I. INTRODUCTION

Top Previous Next References

It has long been recognized that cyclooxygenase (COX) and lipoxygenase enzymes metabolize arachidonic acid (AA) to 5-, 12- and 15-hydroxyeicosatetraenoic acid (HETE), prostaglandins, prostacyclin, thromboxane, and leukotrienes (Fig. 1). These products modulate renal and pulmonary function, vascular tone, and inflammatory responses (155, 293, 303, 393). However, 20 years ago a third pathway for the metabolism of AA emerged. At this time, Capdevila et al. (51), Morrison and Pascoe (317), and Oliw et al. (340) reported that AA is also metabolized via cytochrome P-450 (CYP) enzymes in the liver and kidney to epoxyeicosatrienoic acids (EETs) and dihydroxyeicosatetraenoic acids (DiHETEs). Subsequent work revealed that CYP enzymes in other extrahepatic tissues also catalyze the formation of 19- and 20-hydroxyeicosatetraenoic acids (19- and 20-HETE) and 7-, 10-, 12-, 13-, 15-, 16-, 17-, and 18-HETEs from AA (Fig. 1). These observations were largely ignored for many years due to the focus on the COX and lipoxygenase pathways, the limited availability of CYP metabolites of AA, and the lack of selective inhibitors of CYP enzymes. However, a decade ago, three key observations rekindled interest in the role of CYP metabolites of AA in the control of renal function and vascular tone. First, Iwai and Ingami (210) identified CYP4A2 as a gene that is regulated by salt intake and is overexpressed in the kidney of spontaneously hypertensive rats (SHR). Second, Sacerdoti et al. (406) and Omata and co-workers (342, 343) reported that the production of 20-HETE was elevated in the kidneys of the SHR. Finally, several investigators demonstrated that agents that induce heme oxygenase and reduce the renal formation of 20-HETE prevent the development hypertension in the SHR (104, 268, 409). These observations led to the hypothesis that an elevation in the renal production of 20-HETE may play a role in the development of hypertension (303). This proposal compelled a number of investigators to examine the role of CYP metabolites of AA in the regulation of renal function, vascular tone, and the long-term control of arterial pressure.

View larger version (15K): [in this window] [in a new window]   Fig. 1. Pathways for the metabolism of arachidonic acid. Arachidonic acid is metabolized via the cyclooxygenase, lipoxygenase, and cytochrome P-450 (CYP) enzymes to prostaglandins (PG), prostacyclin (PGI2), thromboxane A2 (TxA2), or a series of hydroxyeicosatetraenoic acids (HETEs), epoxyeicosatrienoic acids (EETs), and dihydroxyeicosatrienoic acids (DiHETEs).

Remarkable progress has been made in the last 5 years. The work has been facilitated by the discovery of selective inhibitors of the formation of EETs and 20-HETE (9, 11, 39, 243, 300, 332, 427, 429, 475, 529) and by Dr. Camille Falck (University of Texas, Dallas, TX) who synthesized CYP metabolites of AA and freely distributed these compounds to investigators around the world. This area of research has entered an exponential growth phase similar to that seen in the nitric oxide (NO) field around 1990. More than 300 papers on biologic actions of CYP metabolites of AA have been published in the last 5 years. It now is now apparent that AA is primarily metabolized by CYP enzymes to EETs, DiHETEs, and 20-HETE in small blood vessels (155, 156, 203) and in the kidney (196, 205, 207, 342, 343), liver (512), lung (34, 211, 301, 513, 520), intestines (510), heart (68, 416, 502, 503), pancreas (416, 509), and white blood cells (29, 172, 402). These metabolites play an important role as paracrine factors and second messengers in the regulation of pulmonary, cardiac, renal, and vascular function (48, 212, 303, 378, 381, 393, 513) and modulate inflammatory (335) and growth responses (69, 70, 162, 272, 320, 321, 464). The focus of this review is to summarize the pathways for the formation of EETs, DiHETEs, and 20-HETE and the evidence that these products contribute to the regulation of renal and cardiovascular function.

	II. METABOLISM OF ARACHIDONIC ACID BY CYTOCHROME P-450 ENZYMES

Top Previous Next References

CYP genes encode for membrane-bound, heme-containing enzymes that catalyze NADPH-dependent oxidation of drugs, chemicals, and carcinogens (331, 416). CYP enzymes require NADPH reductase and CYPb₅ (NADH/NADPH oxidase) as cofactors (53). Because CYPb₅ oxidase is a major source of superoxide ions, there is considerable interest in determining whether the metabolism of AA by CYP enzymes contributes to oxidative stress. Indeed, there is evidence that the metabolism of various substrates by CYP enzymes leads to the formation of O radicals in the liver (37, 373). More recently, inhibitors of CYP2C9 were found to block the formation of O radicals by cultured porcine coronary artery endothelial cells (123). These studies suggest the metabolism of AA by CYP enzymes is a significant source of oxygen radicals in the vasculature.

More than 500 CYP genes have been identified and categorized by sequence homology (331). These genes are subdivided into 78 families. Fourteen of these families (29 subfamilies) are expressed in mammalian tissue. Up to 50 different CYP enzymes are expressed in a given species. Most of these enzymes are expressed in the liver and are involved in the metabolism of drugs and foreign chemicals. As such, CYP enzymes have been largely the province of biochemists and toxicologists. However, there has been a growing appreciation that many CYP enzymes are expressed in extrahepatic tissues and that they metabolize endogenous substrates like vitamins, steroids, and fatty acids, including AA. This review focuses on the formation of and cardiovascular actions of CYP metabolites of AA because most of the published work has focused on these metabolites. It is important to recognize that CYP enzymes metabolize other fatty acids (linolenic, linoleic, palmitic acid) as well or better than AA. The biologic actions of CYP metabolites of other fatty acids are unknown. However, it is likely that they are just as important as CYP metabolites of AA in the control of renal and cardiovascular function and will become a major focus of this field in the future.

Enzymes of the CYP4A, 4B, and 4F families catalyze the -hydroxylation of fatty acids. Several isoforms in these families produce 20-HETE when incubated with AA (144, 160, 240, 241, 370, 441, 431). These closely related gene families arose from ancient gene duplication, since the CYP4A and 4B gene families map to a conserved gene cluster on a single 450-kb YAC clone on chromosomes 4 in the mouse (169) and the homologous regions on chromosome 5 in the rat and chromosome 1 in humans. The 4F family appears to be a more ancient gene family that maps to a different region on chromosome 19 in humans (31).

cDNAs encoding for 15 different isoforms of the CYP4A family have been identified in various species (331). A summary of the isoforms responsible for the -hydroxylation of fatty acids in various species is presented in Table 1. CYP4A11, 4F2 (239, 264, 370), 4F11 (84), and 4F12 (44) are expressed in human liver and kidney. The substrate specificity of the CYP4F11 enzyme is unknown. CYP4F12 catalyzes the conversion of AA to 18-HETE (44). CYP4F2 appears to be the enzyme primarily responsible for the formation of 20-HETE in human kidney (370). This conclusion is based on the finding that a CYP4F2 antibody lowered the formation 20-HETE by microsomes prepared from human kidney by 67%, whereas a CYP4A11 antibody only reduced the formation of 20-HETE by 32% (370). Another member of the CYP4F family, CYP4F3, is highly expressed in human polymorphonuclear leukocytes (PMNs) (237). It catalyzes the -hydroxylation of leukotrienes. However, it remains to be determined if CYP4F3 is the source of 20-HETE produced by human PMNs (29).

View this table: [in this window] [in a new window]   Table 1. -Hydroxylases

Four isoforms of the CYP4A gene family, CYP 4A1, 4A2, 4A3, and 4A8, are expressed in the liver, kidney, and brain of rats (144, 160, 168, 240, 241, 431, 441, 452) (Table 1). Enzymes of the CYP4F family (CYP4F1, 4F4, 4F5, 4F6) have also been cloned from rat tissues (226, 227, 238). CYP4F4-6 enzymes are expressed in the kidney, but the ability of these isoforms to metabolize AA has not been investigated. Recombinant CYP 4A1, 4A2, and 4A3 enzymes, but not 4A8, have been reported to produce 20-HETE when incubated with AA (Table 1) (332, 443, 444, 477). In contrast, there is one recent report indicating that CYP4A8 also produces 20-HETE when incubated with AA (176). The catalytic activity of the CYP4A1 enzyme for AA is 10 times greater than that of CYP4A2 or 4A3 (332). However, the results of RT-PCR and Western blot experiments suggest that CYP4A2 is the isoform that is constitutively expressed in the liver, kidney, and vasculature of rats (205, 452). Message and protein levels for CYP4A1 are relatively low in these tissues unless the expression of this isoform has been induced by fasting (376), diabetes, or fibrates (140, 205, 207). Messages for the CYP4A2 and 4A3 isoforms are also expressed in small arterioles in the kidney (203, 207), brain (140, 156), and skeletal muscle beds (255) of rats. In contrast, CYP4A1 message levels are very low in vessels isolated from the kidney (203, 205), even though it is easy to detect this isoform by RT-PCR in vessels microdissected from the cremaster muscle of the rat (159) and cerebral arteries of rats (140) and cats (141). Thus the isoforms that underlie the formation of 20-HETE in the peripheral vasculature and kidney remain in question. This issue is further clouded since there are no antibodies that can distinguish between CYP4A isoforms. There are also posttranslational modifications of the CYP4A proteins that make definitive identification of the isoforms on Western blots difficult. For example, CYP4A proteins can be glycosylated (256), and there is considerable evidence that the activity of CYP enzymes is regulated by phosphorylation via cAMP-dependent protein kinases (337). Finally, the high degree of homology between the CYP4A isoforms limits the usefulness of cDNA probes in Northern blots for measuring changes in the expression of these isoforms. The only technique that can specifically distinguish between the different CYP4A isoforms is RT-PCR (205).

Unlike CYP4A1 that only catalyzes the -hydroxylation of fatty acids, CYP4A2 and 4A3 can also produce EETs when incubated with AA (332, 477). However, this conclusion remains controversial in that Helvig et al. (168) and Hoch et al. (176) failed to confirm that recombinant CYP4A2 and 4A3 enzymes produce EETs. This discrepancy may be due to polymorphisms of the CYP4A isoforms between strains of rats, since the sequence of the CYP4A cDNAs that have been cloned and expressed by these groups are slightly different.

In rabbits, CYP4A4, 4A5, 4A6, and 4A7 all catalyze the -hydroxylation of fatty acids (Table 1) (360, 391). CYP4A6 and 4A7 also metabolize AA to 20-HETE (391), but the rate of reaction is half that seen when the shorter chain unsaturated fatty acid lauric acid is utilized as the substrate. The ability of CYP4A7 to metabolize AA is greatly enhanced by addition of cytochrome b₅ (284). In the presences of cytochrome b₅, the catalytic activity of CYP4A7 for AA exceeds that for other fatty acids (284). CYP4A5 exhibits little activity for AA, but it readily metabolizes lauric and palmitic acids (185, 391). CYP4A4 exhibits little activity toward lauric acid. It favors AA as a substrate (391). However, the rate of this reaction is only 20% of that seen when prostaglandin E₁ is used as a substrate for the reaction (391).

The expression of CYP4A6 is induced in the liver and kidney of rabbits by fibrates in a manner similar to the effect of fibrates on the CYP4A1 isoform in rats. The expression of CYP4A7 is induced to a lesser extent by fibrates. In this way, CYP4A7 resembles the CYP4A2 and 4A3 isoforms in rats (391). The expression of the CYP4A4 isoform is constitutively expressed at low levels in the liver and lung of control animals, but it is markedly induced by 20-fold during pregnancy (318, 360). It is also induced by progesterone and dexamethasone (298). The functional significance of the remarkable induction in the expression of CYP4A protein and enzyme activity during pregnancy remains to be determined.

CYP4A10 and CYP4A12 are the enzymes primarily responsible for the formation of 20-HETE in the liver and kidney of mice (Table 1) (31). CYP4A14 is also expressed in the kidney (184). However, this isoform favors the metabolism of short-chained, unsaturated fatty acids and exhibits little activity toward AA (184). Kikuta et al. (236) and Cui et al. (83) recently cloned members of the CYP4F family (CYP4F14, 4F15 and 4F16) from the liver of mice that catalyze the -hydroxylation of leukotrienes, prostaglandins, and lipoxins. Two other members of this family, CYP4F15 and 4F16, are expressed in murine liver and kidney and are induced by NO and fibrates (83). CYP4B1 is also another potential fatty acid -hydroxylase that is expressed in the kidney of mice (192). However, the importance of enzymes of the CYP4F and 4B families to the generation of 20-HETE in the liver and kidney has not been determined. CYP2J9 has recently been cloned from the brain of mice (375). This enzyme catalyzes the formation of 19-HETE from AA, and it is highly expressed in Purkinje cells. Because 19-HETE inhibits Ca²⁺ channel activity in Purkinje cells, a role for this enzyme in neurotransmitter release has been postulated (375).

The expression of CYP4A10 and CYP4A14 is induced by fibrates in murine liver and kidney similar to the CYP4A1 and 4A3 in rats (31, 184). CYP4A12 is constitutively expressed in the liver and the kidney of mice. The expression of CYP4A12 is weakly induced by fibrates (31). It is also differentially expressed in the liver and kidney of male versus female mice (169). In this way, its expression resembles that of CYP4A2 in rats (205, 452).

The formation of EETs from AA is very promiscuous. CYP enzymes of the P-450 1A, 2B, 2C, 2D, 2E, 2J, and 4A families (Table 2) have all been reported to catalyze the formation of EETs in various tissues and species (228, 243, 259, 288, 293, 387). CYP2C8, 9, and 19 as well as CYP1A2, 2J2, and CYP2B6 are all expressed in human liver and produce EETs when incubated with AA (46, 86, 387). The CYP2C8 isoform produces 14,15- and 11,12-EETs in a ratio of 1.25:1 (86). CYP2C9 produces 14,15-, 11,12-, and 8,9-EETs in a ratio of 2:1:1, whereas the CYP2J2 isoform produces 5,6-, 8,9-, 11,12-, and 14,15-EETs with equal efficiency (416, 503). Studies using the CYP2C9 inhibitor sulfaphenazole and the CYP1A/3A inhibitor -naphthoflavone suggest that CYP2C9 is responsible for 50% of the epoxygenase activity in human liver and that enzymes of the CYP1A and 3A families contribute 25% to this activity (387). On the other hand, this view has been challenged by the results of a recent study suggesting that CYP2C8 is the primary epoxygenase expressed in human liver (512). In this experiment, immunoprecipitation of human liver microsomes with a CYP2C8 antibody reduced the formation of EETs by 85% (512).

CYP2J2 has emerged as one of the major enzymes involved in the formation of EETs in extrahepatic tissues in humans (503, 510, 511, 513, 515). Wu et al. (503) first reported that this isoform is highly expressed in cardiac myocytes. CYP2J2 is also expressed in the proximal tubule and medullary collecting duct in the kidney (288); -, -, -, and polypeptide-containing cells in the pancreas (509); in the lung (511); in the small intestine, and in the colon (503). This enzyme serves as the primary enzyme that catalyzes the formation of EETs in the heart. EETs diminish cardiac contractility by inhibiting the opening of sodium channels (265) and L-type calcium channels (68). CYP2J2 is also the enzyme primarily responsible for the formation of EETs in the pancreas of humans (509). In this tissue, EETs influence the secretion of insulin and glucagon (106). In the kidney, EETs regulate renal hemodynamics and inhibit sodium transport (213, 291, 381, 399, 412); however, the relative importance of CYP2J2 versus enzymes of the 2C and 4A families in generating EETs in human kidney has yet to be determined. Finally, EETs decrease bronchiolar tone and vascular resistance in human lung (212) (see sect. VII); however, little is known to what extent CYP2J enzymes versus other isoforms contribute to the production of EETs in the lung (383, 519, 521).

Human endothelial cells express enzymes of the CYP2C8, 2C9 (120, 273), 2J2 (335), 3A (178), and 2B1 (119, 178) families that all produce EETs. EETs dilate human coronary arteries (309, 310). They act by hyperpolarizing VSM by opening K⁺ channels and are a prime candidate for endothelial-dependent hyperpolarizing factor (EDHF) (48, 315). The recent findings that the vasodilator responses to acetylcholine and bradykinin are potentiated by an inducer of CYP2C8/9 protein (-naphthoflavone) (369) and are inhibited by sulfaphenazole (295) and antisense CYP2C8/9 oligonucleotides (120) suggest that CYP2C8 and 2C9 are likely the major isoforms responsible for the formation of EETs in human endothelial cells.

Enzymes of the CYP2C11 (159, 180, 451), 2C13 (507), 2C23 (180, 194, 223, 294), 2C24 (180), 2J3 (503), 2J4 (513), 2J10 (507), 2B1 and 2B2 (259), 2D18 (458), 2E1 (258), and 1A1 and 1A2 (387) families all produce EETs in rats (Table 2). CYP2C11 and 2J3 probably serve as the primary epoxygenases in the liver of male rats (perhaps CYP2C12 in females). However, the expression of CYP1A, 2B, and 2E families in the liver can be induced by hydrocarbons, barbiturates, and ethanol, respectively. As a result, these isoforms can become the major source of EETs in the liver and other tissues after exposure of rats to drugs and xenobiotics.

CYP2C11, 2C12, 2C23, 2C24, 2J3, 2J4, and 2J10 are expressed in the kidney of rats (Table 2). Holla et al. (180) concluded that CYP2C23 is the major epoxygenase enzyme expressed in the kidney of rats. More recently, a CYP2J enzyme has been reported to be highly expressed in the proximal tubule and collecting duct of rats (288). It is likely to also contribute to the formation of EETs in the kidney. The recent report that the production of EETs is elevated in the kidney of SHR and that this is associated with an increased expression of a protein that cross-reacts with a CYP2J antibody supports this view (507).

CYP2C11 protein is highly expressed in astrocytes in the brain of rats (7). EETs are formed by this cell type (7, 18), and they have been implicated in mediating reactive hyperemia in the brain (155) and the cerebral vasodilator response to glutamate (6). Recently, Thompson et al. (458) reported that CYP2D18 is also expressed in the brain of rats. The CYP2C18 enzyme that normally converts dopamine to aminochrome also catalyzes the formation of EETs from AA. The functional significance of this observation to control of neural function, reactive hyperemia, and the destruction and death of dopaminergic neurons in Parkinson disease are likely to become important new areas of investigation.

CYP2J3 is expressed in the heart, pancreas, liver, lung, and kidney of the rat. It appears to be the functional homolog of CYP2J2 enzyme expressed in humans. However, CYP2J3 does exhibit some differences with regard to the stereospecificity of the EETs produced. CYP2J3 favors the production of 14(S),15(R)- and 11(R),12(S)-EETs (502), whereas the 2J2 isoform produces 14(R), 15(S)- and a racemic mixture of 11,12-EETs (503). CYP2J4 is also expressed in the small intestine and liver of rats, but the importance of this isoform to the formation of EETs in these organs remains to be established (515).

Despite the fact that much of the work on endothelial-derived vasodilators and EDHF has been performed in rats, little is known about the CYP isoforms that produce EETs in rat endothelial cells. Message for CYP2C11 can be detected in cerebral, renal, splachnic, and skeletal muscle arterioles in rats (personal observation). CYP2J3 protein is also expressed in rat vascular endothelial cells (335). However, no one has determined the relative contribution of the various CYP isoforms to the production of EETs, nor is it known if the expression of CYP isoforms in the endothelium is altered by hypertension, diabetes, or hyperlipidemia. This information is critical since CYP isoforms exhibit different regio- and stereospecificity with regard to the type of EETs formed and the vascular response to EETs in some beds is both regio- and stereospecific (143, 200, 526).

CYP2C29, 2C38, 2C39, and 2C40 enzymes are expressed in the liver, kidney, and brain of mice (Table 2) (287). CYP2C38 is the most active epoxygenase. It preferentially metabolizes AA to 11,12-EET. CYP2C29, 2C39, and 2C40 are less active and produce 14,15-EET. CYP2B19 is expressed in the skin of mice. This enzyme produces EETs and 11-, 12-, and 15-HETE (228). CYP2J5 is highly expressed in the proximal tubule and collecting ducts in the kidney of mice (288). It produces 14,15-, 11,12-, and 8,9-EETs in a ratio of 2:1:1 when incubated with AA. More recently, cDNAs for CYP2J6, 7, 8, and 9 have also been cloned from the mouse. However, little is known about the expression or function of these isoforms (416).

CYP2C1, 2C4, and 2CAA are epoxygenases expressed in the liver of rabbits (Table 2) (85, 260). CYP2B, 2E, 2J, 1A, 2C, and 4A proteins are expressed in the pulmonary artery and peripheral lung tissue of rabbits (519). All of these isoforms produce EETs when incubated with AA. However, the observations that CYP2B4 inhibitors and antibodies completely inhibit the formation of EETs in rabbit lung (244, 513) suggest that CYP2B4 is probably the primary epoxygenase expressed in this tissue.

AA is also metabolized to 7-, 8-, 9-, 10-, 11-, 12-, 13-, 15-, 16-, 17-, and 18-HETE by CYP enzymes (Table 3). Brash et al. (40) reported that 7-, 10-, and 13-HETE are major CYP metabolites of AA produced in liver microsomes. Carroll and co-workers (57, 59) reported that 16-, 17-, and 18-HETE are produced in the kidney. They are incorporated into tissue phospholipid pools in the kidney, and high concentrations (ng/ml) of these compounds are excreted in the urine. Moreover, the release of 16-, 17-, and 18-HETEs from isolated perfused rabbit kidneys is stimulated by ANG II and other vasoconstrictors that stimulate phospholipases (59). Finally, Bednar and co-workers (29, 30) identified 16-HETE as one of the major metabolites of AA produced by human PMNs. The other major metabolite formed is 20-HETE (29, 172). The potential biologic importance of the myriad of HETEs formed by CYP isoforms has been highlighted by the recent discovery that structural analogs of 20-HETE, like 5-, 15-, and 19-HETE (likely 16-, 17-, and 18-HETE as well), serve as competitive antagonists of the vasoconstrictor actions of 20-HETE (9). Because 15-, 16-, and 19-HETE are produced in vivo, they could serve as endogenous inhibitors of the actions of 20-HETE.

A summary of the isoforms responsible for the formation of the bisallylic and subterminal hydroxyl metabolites of AA is presented in Table 3. CYP1A2 and 3A4 enzymes catalyze the formation of 7-, 10-, and 13-HETE from AA in human liver (40, 46). In addition to catalyzing the formation of EETs, CYP2C8 also catalyzes the formation of 11-, 13-, and 15-HETE (46), whereas CYP2C9 catalyzes the formation of 12- and 13-HETE (46). CYP4F8 (45) and CYP4F12 (44) catalyze the formation of 18-HETE in humans. CYP4F12 is expressed in the liver, kidney, small intestine, and heart (44). Finally, CYP2C37 and CYP2C40 have recently been cloned from the liver of the mouse (287). CYP2C37 produces 12-HETE (287), whereas CYP2C40 catalyzes the formation of 16-HETE (462).

A summary of the metabolic fates of EETs, DiHETEs, and 19- and 20-HETEs in various tissues is presented in Figure 2. EETs are converted by soluble or microsomal epoxide hydroxylases (sEH or mEH) to the corresponding DiHETEs. In renal arterioles, EETs are potent vasodilators, whereas the corresponding DiHETEs are much less active (200). Thus the conversion of EETs to DiHETEs is thought to limit the actions of EETs. The exception to this may be in the coronary microcirculation, where DiHETEs and EETs are equipotent as dilators of porcine coronary arteries (109, 341). There are strain differences in the expression of epoxide hydrolase enzymes that affect the ratio of EETs to DiHETEs produced in various tissues. For example, microsomes prepared from the kidneys of SHR produce DiHETEs, whereas those prepared from the kidneys of Wistar-Kyoto (WKY) rats produce EETs (142, 196). More recently, Yu et al. (507) have reported that the expression of sEH is elevated in the kidneys of SHR. They also found that administration of an inhibitor of sEH increases the renal formation of EETs and lowers blood pressure in SHR. Further evidence for a role for sEH and EETs in the control of blood pressure has been provided by Sinal et al. (432). They found that knockout of the sEH gene increased the renal formation of EETs and lowers blood pressure by 13 mmHg in male mice (432).

View larger version (15K): [in this window] [in a new window]   Fig. 2. Metabolic fate of cytochrome P-450 metabolites of arachidonic acid (AA). Once formed, epoxyeicosatrienoic acids (EETs) and 20-hydroxyeicosatetraenoic acid (HETE) are metabolized by -oxidation to 18- and 16-carbon derivatives, which are less biologically active. 5,6-EET, 8,9-EET, and 20-HETE can also be metabolized by cyclooxygenase (COX) to vasoconstrictor endoperoxides or to vasodilator prostaglandin or prostacyclin-like derivatives. EETs, DiHETEs, and 20-HETE are reincorporated into membrane phospholipid pools. They can be released again through the actions of agents that activate phospholipases. All of these fatty acids are also avidly bound by plasma proteins, which limits their biologic activity in the circulation and restricts their distribution to plasma when given via a blood-borne route of administration.

EETs, DiHETEs, and 19- and 20-HETE predominately act as autocrine factors. There are several arguments that support this view. First, these compounds are very lipophilic and are found in higher concentrations within tissue than in plasma, urine, or interstitial fluid. Second, EETs and 20-HETE avidly bind to proteins (492). The binding of these compounds to albumin likely restricts the distribution of these compounds to the vascular space when they are given systemically. In this regard, EETs and 20-HETE readily alter vascular tone and epithelial transport when studied in vitro. However, it is extremely difficult to demonstrate that they have any effect on renal function or vascular tone when given systemically because of protein binding. Finally, EETs, HETEs, and 20-HETE are rapidly incorporated into membrane phospholipids in many cell types (468, 469, 486). Karara and co-workers (221, 222) reported that the concentration of EETs incorporated in lipids extracted from human kidney is in the micromolar range. Similar values have also been reported in kidneys of rabbits (107) and rats (453). High concentrations of EETs have also been reported in phospholipids extracted from human platelets (523), VSM cells (108), vascular endothelial cells (468, 469), human and rat hearts (265, 502, 503), and the brain (100). Carroll et al. (57) demonstrated that 16-, 17-, 18-, 19-, and 20-HETE are incorporated in phospholipids extracted from the kidney of rabbits. EETs and 20-HETE are released from membrane phospholipids in response to vasoactive hormones and other stimuli that activate phospholipases. Weintrab et al. (486) reported that bradykinin releases EETs and DiHETEs from endothelial cells, and they contribute to the vasodilator response to bradykinin in porcine coronary arteries. Similarly, Carroll et al. (57) reported that ANG II increases the release of 16-, 17-, 18-, 19-, and 20-HETE into the urine and renal venous effluent of isolated perfused rabbit kidneys even after blockade of the synthesis of 20-HETE with N-methylsulfonyl-12,12-dibromododec-11-enamide (DDMS). Thus it may take considerable time to fully deplete tissue pools of EETs and 20-HETE after administration of CYP inhibitors. This concept has been largely ignored by most investigators studying the effects of CYP inhibitors on renal function and vascular tone and will likely have major impact on the interpretation of future studies.

A further demonstration that the release of EETs from membrane phospholipids contributes to the vasodilator response to bradykinin in canine coronary arteries has recently been provided by Mombouli et al. (314). They found that the epoxygenase inhibitor miconazole prevented the vasodilator response to AA in canine coronary arteries, indicating the dose of this agent was sufficient to completely block the synthesis of EETs from AA. Miconazole had no effect on the vasodilator response to bradykinin in this same study. The response to bradykinin was blocked by an inhibitor of phospholipase A₂. Moreover, the vasodilator response to repeated challenges with bradykinin gradually fell as the tissue pool of EETs became depleted in vessels treated with miconazole. Mombouli et al. (314) suggested that the failure of miconazole to acutely block the vasodilator response to bradykinin was due to the sustained release of EETs from membrane phospholipid pools. The results of this study are extremely important because they illustrate that finding that a CYP inhibitor has no effect on a vascular response is not sufficient to conclude that CYP metabolites of AA play no role in mediating that response.

EETs and HETEs are rapidly metabolized by -oxidation to 18- and 16-carbon products (108, 110, 112) that are generally less active than the parent compounds. They are also excellent substrates for chain elongation (110). In many tissues, EETs and HETEs can be further metabolized by CYP, lipoxygenase, and cyclooxygenase enzymes. For example, we reported that 11,12- and 14,15-EETs are metabolized by CYP4A enzymes in the kidney of Dahl rats to 20-hydroxy-11,12- and 14,15-EET (290). Others have described the formation of trihydroxy metabolites of AA that can arise from the metabolism of EETs by lipoxygenase or cyclooxygenase enzymes (317, 340).

The metabolism of EETs and 20-HETE by COX also plays a role in modifying the biologic actions of these metabolites. For example, 5,6-EETs can be metabolized by COX to both vasodilator and vasoconstrictor compounds. 5,6-EET constricts the renal circulation of rats both in vivo (453) and in vitro (60, 200). This effect is blocked by COX inhibitors and thromboxane receptor antagonists and is mediated by the formation of an endoperoxide intermediate. Similarly, Schwartzman et al. (421) and Escalante and co-workers (103, 105) have reported that the vasoconstrictor response to 20-HETE in peripheral arteries of rats and rabbits is blocked by indomethacin and endoperoxide receptor antagonists, suggesting it is related to the formation of endoperoxides.

The metabolism of CYP metabolites of AA by COX appears to be both tissue and species specific. For example, 5,6-EET dilates the isolated perfused rabbit kidney (58). This effect is COX dependent and thought to be related to the formation of 20-hydroxy-PGE₁ or PGI₂ by the endothelium. On the other hand, 20-HETE constricts in vitro perfused rabbit afferent arterioles preconstricted with phenylephrine (23). Only a small part of this response was blocked by indomethacin. Arima et al. (23) concluded that the vasoconstrictor response to 20-HETE in the rabbit afferent arteriole is due to a direct effect of 20-HETE on VSM and a minor component that is dependent on the metabolism of 20-HETE by COX to a thromboxane/endoperoxide metabolite.

It is also important to recognize that 5,6- and 8,9-EETs, 20-HETE, and subterminal HETEs (16-, 17-, 18-, and 19-HETE) are good substrates for COX. Thus the biologic responses to CYP metabolites of AA are a function of their direct effects on the target tissue as well as those of any COX metabolites formed that augment or oppose the actions of the parent compound. Moreover, experimental conditions alter the expression of COX in various tissues. This will change the types and amounts of COX metabolites of EETs and 20-HETE formed and modify the overall vascular response to these compounds.

In addition to being metabolized by COX and lipoxygenase enzymes, 20-HETE and EETs can affect the release of COX and/or lipoxygenase metabolites from tissues. For example, 20-HETE has been reported to dilate canine coronary arteries by stimulating the release of prostacyclin from the endothelium (371). Finally, it should be noted that most of the antibodies used to measure COX and lipoxygenase metabolites of AA cross-react with 20-hydroxy metabolites of the parent compounds (personal observations). Thus much of the previous work attributing the actions of COX inhibitors to blockade of the formation of prostaglandins, thromboxane, and prostacyclin needs to be reevaluated to determine if some of the actions of these agents are due to inhibition of the formation of COX metabolites of EETs and 20-HETE.

Many factors influence the expression of the CYP enzymes that metabolize AA. CYP4A1 and 4A3 mRNA and protein are highly expressed in the kidney of neonatal rats, but the levels decline in adulthood (254, 342). In contrast, CYP4A2 protein is not expressed in the kidney of neonates, but the levels increase with age until it becomes the dominant isoform expressed in the kidney of adult rats (342). Because 20-HETE is a potent mitogen (272, 321, 323), the upregulation of CYP4A1 expression (the most active isoform for the formation of 20-HETE) just after birth may contribute to the postnatal growth and development of the kidney.

In the proximal tubule, ANG II increases the formation of EETs (291, 381), whereas epidermal growth factor (272), dopamine (21, 336), and parathyroid hormone (PTH) (345, 386, 430) increase the formation of 20-HETE. In the thick ascending loop of Henle (TALH), a variety of peptide hormones increase the formation of 20-HETE (16, 145, 150, 286, 303, 481). The expression of CYP4A proteins in the kidney and vasculature falls in rats fed a high-salt diet (207, 210, 290, 439, 441), and this can be prevented if circulating ANG II levels are maintained by an intravenous infusion (11). In contrast, a high salt intake increases the expression of CYP2C23 and the formation of EETs in the kidney of rats (55, 294, 358). Because EETs inhibit sodium transport in the proximal tubule and collecting duct (291, 399, 412), upregulation of renal epoxygenase activity has been postulated to play an important role in the chronic adaptation of animals to a high-salt diet (293, 294) and in the development of salt-sensitive hypertension in humans (261) and rats (294). However, we have recently noted that there is considerable heterogeneity among inbred strains of rats regarding the effects of changes in salt intake on the expression of CYP enzymes in the kidney and the formation of EETs and 20-HETE (175). Thus it remains to be determined whether this reflects genetic mutations in CYP genes and whether these mutations are linked to differences in the salt sensitivity of blood pressure in different strains of rats.

The expression of CYP4A protein is increased by glucocorticoids, mineralocorticoids, and progesterone in the kidney and the lung (263, 298, 431). Antilipidemic agents, such as clofibrate, induce the expression of CYP4A1 and 4A3 and the synthesis of 20-HETE in the liver and kidney (20, 144, 160, 398, 431) by interacting with the peroxisome proliferator-activated receptor (PPAR)- receptor (149, 255). There is also evidence that chronic exposure to NO increases the expression of CYP4A mRNA and protein in the liver and kidney (208, 425, 426), even though NO itself inhibits the formation of EETs and 20-HETE (13, 8, 358, 446). In contrast, NO downregulates the formation of CYP2C isoforms that produce EETs (283, 425, 426). Other inducers of CYP enzymes, such as phenobarbital, 3-methylcholantherene, and 3,4-benzo(a)pyrene, do not alter renal CYP4A activity (431). Nearly all vasoconstrictor agents, i.e., ANG II, norepinephrine, vasopressin, and endothelin, activate phospholipases and increase the synthesis and release of 20-HETE in the kidney and VSM cells (57, 59, 82, 303, 350, 354, 470). The expression of CYP enzymes involved in the metabolism of AA is altered in diabetes (25, 193, 255), pregnancy (64, 144, 301, 298, 318), hepatorenal syndrome (407), cyclosporin- induced (43, 408, 506) and cisplatin-induced (325) nephrotoxicity, alcohol-induced liver disease (327), chronic renal disease (267), and various models of hypertension (104, 268, 290, 294, 343, 353, 381, 394, 398, 439, 441, 494). It is also altered by changes in dietary sodium (180, 207, 293, 294, 358) or potassium (62, 480, 514) intake and by fasting (255, 376). However, the role of 20-HETE in mediating the changes in renal function and/or vascular tone associated with these conditions has yet to be established. The patterns of expression of CYP isoforms of the 2C and 4A families differ in male versus female rats (205, 290, 451, 452); however, the functional significance of these observations remains to be determined.

F.  CYP Inhibitors and Inducers

1.  Nonselective inhibitors

The development of selective inhibitors and inducers of the formation of 20-HETE and EETs has been the catalyst driving the exponential growth of information regarding the roles of these substances in the control of cardiovascular function. A summary of the agents that have been used to manipulate the metabolism of AA is presented in Figure 3. The release of AA from membrane phospholipids can be blocked by the relatively nonspecific phospholipase inhibitors quinacrine or heparin or the more specific inhibitor of cytoplasm phospholipase A₂ arachidonyl-trifluoromethyl ketone (AACOF₃).

View larger version (24K): [in this window] [in a new window]   Fig. 3. Summary of agents available to induce or inhibit various pathways for the metabolism of arachidonic acid. The complete chemical names of the various drugs are presented in text. PLA2, phospholipase A2; EETs, epoxyeicosatrienoic acids; 20-HETE, 20-hydroxyeicosatetraenoic acid; DiHETEs, dihydroxyeicosatrienoic acids; NO, nitric oxide; CO, carbon monoxide.

Much of the information regarding CYP metabolites of AA in the control of renal and cardiovascular function has been deduced by comparing the effects of COX (indomethacin, meclofenamate) and lipoxygenase [baicalein, cinnamyl-3,4-dihydroxy--cyano-cinnamate (CDC)] inhibitors with those obtained after blockade of all three pathways for the metabolism of AA with 5,8,11,14-eicosatetraynoic acid (ETYA). Other approaches to reduce the formation of EETs and 20-HETE have been to use -diethyl-aminoethyldiphenylpropylacetate (SKF-525) that binds to heme in CYP enzymes (55) or heme arginate, SnCl₂, or CoCl₂ that induce heme oxygenase (104, 268, 350, 353, 354, 409) and reduce CYP activity by diminishing the availability of heme for incorporation into CYP enzymes. However, these agents are not very specific and reduce the activity of all CYP enzymes, including those involved in steroid metabolism. They also inhibit other heme-containing enzymes, like NO synthase. The interpretation of the results of studies using heme oxygenase inducers is further complicated since these agents stimulate the formation of carbon monoxide (CO), which inactivates both CYP and NOS enzymes and has direct effects on ion channels and vascular tone (479).

More specific suicide-substrate CYP inhibitors like 1-aminobenzotriazole (ABT) were developed (243, 300, 445) about 10 years ago. These agents have an advantage that they should only bind to and inactivate CYP isoforms that utilize these compounds as substrates. In this regard, ABT (50 mg/kg ip) has recently been reported to selectively inhibit enzymes of the CYP4A family and reduce the formation of 20-HETE in the kidney of rats (445). However, others have reported that ABT also blocks the activity of other CYP isoforms (CYP2B, 1A, 2C) that produce EETs in the liver and lung (243, 300). We have found that ABT (50 mg/kg ip) completely blocks the formation of EETs, DiHETEs, and 20-HETE in the kidney of rats within 2 h and reduces the 24-h excretion of 20-HETE by >50% (292). It also reduces the formation of 20-HETE and EETs in the liver and lung by ~50%. Even though ABT is nonselective in blocking the formation of EETs versus 20-HETE, it is still useful for evaluating the role of CYP metabolites of AA in mediating biologic responses in vivo, since it is inexpensive, readily available (Sigma), orally active, and well tolerated by rats for >2 wk (304). It also irreversibly inactivates CYP enzymes so that the organs can be collected and the tissue prepared, and one can assess the degree of inhibition of various CYP isoforms using in vitro assays.

Another useful agent is 17-octadecynoic acid (17-ODYA). This compound was synthesized by Dr. Ortiz de Montellano to be a specific suicide substrate inhibitor of the CYP4A enzymes that metabolize AA to 20-HETE (318, 348, 427). Subsequent studies demonstrated that 17-ODYA irreversibly blocks the formation of 20-HETE at concentrations >1 µM (529). Unfortunately, it is equally as effective in blocking the formation of EETs in various tissues (529).

The endogenous gaseous mediators NO and CO bind to heme and inactivate CYP enzymes. They completely block the formation of EETs and 20-HETE in renal and hepatic microsomes and renal and cerebral arteries at concentrations <100 nM (13, 446, 448). It is believed that they serve as modulators of the activity of CYP metabolism of AA in the kidney and peripheral vasculature in vivo. Indeed, many of the cGMP-independent actions of NO and CO on vascular tone and epithelial transport have been shown to be related to their ability to inhibit the endogenous formation of 20-HETE in various tissues (12, 13, 77, 276, 446, 448).

Dr. Falck synthesized 12,12-dibromododec-11-enamide (DBDD) and DDMS as a second generation of inhibitors to selectively block the formation of 20-HETE (8, 475). These compounds competitively inhibit the formation of 20-HETE by renal microsomes at concentrations <10 µM, whereas epoxygenase activity is only reduced by 10-20% (8). At higher concentrations (50 µM), however, they are equally effective at inhibiting the formation of EETs and 20-HETE. The main limitation in using these inhibitors for in vivo studies is that they are fatty acids that avidly bind to plasma proteins. This property limits their ability to diffuse into tissues and inhibit the formation of 20-HETE when given via a blood-borne route. Thus these compounds are only effective at inhibiting the formation of 20-HETE when added to protein-free solutions in vitro or when directly applied to tissues in vivo.

More recent studies have employed a new inhibitor of the formation of 20-HETE, N-hydroxy-N'-(4-butyl-2 methylphenyl)formamidine (HET-0016) (229, 311). This compound selectively inhibits the formation of 20-HETE at a concentration <10 nM. It has no effect on epoxygenase, COX, or lipoxygenase activity at concentrations up to 1 µM (311). At this concentration, it also has minimal effects on the activity of other CYP isoforms (2CYP2C9, 2D6, 3A4) involved in drug metabolism in humans (311). HET-0016 appears to be the most specific inhibitor of the synthesis of 20-HETE currently available. However, little is yet known about the biologic effects of HET-0016 in vivo or its potential side effects.

A number of analogs of 20-HETE, such as 5-, 15-, and 19-HETE, block the vasoconstrictor actions of 20-HETE in renal (9) and cerebral arteries (140). The most effective analog appears to be 20-hydroxyeicosa-6(Z),15(Z)-dienoic acid [6(Z),15(Z)-20-HEDE or WIT-002]. Unlike 5-, 15-, and 19-HETE, 6(Z),15(Z)-20-HEDE lacks double bonds at the 8,9- and 11,12-positions, so it is not metabolized by COX. It completely blocks the vasoconstrictor response to 20-HETE in renal (9), cerebral (140), and skeletal muscle (128, 132) arterioles at a concentration of 1 µM. It has become an important adjunct to enzyme inhibitors to establish a role for 20-HETE in the myogenic response of renal, cerebral, and skeletal muscle arterioles and the vasoconstrictor response to elevated PO₂ in cerebral and skeletal muscle arterioles (131, 132, 140). However, 6(Z),15(Z)-20-HEDE also binds to plasma protein. This property limits its usefulness for in vivo experiments. Like most fatty acids, it also is a competitive inhibitor of the synthesis of 20-HETE and EETs at high concentrations.

The most recent approach to block the formation of 20-HETE has been the use of antisense cDNA oligonucleotides directed against CYP4A1. Wang and co-workers (476, 478) demonstrated that daily intravenous injections of a 20-mer antisense CYP4A1 oligonucleotide for 5 days reduced the expression of CYP4A1 and 4A2 protein and the production of 20-HETE in renal arterioles of Sprague-Dawley rats. Blockade of this pathway was associated with a modest fall in arterial blood pressure in Sprague-Dawley rats (476) and SHR (478). The advantage of antisense oligonucleotides to block the expression of CYP proteins is that they can be very specific provided there is enough difference in the sequence of the targeted region between CYP isoforms. The recent successes using antisense oligonucleotides suggest that it may be possible to achieve long-term knockdown of the expression of specific CYP isoforms by transfection of animals with viral constructs that produce antisense cRNAs.

A number of compounds have been identified that selectively inhibit epoxygenase activity. These include the antimycotic agents miconazole, econazole, ketoconazole, and clotrimazole (52, 529). Miconazole (1 µM) selectively reduces the formation of EETs by renal and hepatic microsomes by 70% and has much less effect on the formation of 20-HETE (529). At higher concentrations (10 µM), it completely inhibits epoxygenase activity while reducing formation of 20-HETE by 50% (52, 529). However, antimycotic agents are not specific epoxygenase inhibitors. They bind to heme and inhibit enzymes involved in drug metabolism, steroidogenesis (390, 429), and the synthesis of NO (151, 186). They also alter intracellular calcium concentration and ion channel activity in various types of cells (88, 94, 388, 389).

More recently, Wang et al. (475) and Nguyen et al. (332) have reported that 6-(2-propargyloxyphenyl)-hexanamide (PPOH) and methylsuphonyl-6-(2-propargyloxyphenyl)-hexanamide (PPOMS) block CYP4A2- and 4A3-catalyzed epoxidation of AA. They also block epoxygenase activity in renal tissue at concentrations of 1-10 µM and have little effect on the formation of 20-HETE. At higher concentrations (20 µM), however, we found that these compounds competitively reduce the synthesis of 20-HETE. There is also evidence that these compounds can be effective as epoxygenase inhibitors in vivo. In this regard, Brand-Schieber et al. (39) reported that an acute intravenous injection of MS-PPOH (5 mg/kg) in rats selectively lowered renal epoxygenase activity for up to 6 h. This was associated with an increase in urine flow and sodium excretion. These results suggest that endogenously formed EETs may have antidiuretic and antinatriuretic properties in the kidney.

Others have begun to use compounds that are metabolized by CYP isoforms that produce EETs as competitive inhibitors. These compounds include sulfaphenazole (CYP2C9) (295, 411), metyrapone (2B) (134, 244), N--methylbenzyl-1-aminobenzotriazole 7-ethoxyresorufin (CYP1A), 7-pentoxyresorufin (CYP2B1), -naphthoflavone (CYP1A1, 1A2, 2C8, 2C9) (66, 411), diazepam (CYP2C19), diethyldithiocarbamate (CYP2E1, 1A1, 1A2, 2A6, 2B6, 2C8, 3A3, 3A4) (66, 134, 411), quinidine (CYP2D6) (411), orpenadrine (CYP2B6) (411), triacetyloleandomycin (CYP3A3, 3A44, 3A5) (66), toleandomycin (134), methohexitone, or thiopentone (274). In theory, this approach of providing a competing substrate to a CYP isoform should reduce its ability to produce EETs depending on the relative affinity of the isoform for AA versus the inhibitor. However, the specificity of these agents has yet to be established, nor has anyone documented that these competitive substrate inhibitors actually reduce the concentration of EETs in the targeted tissues.

The conversion of EETs to DiHETEs can be inhibited using trichloropropylene oxide (TCPO) (73) or 4-phenylchalcone oxide (4-PCO) (341) that block mEH or N,N-dicyclohexylurea (DCU) (110, 508) that blocks sEH. Triacsin C (486) blocks the reincorporation of EETs, DiHETEs, and HETEs into membrane phospholipid pools by acyl CoA synthase. This augments the concentration of EETs released from tissues and potentiates the vascular responses to EETs and DiHetes (108, 486).

EETs receptor antagonists have yet to be identified. However, 11(R),12(S)-EET dilates rat renal arterioles and activates K⁺ channels in renal VSM cells, whereas, the S,R-isomer is inactive (200, 526). Given the recent findings that inactive 20-HETE analogs serve as competitive antagonists of the vasoconstrictor actions of 20-HETE, one might predict that S,R-isomers of EETs may block the vasodilator responses to the active R,S-isomers.

Recently, Fisslthaler and co-workers (119, 120) and others (36, 123) demonstrated the utility of using molecular approaches to obtain isoform-specific inhibition of the formation of EETs. They found that incubating porcine coronary arteries with antisense oligionucleotides to human CYP2C8/9 reduced the expression of an immunoreactive CYP2C8/9-like protein and the vasodilator response to AA and bradykinin. They concluded that a homolog of human CYP2C8/9 is likely responsible for the formation of EETs that serves as EDHF in these arteries.

	III. 20-HYDROXYEICOSATETRAENOIC ACID AND VASCULAR TONE

Top Previous Next References

The concept that small arteries throughout the body express CYP4A protein and produce 20-HETE and that this substance plays a central role in the regulation of vascular tone and growth emerged from work done 10 years ago (225). A summary of these experiments is presented in Figure 4. The effects of inhibitors of AA metabolism on the response of canine renal arcuate arteries to elevations in transmural pressure were examined. These vessels constrict in response to elevations in transmural pressure, and AA markedly enhanced the myogenic response of these arteries (225). The effects of AA were enhanced by indomethacin and were blocked by several mechanistically different CYP inhibitors, including SKF-525, miconazole, and 12-hydroxy-16-heptadecynoic acid (12-HHDYA). These studies were the first to suggest a role for endogenously formed CYP metabolites of AA in modulating the myogenic response of renal arterioles.

View larger version (18K): [in this window] [in a new window]   Fig. 4. Effects of arachidonic acid (AA; 50 µM) and cytochrome P-450 (CYP) inhibitors (ketoconazole, SKF 525, ABT, or 12-hydroxy-16-hepatadecynoic acid) on the pressure-diameter relation in dog renal arcuate arteries. [Redrawn from Kauser et al. (225).]

CYP inhibitors also impair the vasodilator response to acetylcholine (225, 289). This observation was ignored for many years because it was hard to understand how CYP inhibitors could block a NO-mediated response. However, it has since become apparent that a NO-induced fall in 20-HETE levels mediates the cGMP-independent effect of NO on K⁺ channel activity and vascular tone in renal and cerebral arteries (8, 12, 13, 351, 446, 447) and that endothelial-derived CYP metabolites of AA serve as EDHF at least in some vascular beds (48, 120, 139). These observations likely explain why inhibitors of CYP activity attenuate the response to endothelial-dependent vasodilators.

The metabolites of AA produced by canine renal interlobular arteries (289), rat renal (203, 527) and cerebral arterioles (140, 448), and cat cerebral arterioles (141, 156) were identified using HPLC and gas chromatography-mass spectroscopy (GC-MS). In one of these studies (Fig. 5), renal microvessels were isolated from the kidney of rats with a magnet after filling the vasculature with iron oxide (203, 527). These vessels produced 20-HETE (retention time 10 min) and lesser quantities of 14,15- and 11,12-EETs (17-20 min) and the corresponding DiHETEs (7-9 min) (203, 527) when incubated with AA. Similar results were obtained when cerebral arteries of rats (140, 448) and cats (156) and microsomes prepared from canine renal arteries (289) were incubated with AA.

View larger version (48K): [in this window] [in a new window]   Fig. 5. Metabolism of 14C-labeled arachidonic acid by rat renal microvessels. Left panel: appearance of renal microvessels isolated using a magnet after filling the kidney with iron oxide. Right panel: typical reverse-phase HPLC chromatogram depicting the metabolites formed when renal microvessels were incubated with 14C-labeled arachidonic acid (AA; 1 µCi/ml, 10 µM) for 30 min at 37°C in a 0.1 M KPO4 buffer containing 1 mM NADPH. The peaks eluting at 9.5 min and 25 min coelute with 20-HETE and AA standards, respectively. Those appearing at 8 and 9 min coelute with 11,12- and 8,9-dihydroxyeicosatrienoic acids, respectively. Peaks appearing between 12 and 17 min coelute with the hydroxyeicosatrienoic acids (18-, 16-, 15-, 12-, 5-HETEs). [Redrawn from Alonso-Galicia et al. (13) and Zou et al. (527).]

EETs are produced by CYP enzymes of the 2C and 2J families in vascular endothelial cells (119, 120, 335, 416). They are potent vasodilators that hyperpolarize VSM by increasing the activity of large-conductance, Ca²⁺-activated K⁺ channels (47, 49, 95, 97, 139, 143, 164, 187, 270, 526) and have been proposed to be an EDHF (48, 120, 165, 314). In contrast, VSM cells in renal, cerebral, pulmonary, and skeletal muscle arterioles (34, 140, 156, 203, 205, 256) express CYP4A mRNA and protein and appear to be the source of the 20-HETE produced in these vessels.

C.  Actions

1.  Vascular tone

20-HETE constricts renal (9, 203, 289, 446, 447), cerebral (140, 155, 156), mesenteric (478), and skeletal muscle arterioles (158, 257). An illustration of the effects of 20-HETE on renal afferent arterioles of the rat (203) is presented in Figure 6. In this experiment, indomethacin, baicalein, and 17-ODYA were added to the bath to block the endogenous formation of metabolites of AA and the metabolism of 20-HETE via the COX, lipoxygenase, and CYP pathways. 20-HETE at a concentration of 10⁸ M significantly reduces the diameter of rat afferent arterioles. This concentration of 20-HETE is similar to those reported to constrict cerebral arteries of cats (156), skeletal muscle arterioles of the rat (158), and the afferent arteriole of rabbits (23). It should be noted that the endogenous level of 20-HETE in microvessels is ~100 nM (140, 446). This probably explains why most investigators find that high concentrations of 20-HETE (1-10 µM) are required to constrict renal and cerebral arterioles when the endogenous formation of this substance is not blocked.

View larger version (13K): [in this window] [in a new window]   Fig. 6. Effects of 20-hydroxyeicosatetraenoic acid (20-HETE) on the diameter of rat afferent arterioles studied using the in vitro perfused juxtamedullary nephron preparation. *Significant difference from the control diameter at 90 mmHg. [Redrawn from Imig et al. (203).]

Although 20-HETE is a potent constrictor of small arteries and arterioles (<100 µm), it has little or no effect on larger arteries or the aorta. The reason for the differential response to 20-HETE in large arteries versus small arterioles remains to be determined. 20-HETE seems to have equivalent mitogenic effects in VSM cells isolated from the aorta and small vessels of the rat (321, 323), so the lack of effect of 20-HETE on vascular tone in large vessels is surprising. It may have something to do with differences in the expression of K⁺ channel types or the signal transduction pathways expressed in large versus small blood vessels.

The initial studies on the mechanism of action of 20-HETE in canine renal arteries revealed that it is associated with a 10-mV depolarization of membrane potential and a rise in intracellular Ca²⁺ concentration in VSM cells (289). Because membrane potential is dependent on K⁺ conductance in VSM cells (154), our lab in collaboration with Drs. Harder and Gebremedhin studied the effects of 20-HETE on K⁺ channel activity in VSM cells isolated from renal and cerebral arterioles (141, 156, 525). The results of a typical experiment in VSM cells isolated from cat cerebral arteries (156) are presented in Figure 7. Inhibition of the formation of 20-HETE with 17-ODYA activates a large-conductance, Ca²⁺-activated K⁺ (K_Ca) channel recorded from VSM cells isolated from cat cerebral arteries (156). This effect is completely reversed by the addition of 1 nM 20-HETE to the bath. Similar results were obtained using VSM cells isolated from rat renal arterioles (525). Further studies indicated that the vasoconstrictor response to 20-HETE in renal arterioles is mimicked by blocking K_Ca channels with TEA or iberiotoxin and that 20-HETE has no effect on vascular tone after administration of these agents (203). These findings indicate that 20-HETE is normally produced by VSM cells, where it serves as an endogenous inhibitor of the K_Ca channel.

View larger version (29K): [in this window] [in a new window]   Fig. 7. Effects of blockade of the endogenous formation of 20-HETE on K+ channel activity in vascular smooth muscle cells isolated from a middle cerebral artery of a cat. Top panel: tracing of signal channel activity recorded from a cell-attached patch from a cell bathed in normal physiological salt solution at resting membrane potential. Middle panel: marked increase in K+ activity after blocking the endogenous formation of 20-HETE with 17-octadecynoic acid (17-ODYA; 10 µM). Bottom panel: ability of 1 nM 20-HETE to reverse the effects of 17-ODYA on K+ channel activity in this cell. [Data from Harder et al. (156).]

A summary of the role of 20-HETE in the regulation of vascular tone is presented in Figure 8. Membrane stretch and a number of vasoactive hormones and mitogens activate phospholipase C that increases the synthesis of inositol trisphosphate (IP₃) and triggers the release of intracellular Ca²⁺. The rise in intracellular Ca²⁺ concentration activates K_Ca channels, leading to membrane hyperpolarization, which oppose vascular contraction by limiting Ca²⁺ influx through voltage-sensitive Ca²⁺ channels. Therefore, some intrinsic mechanism must exist that buffers against activation of K_Ca channels after elevations in intracellular Ca²⁺ concentration. This is where 20-HETE likely plays a critical role, since elevations in intracellular Ca²⁺ concentration are known to activate Ca²⁺-sensitive phospholipase A and diacylglycerol (DAG) lipase to release AA. AA is then converted to 20-HETE, which blocks K_Ca channels. Blockade of this channel depolarizes VSM and enhances Ca²⁺ entry and vascular contraction by enhancing the opening of voltage-sensitive Ca²⁺ channels. There is also evidence that 20-HETE has an additional effect to enhance the opening of L-type Ca²⁺ channels (141).

View larger version (36K): [in this window] [in a new window]   Fig. 8. Summary of the role of cytochrome P-450 metabolites of arachidonic acid (AA) in the control of vascular tone. Membrane stretch and vasoactive agents [angiotensin II (ANG II), norepinephrine (NE)] activate phospholipase C (PLC) to release inositol trisphosphate (IP3) and diacylglycerol (DAG). IP3 triggers the release of intracellular Ca2+ from the endoplasmic reticulum. Elevations in intracellular Ca2+ concentration activate Ca2+-sensitive phospholipase A2 (PLA) and diacylglycerol (DAG) lipase to release AA and stimulate the formation of 20-hydroxyeicosatetraenoic acid (20-HETE). 20-HETE blocks the large-conductance, calcium-activated potassium (KCa) channel in vascular smooth muscle (VSM) cells, leading to a fall in membrane potential (Em), which enhances Ca2+ influx via L-type, voltage-sensitive Ca2+ channels. Epoxyeicosatrienoic acids (EETs) are produced by the endothelium and are potent vasodilators that hyperpolarize VSM cells by increasing the activity of the KCa channel. Acetylcholine and bradykinin stimulate the release of EETs from the endothelium, and they serve as endothelial-derived hyperpolarizing factor (EDHF) in some vascular beds.

The mechanism by which 20-HETE blocks the K_Ca channel remains to be determined. No one has yet identified a 20-HETE receptor or demonstrated that it alters the production of second messengers (DAG, IP₃, cAMP, or cGMP). However, the finding (9) that structural analogs of 20-HETE block the vasoconstrictor response to 20-HETE suggests that there is some type of receptor in VSM cells. In addition, the observations that 20-HETE blocks K⁺ channel activity in cell-attached, but not in excised membrane patches, suggest that the inhibitory effect of 20-HETE on the K_Ca channel requires an intact intracellular signal transduction pathway (262, 447).

20-HETE activates protein kinase C (PKC) in VSM cells isolated from cat cerebral arteries (262) and in proximal tubule cells (336). Inhibitors of PKC block the effects of 20-HETE on K_Ca channel activity and vascular tone in cat cerebral arteries (262). These results indicate an obligatory role for PKC in mediating the vasoconstrictor response in this vascular bed.

20-HETE also activates PKC in rat renal VSM cells (447). However, PKC inhibitors do not block the effects of 20-HETE on K_Ca channel activity or vascular tone in renal interlobular arteries of the rat (447). 20-HETE also activates the mitogen-activated protein (MAP) kinase-signaling cascade and the expression of the phosphorylated forms of extracellular signal-related kinase (ERK) in rat renal arteries (447). However, inhibition of MEK kinase with PD98059 has no effect on the vasoconstrictor response to 20-HETE or its effects on K_Ca channels in rat renal interlobular arteries (447). The tyrosine kinase inhibitors genistein and tyrphostin block the vasoconstrictor response to 20-HETE in rat renal interlobular arteries (447). These studies suggest that activation of a tyrosine kinase, proximal to MEK in the MAP kinase cascade, contributes to the effects of 20-HETE on K⁺ channel activity and vascular tone in the renal circulation of the rat.

Uddin et al. (464) and Muthalif and co-workers (321, 323) directly demonstrated that 20-HETE activates the MAP kinase pathway in rat aortic VSM cells. They found that blockade of the formation of 20-HETE prevents the effects of norepinephrine (324, 464) and ANG II to activate the MAP kinase system and promote growth in aortic VSM cells (323). In the case of norepinephrine and 20-HETE, the activation of MAP kinase activity involves translocation of the small G protein Ras to the membrane (324). Moreover, Muthalif et al. (324) recently found that both norepinephrine- and 20-HETE-induced proliferation of VSM cells and activation of MAP kinase activity is attenuated by a Ras farnesyl transferase inhibitor and in cells transfected with a dominant negative Ras. Norepinephrine-induced activation of phospholipase D activity in rabbit VSM cells is also coupled to the formation of 20-HETE and activation of MAP kinase pathway via a Ras-dependent mechanism (361). Thus 20-HETE appears to couple receptors that activate phospholipases to activation of MAP kinase signal transduction cascade in VSM cells via Ras-dependent manner similar to the mechanism previously described by Dulin et al. (91) and Jiao et al. (215) for other fatty acids in rabbit renal proximal tubule cells.

A summary of a working hypothesis for the mechanism by which 20-HETE activates the MAP kinase system and alters vascular tone and growth is presented in Figure 9. After activation of a receptor coupled to phospholipase, AA and 20-HETE levels increase within a cell. 20-HETE then promotes phosphorylation of epidermal growth factor (EGF) receptor (perhaps by interaction with PYK or Src), which acts as a receptor tyrosine kinase and phosphorylates the adaptor protein, Shc, which associates with the membrane. Shc binds to the adapter protein Grb2 in a complex with guanine nucleotide exchange factor Sos. The recruitment of Sos to the membrane activates the small GTPase protein p21ras, which in turn activates the small G protein raf. Raf phosphorylates residues in K⁺ and Ca²⁺ channels that increase the opening of Ca²⁺ channels and close K⁺ channels. Overall, 20-HETE likely interacts directly with intracellular proteins (perhaps, PKC, Src, PYK, and/or the EGF receptor) that activate both the MAP kinase and PKC signal transduction pathways to alter ion channel activity, vascular tone, and growth. The signal transduction pathway that predominates is probably species and cell-type dependent and likely reflects diversity in the expression of the signaling transduction proteins and the types of ion channels expressed within a tissue.

View larger version (35K): [in this window] [in a new window]   Fig. 9. Proposed mechanisms for the vasoconstrictor and mitogenic effects of 20-HETE. 20-HETE activates protein kinase C leading to stimulation of the small G protein rho. Rho increases sensitivity of the contractile mechanism to Ca2+. PKC, acting directly or through activation of the small G protein raf, also increases the activity of the KCa channel leading to a fall in membrane potential (Em) and enhanced Ca2+ entry through voltage-sensitive Ca2+ channels. 20-HETE interacts with other intracellular proteins (perhaps Src or PYK) to enhance phosphorylation of the epithelial growth factor (EGF) receptor, which then acts as a tyrosine kinase to activate and recruit the Shc, GRb, Sos complex to the membrane. The recruitment of Sos to the membrane activates the small GTPase protein p21ras, which in turn activates the small G protein raf. Raf activates the mitogen-activated protein kinase (MAPK) system to facilitate cell growth and inhibit the KCa channel.

20-HETE plays a significant role in the myogenic response of renal, cerebral, mesenteric, and skeletal muscle arterioles to elevations in transmural pressure (131, 140, 156, 195, 197, 202, 225, 478) and autoregulation of renal and cerebral blood flow in rats in vivo (140, 349, 527). Some of the evidence supporting this view is presented in Figure 10. These experiments were performed in vitro using rat middle cerebral arteries (140). The diameter of these vessels fell 20% as transmural pressure was increased from 60 to 140 mmHg. Blockade of the synthesis of 20-HETE with DDMS blocked pressure-induced contraction of these vessels. Similar results were obtained when the vasoconstrictor action of 20-HETE was blocked using 15-HETE or 6(Z),15(Z)-20-HEDE (140). Elevations in transmural pressure from 20-140 mmHg produced a fourfold rise in the concentration of 20-HETE in rat middle cerebral arteries (140). These results indicate that elevations in transmural pressure stimulate the release of 20-HETE in cerebral arteries and that blockade of the formation or vasoconstrictor actions of 20-HETE prevents pressure-induced contraction of these vessels that is necessary for autoregulation of cerebral blood flow.

View larger version (29K): [in this window] [in a new window]   Fig. 10. Effects of an inhibitor of the formation of 20-HETE, N-methylsulfonyl-12,12 dibromododec-11-enamide (DDMS; 10 µM), the 20-HETE antagonists 20-hydroxyeicosa-6(Z),15(Z)-dienoic acid (20-HEDE; 1 µM) and 15-HETE (1 µM), and Ca2+-free physiological salt solution on the pressure-diameter relation of rat middle cerebral arteries studied in vitro. Addition of DDMS, 20-HEDE, or 15-HETE to the bath eliminated pressure-induced constriction of these arteries, whereas these vessels exhibited passive dilation in Ca2+-free media. * Significant difference in the control diameter versus DDMS, 20-HEDE, and 15-HETE (n = 5 or 6 vessels/group). [From Gebremedhin et al. (140). Copyright 2000 Lippincott Williams & Wilkens.]

While most studies indicate that CYP inhibitors block pressure-induced constriction of small arteries in vitro, controversy still exists as to whether 20-HETE is the primary mediator of the myogenic response or it simply modulates the responsiveness of the vessel to pressure (131, 417). For example, the data presented in Figure 10 demonstrate that cerebral vessels are not passive after blockade of the synthesis of 20-HETE. They still exhibit myogenic tone and maintain a diameter less than that seen after removal of Ca²⁺ from the bath. There are several possible explanations for these observations. First, while CYP inhibitors block the synthesis of 20-HETE, they do not prevent the release of 20-HETE from membrane phospholipid pools. It may take some time to fully deplete the pool of 20-HETE in vessels and eliminate myogenic tone. Second, 20-HETE activates the PKC and MAP kinase signal transduction cascades. These pathways phosphorylate ion channels and have long-lasting effects on vascular tone. Thus reversal of the vasoconstrictor effects of 20-HETE after acute blockade of its synthesis may be delayed or incomplete. Third, most of the CYP inhibitors used to block the synthesis of 20-HETE also inhibit the formation of EETs. EETs promote Ca²⁺ entry in endothelial cells (146-148, 177, 178, 316). Blockade of epoxygenase activity should decrease Ca²⁺ levels in endothelial cells and reduce the formation of NO and prostacyclin. Thus a fall in NO and prostacyclin release after administration of CYP inhibitors may attenuate the vasodilator response. The recent finding that inhibitors of the formation of 20-HETE have no effect on the diameter of renal and cerebral vessels with an intact endothelium but readily dilate these arteries after the endothelium is removed supports this view (13, 448). Fourth, it is possible that 20-HETE only modulates myogenic tone in vessels. This view is consistent with the previous suggestion that the myogenic response is triggered by the influx of Ca²⁺ through stretch-activated channels (87). The rise in intracellular Ca²⁺ concentration should stimulate the release of AA and promote the formation of 20-HETE. 20-HETE blocks the K_Ca channel and prevents the hyperpolarization of VSM cells normally associated with elevations in intracellular Ca²⁺ concentration. By maintaining a more depolarized membrane potential, 20-HETE potentiates Ca²⁺ influx and myogenic tone. A modulator role for 20-HETE is also consistent with previous observations that the myogenic response and vascular responses to vasoconstrictors and hyperoxia are elevated in SHR in which the formation of 20-HETE is elevated (142, 196, 257).

There is considerable evidence that 20-HETE plays an important role as an oxygen sensor in the microcirculation. This conclusion is based on the results of studies indicating that the production of 20-HETE in renal and cerebral arterioles, renal microsomes, isolated glomeruli, and pulmonary arteries is highly oxygen sensitive (158, 207, 518). The Michaelis constant (K_m) of this reaction is ~50 Torr, which is higher than the PO₂ found in many tissues throughout the body. The formation of EETs in microvessels is also oxygen dependent, but the K_m of the reaction is much lower than that needed to form 20-HETE (158). In contrast, the formation of lipoxygenase and cyclooxygenase metabolites of AA in vessels is not sensitive to oxygen until PO₂ is reduced below 10 Torr (158). Although it seems unusual for the CYP4A enzyme to have such a high requirement for oxygen, there is precedent for this since other CYP-catalyzed reactions are oxygen dependent in the physiological range. For example, Raff and Jankowsik (379) have reported that the K_m for CYPaldo that converts corticosterone to aldosterone is >50 Torr.

Several in vivo studies have demonstrated that 20-HETE plays an important role in mediating the vasoconstrictor response to hyperoxia. Experimental results from one of these studies (127) are presented in Figure 11. Under control conditions, the diameter of third-order arterioles in the cremaster muscle of rats fell by 40% as PO₂ was increased from 30 to 45 Torr by increasing the oxygen concentration of the superfusion solution. Blocking the formation of 20-HETE with 17-ODYA or DDMS reduced the vasoconstrictor response to elevations in tissue PO₂ by 80%. Similar results have been reported in the rat cremaster muscle (127, 158, 257) and in the cheek pouch and cremaster muscle of hamsters (281). DDMS has also been shown to reduce the fall in retinal blood flow in the newborn pig associated with elevating inspired PO₂ (522).

View larger version (22K): [in this window] [in a new window]   Fig. 11. Effects of inhibitors of the formation of 20-HETE, DDMS (10 µM), and 17-ODYA (10 µM) on the vasoconstrictor response of cremaster arterioles to elevations in tissue PO2 produced by superfusion of the preparation with physiological salt solution equilibrated with 21% O2. Blockade of the formation of 20-HETE reduced the vasoconstrictor response to elevation in tissue PO2 by 80%. [Redrawn from Frisbee et al. (127).]

Previous studies have indicated that mRNA for the CYP4A1, 2, and 3 enzymes that produce 20-HETE can be detected in the rat cremaster muscle, but it was unclear whether 20-HETE is produced by the vessels or the surrounding skeletal muscle tissue. Kunert et al. (256) recently answered this question. They found that CYP4A protein is expressed in both the wall of arterioles and in skeletal muscle fibers in the rat cremaster muscle. This observation suggests that CYP4A enzymes serve as a tissue oxygen sensor in both locations. This redundancy in the expression of oxygen-sensing systems may provide skeletal muscle the ability to maintain constant flow when cardiac output is elevated but allow for vasodilation during exercise when the oxygen demand of skeletal muscle is increased.

In contrast to the consistent finding that 20-HETE contributes to the vasoconstrictor response to elevations in tissue PO₂, the contribution of this system to the vasodilator response to hypoxia is unclear. The PO₂ in the lumen of small arterioles (92, 93) is near the lower limit of PO₂ required for 20-HETE formation (158). Tissue PO₂ in skeletal muscle is below the K_m for the formation of 20-HETE. Thus the production of 20-HETE is likely oxygen-limited in many tissues under baseline conditions in vivo, and blocking the formation of 20-HETE during hypoxia may have little additional effect. Indeed, most of the available evidence suggests that endothelial-derived prostaglandins and NO are the primary mediators of the vasodilator response to hypoxia in different vessels (126, 231, 282, 307). Nevertheless, some studies (130, 132) have demonstrated that blocking the synthesis of 20-HETE with DDMS, 17-ODYA, or the vasoconstrictor actions of 20-HETE with 6(Z),15(Z)-20-HEDE attenuates (but does not eliminate) the dilation of skeletal muscle resistance arteries after a reduction in PO₂ in vitro. The relevance of these findings to the in vivo situation is still unclear, since vessels studied in vitro are typically bathed with solutions equilibrated with room air (PO₂ = 150 Torr). Thus the baseline production of 20-HETE is high under these conditions, and reducing bath PO₂ from 150 to 20 Torr likely lowers the production of 20-HETE in vessels studied in vitro and removes its influence on vascular tone. The remaining response of isolated vessels to hypoxia (PO₂ <20 Torr) is 20-HETE independent and is blocked by indomethacin (130). These results suggest that dilation associated with severe hypoxia is largely mediated by release of prostanoids from the vascular endothelium (130, 132), and the dilation seen over the range of PO₂ from 150 to 20 Torr has a component that is 20-HETE dependent.

On the other hand, Kerkhof et al. (231) reported that blockade of the formation of CYP metabolites of AA potentiates the vasodilator response to severe hypoxia (5-10 Torr) in vessels isolated from the cremaster muscle of rats. This response was blocked by removal of the endothelium and inhibition of NO synthase with N^G-nitro-L-arginine methyl ester (L-NAME). They concluded that the vasodilator response to severe hypoxia is mediated by NO released from the endothelium and that blockade of the formation of CYP metabolites of AA potentiates Ca²⁺ entry and promotes the formation of NO. They concluded that 20-HETE exerts a tonic inhibitory influence on K_Ca channels in endothelial cells. Thus blockade of the formation of 20-HETE hyperpolarizes endothelial cells and increases Ca²⁺ influx and the formation of NO.

Given the evidence that increases in the production and/or activity of the 20-HETE contributes to the enhanced vasoconstrictor responses to elevated PO₂ in hypertension (127, 129, 132), it is tempting to speculate that this system contributes to the impaired vasodilator response to hypoxia observed in many models of hypertension (132, 277, 473). Specifically, in hypertensive Dahl S rats fed a high-salt diet, inhibition of the synthesis or actions of 20-HETE reduces the vasodilator response of skeletal muscle resistance arteries to hypoxia. However, inhibition of the formation or action of 20-HETE has very little effect on vascular tone of these same vessels in normotensive Dahl S rats fed a low-salt diet (132). Thus it appears that upregulation of the formation of 20-HETE and/or an increase in the sensitivity of the vasculature to this compound contributes to the changes in the sensitivity of the vasculature to hypoxia in Dahl S rats fed a high-salt diet. On the other hand, in reduced renal mass hypertension, the impaired relaxation of resistance arteries in response to reduced PO₂ seems to involve other mechanisms, namely, a reduced vascular sensitivity to prostacyclin, rather than an increased contribution of 20-HETE constrictor pathways to vascular tone (277).

The concept that 20-HETE plays a role in determining vascular responses to changes in tissue PO₂ has important implications for the development of hypertension. The vascular response to changes in tissue PO₂ underscores the concept of whole body autoregulation of blood flow that is thought to mediate the elevation in total peripheral resistance in volume-dependent forms of hypertension. There is also evidence that vascular responsiveness to PO₂ is upregulated in SHR and in other forms of hypertension (127, 257, 279, 280, 380). Recent studies (127, 257) have examined the role of 20-HETE in mediating the increased reactivity of arterioles to elevated PO₂ in SHR and in rats with reduced renal mass (RRM) hypertension. Arteriolar constriction in response to elevated PO₂ is enhanced in both these models (127, 257, 281, 280) as well as in Dahl salt-sensitive rats (380). Blockade of the formation of 20-HETE caused a greater reduction of the constrictor response to elevated PO₂ in arterioles of SHR and RRM hypertensive rats compared with those seen in normotensive controls. It also eliminated the difference in the magnitude of oxygen-induced constrictor responses seen in hypertensive and normotensive animals (127, 257). Inhibition of the formation of 20-HETE with DDMS also prevents the ability of elevated PO₂ to attenuate flow-induced dilation in the rat cremaster muscle (129). Taken together, these findings indicate that upregulation of 20-HETE-dependent constrictor pathways play a role in elevating total peripheral vascular resistance in the SHR and RRM models of hypertension. Further work is needed to determine whether this is due to increases in the expression of CYP4A protein and the production of 20-HETE or changes in the sensitivity of the vasculature to 20-HETE. It will also be important to determine whether the upregulation of the 20-HETE constrictor pathways is triggered by transient elevations in tissue PO₂, intravascular pressure, sympathetic tone, circulating hormones, and/or other factors.

A recent report by Croft et al. (82) provides direct support for this hypothesis. ANG II increased 20-HETE levels in preglomerular arterioles isolated from the kidney of rats. Others have reported that inhibitors of the formation of 20-HETE attenuate the vasoconstrictor responses to vasopressin, ANG II, endothelin, and norepinephrine in renal and/or mesenteric arteries of the rat (11, 59, 74 195, 201, 353, 354, 450, 470). In the case of endothelin, blockade of the synthesis of 20-HETE attenuated the sustained rise in intracellular Ca²⁺ concentration in renal VSM cells that is dependent on Ca²⁺ influx through voltage-sensitive channels (201). Similar findings have recently been reported for the effects of ANG II on intracellular Ca²⁺ concentration in renal VSM cells (450).

H.  Response to Vasodilators

1.  NO

Previous studies have established that NO mediates the effects of endothelial-dependent vasodilators and serves as the major paracrine system that buffers the vasoconstrictor and hypertensive actions of pressor hormones throughout the body (80, 315, 329). A summary of the potential mechanisms for the vasodilator response to NO are presented in Figure 12. It has generally been assumed that the effects of NO are secondary to activation of guanylyl cyclase and elevations in intracellular levels of cGMP (80, 359). Previous observations that NO increases cGMP levels in VSM cells, that cGMP analogs mimic the vasodilator response to NO, and that inhibitors of guanylyl cyclase and protein kinase G completely block the vasodilator response to NO in many large arteries support a primary role of cGMP in mediating this response (80, 359). However, there are many reports that NO-induced vasodilation can be dissociated from elevations in cGMP and that the effects of NO are not completely blocked by inhibitors of guanylyl cyclase in the microcirculation (8, 12, 13, 35, 56, 80, 117, 182, 296, 346, 446, 488). These observations have led to a search for alternative pathways by which NO promotes vasodilation. One possible mechanism has been described by Bolotina and Cohen (35). They reported that NO directly activates K_Ca channels in VSM isolated from rabbit aorta and that this contributes to the cGMP-independent effects of NO on vascular tone. NO also inhibits NO synthase (151, 186) and CYP enzymes of the 1A, 2B, 3A, and 2C families by forming an iron-nitrosyl complex at the catalytic heme binding sites in these enzymes (234, 235, 283, 454, 495). There are also data attributing the inhibitory effect of nitrates and lipopolysaccharide on drug metabolism to inhibition of hepatic CYP enzymes by NO (96, 283, 418). Given the central role of 20-HETE in the regulation of K⁺ channel activity and vascular tone, it was logical to suspect that this system interacts with NO. Several studies have since examined whether NO inhibits CYP enzymes in renal and cerebral arteries and if an NO-induced fall in 20-HETE levels contributes to the cGMP-independent effects of NO on K⁺ channel activity and vascular tone (8, 12, 13, 446, 448). In support of this hypothesis, Sun et al. (446) demonstrated that addition of nanomolar concentrations of NO to microsomes prepared from renal arterioles increased absorption at 440 nm, which is characteristic of the formation of ferric and ferrous-nitrosyl complexes at the heme binding sites of CYP enzymes (339). The effects of NO on the metabolism of AA in renal and cerebral arteries have also been examined. The results of an experiment performed in renal arterioles (13) are presented in Figure 13. These vessels produced 20-HETE, DiHETEs (retention time 8-10 min), and EETs (16-18 min) when incubated with [¹⁴C]AA. The NO donor sodium nitroprusside (SNP) completely inhibited the formation of 20-HETE and DiHETEs (Fig. 12) (446). Similar results have since been reported for the effects of NO donors on the formation of 20-HETE in renal and cerebral arterioles (12, 13, 446, 448), isolated glomeruli (208), and renal microsomes (8, 358). It is important to realize that low concentrations of NO also inhibit the formation of EETs and DiHETEs (Fig. 13). This may explain why blockade of the formation of NO is usually required to uncover the influence of EETs (EDHF) on vascular tone.

View larger version (29K): [in this window] [in a new window]   Fig. 12. Potential mechanisms for the effects of nitric oxide (NO) on K+ channel activity and vascular tone in renal and cerebral arteries. NO released from the endothelium activates guanylyl cyclase to increase cGMP levels. The increase in cGMP levels activates protein kinase G, which enhances Ca2+ reuptake into intracellular stores and decreases Ca2+ sensitivity of the contractile mechanism. NO also inhibits the formation of 20-HETE. The fall in 20-HETE levels increases the activity of the large-conductance, Ca2+-activated, K+ channel. This hyperpolarizes the cell and reduces Ca2+ influx through voltage-sensitive channels. NO also directly activates K+ channels.

View larger version (32K): [in this window] [in a new window]   Fig. 13. Effects of NO on the metabolism of arachidonic acid (AA) by microsomes prepared from renal preglomerular arteries of the rat. A and B: representative reverse-phase HPLC chromatograms depicting the metabolites of AA formed under control conditions and after addition of NO donors to the incubation media. Total counts detected averaged 58,988 cpm in A and 72,025 cpm in B. SNP, sodium nitroprusside. [From Alonso-Garcia et al. (13).]

Representative results comparing the effects of blockade of cGMP and the 20-HETE pathways on the response of K⁺ channels to NO in VSM cells isolated from renal arteries (446) are presented in Figure 14. NO selectively activates the large-conductance K_Ca channel in VSM cells isolated from either renal interlobular and middle cerebral arteries of rats in a concentration-dependent manner (446, 448). These effects are cGMP independent since inhibitors of guanylyl cyclase or PKG have no effect on the response of K⁺ channels to the NO donors (446, 448, 449) (Fig. 14). In contrast, the effects of NO on K⁺ channel activity are completely blocked by preventing the NO-induced fall in 20-HETE levels by adding nanomolar concentrations of 20-HETE to the bath. In other studies, blockade of the endogenous formation of 20-HETE with DDMS or 17-ODYA mimicked the effects of NO to activate the K_Ca channels in VSM cells isolated from renal and cerebral arteries of rats (446, 448). Moreover, NO has no effect on K_Ca channel activity in VSM cells treated with an inhibitor of the formation of 20-HETE. These studies suggest that even though cGMP analogs and high concentrations of NO directly activate K_Ca channels, the increase in the activity of these channels in response to NO in renal and cerebral arteries seems to be exclusively mediated by a fall in 20-HETE levels.

View larger version (21K): [in this window] [in a new window]   Fig. 14. Representative tracings depicting the effects of blockade of cGMP and 20-HETE pathways on the response of K+ channels to NO in cell-attached patches on vascular smooth muscle cells isolated from rat middle cerebral arteries. The cells were bathed with a high-K+, low-Ca2+ solution to null membrane potential and limit Ca2+ influx. Currents were recorded at a pipette potential of 40 mV before and after addition of 105 M sodium nitroprusside (SNP) to the bath. Cells were treated with vehicle in A, an inhibitor of guanylyl cyclase, 1H-[1,2,4]oxadiazole[4,3-a]quinoxalin-1-one (ODQ; 10 µM) in B, or 20-HETE (100 nM) in C. Horizontal bars indicate baseline current (closed channels). [From Sun et al. (448).]

The relative contributions of a fall in the levels of 20-HETE versus elevations in cGMP to the vasodilator effects of NO donors in renal (8, 13, 446) and cerebral (12, 448) arteries have been compared. The results of a study performed in rat cerebral arteries (448) are presented in Figure 15. The NO donor diethylammonium (Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate (DEA-NONOate) dose-dependently dilated middle cerebral arteries of rats to 80% of control. Blockade of guanylyl cyclase with 1H-[1,2,4]oxadiazole[4,3-a]quinoxalin-1-one (ODQ), or protein kinase G with KT-5823, inhibited the vasodilator response to NO by 35%. In contrast, preventing the fall in 20-HETE levels by adding it to the bath reduced the vasodilator response to NO by 70%. Blockade of both pathways eliminated the vasodilator response to NO. Moreover, inhibition of the formation of 20-HETE with 17-ODYA or DDMS mimics the vasodilator effect of NO in renal and cerebral arterioles and prevents the response to NO donors (12, 13, 448). These findings indicate that a fall in the endogenous formation of 20-HETE plays a major role in the activation of K⁺ channels and the vasodilator response to NO in renal and cerebral arteries in vitro.

View larger version (26K): [in this window] [in a new window]   Fig. 15. Cumulative concentration-response curves depicting the effects of an NO donor on the diameter of rat middle cerebral arteries studied in vitro before and after blockade of guanylyl cyclase with ODQ (10 µM), after intracellular 20-HETE levels were fixed by adding 20-HETE (100 nM) to the bath, or after blockade of both pathways with ODQ and 20-HETE. Data are expressed as the percent increase in the inner diameter of the vessels after preconstriction with serotonin (107 M). Numbers in parentheses indicate the number of vessels studied. *Significant difference from control. [From Sun et al. (448).]

Other studies evaluated the contribution of a fall in the production of 20-HETE to the vasodilator response to NO in rats in vivo. DDMS attenuates the fall in blood pressure and elevations in cerebral and renal blood flow produced by NO donors in rats in vivo (8, 12, 13). DDMS also blunts the hypertensive effects and fall in RBF produced by administration of an inhibitor of NO synthase by 50% (8). Similarly, Oyekan and McGiff (351) reported that an inhibitor of the formation of 20-HETE blunts the effects of L-NAME on blood pressure, renal vascular resistance, urine flow, and sodium excretion in rats by 40-70%. These investigators also found that induction of NO production with lipopolysaccharide attenuates the vasodilator and vasoconstrictor responses to agents that alter the synthesis and/or release of EETs or 20-HETE (352).

Thus the available evidence suggests that a fall in 20-HETE levels contributes to the vasodilator and renal response to NO in rats in vivo. Further studies are now needed to determine the contribution of 20-HETE to the changes in renal function and the development of hypertension associated with chronic blockade of NO synthase and the impairment of vascular responsiveness in septic shock models. It will also be interesting to determine to what extent upregulation of the vascular production of 20-HETE may contribute to the increase in vascular reactivity and the development of hypertension associated with chronic administration of pressor hormones and/or endothelial dysfunction associated with diabetes and salt-dependent forms of hypertension.

A similar interaction seems to be emerging between 20-HETE and the endogenous CO/heme oxygenase systems. Heme oxygenase metabolizes heme to biliverdin and CO. There is now considerable evidence suggesting that CO, like NO, also serves as an endogenous vasodilator in the microcirculation (471, 479). Indeed, blockade of heme oxygenase reduces the formation of CO and constricts isolated arterioles in vitro (217, 252). Acute administration of heme oxygenase inhibitors produces substantial hypertension in experimental animals in vivo. These effects are reversed by administration of CO (216). It has been generally assumed that CO, like NO, acts by stimulating the formation of cGMP (479). However, the ability of CO to increase K⁺ channel activity has been disassociated from elevations in cGMP in some tissues (479). Because CO binds to heme and inhibits of CYP enzymes, inhibition of the formation of 20-HETE may contribute to the effects of CO. In this regard, Coceani and co-workers (76-78) have reported that closure of the ductus arteriosus is dependent on the formation of CYP metabolites of AA and that the inhibitory effects of CO on this response are due to inhibition of the formation of these metabolites. Inhibitors of the formation of 20-HETE also block the effects of CO to activate K⁺ channels in the TALH cells in the kidney (276) and attenuate the increase in vascular tone in renal interlobular arteries after inhibition of the formation of CO (220).

Human polymorphonuclear leukocytes (PMN) express CYP mRNA of the 4F3 family (237). Human and canine PMNs avidly produce 20- and 16-HETE when incubated with AA (29, 30, 171, 172, 402). 20-HETE is also produced by stromal cells in bone marrow where it stimulates erythropoiesis (1). 20-HETE and 16-HETE are released from PMNs after stimulation with calcium ionophore (29), thrombin, platelet activating factor, and other stimuli that activate phospholipases (523). 20-HETE is an inhibitor of AA- and thromboxane-induced platelet aggregation (171). It also antagonizes the aggregation of PMNs induced by a thromboxane mimetic. Thus 20-HETE blocks platelet and PMN aggregation by competitively blocking the thromboxane/endoperoxide receptor. Whether this action is a direct effect of 20-HETE or is secondary to the formation of a COX metabolite, such as 20-hydroxy-PGE₂ or 20-hydroxy-PGI₂, remains to be determined.

16-HETE inhibits adhesion of PMNs to the endothelium and the synthesis of leukotrienes (30). It also prevents the rise in cerebrospinal fluid pressure (an index of tissue damage and swelling) in a thromboembolic model of stroke in rabbits (30). The mechanism by which 16-HETE reduces infarct size and tissue injury in this model remains to be determined.

In addition to regulating vascular tone, 20-HETE contributes to the mitogenic actions of vasoactive agents and growth factors. 20-HETE increases thymidine incorporation in a variety of cells (272, 321, 323, 464). 20-HETE also mediates the activation of the MAP kinase system and growth responses in rat aortic VSM cells in response to norepinephrine and ANG II (323, 464). CYP inhibitors attenuate the growth responses to serum, vasopressin, EGF, and phorbol esters in cultured glomerular mesangial cells (70, 124, 162, 183, 424). 20-HETE (10⁹ M) promotes the growth of cultured LLC-PK₁ and opossum kidney cells, and CYP inhibitors block the mitogenic actions of EGF in primary cultures of rat proximal tubular cells (272). Overall, these studies suggest that 20-HETE may play a role in angiogenesis and the proliferation of mesangial cells in diabetes and hypertension-induced glomerulosclerosis.

	IV. EPOXYEICOSATRIENOIC ACIDS AND VASCULAR TONE

Top Previous Next References

Renal (13, 158, 203, 527), cerebral (156, 159), pulmonary (519), and coronary arteries (47, 400, 403) produce EETs when incubated with [¹⁴C]AA. EETs are produced by endothelial cells. This conclusion is based on the observations that cultured endothelial cells express mRNA encoding for epoxygenase enzymes of CYP1A, 2B, 2C, and 2J families (119, 120, 178, 273, 335). Endothelial cells exhibit CYP monooxygenase activity (2, 24, 219, 368) and produce EETs when incubated with AA (218, 385, 401, 468, 469). There is only one report in the literature that VSM cells produce EETs when incubated with AA (163). EETs are also produced by astrocytes in the brain (7, 18, 320), cardiac myocytes (502, 503), airway and parenchymal tissues in the lung (34, 244, 513, 519), the gastrointestinal tract (510, 515), pancreas (511), and by the glomerulus (207), proximal tubule (205, 343), TALH (63, 102, 101, 206), and collecting duct in the kidney (173).

The isoforms responsible for the formation of EETs in endothelial cells remain to be identified. An enzyme of the CYP2C8/34 family is the primary source of EETs in human and porcine coronary arteries (119, 120, 123) and hamster gracillus arteries (36). This conclusion is based on the observations that endothelium-dependent vasodilator responses associated with the formation of EETs are blocked by antisense oligionucleotides directed against human CYP2C8 (36, 119, 120, 123). Moreover, these responses are potentiated with -naphthoflavone (369) (a strong inducer of CYP1A and a weak inducer of CYP2C8 enzymes) and inhibited by sulfaphenazole (119, 120, 123), which is a selective inhibitor of CYP2C9 enzymes (295). These results suggest that the other epoxygenase enzymes that are expressed in endothelial cells (CYP1A, 2B, 2C, and 2J) only play a minor role in generating EETs under basal conditions. However, the expression of these other isoforms is induced by hydrocarbons, anesthetics, antibiotics, drugs, hormones, and dietary manipulation. Thus they could become more important in the formation of EETs in endothelial cells under different experimental conditions.

EETs dilate renal arteries of rats and rabbits (58, 200, 526); the caudal artery of rats (60); rat mesenteric arteries (372); cerebral arteries of cats, rats, pigs and rabbits (19, 100, 143, 266); canine pulmonary arteries (442); and canine (341, 400), bovine (47, 139, 400, 401), porcine (98, 109, 111, 486), guinea pig (97), and rat (27) coronary arteries. In bovine and porcine coronary arteries, all the regioisomers (5,6-, 8,9-, 11,12-, and 14,15-EETs) are equipotent as vasodilators (47, 341). However, this is not the case in other vascular beds. For example, in cat cerebral arteries, the vasodilator response to 11,12-EETs is greater than that seen with 8,9- or 5,6-EETs (143). In cerebral arteries and caudal arteries of the rat, 5,6-EET is a more effective dilator than the other regioisomers (60, 100). In both of these vascular beds, the vasodilator response to 5,6-EET is blocked by indomethacin, indicating that it is dependent on the formation of a COX metabolite of 5,6-EET or that 5,6-EET promotes the release of prostacyclin or PGE₂.

Differences in the potency and actions of the regioisomers of EETs have also been reported in the renal circulation. 11,12-, 14,15-, and 8,9-EETs dilate the renal microcirculation of rats in vitro (200); however, 5,6-EET is a vasoconstrictor (60, 200). The renal vasoconstrictor response to 5,6-EET is blocked by COX inhibitors and thromboxane receptor antagonists, indicating that this response is mediated by the formation or release of an endoperoxide (60, 200). Infusion of 8,9- and 5,6-EET directly into the renal artery of rats in vivo lowers renal blood flow and GFR (224, 453). These effects are blocked by indomethacin. In the isolated perfused rabbit kidney, 5,6-, 8,9-, and 11,12-EETs are vasodilators, but 14,15-EET is a vasoconstrictor (61). The vasodilator action of 5,6-EET in the isolated perfused rabbit kidney is blocked by indomethacin, indicating the involvement of a vasodilator COX metabolite in this response (58). Finally, EETs constrict pulmonary arteries of the rabbit in vitro (519). This effect was blocked by indomethacin. However, the vasoconstrictor response to EETs in rabbit pulmonary arteries appears to be dependent on release of a vasoconstrictor COX metabolite from the endothelium rather than the formation of a COX metabolite of EETs (519). Overall, it appears that EETs are direct vasodilators in most vascular beds, but 5,6- and 8,9-EETs can be metabolized by COX to either vasodilator or vasoconstrictor COX metabolites that modify their actions. The ability of COX to metabolize 5,6- and 8,9-EETs to vasoconstrictor metabolites may explain why 11,12- and 14,15-EETs has generally been found to be more potent vasodilators than the other regioisomers.

EETs also increase intracellular Ca²⁺ concentration in endothelial cells (146-148, 177, 178, 316). The rise in intracellular Ca²⁺ concentration stimulates the formation and release of NO, PGI, PGE₂, thromboxane, and PGF₂. The release of these endothelial factors can either augment or oppose the direct effects of EETs on vascular tone.

Several studies have suggested that EETs are more potent vasodilators than the corresponding DiHETEs (143, 200). This observation has led to the view that epoxide hydrolase activity limits the biologic actions of EETs. However, the effects of DiHETEs on vascular tone have not been systemically evaluated in most studies. Moreover, Fang et al. (111) and Oltman et al. (341) have shown that DiHETEs and chain-shortened, -oxidation metabolites of DiHETEs are just as potent vasodilators in canine and porcine coronary arteries as EETs. Thus the vasodilator response to EETs is not diminished by conversion to DiHETEs and shorter-chain metabolites at least in coronary arteries. It remains to be determined whether DiHETEs contribute to the regulation of vascular tone in other vascular beds.

A number of studies have examined the mechanism by which EETs reduce vascular tone. EETs hyperpolarize VSM cells (47, 97, 134, 143, 166) by increasing the open-state probability of K_Ca channels (47, 49, 95, 97, 139, 143, 164, 187, 270, 526). This is thought to attenuate Ca²⁺ entry via voltage-sensitive channels. Previous findings that depolarization of VSM cells with a high-K⁺ media and blockade of K_Ca channels with iberiotoxin, charybdotoxin, or TEA prevent the vasodilator response to EETs, provide strong support for a primary role of K⁺ channels in mediating the vasodilator response to EETs (47, 97, 143, 526). However, no studies have documented that the vasodilator response to EETs is associated with a fall in Ca²⁺ influx or intracellular Ca²⁺ concentration in VSM cells; rather, there appears to be unequivocal evidence that EETs increase intracellular Ca²⁺ concentration in endothelial cells (146-148, 177, 178, 316). This is likely due to hyperpolarization of the cells, which enhances the electrochemical gradient for Ca²⁺ influx, since EETs activate K_Ca channels in pig coronary artery endothelial cells (26). EETs have also been reported to increase intracellular Ca²⁺ concentration in VSM isolated from porcine coronary arteries (113) and bovine aorta (316). They constrict porcine coronary arteries (113) under baseline conditions. They also potentiate vasopressin-induced elevations in intracellular Ca²⁺ concentration in A7R5 aortic smooth muscle cells (457). Thus one of the remaining challenges in this field is to explain how EETs dilate blood vessels if they increase intracellular Ca²⁺ concentration in VSM cells.

The mechanism by which EETs alter K_Ca channel activity also remains controversial. As presented in Figure 16, 11(R)-, 12(S)-, but not 11(S),12(R)-EET dilates rat renal arcuate and afferent arterioles (200, 526) and activates the K_Ca channel in cell-attached patches in VSM cells isolated from these arteries (526). These results indicate that the vascular response to EETs is stereospecific and suggest that there may be receptors for EETs. Further support for the existence of an EET receptor is the findings that EETs specifically bind to U-937 cells and guinea pig monocytes (499-501). However, similar studies have yet to be performed using VSM or endothelial cells.

View larger version (20K): [in this window] [in a new window]   Fig. 16. Top: representative tracings depicting the effects of stereoisomers of 11,12-EETs on potassium channel activity in cell-attached patches on vascular smooth muscle cells isolated from rat renal interlobular arteries. The cells were bathed with a high-K+, low-Ca2+ solution to null membrane potential and limit Ca2+ influx. Currents were recorded at a pipette potential of 40 mV before and after addition of 11(R),12(S)- and 11(S),12(R)-EETs (100 nM) to the bath. c indicates baseline current (closed channels). Bottom: vasodilator effects of different concentrations on 11(R),12(S)- and 11(S),12(R)-EETs on the diameter of renal interlobular arteries. Data are expressed as the percent increase in the inner diameter of the vessels after preconstriction with phenylephrine (107 M). Numbers in parentheses indicate the number of vessels studied. * Significant difference from control. [From Zou et al. (526).]

Several investigators found that EETs have no effect on the activity of the K_Ca channel in inside-out detached membrane patches excised from VSM cells (143, 270, 526). This suggests that EETs do not directly activate the K_Ca channel in VSM as has been described for many other fatty acids (347). Rather, EETs require some sort of intracellular signal transduction system to increase K⁺ channel activity. In the presence of GTP, which dissociates the - and -subunits of G proteins, EETs activate the K_Ca channel in patches excised from bovine coronary artery VSM cells and human embryonic kidney 293 (HEK 293) cells expressing a K_Ca channel (135, 164, 270). Li et al. (270) demonstrated the ability of 11,12-EETs to activate this channel in excised membrane patches is dependent on G_s by showing that the response was blocked by an anti-G_s antibody. This finding has recently been confirmed by two other labs (135, 164).

In further studies, EETs were found to stimulate the ADP-ribosylation of G_s in VSM cells isolated from bovine coronary artery (271) in a manner similar to that described by Seki et al. (423) in the liver. Inhibition of mono-ADP-ribosyltransferases with novobiocin, vitamin K, 3-aminobenzamide, or m-iodobenzylguanidine blocks the ability of EETs to activate K_Ca channels in bovine coronary artery (271) and in HEK 293 cells overexpressing the K_Ca channel (135). These studies indicate that EETs likely interact with an intracellular receptor to activate mono-ADP-ribosyltransferases that catalyze the ADP-ribosylation of G_s. G_s then promotes phosphorylation and activation of the K_Ca channel.

The mechanism by which G_s activates the K_Ca channel remains to be established. G_s stimulates adenylyl cyclase, which increases the formation of cAMP. cAMP activates protein kinase A (PKA), which increases K_Ca channel activity (438) via phosphorylation of the serine-869 residue (328). However, the vasodilator effect of EETs in bovine coronary arteries has been reported to occur in the absence of changes in intracellular levels of cAMP or cGMP (47). Moreover, inhibitors of adenylyl cyclase or PKA do not alter the effects of EETs on the K_Ca channel expressed in HEK 293 cells (135). Thus in HEK 293 cells and porcine and bovine coronary artery VSM cells, EETs can activate the K_Ca channel via a cAMP-independent, membrane-delimitated action of G_s.

In contrast, Imig et al. (199) reported that vasodilator response to EETs in the renal circulation of rats is entirely dependent on activation of adenylyl cyclase and PKA, similar to the effects of prostacyclin and PGE₂. Moreover, 11,12-EETs have been shown to increase cAMP levels in ventricular myocytes and modulate L-type Ca²⁺ channel activity through activation of PKA (504). Thus it is difficult to understand why EETs, which promote ADP-ribosylation of G_s, do not also increase cAMP levels in VSM cells. Overall, the available evidence suggests that the relative contribution of PKA-dependent versus PKA-independent mechanisms for activation of K_Ca channels must vary between species and vascular beds. This could reflect differences in the expression of isoforms of adenylyl cyclase and PKA and the coupling between G_s and adenylyl cyclase in various cell types.

In addition to NO, prostacylin, and prostaglandins, the endothelium releases other autacoids that contribute to the vasodilator response to ACh, bradykinin, and AA in vessels treated with inhibitors of NO and COX (48, 80, 165, 315). This substance has been termed EDHF (71, 118, 188, 246) since it hyperpolarizes VSM cells. The vasodilator and hyperpolarization response to EDHF is blocked by elevating extracellular K⁺ concentration and by K⁺ channel blockers. Three groups have established that EDHF is a diffusible factor that hyperpolarizes VSM by activating a K_Ca channel using patch-clamp techniques in combination with a bioassay system in which the effluent from an isolated vessel was superfused over VSM cells (47, 139, 369).

The findings that EETs are made by the endothelium (218, 385, 401, 468, 469) and are potent vasodilators that hyperpolarize VSM cells (47, 97, 134, 143, 166) by activating K_Ca channels (47, 49, 95, 97, 139, 143, 164, 187, 270, 526) led many investigators to examine whether EETs are EDHF (47, 166, 164, 378, 139, 526). In renal (137, 198, 355, 356), pulmonary (116), coronary (47, 48, 97 119, 120, 122, 123, 136-139, 164, 166, 309, 310, 314, 333, 368, 369, 374, 378, 400, 486, 491), skeletal muscle (36), carotid (89, 274), subcutaneous tissue (75), and mesenteric arteries (3, 67), CYP inhibitors attenuate the vasodilator responses to AA, bradykinin, and ACh. Moreover, Campbell et al. (47) demonstrated that the vasodilator response to methacholine in bovine coronary arteries is associated with the release of EETs into the perfusate.

The hypothesis that EETs serve as EDHF is supported by the observations that the inducer of the CYP1A and 2C enzymes, -naphthoflavone, enhances EDHF responses (67, 369) and that many anesthetics (thiopental, isoflurane, etomidate, methohexitone) that inhibit CYP2C enzymes attenuate EDHF-mediated responses (147, 275, 274, 278). This concept is also consistent with the findings that NO, which blocks the synthesis of EETs (8, 233) and decreases the expression of CYP enzymes of the 2C and 3A families (425, 426, 454), also diminishes EDHF-mediated responses (27, 253, 352).

In contrast, there is convincing evidence that epoxygenase inhibitors do not block the response to endothelial-dependent dilators in guinea pig carotid, basilar, mesenteric, mammary, and coronary arteries (65, 81, 97, 364, 365, 456, 461), rat hepatic (531) and mesenteric arteries (134, 465-467), and in the aorta, hindlimb, and cerebral circulation of mice (38). It is possible that the vasodilator response to endothelium-dependent dilators is more dependent on the release of EETs from membrane stores in these vessels than newly synthesized EETs. However, this seems unlikely, since EETs themselves have been reported to have no effect on membrane potential or vascular tone in the guinea pig basilar artery (365) and in hepatic and mesenteric artery of rats (134, 531). Nevertheless, these vessels still exhibit robust EDHF-mediated dilation to ACh and bradykinin. Others have emphasized that the effects of EETs on K⁺ channels differ from those of native EDHF in these vascular beds. For example, the vasodilator and hyperpolarizing effects of EETs are blocked by iberiotoxin, but the effects of EDHF are iberiotoxin insensitive and blocked by charybdotoxin and apamin (different type of K_Ca channel blockers) in guinea pig coronary and carotid arteries and porcine and coronary arteries (97-99). Moreover, most large-conduit vessels, such as rabbit carotid artery and the aorta of rats and rabbits, exhibit EDHF-mediated responses, but they do not produce EETs when incubated with AA (48, 315, 357, 366, 367). These observations have led to suggestions that the effects of CYP inhibitors to block EDHF actions may be an artifact (134, 357, 466, 467) because some of the compounds used to block CYP activity, such as SKF-525 and imidazoles (miconazole, ketoconazole), have direct effects on K⁺ and Ca²⁺ channels (15, 88, 388, 389, 472). However, this is not a valid criticism for most of the other studies in which CYP activity was blocked with 17-ODYA, sulfaphenazole, or antisense oligionucleotides that do not have direct effects K⁺ or Ca²⁺ channels.

In summary, the question as to whether EETs serve as EDHF and play a major role in the regulation of vascular tone remains unresolved. EETs are synthesized and released by the endothelium, activate K_Ca channels, and are vasodilators. In the renal and coronary arteries, EETs appear to serve as EDHF and mediate the vascular responses to AA and bradykinin in the presence of inhibitors of NO synthase and COX. However, as noted above, many large vessels that exhibit EDHF-mediated responses do not produce EETs. In guinea pig coronary and carotid artery and rat mesenteric and hepatic arteries, it seems clear that a factor other than EETs serves as EDHF. Other candidates for EDHF include K⁺ (98, 99), anandamide (382), trihydroxy metabolites of AA (366, 367), and even NO itself (79, 80).

Previous studies have indicated that human endothelial cells express CYP enzymes of the 2C8/9 and 2J2 families and produce EETs when incubated with AA (119, 120, 122, 147, 315, 335). In addition to regulating vascular tone, EETs have anti-inflammatory actions and prevent adhesion of activated leukocytes to the vascular wall (335). Nanomolar concentrations of 11,12-EET or overexpression of CYP2J2 prevent the upregulation of the expression of the cell adhesion molecules vascular cell adhesion molecule-1 (VCAM-1) and E-selectin in cultured endothelial cells in response to cytokines (tumor necrosis factor, interleukin-1) and lipopolysaccharide (123, 335). These effects are associated with inhibition of the actions of the transcription factors NF-B and IB kinase. The mechanism by which EETs inhibit the actions of these transcription factors remains to be determined. EETs activate tyrosine kinase and the MAP kinase signal transduction cascades in both endothelial (123, 147, 316) and VSM cells (122, 316). Activation of these pathways may mediate the anti-inflammatory effects of EETs.

EETs along with 20-HETE and 16-HETE are released after activation of platelets with thrombin and platelet activating factor (523). EETs inhibit platelet aggregation induced by AA and vascular injury (42, 121, 167). Some of the isomers of EETs released (8,9- and 14R,15S-EETs) inhibit platelet aggregation by competing with AA and inhibiting the formation of thromboxane (121). However, some of the other regioisomers of EETs that do not inhibit cyclooxygenase activity also prevent platelet aggregation by an unknown mechanism (121). Moreover, the recent finding that 14,15-EET reduces platelet aggregation in pial arteries of mice (167) suggests that EETs released from the endothelium or from elements of blood during hemostasis modulate platelet function in vivo.

There is considerable evidence that EETs are potent mitogens and contribute to the effects of growth factors on the renal epithelial and mesangial cells as well as endothelial and VSM cells in the vasculature. EETs increase thymidine incorporation and cell proliferation in rat glomerular mesangial cells (183, 424), renal epithelial cells (69, 70), and cerebrovascular endothelial cells (320). CYP inhibitors attenuate the growth responses to serum, EGF, and vasopressin in cultured renal epithelial and mesangial cells (69, 70, 124, 272, 424). The mitogenic effects of EETs in renal epithelial cells are associated with a transient increase in intracellular Ca²⁺ concentration, an increase in cell pH, and activation of tyrosine kinase and MAP kinase signal transduction cascades in a manner similar to that described earlier for AA and 20-HETE (69, 70). EETs also increase intracellular Ca²⁺ concentration and activate the MAP kinase pathway in endothelial (123, 147, 315) and VSM cells (122, 315).

The similarity of signal transduction responses to EETs in renal epithelial and mesangial cells to that seen in vascular endothelial and smooth muscle cells suggests that EETs may also be mitogens in the vascular system and play a role in angiogenesis or vascular injury and fibrosis. In support of this view, Munzenmaier and Harder (320) found that coculture of brain microvascular endothelial cells with astrocytes (that produce EETs) increased cell growth and the formation of endothelial tubes (a precursor to capillaries). The effects of astrocytes on tube formation by endothelial cells were blocked by epoxygenase inhibitors. These results have provided the first direct evidence for a potential role of EETs in promoting angiogenesis.

	V. KIDNEY

Top Previous Next References

The role of CYP metabolites of AA as second messengers in the regulation of renal hemodynamics and sodium transport in the kidney has emerged as a dynamic new field. As can be seen in the sections of rat kidney immunostained for CYP4A protein (205) in Figure 17, CYP4A protein is expressed in preglomerular renal arterioles, glomeruli, proximal tubules, and the TALH (205). All of these structures produce CYP metabolites when incubated with AA. 20-HETE and EETs are both produced in the preglomerular arterioles (13, 82, 203, 527), glomerulus (207), proximal convoluted tubule, and pars recta (343, 377). 20-HETE is the primary metabolite of AA produced by the TALH (63, 101, 206). EETs are the primary CYP metabolites of AA produced in the collecting duct (173).

View larger version (101K): [in this window] [in a new window]   Fig. 17. Immunohistochemical localization of cytochrome P-4504A protein in the kidney of rats. A: low-power view of the renal cortex. CYP450A protein (brown staining) is localized in the afferent and efferent arterioles, interlobular artery, glomeruli, proximal tubule, and thick ascending loop of Henle. B: higher power view of the same field illustrating CYP450A protein in the macula densa. C: section through the outer medulla demonstrating staining for CYP450A protein in the thick ascending loop of Henle. D: section through inner medulla demonstrating staining for CYP450A protein in vasa recta capillaries. [From Ito et al. (205).]

CYP4A protein is expressed in macula densa cells (205), suggesting a role for 20-HETE in modulating tubuloglomerular feedback responses (Fig. 17). It is also found in the pericytes surrounding vasa recta capillaries (Fig. 17). This latter finding is consistent with previous observations that inhibitors of the formation of 20-HETE selectively increase renal medullary blood flow (529, 530). As discussed earlier, enzymes of the 2C23, 2J, and 2C families are expressed in the kidney, but the distribution of these enzymes along the nephron has yet to be studied.

Given the central role of 20-HETE as a second messenger in the regulation of renal vascular tone, a number of investigators have focused on defining the contribution of 20-HETE to the renal hemodynamic responses to ANG II, endothelin, and NO. Recent studies have indicated that endothelin increases the release of 19- and 20-HETE from the isolated perfused kidney of the rat (350). Blockade of the formation of 20-HETE with DBDD or induction of heme oxygenase activity with CoCl₂ reduces the renal vasoconstrictor response to endothelin-1 (ET-1) in the isolated perfused rat kidney by 30% (350). In renal interlobular arteries, blockade of the formation of 20-HETE with DBDD or 17-ODYA reduces the vasoconstrictor response to ET-1 by 50% (170). Inhibition of the formation of 20-HETE with DBDD or CoCl₂ diminishes the fall in glomerular filtration rate and the natriuretic response to intravenous infusion of ET-1 in rats (354). These studies indicate that ET-1 enhances the formation of 20-HETE in the kidney and that 20-HETE contributes to the vasoconstrictor and natriuretic actions of this peptide.

More recently, Imig et al. (201) explored the mechanism by which 20-HETE contributes to the vasoconstrictor response to endothelin. They found that blockade of 20-HETE formation had no effect on the transient rise in intracellular Ca²⁺ concentration in renal VSM cells produced by ET-1. However, it markedly enhanced the sustained response that is dependent on Ca²⁺ influx through voltage-sensitive channels. These findings are consistent with the view that 20-HETE potentiates the vasoconstrictor response to ET-1 by preventing activation of K_Ca channels after receptor-mediated release of intracellular Ca²⁺ stores. On the other hand, selective blockade of the formation of EETs with MS-PPOH enhanced the vasoconstrictor effect of ET-1 on the afferent arteriole of the rat. This indicates that ET-1 stimulates the formation of EETs in renal arterioles and that EETs oppose the vasoconstrictor response to ET-1 (201).

A similar story seems to be emerging regarding the role of CYP metabolites of AA in modulating the renal vasoconstrictor response to ANG II. ANG II stimulates the release of 20-HETE by isolated perfused rabbit kidney (57, 59). Croft et al. (82) demonstrated that ANG II increases the production of 20-HETE in rat preglomerular arterioles. In renal afferent arterioles, blockade of the formation of EETs with miconazole enhances the vasoconstrictor response to ANG II (195, 245). This indicates that ANG II promotes the release of EETs from the endothelium that opposes the vasoconstrictor response. Further work by Arima et al. (22) established that this effect is mediated by stimulation of the AT₂ receptor on the endothelium. Recent work in our laboratory has indicated that inhibition of the formation of 20-HETE reduces the vasoconstrictor response to ANG II by ~50% in rat renal interlobular arteries in which the endothelium has been removed (450). We also found that ANG II inhibits K_Ca channel activity in renal VSM cells via stimulation of 20-HETE and that this enhances the sustained rise in intracellular Ca²⁺ concentration in these cells that is dependent on Ca²⁺ influx through voltage-sensitive channels (450).

20-HETE also contributes to the vasodilator response to NO in the renal circulation. NO inhibits the formation of 20-HETE in renal VSM cells (8, 13, 446). Other studies have established that the fall in 20-HETE levels mediates the activation of K_Ca channels and ~70% of the vasodilator response to NO donors in isolated renal interlobular arteries (13, 446). Inhibition of the formation of 20-HETE with DDMS reduces the fall in arterial pressure and renal vasodilator response to an intravenous infusion of an NO donor in rats by ~50% (8). Blockade of the formation of 20-HETE also attenuates the pressor response to inhibitors of NO synthase and the fall in renal blood flow by ~50-70% (8, 351). These studies indicate that 20-HETE plays a role as a second messenger that modulates renal hemodynamic responses to NO.

CYP metabolites of AA play an important role in the autoregulation of renal blood flow both in vivo and in vitro. In this regard, Imig et al. (202) first examined the role of CYP metabolites of AA in mediating autoregulatory responses of the afferent arteriole using the rat juxtamedullary preparation. They found that the diameter of the afferent arteriole fell by 8% when renal perfusion pressure was increased from 80 to 160 mmHg. Nonselective inhibition of the formation of EETs and 20-HETE with 7-ethoxyresorufin, 17-ODYA, and miconazole completely blocked the vasoconstrictor response of the afferent arteriole to elevations in perfusion pressure and markedly impaired autoregulation of glomerular capillary pressure (202). In subsequent studies, selective blockade of the formation of 20-HETE with DDMS was found to prevent the vasoconstrictor response of the afferent arteriole to elevations in perfusion pressure, whereas blockade of epoxygenase activity with PPOMS enhanced this response (197). These findings suggest that 20-HETE plays an essential role in mediating changes in afferent arteriolar tone in response to changes in perfusion pressure and that shear-related release of epoxygenase metabolites from the endothelium normally buffers against excessive afferent arteriolar constriction.

The results of an experiment that illustrates the importance of 20-HETE in the autoregulation of renal blood flow in rats in vivo (527) are summarized in Figure 18. Renal blood flow is well autoregulated over a range of renal perfusion pressures from 100 to 150 mmHg. Infusion of 17-ODYA, an inhibitor of the formation of EETs and 20-HETE, into the renal artery of the rats markedly impaired autoregulation of renal blood flow. The autoregulatory index rose from 0.1 to 1 after infusion of 17-ODYA (527). In contrast, intrarenal infusion of the more selective epoxygenase inhibitor miconazole had no significant effect on the autoregulation of renal blood flow (527).

View larger version (18K): [in this window] [in a new window]   Fig. 18. Effects of blockade of the formation of 20-HETE on autoregulation of renal blood flow in the rat. Renal blood flow was measured in rats using an electromagnetic flowmeter as renal perfusion pressure was altered using a clamp on the abdominal aorta before and after an intrarenal infusion of 17-ODYA (33 nmol/min). Renal blood flow is expressed as percent of the control blood flow measured at the animal's spontaneous level of arterial pressure (n = 8-10 animals/group). [Redrawn from Zou et al. (527).]

The observation that inhibitors of the formation of 20-HETE completely block autoregulation of renal blood flow in rats in vivo (527) was unexpected, since autoregulation in the kidney is mediated by tubuloglomerular feedback acting in concert with the myogenic response of afferent arterioles. This led to a reexamination of the role of AA metabolites as modulators of tubuloglomerular feedback (TGF) responses in the kidney. Indeed, Franco et al. (125) first reported that perfusion of the loop of Henle with AA potentiates TGF responses in the rat in vivo. Because the effects of AA were not blocked by COX or lipoxygenase inhibitors, they concluded that some unknown metabolite of AA plays a role in TGF. A few years later, Zou et al. (528) explored the role of 20-HETE in mediating the effects of AA on TGF. Stop-flow pressure was measured as an index of glomerular capillary pressure, while the loop of Henle was perfused at various rates with AA and CYP inhibitors. As presented in Figure 19, stop-flow pressure fell from 45 to 30 mmHg when flow in the loop of Henle was increased from 0 to 50 nl/min. These studies confirmed the findings of Franco et al. (125) that AA potentiates the TGF response. Addition of the CYP inhibitor 17-ODYA to the tubular fluid fully reversed the effects of AA and completely blocked TGF responses (528). Moreover, perfusion of the loop of Henle with a solution containing 20-HETE restored TGF responses after blockade of the formation of this compound. This finding indicates that 20-HETE can diffuse across the macula densa and constrict the afferent arteriole. Subsequent studies established that the enzyme responsible for the formation of 20-HETE is expressed in both the afferent arteriole and macula densa (205). Overall, the available evidence suggests that 20-HETE either serves as the vasoconstrictor mediator of TGF that is released by macula densa cells or it acts as a second messenger at the level of the afferent arteriole to transduce the vasoconstrictor response to some other mediator released by the macula densa, perhaps adenosine or ATP (204). The concept that 20-HETE serves as a mediator or modulator of TGF is also consistent with the known effects of NO and ANG II on TGF. In this regard, NO inhibits the production of 20-HETE (8, 13, 446) and attenuates TGF (493), whereas ANG II that stimulates the formation of 20-HETE (82) potentiates TGF (329).

View larger version (17K): [in this window] [in a new window]   Fig. 19. Effects of an inhibitor of the formation of 20-HETE and AA on tubular glomerular feedback response in rats. Stop flow pressure (an index of glomerular capillary pressure) was recorded as loop of Henle was perfused at different rates (0-50 nl/min) under control conditions and after addition of AA (10 µM) plus 17-ODYA (10 µM) and AA only (10 µM) to the tubular perfusate. *Significant difference from the control response. AA enhanced while AA with 17-ODYA inhibited tubuloglomerular feedback responses. [Redrawn from Zou et al. (528).]

CYP4A protein is expressed in pericytes surrounding vasa recta capillaries (205). The localization of CYP4A at this site is consistent with the finding that inhibitors of the formation of 20-HETE selectively increase renal medullary blood flow (529, 530). 20-HETE may also contribute to the actions of other mediators on intrarenal distribution of blood flow. In this regard, Hercule and Oyekan (170) reported that endothelin increases medullary blood flow and that NO synthase inhibitors or blockade of the formation of 20-HETE prevents the increase in medullary blood flow produced by ET-1. They concluded that endothelin raises medullary blood flow by increasing the production of NO that reduces the levels of the medullary vasoconstrictor 20-HETE (170).

CYP metabolites of AA play an important role in the regulation of tubular reabsorption of sodium. A summary of the nephron segments that produce EETs and 20-HETE and the effects of these compounds on tubular function are presented in Figure 20. Both EETs and 20-HETE are produced in the proximal tubule (207, 343). 20-HETE inhibits renal Na⁺-K⁺-ATPase activity (422). Quigley et al. (377) demonstrated that 20-HETE also inhibits sodium transport in isolated perfused rabbit proximal tubules. 19-HETE has the opposite effect and stimulates sodium transport in the proximal tubule. This latter finding is consistent with the idea that 19-HETE is a competitive antagonist of the effects of 20-HETE (9). Thus the stimulatory effect of 19-HETE on proximal sodium reabsorption may be due to blockade of the action of endogenously formed 20-HETE.

View larger version (31K): [in this window] [in a new window]   Fig. 20. Summary of the effects of cytochrome P-450 metabolites of AA on sodium reabsorption in the proximal tubule and thick ascending loop of Henle (TALH). 20-HETE inhibits sodium reabsorption in the proximal tubule by inhibiting Na+-K+-ATPase through a protein kinase C (PKC)-dependent phosphorylation. Angiotensin II (ANG II), parathyroid hormone (PTH), and dopamine inhibit Na+-K+-ATPase activity in the proximal tubule by activating phospholipase A2 (PLA) to release AA and stimulate the formation of 20-HETE. EETs also inhibit sodium transport by inhibiting the translocation of Na+/H+ exchanger to the apical membrane of the proximal tubule. 20-HETE also is produced in the TALH and inhibits sodium transport by blocking a 70-pS K+ channel in the apical membrane. Blockade of this channel limits K+ availability for transport via the Na+-K+-2Cl cotransporter and reduces the lumen-positive transepithelial potential that provides the driving force for passive reabsorption of cations in this portion of the nephron.

In subsequent studies, the inhibitory effects of PTH (345, 386, 415, 430), AA (269), dopamine (21, 336), and ANG II (414) on Na⁺-K⁺-ATPase activity and sodium transport in the proximal tubule have been shown to be dependent on the formation of 20-HETE. Recent work by Nowicki et al. (336) revealed that 20-HETE inhibits Na⁺-K⁺-ATPase activity by stimulating PKC to phosphorylate the -subunit of the Na⁺-K⁺-ATPase.

CYP enzymes of the 2C and 2J family that catalyze the formation of EETs (55, 194, 223, 260, 288, 293, 294, 507) are expressed in the kidney of rats. Enzymes of the 2J family are highly expressed in the proximal tubule of the mouse kidney (288). EETs activate the Na⁺/H⁺ exchanger in cultured rat glomerular mesangial cells (162), which suggests they may play some role in the regulation of glomerular filtration rate. EETs also serve as second messengers for the natriuretic actions of ANG II in the proximal tubule (291, 381, 399), perhaps through an action to potentiate the release of intracellular Ca²⁺ release and inhibit translocation of the Na⁺/H⁺ exchanger.

The findings that 20-HETE and EETs play an important role in the control of sodium excretion suggest that they probably influence pressure-natriuresis and long-term control of arterial pressure. In this regard, Zhang et al. (517) have recently provided direct evidence for a role of CYP metabolites of AA in modulating the pressure-natriuresis response. They found that pretreatment of rats with CoCl_2, which induces heme oxygenase activity and reduces the formation of EETs and 20-HETE, blocks the inhibitory effects of elevated perfusion pressure on Na⁺-K⁺-ATPase activity and proximal tubule sodium transport (517).

20-HETE also plays a major role in the regulation of chloride transport in the TALH (Fig. 20). Escalante et al. (101, 102) and Carroll et al. (63) first reported that 20-HETE was one of the major metabolites of AA produced in TALH cells and that it inhibits Na⁺-K⁺-2Cl cotransport in this nephron segment. Subsequent patch-clamp studies by Wang and co-workers (482-484) revealed that 20-HETE blocks a 70-pS K⁺ channel in the apical membrane of TALH cells. Blockade of this channel limits K⁺ availability for transport via the Na⁺-K⁺-2Cl cotransporter. It also reduces the lumen-positive transepithelial potential that serves as the main driving force for the passive reabsorption of cations (Na⁺, K⁺, Ca²⁺, and Mg²⁺) in the TALH (481). The findings that CYP4A inhibitors increase and 20-HETE decreases transepithelial potential and chloride transport in the TALH perfused in vitro (206) and the loop of Henle perfused in vivo (524) strongly support this view. Others have reported that the inhibitory effects of ANG II (16, 17, 145, 286), bradykinin (150), and elevations in intracellular Ca²⁺ concentration (483, 484) on sodium transport in the TALH are blocked by CYP inhibitors and are mediated by 20-HETE. There is also evidence suggesting that the effects of NO (285) and CO (276) on K⁺ channel activity and sodium transport in the TALH are secondary to inhibition of the formation of 20-HETE (303). Moreover, the elevation in chloride transport seen in the TALH of Dahl S rats is associated with a diminished formation of 20-HETE (206, 396, 524). Overall, these studies indicate that 20-HETE serves as the major intracellular second messenger system controlling sodium reabsorption and K⁺ recycling in the TALH.

Hirt et al. (173) demonstrated that EETs are esterified into membrane phospholipids of the rabbit cortical collecting duct (CCD). They found that all four DiHETEs were potent inhibitors of the hydrosmotic effects of vasopressin in the CCD. These compounds block the cAMP-induced changes in water permeability in the rabbit CCD. These findings suggest that DiHETEs likely interfere with the process by which aquaporins are inserted into the apical membrane of the CCD. EETs also inhibit sodium transport in the CCD (213, 412). 5,6-EET reduces transepithelial voltage and sodium transport in the isolated perfused rabbit CCD. This effect is associated with an elevation in intracellular Ca²⁺ concentration and an increase in the synthesis of PGE₂ in the CCD. The other regioisomers (8,9-, 11,12-, and 14,15-EETs) are inactive (412). The inhibitory effect of 5,6-EET on transepithelial potential is blocked by COX inhibitors; however, COX metabolites of 5,6-EET were inactive (412). These findings suggest that the inhibitory effects of 5,6-EET on sodium transport are secondary to stimulation of the formation of PGE₂ rather than mediated by the formation of a COX metabolite of 5,6-EET.

The expression of CYP4A enzymes and formation of 20-HETE are elevated in the kidney in diabetes (25, 193, 255) and cyclosporin-induced (408, 506) and cisplatin-induced nephrotoxicity (325). The renal excretion of 20-HETE is elevated in Bartter's syndrome (485) and in hepatorenal syndrome (407). The renal synthesis of 20-HETE is elevated in the TALH of rats fed a low-K⁺ diet (480). An elevated production of 20-HETE in the TALH is consistent with the inhibition of sodium transport and urinary concentrating defect seen in Bartter's syndrome and after dietary K⁺ restriction. Similarly, an elevated renal production of 20-HETE in the glomerulus and renal vasculature might contribute to the elevated renal vascular resistance seen in cyclosporin- and cisplatin-induced nephrotoxicity and hepatorenal syndrome. However, the contribution of 20-HETE to the changes in renal hemodynamics and sodium and water excretion associated with these conditions and the nephron segments in which the renal formation of 20-HETE is altered in these disease states remains to be determined.

	VI. BRAIN

Top Previous Next References

Enzymes of the CYP2C, 2D, 2E, 4A, and 4F families that metabolize fatty acids are expressed in the brain (7, 140, 156, 226, 227, 338, 444, 458). Microsomes prepared from whole brain produce EETs (143, 458), and cerebral tissue contains high concentration of EETs (19, 100). CYP2C11 is expressed in astrocytes in the brain, and these cells produce EETs when incubated with AA (7, 18, 155). A CYP2D19 enzyme is also expressed in dopaminergic neurons in the brain that catalyze the formation of EETs (458). The levels of CYP4A1, 2, 3, and 8 mRNA and protein in whole brain are too low to be detected using Northern or Western blot techniques. The expression of these isoforms can only be detected by RT-PCR (444). Nevertheless, the expression of CYP4A1, 2, 3, and 8 mRNA and protein in the cerebral vasculature is higher than that found in renal vessels (140, 141, 156, 262, 338). Moreover, cerebral microvessels avidly produce 20-HETE when incubated with AA (141, 156). Messages encoding for enzymes of the CYP4F3-6, 15, and 16 families are also expressed in the brain (83, 227). Some members of the CYP4F family of enzymes metabolize AA and produce 20-HETE (264) and 18-HETE (44, 45). However, most members of this family seem to prefer prostaglandins and leukotrienes as substrates. Thus it remains to be determined whether the CYP4F isoforms expressed in the brain contribute to the formation of 20-HETE or other biologically active metabolites of AA.

Autoregulation of cerebral blood flow is mediated by pressure-induced myogenic constriction of cerebral arteries acting in concert with the release of vasoactive metabolites by cerebral tissue (metabolic autoregulation) (114). The nature of the myogenic response has been the subject of intense investigation since it was first described by Bayliss (28) nearly a century ago. Pressure-induced contraction of cerebral arteries is associated with inhibition of K⁺ channel activity resulting in membrane depolarization and Ca²⁺ entry through voltage-sensitive channels (87, 154, 157, 395). However, the mechanism by which this occurs has not been determined.

Recent findings have indicated that cerebral arteries produce 20-HETE when incubated with AA (156). This substance serves as an endogenous vasoconstrictor that activates PKC (262) and depolarizes cerebral VSM cells by inhibiting the opening of the large-conductance K_Ca channel (141, 156, 262) and enhancing Ca²⁺ influx through voltage-sensitive Ca²⁺ channels (141). The striking similarities between the vasoconstriction actions of 20-HETE and the myogenic response (156, 225, 289) led to the hypothesis that pressure-induced activation of the synthesis of 20-HETE may contribute to the myogenic response of cerebral arteries and autoregulation of cerebral blood flow. The recent observations that elevations in transmural pressure increase 20-HETE levels (140) and that CYP inhibitors prevent pressure-induced contraction of rat middle cerebral arteries indicate that 20-HETE plays a critical role in the myogenic response of these vessels (140).

To determine the functional significance of 20-HETE to autoregulation of cerebral blood flow in vivo, Gebremedhin et al. (140) examined the effects of inhibitors of the synthesis or actions of 20-HETE on cerebral blood flow responses in rats after elevations in arterial pressure. Elevations in arterial pressure from 100 to 150 mmHg had little effect on cerebral blood flow measured using laser Doppler flowmetry in control rats. The autoregulatory index (percentage change in flow/percentage change in pressure) averaged 0.1 ± 0.06 in these animals. After blockade of the formation of 20-HETE with DDMS, the autoregulatory index rose to 0.92 ± 0.09 (140). This is indicative of a system with fixed vascular resistance. The effects of DDMS on autoregulation of cerebral blood flow were fully reversible (140). Similar results were obtained when the vasoconstrictor effects of 20-HETE were blocked using inactive 20-HETE analogs (15-HETE and 20-HEDE) (140). These studies provide the first direct evidence that elevations in transmural pressure increase 20-HETE levels in cerebral arteries and that this substance augments myogenic constriction of these vessels. They also indicate that 20-HETE plays a critical role in determining the efficiency of autoregulation of cerebral blood flow in vivo. One would predict that elevations in the formation of 20-HETE in the cerebral vasculature might compromise cerebral perfusion and contribute to the pathogenesis of cerebral vascular diseases, such as stroke. In this regard, it is interesting to note that the CYP4A genes lie within a region of chromosome 5 that cosegregates with infarct size after occlusion of the middle cerebral artery in an F₂ cross of stroke-prone and WKY rats (214).

NO plays an important role in the tonic regulation of cerebral vascular tone and mediates the vasodilator responses to acetylcholine, bradykinin, substance P, ₂-adrenergic receptor agonists, and ANG II in the cerebral circulation (41, 114, 247). NO also contributes to the changes in cerebral blood flow produced by inhalational anesthetics and hypercapnia (114). Cerebral vasospasm after subarachnoid hemorrhage is associated with a reduced vascular response to NO (437).

Despite the importance of NO in the control of cerebral vascular tone, considerable uncertainty still exists regarding its mechanism of action. The results of in vivo studies indicating that an inhibitor of guanylyl cyclase, ODQ, blocks 80% of the vasodilator response to NO donors in pial arteries of several species (115, 435) and in the basilar artery of the rat (436) support a primary role for cGMP in mediating the response to NO. However, inhibitors of guanylyl cyclase and PKG do not fully block the vasodilator responses to NO donors in cerebral arteries studied in vitro (12, 296, 346, 448). This observation has led to a search for cGMP-independent mechanisms by which NO reduces vascular tone. Given the important role of 20-HETE in the regulation of vascular tone in the cerebral circulation (140, 141, 156) and the findings that NO inhibits the formation of 20-HETE (8, 446), we studied whether a fall in 20-HETE levels contributes to effects of NO in the cerebral circulation (12, 448). The results of these studies indicate that blockade of guanylyl cyclase activity or PKG has no effect on the ability of NO to activate K_Ca channels in VSM isolated from the middle cerebral arteries of the rat (448). However, preventing the fall in 20-HETE levels in cerebral VSM cells by adding nanomolar concentrations of 20-HETE to the bath completely blocks the ability of NO to activate K_Ca channels (448). In pressurized middle cerebral arteries studied in vitro, blockade of guanylyl cyclase activity reduces the vasodilator response to NO by 40% (12, 448). Preventing the fall in 20-HETE levels reduces the vasodilator response to NO by 60% (12, 448). Simultaneous blockade of both pathways eliminates the vasodilator response to NO in rat middle cerebral arteries. The component of the vasodilator effect of NO that is cGMP dependent is K⁺ channel independent, since the vasodilator response to cGMP is not blocked by a high-K⁺ media or K⁺ channels blockers (448, 449). Further experiments using Ca²⁺-permeabilized vessels suggest that the cGMP-dependent component of the vasodilator response to NO in cerebral arteries is associated with activation of PKG and desensitization of the contractile mechanism to Ca²⁺ (449). The remainder of the vasodilator response to NO is due to inhibition of the endogenous formation of 20-HETE and is secondary to activation of K_Ca channels (448).

The significance of the NO-20-HETE interaction in the control of cerebral blood flow in vivo has also been evaluated. Blockade of the formation of 20-HETE with DDMS reduces the cerebral vasodilator response to central administration of an NO donor in rats by 80% (12). DDMS also reduces the cerebral hyperemic response to an intracerebral ventricular injection of lipopolysaccharide in rats (14). Thus most of the available evidence suggests that a fall in 20-HETE levels in VSM contributes to the cGMP-independent component of the vasodilator response to NO in the cerebral circulation both in vivo and in vitro.

The function of neurons relies on oxidative metabolism and is dependent on cerebral blood flow to maintain adequate oxygen levels. After an increase in neural activity, blood flow is recruited to the active area. This process is termed reactive hyperemia. Recently, astrocytes have been found to express high levels of CYP2C11 protein and produce EETs when incubated with AA (7, 18). The formation of EETs by astrocytes is stimulated by excitatory neurotransmitters (glutamate) (6). This has led to the development of a novel hypothesis for reactive hyperemia. Harder et al. (155) proposed that astrocytes are stimulated by spillover of excitatory neurotransmitters from active neurons. This raises intracellular Ca²⁺ concentration in the astrocytes and stimulates the synthesis and release of EETs. EETs then diffuse across the astrocytic foot processes to cerebral VSM cells. There, EETs activate K⁺ channels, dilate cerebral arteries, and recruit blood flow to the active regions of the brain.

The finding that subdural infusion of the selective epoxygenase inhibitor miconazole reduces baseline blood flow in the brain of rats by 20% supports the view that EETs like NO contribute to the tonic regulation of cerebral vascular resistance (5). Miconazole also reduces the vasodilator response to intracerebroventricular administration of glutamate in rats (6). Similarly, Bhardwaj et al. (33) found that the increase in cerebral blood flow produced by N-methyl-D-aspartate (NMDA) is blocked by the epoxygenase inhibitors MS-PPOH and miconazole. Interestingly, the response to NMDA is also blocked by an inhibitor of NO synthase. This suggests there is likely a complex interaction between EETs and NO in astrocytes. In addition to their direct effects on K⁺ channels in VSM cells, EETs may serve as a calcium influx factor in astrocytes (404) and increase the synthesis of NO. Thus EETs may dilate cerebral arteries via a direct EDHF-like effect and indirectly by stimulating the release of NO from astrocytes. This hypothesis is attractive since it can explain why both epoxygenase inhibitors and inhibitors of NO synthase are equally effective in attenuating the vasodilator response to NMDA (33). Clearly, more experiments demonstrating that selective blockade of the formation of EETs can prevent the redistribution of cerebral blood flow after evoked responses are needed to confirm or refute the hypothesis that astrocyte-derived EETs serve as a mediator of reactive hyperemia in the brain. However, the initial results that have been reported are promising.

	VII. LUNG

Top Previous Next References

The lung metabolizes AA via the COX, lipoxygenase, and CYP pathways (181, 302, 490). COX and lipoxygenase metabolites of AA have long been recognized as important mediators of pulmonary function. PGE₂ is a bronchodilator with anti-inflammatory properties (299, 363). Prostacyclin is a vasodilator and bronchodilator that inhibits platelet aggregation (161, 190). In contrast, PGF₂ and thromboxane are vasoconstrictors and bronchoconstrictors that enhance platelet aggregation (190, 299). Lipoxygenase products mediate much of the bronchoconstriction, increased mucous secretion, and elevated vascular permeability in asthma (4, 90). Indeed, 5-lipoxygenase inhibitors and leukotriene receptor antagonists are now mainstays in the treatment of asthma (90).

While CYP enzymes have long been known to be important in the metabolism of xenobiotics in the lung (513), in the last few years it has become clear that CYP metabolites of AA also play an important role in the regulation of airway resistance and pulmonary vascular tone (34, 513, 518-521). Microsomes prepared from the lungs of rabbit, rat, guinea pig, dog, and humans produce EETs and 20-HETE when incubated with AA (34, 244, 442, 511, 518-521). The concentrations of EETs and DiHETEs in homogenates of human and rabbit lung and in bronchoalveolar lavage fluid are relatively high (242, 511). Immunohistochemical studies have shown that the CYP4A enzymes that produce 20-HETE are expressed in nonciliated cells in the airways and capillary endothelial cells (211, 298, 511). mRNA and protein for the CYP2J enzyme that produces EETs are expressed in ciliated and nonciliated airway epithelial cells, bronchial and vascular smooth muscle cells, and capillary endothelial cells (416, 502, 503, 513). Message for other epoxygenase enzymes, including CYP1A2, CYP2B6/7, CYP2E1, CYP2F1, CYP3A5, and CYP4B1, have been detected in human lung. CYP2B, CYP2C, CYP2E, and CYP1A proteins have been detected in peripheral lung tissue and in pulmonary artery microsomes in human and rabbit lung (383, 511, 518).

AA inhibits Cl secretion in the human airway (189). EETs also reduce transepithelial voltage (V_T) and short-circuit current (I_SC) in tracheal epithelial cells (362). The changes in V_T and I_SC are selective for 11(R),12(S)-EET. They are mediated by blockade of a Cl channel (362). Indeed, Salvail et al. (413) found that EETs inhibit Ca²⁺-sensitive Cl currents in tracheal epithelial cells. Moreover, the epoxygenase inhibitor ketoconazole activates Cl conductance in cultured cystic fibrosis cells (232). Together, these data suggest that EETs cause a net reduction in Cl transport and fluid secretion by airway epithelium.

The effects of 20-HETE in the pulmonary circulation are opposite to those seen in the rest of the body. 20-HETE is a dilator of human and rabbit pulmonary arteries (34, 211, 520). The vasodilator response to 20-HETE is dependent on an intact endothelium and is COX dependent (211). This suggests that 20-HETE is metabolized by COX to a vasodilator, or it stimulates the release of vasodilator COX metabolites from the endothelium (PGE₂ or PGI). Together, these data suggest that 20-HETE acts to decrease the tone of pulmonary arteries.

5,6-EET decreases the pressure of isolated perfused rabbit lungs preconstricted with a thromboxane mimetic (455). Stephenson et al. (442) reported that 5,6-EET also decreases pulmonary vascular resistance in isolated perfused lungs of dogs preconstricted with U-46619. This effect is associated with reduced venous rather than arterial vascular tone. Overall, these studies indicate that EETs dilate the pulmonary circulation of the dog. They also regulate vascular permeability in the lung. This conclusion is based on the finding that the CYP inhibitors ketoconazole and 17-ODYA prevent the increase in the capillary permeability coefficient of canine lung produced by raising intracellular Ca²⁺ with thapsigargin (209).

On the other hand, all regioisomers of EETs were found to constrict pulmonary arteries of rabbits in vitro (519). This effect is both endothelial and COX dependent (519). Because 14,15- and 11,12-EETs are not metabolized by COX, the COX dependency of this response indicates that EETs likely release a vasoconstrictor COX metabolite from the endothelium of rabbit pulmonary arteries. Epoxygenase inhibitors also shift the concentration-response curve to phenylephrine to the right in rabbit pulmonary arteries (519). This finding suggests that epoxygenase inhibitors block the formation of a vasoconstrictor metabolite of AA (perhaps EETs?). Thus the available evidence demonstrates that the effects of EETs on the pulmonary circulation differ between species (dogs versus rabbits) and probably depend on the types of COX products that EETs release from the endothelium.

Given the observations that 20-HETE dilates the pulmonary circulation (34, 116, 212) and the formation of 20-HETE is oxygen dependent (158, 520), Zhu et al. (518) examined whether inhibition of the formation of 20-HETE contributes to hypoxia-induced vasoconstriction in the pulmonary circulation. They found that 17-ODYA and DDMS increased baseline perfusion pressure in isolated perfused rabbit lungs and doubled the increases in pressure seen in response to hypoxia (518). These findings indicate that a fall in the formation of 20-HETE does not mediate hypoxia-induced vasoconstriction. Rather, the results suggest that 20-HETE opposes hypoxic-induced vasoconstriction in the lung. On the other hand, ABT and methoxsalen were recently found to inhibit hypoxia-induced pulmonary vasoconstriction in the rabbit lung (489). These results suggest that the release of EETs may contribute to hypoxia-induced pulmonary vasoconstriction in rabbits.

EETs are well equipped to serve as modulators of airway tone given the abundance of CYP epoxygenases expressed in the airways and the capacity of EETs to relax bronchial rings (511, 521). The biologic imperative to maintain sufficient flow to the lungs indicates that there should be redundant mechanisms to counteract bronchoconstriction. Both 5,6- and 11,12-EETs hyperpolarize rabbit tracheal smooth muscle cells and directly activate large-conductance, Ca²⁺-activated K⁺ channels in airway smooth muscle cells (95). 5,6- and 8,9-EETs relax guinea pig and human bronchi preconstricted with histamine (511).

20-HETE relaxes rabbit bronchi preconstricted with histamine or KCl. The bronchodilator effect of 20-HETE can be eliminated by removal of the epithelial lining of the airway or blockade of tissue COX activity with indomethacin (211). Similarly, 20-HETE relaxes human bronchi preconstricted by histamine (521). These studies indicate that 20-HETE is metabolized by COX to a bronchodilator, or it stimulates the release of vasodilator COX metabolites of AA from the epithelial lining of the lung. On the other hand, 20-HETE is a bronchoconstrictor in guinea pig lung. Thus the effects of 20-HETE on the airways are species dependent and probably reflect differences in the types of COX metabolites produced.

	VIII. HEART

Top Previous Next References

CYP enzymes of the 1A, 2B, 2C, and 2J families are expressed in endothelial cells in coronary arteries. Coronary endothelial cells and intact coronary arteries produce EETs when incubated with AA (218, 401, 403). No one has examined whether CYP4A mRNA or protein are expressed in VSM isolated from coronary arteries. There is one report that bovine coronary arteries do not produce 20-HETE when incubated with AA (403). However, these studies were performed without adding NADPH or maintaining high levels of oxygen in the media. Under similar experimental conditions, even isolated glomeruli, renal and cerebral arterioles that avidly express CYP4A protein will not produce 20-HETE (158, 207). Thus further studies are needed to determine whether 20-HETE is formed and plays a role in the regulation of vascular tone in the coronary circulation.

Recently, Wu and co-workers (502, 503) reported that human and cardiac myocytes express CYP2J2/3 enzymes and produce 14,15- and 11,12-EETs when incubated with AA. The tissue concentration of EETs in the heart is very high (503), and cardiac myocytes incorporate large quantities of EETs and DiHETEs into membrane phospholipid pools (108, 112, 265). These pools are liberated by hypoxia and agonists that activate phospholipases, and the EETs that are released influence cardiac function.

The observations that CYP inhibitors attenuate the vasodilator response to AA and bradykinin in coronary arteries (48, 97, 119, 120, 122, 123, 136-139, 164, 166, 309, 310, 314, 333, 368, 369, 374, 378, 400, 486, 487, 491) strongly support the view that EETs serves as EDHF in this vascular bed. The exceptions have been in porcine (99, 314, 487) and guinea pig coronary arteries (97). However, even in these studies the vasodilator responses to bradykinin and ACh were blocked by an inhibitor of phospholipase A₂ (314, 487). This suggests that the sustained vasodilator responses to bradykinin in porcine (99, 314, 487) and guinea pig (97) coronary arteries treated with epoxygenase inhibitors may be due to the release of EETs from membrane stores.

Despite the overwhelming evidence that EETs serve as EDHF in the coronary circulation, the importance of this system to the control of coronary blood flow and cardiac function in vivo remains to be determined. Only two studies have examined the effects of EETs on coronary blood flow in intact animals. The results indicate that epoxygenase inhibitors block the vasodilator response of canine coronary arteries to bradykinin after administration of inhibitors of NO and COX (333, 491). However, restoration of NO levels after blockade of NO and COX eliminated the EDHF-mediated vasodilator response to bradykinin. This finding suggests that NO normally inhibits the formation of EETs (EDHF) in the coronary circulation. If true, then EETs (EDHF?) probably do not play much of a role in the regulation of coronary blood flow unless the synthesis of NO is compromised. Clearly, more work is needed to better define the experimental conditions under which EETs (EDHF?) may contribute to the regulation of coronary blood flow.

EETs formed in cardiac myocytes affect ion channel activity, intracellular Ca²⁺ concentration, and the contractile function of the heart. EETs block Na⁺ channels in rat cardiac myocytes (265). Because hypoxia raises intracellular Ca²⁺ concentration and activates phospholipases in myocardial tissue, it is likely that EETs are released and modulate electrophysiology of the heart during ischemia (reduce contractile force and heart rate). Thus the release of EETs from cardiac tissue during ischemia may contribute to cardiac preconditioning.

EETs also inhibit the opening of L-type cardiac Ca²⁺ channels (68). Release of EETs during cardiac ischemia would be expected to reduce contractile force and oxygen utilization through this mechanism. The recent finding that pretreatment of rat hearts with 11,12-EETs improved recovery of cardiac function after 20 min of ischemia and reperfusion is consistent with this view (502). In contrast, Xiao et al. (504) reported that 11,12-EET enhances Ca²⁺ current in rat ventricular myocytes. This was associated with an elevation in intracellular cAMP levels and intracellular Ca²⁺ concentration. These results are consistent with a previous report that EETs increase intracellular Ca²⁺ concentration and contraction of guinea pig cardiac myocytes (312). EETs also delayed the recovery of myocardial function in guinea pig hearts after ischemia and reperfusion (312). The reasons for the differences in the results on the effects of EETs on the recovery of cardiac function after ischemia in rats and guinea pigs are unknown. They may reflect species differences or differences in the severity of the ischemic insults used in the two studies. Overall, the current data indicate that EETs are released from cardiac myocytes after ischemia and after administration of vasoactive agents that activate phospholipases. EETs alter Na⁺ and Ca²⁺ currents in cardiac myocytes and likely contribute to changes in heart rate and contractile force during ischemia. However, further work is needed to determine whether EETs enhance or diminish recovery of cardiac function after ischemia and if they play a role in cardiac preconditioning.

	IX. PANCREAS

Top Previous Next References

Glucose-stimulated insulin secretion is coupled to elevations in intracellular Ca²⁺ concentration (308, 497, 498), activation of phospholipases, and the release of AA (230, 474). Pancreatic islets produce PGE₂, PGF₂, thromboxane, and 12-HETE when incubated with AA (463, 505). Blockade of COX activity, however, has little effect on insulin secretion. In contrast, lipoxygenase inhibitors attenuate glucose- and arginine-stimulated insulin release (463, 505). However, the degree of inhibition produced by lipoxygenase inhibitors is not complete, and 12-HETE does not fully reverse the effects of lipoxygenase inhibitors on insulin secretion. These findings suggest that other metabolites of AA might be involved in stimulus-secretion coupling in the pancreas.

Recently, Zeldin et al. (509) reported that CYP2J2/3 enzymes are expressed in glucagon-, insulin-, somatostatin-, and pancreatic polypeptide-containing cells in pancreas of humans and rats. They further documented that EETs are endogenously produced and incorporated into membrane phospholipids in the pancreas. Low concentrations of 5,6-EETs stimulate insulin secretion, whereas 8,9-, 11,12-, and 14,15-EETs stimulate glucagon secretion from the pancreas (106). Given the importance of elevations in intracellular Ca²⁺ concentration in insulin and glucagon secretion (308, 497, 498) and the capacity of EETs to elevate intracellular Ca²⁺ concentration in most cell types (146-148, 177, 178, 315), EETs have been suggested to play a role in stimulus-secretion coupling in the pancreas. If true, then genetic abnormalities in the function or expression of CYP2J enzymes in the pancreas may play a role in diabetes. The recent observation that diabetes and renal end organ damage cosegregate with a region on rat chromosome 5 in an F₂ cross of Goto-Kakizaki (GK) diabetic and Brown Norway (BN) rats that includes the CYP2J3 gene (334) provides direct evidence that supports this possibility.

	X. OTHER ACTIONS

Top Previous Next References

There are a number of other miscellaneous actions of EETs. One of the more interesting is the role of EETs in temperature control. Endogenous pyrogens (interleukin-1 and -4) release AA leading to the formation of prostaglandins. Aspirin is thought to inhibit fever and inflammation by blocking the formation of PGE₂. However, aspirin induces CYP enzymes of the 4A1 and 2E1 pathways (250), indicating there may be involvement of CYP metabolites of AA in fever. In this regard, EETs are formed in the hypothalamus, and CYP metabolites of AA appear to serve as an endogenous antipyretic system. The epoxygenase inhibitors econazole, miconazole, clotrimazole, SKF-525, 1-ABT, and 17-ODYA enhance lipopolysaccharide- and interleukin-1-induced fever in rats and mice (248, 250, 251, 326). Induction of CYP enzymes with fibrates or DHEA has potent antipyretic effects (249). Moreover, central administration of 8,9-, 11,12-, and 14,15-EETs reduces lipopolysaccharide-induced fever (249). These studies indicate that CYP metabolites of AA play a role in fever. It remains to be determined whether this is due to anti-inflammatory properties of these metabolites or direct effects of these compounds on central thermoregulatory pathways.

CYP enzymes are expressed in the hypothalamus (419) and pituitary gland. EETs are synthesized in the posterior pituitary (54). They potentiate the release of somatostatin, vasopressin, oxytocin, growth hormone, and luteinizing hormone from the pituitary by facilitating Ca²⁺ release and entry (50, 330, 433, 434). In this way, they serve as second messengers for the release of these hormones by trophic release factors.

CYP metabolites of AA also play a role in bone reabsorption. Clotrimazole, an inhibitor of the formation of EETs, reduced the interleukin-1-induced Ca²⁺ responses in osteoclasts and reduced bone reabsorption in vivo (72).

Finally, there are reports linking alterations in the expression of CYP2C enzymes to sudden infant death syndrome (SIDS) (32, 459, 460). In patients that have died of SIDS, there is elevated expression of CYP2C8 and 2C9 mRNA and protein and the formation of 14,15- and 11,12-EETs in the liver. The upregulation of CYP2C expression may be secondary to a viral infection and stimulation of gene transcription by interferon- (32). The mechanism by which upregulation of the expression of CYP2C8 and 2C9 enzymes lead to death is unknown. There is some speculation that these enzymes may also be induced in the vasculature, and the increased formation of EETs could lead to hypotension and apnea during sleep (459). However, there is no experimental evidence that supports or refutes this hypothesis.

	XI. HYPERTENSION

Top Previous Next References

The findings that CYP metabolites of AA serve as second messengers in the regulation of renal function and vascular tone have led to considerable interest in defining the role of this pathway in the pathogenesis of hypertension. There is now overwhelming evidence that the renal production of 20-HETE and EETs is altered in many models of hypertension and that blockade of the formation of CYP metabolites of AA alters blood pressure in several of these models. However, it is difficult to predict the consequences of changes in the synthesis of EETs or 20-HETE on blood pressure, since as summarized in Figure 21, CYP metabolites of AA have both pro- and antihypertensive properties. At the level of the renal tubule, 20-HETE and EETs inhibit sodium transport and oppose the development of hypertension. However, in the vasculature, 20-HETE promotes vasoconstriction and hypertension. On the other hand, EETs are endothelial-derived vasodilators that have antihypertensive properties.

View larger version (35K): [in this window] [in a new window]   Fig. 21. Summary of the pro- and antihypertensive actions of 20-HETE. In the proximal tubule and thick ascending loop of Henle of the kidney, 20-HETE inhibits sodium transport and lowers plasma volume and blood pressure. In the renal vasculature and glomerulus, 20-HETE is a constrictor that lowers glomerular filtration rate, promotes sodium retention, and increases arterial pressure. In the peripheral circulation, 20-HETE increases vascular tone and increases blood pressure. TGF, tubuloglomerular feedback; TPR, total peripheral resistance.

B.  Experimental Models of Hypertension

1.  DOCA-salt hypertension

Recent studies have indicated that endothelin contributes to the pathogenesis of hypertension in animals treated with DOCA and a high-salt diet. This model is characterized by elevations in plasma endothelin levels. Endothelin receptor antagonists reduce blood pressure and renal and vascular end organ damage in this model. Previous studies have also indicated that endothelin stimulates the release of 20-HETE from the kidney and that 20-HETE mediates the renal vasoconstrictor and natriuretic actions of endothelin (350, 354). These observations led to an examination of the effects of inhibitors of the formation of 20-HETE on the development of hypertension and end organ damage in DOCA-salt hypertension in rats (353). The results of these experiments are presented in Figure 22 (353). DOCA-salt treatment produced a fourfold elevation in the renal excretion of endothelin and 20-HETE in rats. Chronic administration of CoCl₂, which induces heme oxygenase activity and lowers the production of 20-HETE, reduced blood pressure in rats treated with DOCA and salt (353). Induction of heme oxygenase activity also reduced vascular damage, renal injury, and proteinuria. To determine whether an elevation in the production of CO or inhibition of the formation of CYP metabolites of AA was responsible for the antihypertensive effects of CoCl₂, experiments were performed using ABT, which is a more specific inhibitor of the renal synthesis of 20-HETE and EETs. ABT also prevented the development of hypertension in DOCA-salt-treated rats (353). However, it did not reduce the degree of renal hypertrophy and proteinuria. These studies indicate that 20-HETE contributes to the development of DOCA-salt hypertension presumably by potentiating the vasoconstrictor actions of endothelin.

View larger version (22K): [in this window] [in a new window]   Fig. 22. Effects of a heme oxygenase inducer CoCl2 (24 mg/kg every other day) or the cytochrome P-450 inhibitor 1-aminobenzotriazole (ABT; 50 mg/kg every other day) on the development of hypertension in uninephrectomized rats treated with DOCA and salt (1% NaCl in the drinking water). MAP, mean arterial pressure. * Chronic treatment of rats with CoCl2 or ABT significantly lowered arterial pressure (n = 4-6 rats/group). [Data from Oyekan et al. (353).]

More recently, Mutahalif et al. (322) have confirmed that chronic treatment of rats with ABT reduces blood pressure and renal and cardiac end organ damage in rats treated with DOCA and salt. The antihypertensive effects of ABT were associated with inhibition of renal -hydroxylase activity and a decrease in the activity of the Ras/Map kinase system in VSM cells (322).

In contrast, Honeck et al. (184) reported that the production of 20-HETE is reduced rather than elevated in DOCA-salt hypertensive mice. They found that induction of the renal formation of 20-HETE with fibrates prevented the development of hypertension in mice treated with DOCA and salt. The antihypertensive effects of fibrates in DOCA-salt hypertensive mice were associated with induction of the renal expression of CYP4A14 protein in the proximal tubule, elevated renal production of 20-HETE (184), and an increase in renal blood flow (152). This finding is also consistent with the results of a recent study indicating that knockout of CYP4A14 gene produces hypertension in mice (179). Given the striking discrepancies between the available data in mice and rats, more work needs to be done to determine the role of CYP metabolites of AA as pro- or anti-hypertensive factors in the DOCA-salt model of hypertension.

ANG II stimulates the formation of 20-HETE in the renal circulation of the rat (82) and in the isolated perfused rabbit kidney (59). Blockade of the synthesis of 20-HETE attenuates the vasoconstrictor actions of ANG II in isolated renal arteries in vitro and the pressor response to intravenous infusion of ANG II in rats in vivo (11). These results suggest that elevations in the vascular production of 20-HETE may contribute to the development of ANG II-induced hypertension. On the other hand, CYP metabolites of AA also mediate the inhibitory effects of ANG II on sodium transport in the proximal tubule (291, 414) and TALH (145). Thus blockade of the production of 20-HETE might promote sodium retention and potentiate the hypertensive actions of ANG II. To resolve this dilemma, Mutahalif et al. (323) examined the role of CYP metabolites of AA in the development of hypertension in rats infused with ANG II. Chronic blockade of the renal formation of EETs and 20-HETE with ABT reduced blood pressure from 171 ± 3 to 113 ± 8 mmHg in rats infused with ANG II. In other studies, chronic administration of ABT or DDMS attenuated the development of ANG II-induced hypertension in rats (11). The mechanism by which CYP inhibitors attenuate the development of ANG II-induced hypertension is unknown. We suspect that since DDMS binds to plasma proteins it is poorly filtered and it does not effectively inhibit renal CYP4A activity. Thus it may selectively lower the formation of 20-HETE in the vasculature and reduce vascular tone without blocking the natriuretic actions of 20-HETE. On the other hand, ABT effectively blocks the synthesis of EETs and 20-HETE in the kidney. However, it is not known whether ABT blocks the formation of these compounds in the vasculature. Selective blockade of the formation of EETs and 20-HETE in the kidney should enhance sodium reabsorption and oppose the antihypertensive actions of ABT. Nevertheless, we have found that ABT is just as effective as DDMS in reducing blood pressure in ANG II-infused rats (unpublished observations). All that can be concluded from the available evidence is that CYP inhibitors lower blood pressure in the ANG II model of hypertension, but the mechanism of action of these compounds remains to be determined.

C.  Genetic Models of Hypertension

1.  SHR

Iwai and Inagami (210) first reported that the P-4504A2 gene is regulated by salt and is overexpressed during the development of hypertension in the kidney of SHR. About this same time, Sacerdoti et al. (406) and others (104, 196, 342, 343, 406, 420, 441, 445) reported that the production of 20-HETE and DiHETEs is elevated, and the formation of EETs is reduced in the kidney of SHR. These provocative findings led several investigators (104, 268, 409, 420) to explore the role of 20-HETE in the development of hypertension in SHR. The results of one of these studies (104) are presented in Figure 23. Chronic administration of SnCl₂ that induces heme oxygenase and lowers the renal production of CYP metabolites of AA (104) attenuated the development of hypertension in SHR. Surprisingly, the beneficial effects of SnCl₂ were quite prolonged. The 13-wk-old SHR remained normotensive after withdrawal of SnCl₂ for the duration of the study (Fig. 23). These initial findings have since been confirmed by other investigators using other less toxic heme oxygenase inducers (268, 297, 420).

View larger version (30K): [in this window] [in a new window]   Fig. 23. Effect of chronic treatment with a heme oxygenase inducer, SnCl2, on the development of hypertension in spontaneously hypertensive rats (SHR). Striped area represents the mean range of systolic pressure measured in control and SnCl2-treated Wistar-Kyoto control rats. Treatment was initiated at 5 wk of age. Arrow indicates the time (13 wk) when SnCl2 treatment was discontinued. [From Escalante et al. (104).]

Induction of heme oxygenase also stimulates the production of the potent vasodilator CO. This problem has led to a reexamination of the role of CYP metabolites of AA in the development of hypertension in SHR. In recent studies, Su et al. (445) found that blockade of the renal formation of EETs and 20-HETE with ABT lowered blood pressure acutely in conscious SHR. Similarly, Wang et al. (478) reported that treatment with SHR for 5 days with antisense oligonucleotides directed against CYP4A1 and 4A2 reduced mean arterial pressure by 15 mmHg. Similar reductions in blood pressure were seen in normotensive Sprague-Dawley rats treated with CYP4A antisense oligonucleotides (476). The antihypertensive effect of CYP4A antisense oligonucleotides in Sprague-Dawley rats was associated with a reduction in the expression of CYP4A protein and the synthesis of 20-HETE in renal microvessels (476). Treatment with CYP4A antisense oligonucleotides also reduced vascular reactivity in vessels obtained from the splachnic bed of SHR (478).

The expression of CYP4A protein and 20-HETE production are elevated in the kidney of SHR (142, 195, 342, 343, 410, 441). Because 20-HETE is a renal vasoconstrictor, 20-HETE may contribute to resetting of the pressure-natriuresis relationship and the development of hypertension in SHR by elevating renal vascular resistance and TGF responses (142, 196). In support of this view, we found that inhibitors of the formation of 20-HETE normalize the elevated renal medullary vascular resistance in the kidneys of SHR (142, 196).

Nevertheless, the contribution of CYP metabolites of AA to the development of hypertension in the SHR remains unsettled for several reasons. First, the relative importance of elevations in CO production versus inhibition of the formation of 20-HETE in mediating the antihypertensive effects of inducers of heme oxygenase needs to be better resolved. Second, the elevated renal production of 20-HETE in SHR is due to enhanced expression of CYP4A protein in proximal tubule (343). An elevated production of 20-HETE in this portion of the nephron should inhibit sodium transport and oppose, rather than promote, the development of hypertension. Finally, Shatara et al. (428) found that induction of the renal formation of 20-HETE with fibrates attenuates the development of hypertension in SHR. This finding fits with the view that an elevated production of 20-HETE in the proximal tubule should inhibit sodium transport, potentiate pressure natriuresis, and lower arterial pressure. Thus the mechanism by which elevations in renal formation of 20-HETE contribute to the development of hypertension in SHR remains unclear, unless the renal vasoconstrictor effects of 20-HETE somehow overcome its natriuretic actions.

To determine whether mutations in CYP4A genes play a causal role in the development of hypertension in SHR, a genetic cosegregation analysis in an F₂ cross of SHR and BN rats has been performed (441). The results of these studies are presented in Figure 24. CYP4A genotype had no influence on mean arterial pressure in this study. Mean arterial pressure averaged 122 ± 1 mmHg in the F₂ rats homozygous for the SHR CYP4A alleles, 125 ± 1 mmHg in heterozygotes, and 123 ± 2 mmHg in rats homozygous for the BN alleles. These findings indicate that inherited mutations in the CYP4A gene are not the primary cause of hypertension in SHR. They also indicate that the upregulation of the production of 20-HETE seen in the kidney of SHR is probably secondary to strain differences in the neural and humoral background to the kidney rather than a result of a mutation in one of the CYP4A genes.

View larger version (26K): [in this window] [in a new window]   Fig. 24. Effect of cytochrome P-4504A genotype on mean arterial pressure in an F2 population of male rats derived from a cross of spontaneously hypertensive (SHR) and Brown Norway (BN) rats. Blood pressures were measured in conscious rats through a catheter implanted in the femoral artery (n = the number of rats studied in each group). There was no significant difference in the mean arterial pressures measured in these groups of F2 rats. [Redrawn from Stec et al. (441).]

The contribution of CYP metabolites of AA to the altered renal function in Lyon hypertensive (LH) rats has also been evaluated. This strain resembles the SHR model of hypertension in that the pressure-natriuresis relationship is shifted to higher pressures. They also exhibit an enhanced preglomerular vascular tone that lowers renal blood flow and glomerular filtration rate (305, 306). Intrarenal infusion of 17-ODYA and miconazole enhances the pressure-natriuresis relationship and increases renal blood flow in LH rats (305, 306). Blockade of the renal formation of EETs and 20-HETE has no effect on renal function in the Lyon normotensive strain of rats. The production of 20-HETE and EETs is lower in the kidney of LH rats compared with that seen in the control strains (305, 306). Thus it is unlikely that an enhanced production of 20-HETE is the cause of the renal vasoconstriction in LH rats. Moreover, the vasodilator response to CYP inhibitors in LH rats is prevented by COX inhibitors and thromboxane receptor antagonists (305, 306). These observations suggest that CYP metabolites of AA may be converted by COX to thromboxane and endoperoxides in LH (but not in LN rats), and these products may contribute to the renal vasoconstriction and the development of hypertension in LH rats.

The pressure-natriuresis relation is shifted in Dahl S rats so that the kidney requires a higher perfusion pressure to excrete the same amount of sodium as normotensive rats (Fig. 25) (392, 397). In addition, micropuncture (396) and in vivo and in vitro tubular microperfusion experiments (206, 524) have established that this is due to elevated Cl transport in the TALH (Fig. 25). Given the evidence that 20-HETE is an endogenous regulator of Cl transport in the TALH (102, 101), several investigators have examined whether production of this substance is altered in Dahl S rats (290, 294, 344).The production of 20-HETE in the renal cortex is not different in Dahl S rats compared with normotensive strains (290, 294, 344). However, as presented in Figure 26, the production of 20-HETE and expression of CYP4A protein are reduced in the outer medulla and TALH of Dahl S rats relative to Dahl R (290) and other normotensive strains (439). This finding led to the hypothesis that a deficiency in the formation of 20-HETE may contribute to the elevation in Cl transport and development of hypertension in Dahl S rats. Along these same lines, Makita et al. (294) suggested that a deficiency in EETs production might also contribute to the development of hypertension in the Dahl S rat, since elevations in salt intake increase the production of EETs in normotensive strains of rats (55, 293, 294), but not in Dahl S rats.

View larger version (17K): [in this window] [in a new window]   Fig. 25. Left: relationship between sodium excretion and renal perfusion pressure in inbred Dahl salt-sensitive rats (Dahl S) and salt-resistant (Dahl R) rats maintained from birth on a low-salt diet. Mean values ± SE are presented. * Significant difference from the control value at a renal perfusion pressure equal to the mean arterial pressure of the animal.  Significant difference from the corresponding value in Dahl R rats. [Redrawn from Roman and Kaldunski (397).] Right: chloride reabsorption in the loop of Henle of Dahl S rats (27 tubules, 17 rats) and Dahl R rats (29 tubules, 16 rats).  Significant difference from the value obtained in Dahl S rats. [Redrawn from Zou et al. (524).]

View larger version (19K): [in this window] [in a new window]   Fig. 26. Left: comparison of the expression of CYP4A protein in the renal outer medulla of Dahl S and Lewis rats (left). Lanes 1-6, outer medullary microsomes from Dahl S rats; lane 7, microsomes prepared from the liver of a clofibrate-treated rat that expresses CYP4A1 and 4A3; lanes 8-13, outer medullary microsomes from normotensive Lewis rats. Right: production of 20-HETE by microsomes prepared from the renal outer medulla of Dahl S and Lewis rats. * Significant difference from the corresponding value in Lewis rats. [Replotted from Stec et al. (439).]

To determine whether a deficiency in the formation of 20-HETE contributes to the development of hypertension in Dahl S rats, a genetic cosegregation study was performed (439). A genetic marker spanning a repeated element in intron 11 of the CYP4A2 gene was used to genotype 151 rats derived from an F₂ cross of Dahl S and Lewis rats. The results of this analysis are presented in Figure 27. CYP4A2 genotype cosegregates with blood pressure in this cross. Systolic blood pressure averaged 201 ± 6 mmHg in rats with the SS genotype, 192 ± 4 mmHg in heterozygotes, and 169 ± 3 mmHg in rats with the LL genotype (439). More recently, we have confirmed that transferring the region of chromosome 5 that contains the CYP4A alleles from Lewis rats into a congenic strain of Dahl S rats reduces blood pressure by ~20 mmHg relative to that seen in the parental Dahl S rats (174). It also reduces the degree of proteinuria and renal end organ damage.

View larger version (29K): [in this window] [in a new window]   Fig. 27. Effect of CYP4A genotype on systolic arterial pressures in an F2 population derived from a cross of Dahl S and Lewis rats. Rats were fed a high-salt diet (2% NaCl) for 6 wk. Blood pressures were measured using the tail cuff method (n = number of rats with the corresponding genotype). * Significant difference from the corresponding value in F2 rats with the LL genotype. [Redrawn from Stec et al. (439).]

Other studies examined whether a deficiency in the renal formation of 20-HETE production contributes to the elevation in Cl transport in the TALH of Dahl S rats. Transepithelial potential is elevated in the isolated perfused TALH of Dahl S rats, and it reabsorbs more Cl than TALH obtained from Dahl R, Lewis, or SD rats (206). Exogenous administration of 20-HETE normalizes Cl transport in TALH of Dahl S rats that are deficient in the expression of CYP4A protein, but it has no effect on Cl transport in the TALH of salt-resistant strains of rats (206). Inhibitors of the formation of 20-HETE increase Cl transport and transepithelial potential in the TALH of normotensive rats, but they have no effect in Dahl S rats (206). These findings demonstrate that 20-HETE regulates Cl transport in TALH of the rat and support the hypothesis that a deficiency in the formation of 20-HETE contributes to the elevation in Cl transport in Dahl S rats.

Studies have also been performed to confirm that altering the levels of 20-HETE in the kidney is sufficient to change arterial pressure in rats. The results of these experiments are summarized in Figure 28. Induction of the expression of CYP4A protein in the kidney of Dahl S rats with fibrates normalizes the pressure-natriuretic relationship (10) and prevents the development of hypertension (394, 398, 494) in Dahl S rats. It also reduces the degree of glomerular injury and proteinuria and improves glomerular filtration rate in this model of hypertension (10, 494). In other studies, chronic renal medullary infusion of an inhibitor of the formation of 20-HETE (17-ODYA) increased mean arterial pressure from 115 ± 2 to 142 ± 2 mmHg in normally salt-resistant Lewis rats fed a high-salt diet (see Fig. 28). 17-ODYA reduced the synthesis of 20-HETE in the renal outer medulla of the Lewis rats to the same level seen in Dahl S rats (440). It had no effect on the production of 20-HETE in the renal cortex. These results confirm that chronic reduction in the synthesis of 20-HETE in the outer medulla of the kidney is sufficient to induce salt-sensitive hypertension in normotensive rats and that elevations in renal 20-HETE production improve renal function and prevent the development of hypertension in Dahl S rats.

View larger version (13K): [in this window] [in a new window]   Fig. 28. Left: effect of induction of the renal production of 20-HETE and expression of CYP4A protein with clofibrate (80 mg/day, n = 19) on mean arterial pressure (MAP) of Dahl S rats fed a high-salt diet (8% NaCl) for 4 wk. [Redrawn from Roman et al. (398).] Right: effect of a chronic renal medullary interstitial infusion of 17-ODYA, an inhibitor of the formation of 20-HETE on MAP measured in conscious Lewis rats fed a high-salt diet (8% NaCl) for 5 days. [Redrawn from Stec et al. (440).]

A) ANG II TYPE 2 RECEPTOR. Most of the renal actions of ANG II are mediated by the AT₁ receptor. However, recent studies have indicated that ANG II acts on AT₂ receptors on renal vascular endothelial cells to release EETs that buffer AT₁-induced renal vasoconstriction (22, 195, 319). AT₂ receptors may influence pressure natriuresis and the long-term control of arterial pressure. In this regard, AT₂ receptor knockout mice develop hypertension (153, 191). This is associated with blunted pressure natriuresis and reduced renal blood flow and glomerular filtration rate. Gross et al. (153) found that these mice completely lack the ability to produce 20-HETE in the kidney. A deficiency in the renal formation of 20-HETE may be responsible for resetting of pressure natriuresis in AT₂ receptor knockout mice, since CYP inhibitors have been reported to block pressure natriuresis in the proximal tubule (517). Moreover, a deficiency in the renal formation of 20-HETE has been linked to the blunted pressure-natriuresis relation and elevated sodium transport in the TALH of Dahl S rats (10, 206).

B) CYP4A14 KNOCKOUT MICE. Recently, Holla et al. (179) examined the role of CYP4A genes in the development of renal and cardiovascular disease using targeted knockout approach in 129/SvJ mice. They found that disruption of the CYP4A14 gene produced hypertension in male but not female mice. Mean arterial pressure was 30 mmHg higher in male knockout mice than in the wild-type controls. The hypertension in the male CYP4A14 knockout mice is androgen sensitive in that castration normalized blood pressure in these animals. Interestingly, knockout of the CYP4A14 gene increased rather than decreased the renal production of 20-HETE. This was due to an apparent compensatory increase in the expression of CYP4A12 mRNA. Holla et al. (179) also found that the diameter of the afferent arteriole of CYP4A14 knockout mice was reduced, and they exhibited impaired myogenic response to elevations in renal perfusion pressure. Holla et al. (179) concluded that knockout of the CYP4A14 gene may produce hypertension by upregulating the expression of CYP4A12 and the formation of the vasoconstrictor 20-HETE in the kidney. This conclusion, however, remains tenuous without additional experimental data demonstrating that chronic blockade of the synthesis of 20-HETE lowers blood pressure in the CYP4A14 knockout mice. Moreover, the mechanism by which an elevation in 20-HETE levels in the kidney promotes sodium retention and resets the pressure-natriuretic relationship to higher pressures in CYP4A14 knockout mice needs to be examined, since 20-HETE is thought to inhibit sodium transport in the proximal tubule and TALH.

	XII. SUMMARY AND CONCLUSIONS

Top Previous Next References

New information on the role of CYP metabolites of AA in the regulation of renal function and vascular tone is growing exponentially, but some general concepts have already emerged. VSM cells produce 20-HETE, and this substance serves as a second messenger that plays a critical role in the myogenic response, TGF, vascular hypertrophy, and the vascular responses to vasoconstrictors and dilators by regulating K⁺ channel activity. EETs are produced by endothelial cells. They are potent vasodilators that hyperpolarize VSM cells by opening K_Ca channels and appear to serve as EDHF in coronary arteries and in other vascular beds. CYP metabolites of AA are avidly produced in the proximal tubule and the TALH in the kidney where they serve as inhibitors of sodium transport. The formation of CYP metabolites of AA is altered in experimental and genetic models of hypertension, diabetes, cyclosporin- and cisplatin-nephrotoxicity, hepatorenal syndrome, and pregnancy. Because CYP metabolites of AA are so intertwined with the control of cardiovascular function, it is likely that they contribute to the pathogenesis of some of these disorders and that drugs that modify the formation of EETs and 20-HETE may have therapeutic benefits.