidase, blocks the rate of production of hydrogen peroxide by the mitochondria. Further work showed that mitochondrial respiration is impaired in angiotensin-treated mitochondrial cells, and that treatment with an NADPH oxidase-inhibitor lessens this impairment. These data support the concept that angiotensin can stimulate NADPH oxidase to generate ROS, which can then damage mitochondrial function.

Role of NOS uncoupling in vascular disease

The uncoupling of NOS involves interruption of electron flow to arginine hydroxylation, and the electrons flow from NADPH to flavin binding domains. The transfer from flavins to heme is calcium/calmodulin-dependent, and in the absence of tetrahydrobiopterin superoxide is made.

Work by Landmesser in Harrison’s lab in DOCA-salt hypertensive, DOCA-eNOS and control mice showed that in the setting of hypertension the ratio of tetrahydrobiopterin to dihydrobiopterin is dramatically altered, with a ratio of 1:1, compared to 8:1 in the control mice. This oxidation is dependent on the formation of peroxinitrite, which requires both NO and superoxide. If this experiment was done in p47^phox knockout mice, less superoxide is generated and there is much less oxidation of tetrahydrobiopterin. In the eNOS knockout mice, the oxidation of tetrahydrobiopterin is also much less severe.

Oral therapy with tetrahydrobiopterin was also shown to affect blood pressure and vascular function. In the DOCA-salt mice, blood pressure continues to rise over time but tetrahydrobiopterin blunts this progressive increase, only the initial rise in blood pressure is seen—suggesting that recoupling of eNOS may prevent progression of hypertension. This has been confirmed in other animal models and in other kinds of diseases like insulin resistance and diabetes.

Oral treatment with tetrahydrobiopterin might be effective in lowering blood pressure and improving vascular function in humans. A dose-dependent reduction in blood pressure with 6-week treatment with tetrahydrobiopterin was found in persons with hypertension, in other work by Harrison’s group, with an average reduction from 143 mmHg systolic to 128 mmHg systolic with the 5 mg dose. Notably, when tetrahydrobiopterin was stopped at the end of the 6 weeks, there was a progressive increase blood pressure. Further, endothelium-dependent vasodilation, measured in the brachial artery, was dramatically improved during the 6-week treatment period but returned to the abnormal baseline function after treatment was stopped.

Exercise training has been shown to increase tetrahydrobiopterin in mice in other work by Harrison and colleagues. This effect may underlie promotion of vascular health with exercise training.

Role of glutathione

Glutathione is another important intracellular antioxidant in endothelial cells, and is the major intracellular small molecule antioxidant in mammalian cells. An important role of glutathione is scavenging hydrogen peroxide by acting as a substrate for glutathione peroxidase. This reaction involves conversion of glutathione to oxidized glutathione disulfide, and in that process hydrogen peroxide is converted to water.

Multidrug resistance protein 1 (MRP1) is the major transporter of glutathione disulfide in human aortic endothelial cells, as shown by studies by Harrison and colleagues. The normal substrate for MRP1 is glutathione disulfide, and thus this transporter very avidly transports oxidized glutathione from the cell. In an earlier study they showed that this glutathione loss from the cell was deleterious, with cells undergoing apoptosis. Preventing the export of oxidized glutathione improved endothelial cell viability.

Recent work by Harrison and colleagues showed in vivo that the export of glutathione is mediated by MRP1 at baseline and in response to angiotensin II. Hypertension increases the export of glutathione from the vessel. In MRP1 knockout mice, the export of glutathione is markedly low. Treatment with angiotensin II for 2 weeks did not increase this export.

MRP1 mediates the intracellular depletion of glutathione. In wild-type FVB mice, angiotensin caused a reduction in total glutathione and a reduction in the reduced form of glutathione. In the MRP1 knockout mice at baseline, the glutathione level is about 25% higher than that in the FVB mice, and there was no change in intracellular glutathione in response to angiotensin.

A hallmark of hypertension is an increase in vascular superoxide production, which was seen in the FVB mice, but not in MRP1 knockout mice. Notably, preservation of intracellular glutathione can prevent the increase in superoxide caused by angiotensin II. 

In the FVB mice, angiotensin II treatment caused a loss of tetrahydrobiopterin and changed the ratio of reduced oxidized tetrahydrobiopterin to 1:1, while these changes were not seen in MRP1 knockout mice. The ability to maintain intracellular tetrahydrobiopterin also promotes NO production in the vessels. In the FVB mice NO production was reduced in response to angiotensin II treatment, while in the MRP1 knockout mice NO production is unchanged from baseline levels. NO synthase levels are not changed in any group, which is probably related to preservation of intracellular tetrahydrobiopterin.

Dose-dependent vasorelaxation in response to acetylcholine was seen, while this endothelium-dependent relaxation was impaired in the angiotensin-treated animals. However, in the MRP1 knockout mice the normal relaxation seen at baseline was unchanged in response to angiotensin II.

Blood pressure levels were the same at baseline in the FVB and MRP1 knockout mice. However, in the wild-type mice, the degree of hypertension (mean arterial, systolic, diastolic) was much more severe in response to angiotensin compared to the MRP1 knockout mice.

MRP1 as a therapeutic target might be a new approach to the treatment of hypertension. MRP1 is the major transporter of oxidized glutathione in humans and in murine endothelial cells and vascular cells. Its inhibition results in maintenance of intracellular glutathione, reduced superoxide production, preservation of intracellular tetrahydrobiopterin, improvement of endothelial function, and lowering blood pressure.

ROS and antioxidant trials

Why have the antioxidant trials failed if reactive oxygen species are so important in pathophysiology? Multiple clinical trials, using a variety of antioxidants, have shown essentially no benefit. The Hope-Too study actually showed an increase in the incidence of heart failure with vitamin E. A meta-analysis showed a trend toward high-dose vitamin E being harmful. Based on these data, vitamin E is not indicated and should not be given to persons with established coronary artery disease. 

The entire concept of oxidative stress needs to be addressed to determine why these antioxidant trials have failed. Research by Griendling and colleagues at Emory indicates that at cytoplasmic sites there is production of superoxide, hydrogen peroxide, and other reactive oxygen species at near millimollar concentrations. Quite likely because of the location at which the ROS are being produced they play an important role, for example, modulation of signal transduction, formation and maintenance of the actin cytoskeleton, and maintenance of focal adhesion contacts. ROS in the nucleus are very likely important in regulating gene transcription and other events relating to the genome. Thus, the concept that an antioxidant vitamin could scavenge only the bad radicals without scavenging good radicals is difficult to understand. Further, it is very difficult to target antioxidant vitamins to specific domains and scavenge radicals that are having an untoward effect, and not scavenge radicals that are having beneficial effects.

Harrison proposed a revised concept of oxidative stress. In a metabolic pathway, radical X donates an unpaired electron to Y which also has an unpaired electron and thereby forms Y with an unpaired electron and also X. Then the Y radical with the unpaired electron donates it to Z and forms a Z radical and a byproduct of Y.  The Z radical is a terminal product.

However, if an abnormal molecule is inserted into this metabolic pathway such that the electrons cannot be transferred to form the Z radical, it forms an abnormal product (Q). Harrison believes this is a common abnormality in cell biology. For example, in uncoupling of NOS, the transfer of electrons is no longer to arginine but to oxygen to form superoxide. When the mitochrondria are damaged, the electron transport through the mitochondria becomes abnormal such that that they are no longer transferred to ATP and water but form superoxide. Abnormal signaling can occur between cells so that phosphotases are not activated appropriately, among other abnormal signaling effects. Harrison noted that it may be that vitamin E and other antioxidants might serve the role of Q and divert electrons away from its normal products to form an abnormal product.

In conclusion, not all ROS are deleterious and they have some important signaling properties, Harrison stated. There are localized sites of generation of ROS and it is difficult to target “antioxidants” to the proper site in the cell. “Reconnecting” disturbed electron transfer rather than simply scavenging radicals might be a superior approach to therapy. Blocking enzymes that are pathologically activated could provide a useful therapeutic target.