Acti Ons of Epoxygenase M Eta Bolites O N the Preglomerular Vasculature1

Acti ons of Epoxygenase M eta bolites o n the Preglomerular Vasculature1

John D. lmig,2 L.G. Navar, Richard J. Roman, K. Kishta Reddy, and John P. Falck

methacin. In contrast, the vasoconstrictor response to J.D. Imig. L.G. Navar. Department of Physiology. Tu- 5,6-EET was abolished by both removal of the endo- lane University School of Medicine, New Orleans. Lou- thellum or cyclooxygenase inhibition. The thrombox- isiana ane/enderoperoxide receptor inhibitor, SQ 29,548, R.J. Roman. Department of Physiology. Medical Col- resulted in a 60% attenuation of the afferent arteriolar lege of Wisconsin, Milwaukee. Wisconsin vasconstriction to 5,6-EEl. These results indicate that the preglomerular vasoconstriction to 5,6-EEl is cyclo- K.K. Reddy. JR. Falck. Department of Molecular Ge- oxygenase dependent and requires an intact endo- netics University of Texas Southwestern Medical Center Dallas. Texas thelium, whereas the vasodilation to 1 1,12-EET is stereoselective and is the result of direct action of the (J. Am. Soc. Nephrol. 1996; 7:23o4-2370) epoxide on the preglomerular vascular smooth muscle.

ABSTRACT Key Words: Epoxyeicosatrienoic acid. kidney. afferent arte- Epoxygenase metabolites of arachidonic acid are riole. endothellum. cyclooxygenase produced by the kidney and have been implicated in the control of renal blood flow. This study examined A rachidonic acid metabolites of the cytochrome the preglomerular actions of various epoxyeicosatrie- P450 pathway include epoxyeicosatnienoic acids noic acids (EET). By use of the in vitro blood-perfused (EET) and dihydroxyeicosatrienoic acids (DIHETE) juxtamedullary nephron preparation, interlobular formed via epoxygenase enzymes and hydnoxyeicosa- and afferent arteriolar diameter responses to 5,6-EEl, traenoic acids (HETE) formed via P450 w-hydroxybase 8,9-EET, 1 1,12-EEl, and 14,15-EEl were determined. Di- enzymes. The kidney produces all four regioisomenic EET (5,6-; 8,9-; 11,12-; 14,15-), corresponding DI- ameters of interlobular and afferent arterioles precon- HETE, HETE (16-, 17-, 18, 19-, 20-), and 2O-COOH stricted with 0.5 M norepinephrine averaged 24 ± 1 anachidonic acid ( 1-3). Stereoselective production of m (N = 27) and 17 ± 1 m (N = 32), respectively, at various regioisomenic EET (4,5) occurs in the kidney. a renal perfusion pressure of 100 mm Hg. Superfusion In the human kidney, the predominant epoxide with 0.0 1 to 100 nM 1 1 , 1 2-EEl caused graded in- formed Is 14, 15-EET and 66% of this regiolsomer is in creases in diameters of the interlobular and afferent the (R,S) configuration (5), whereas 14, 15(S,R)-EET is arterioles. At a dose of 100 nM, 1 1,12-EET increased the the predominant stereoisomer and 1 1 , 1 2(R,S)-EET is diameters of the interlobular and afferent arterioles by the major epoxide produced in the rat kidney (4). Alterations in the renal production of cytochnome 18 ± 2% (N = 10) and 20 ± 3% (N = 9), respectively. P450 metabobites occurs during the development of The vasodilatory response to 11, 12-EEl was stereose- hypertension ( 1 ,6), after increased dietary salt intake lective because 1 1, 12(R,S)-EET but not 1 1, 12(S,R)-EET (7,8), after uninephrectomy (9), and after induction of increased the diameters of the interlobular and affer- diabetes mellitus ( 1 0). EET have been reported to alter ent arterioles. 14,15-EEl had a much smaller effect renal vascular tone (3, 1 1) and tubular transport (12- and increased the diameters of the these vessels by 14). Endotheliab cells synthesize epoxygenase metab- 10%; 8,9-EET did not significantly affect vascular diam- obites from arachidonic acid (15-17) and these metab- eters. In contrast, 5,6-EEl constricted the interlobular olites have been reported to be vasodilators or and afferent arterioles by 16 ± 3% (N = 6) and 21 ±3% vasoconstrictors depending on the route of adminis-

(N = 7), respectively. The corresponding diols, 5,6- tration (18). DIHETE and 1 1, 12-DIHETE, had no effect on diameters We have recently demonstrated that arachidonic of the interlobular and afferent arterioles at concen- acid decreased the diameter of the affenent arteriole without affecting the postgbomerular vasculature (19). trations up to 1 M. The vasodilatory response to Infusion of arachidonic acid in the presence of indo- 1 1 ,12-EET was not affected by removal of the endo- methacin resulted in vasodibation of the pregbomeru- thelium or by inhibition of cyclooxygenase with indo- bar vasculature that was attributed to an endotheliab- derived cytochrome P450 epoxygenase metabolite

Received March 18. 1996. Accepted July 8, 1996. ( 1 9). Depending on the experimental conditions and 2 Correspondence to Dr. J.D. Imig, Department of Physiology, SL39, Tulane species studied, either renal vasoconstnictive (3,20) or University School of Medicine, 1430 Tulane Avenue. New Orleans, LA 70112. vasodibatory ( 1 1 , 2 1 ) responses to epoxygenase metab- 1046.6673/071 1-2364$03.00/0 olites have been observed. Infusion of 5,6-EET or Journal of the American Society of Nephrology Copyright C 1996 by the American Society of Nephrology 8,9-EET into the renal artery of the rat was reported to

2364 Volume 7 . Number 1 1 - 1996 Imig et al

increase renal vascular resistance. The vasoconstric- reconstituted blood, and renal artery perfusion pressure tion observed may be related to the formation of a measured at the tip of the cannuba was set to 100 mm Hg. cycbooxygenase compound because 5,6-EET in- The organ chamber was warmed and the tissue surface was continuously superfused with a Tyrode’s solution containing creased renal blood flow during cyclooxygenase inhi- 1 % albumin at 37#{176}C. bition (3). In contrast, 5,6-EET, 8,9-EET, and 1 1, 12- Interbobular and afferent arteriobar diameters were mea- EET dilated the isolated perfused rabbit kidney sured via videomicroscopy techniques. The tissue was trans- preconstnicted with phenylephnlne, and the vasodiba- illuminated on the fixed stage of a Leitz Laborlux microscope tion to 5,6-EET was abolished by cycbooxygenase in- equipped with a 75-watt xenon lamp and a water immersion hibition (1 1). In the rat, 1 1,12(R,S)-EET has been objective (40 x ). Video images of the tissue under study were shown to dilate renal arteries preconstricted with generated by a Newvlcon camera (Model NC-67 M; Dage-MTI, phenybephnine and to activate Ca2-activated K Michigan City, IN), passed through a time date generator channels in vascular smooth muscle cells isolated (Model WV-810; Panasonic, Secaucus, NJ) and image en- from renal arteries (22). Collectively, these studies hancer (MFJ Enterprises, Starkville, MS), displayed on a monitor (Model WV-5410, Panasonic), and videotaped for suggest that renal blood flow is affected by epoxygen- later analysis (super VHS VCR, Panasonic). The final magni- ase metabolites; however, the intrarenal sites of action fication at the video monitor was X950 to X 1200. of these metabolites and their ability to exert direct effects on the pregbomerular vasculature have not Vascular Response to Epoxygenase been demonstrated directly. Accordingly, this study was undertaken to Investigate the renal pregbomerubar Metabolites actions of epoxygenase metabolites. Preglomerular After a 20-mm equilibration period, preglomerular inter- vascular responses to superfusion of all four regiolso- lobular and afferent arterioles were chosen for study. Mea- menic EET were observed by using the in vitro blood- surements of interbobubar and afferent arteriolar diameters perfused juxtamedublary nephron preparation. Ad- were made at beast 50 m from any branch points. Norepi- ventitial administration to the renal microvasculatune nephnine (0.5 M) was added to the perfusate to elevate basal permits investigation of the panacnine actions of ep- vascular tone. The diameters of interlobubar arterioles and afferent arterioles decreased from 29 ± 1 j.tm to 24 ± 1 m oxygenase metabolites. Cycbooxygenase, thrombox- and from 20 ± 1 m to 16 ± 1 m, respectively, after ane/enderoperoxide (TP) receptor, and endothelium noreplnephrine administration. After the diameter measure- dependency of the interlobulan and affenent arterioban ment in the presence of norepinephrine, cumulative concen- vasodilation to 1 1 , 1 2-EET and vasoconstriction to tration response curves during superfusion of either 5,6- 5,6-EET were also investigated. EET, 8,9-EET, 11,12-EET, or 14,15-EET (0.01 to 100 nM) were obtained. Superfusion directly over the adventitial side MATERIALS AND METHODS of pregbomerular vessels permits investigation of the para- Vascular Preparation crime actions of epoxygenase metabolites. Four regioisomeric EET were synthesized according to the method of Falck et al. Experiments were performed on male Sprague Dawley (25), and the stereoisomers were separated by chiral phase CD-VAF rats (Charles River Laboratories, Wilmington, MA) high-performance liquid chromatography (26) in the labora- weighing an average of 339 ± 5 g. All experiments were tory of Dr. Falck at the University of Texas Southwestern approved by the Tulane University Animal Care and Use Medical Center. Stock solutions of epoxygenase metabolites Committee. The rats were anesthetized with sodium pento- in ethanol were kept in sealed vials and stored under nitro- barbital (40 mg/kg body wt intraperitoneably lip)), the right gen at -80#{176}Cuntil the experiment. Immediately before use, carotid artery was cannulated, and a midline abdominal the stock solution was added to superfusion solutions and incision was made. The right renal artery of the kidney was the final concentration of ethanol vehicle was <0.05% (vol/ cannulated via the superior mesenteric artery and perfusion vob). A previous study has demonstrated that 5,6-EET is of the kidney was initiated immediately. The kidney was unstable and undergoes extensive conversion to 5,6-DIHETE perfused with a Tyrode’s solution containing 6% albumin and its corresponding 8-lactone (3). Consequently. the (Sigma Chemical Co., St. Louis, MO) and a mixture of L- methyl ester of 5,6-EET (5,6-EET-Me) was used in these amino acids (23). experiments because of its greater stability. Because prelim- Blood was collected through the carotid artery cannula mary studies determined that the response to epoxygenase into a heparinized syringe (2000 U). Erythrocytes were sep- metabolites was maximal by 1 to 2 mm and sustained over a arated from plasma and leukocytes by centnifugation, as 5-mm period, all subsequent experiments monitored the described previously (23). The supernatant was removed and vascular diameter over a 5-mm period. washed twice with 0.9% NaCl containing 0.2% dextrose (pH Because 1 1 , 1 2-EET and 5,6-EET exerted the greatest ef- 7.0). The erythrocytes were resuspended In Tyrode’s solution fects on vascular diameter, the mechanisms by which they containing 6% albumin to yield a hematocrit of 20%. The exert their actions were investigated further. In these exper- reconstituted blood solution was filtered and stirred contin- iments, the effects of the corresponding diols and the cyclo- uously in a closed reservoir that was pressurized by a 95% oxygenase and endothelium dependency of the pregbomeru- 02-5% CO2 tank. The kidney was removed and maintained In bar responses were examined. After control measurements in an organ chamber at room temperature throughout the the presence of norepinephrine, cumulative concentration isolation and dissection procedure. The juxtamedullary ml- response curves for either 5,6-DIHETE or 1 1 , 1 2-DIHETE crovasculature was isolated for study as described prey- (0.01 to 100 nM) were obtained as described above. To ously (24). A portion of the renal microvasculature on the examine the cycbooxygenase dependency, the pregbomerular inner cortical surface was isolated by tying off branches of vascular responses to 100 nM 5,6-EET-Me or 1 1,12-EET the renal artery. The Tyrode’s solution was then replaced by were examined before and 30 mm after addition of indometh-

Journal of the American Society of Nephrology 2365 Afferent Arteriolar Responses to EEl

acm ( 1 0 M, Sigma Chemical Co. , St. Louis, MO) (27) to both AFFERENT ARTERIOLE INTERLOBLYLAR ARTERIOLE the perfusate and superfusate. The role of the TP receptor In the afferent arteriolar vasoconstriction to 5,6-EET was also determined. The afferent arteriolar responses to 5,6-EET (100 nM) and the thromboxane mimetic, U-46619 (100 nM; Biomol. Plymouth Meeting, PA), were examined before and 30 mm after addition of the TP receptor blocker, SQ 29,548(5 j.tM; Biomol, Plymouth Meeting, PA) (28), to both the perfusate and superfusate. In subsequent experiments the robe of

the endothelium in the vasodilatory response to 5,6-EET and -l5 -5 45.6’ *8.9-EEl 1 1 , 1 2-EET was determined. The endothelium was removed 8.9-EET S ::i 1.12-EEl 0 I I.12.EEF .c. by switching the perfusate to a Tyrode’s solution containing -25 014.15-EEl S -25 0 6% albumin and 0.05% 3-l(3-cholamidopropyl)-dimethylam- 10.u -ii 10.11 ur’ 10’ 1O iO’ moniol- 1 -propanesubfonate (CHAPS; Sigma Chemical Co., CONCENTRATION (MJ CONCENTRATION tMj

St. Louis. MO) for 5 mm (19). Elimination ofthe vasodibatory Figure 1. Effect of epoxygenase metabolites on the afferent response to 1 tM acetylcholine and a maintained vasodila- (left panel) and Interlobular arteriole (right panel). Interlob- tion to 10 MM nitroprusside in vessels preconstricted with ular arteriolar diameter averaged 24 ± 1 m (N = 27) and 0.5 gM norepinephrine were used as criteria to confirm afferent arteriolar diameter averaged 17 ± 1 m (N = 32) in endothelial removal and intact vascular smooth muscle re- the presence of norepinephrine (0.5 .M). indicates signif- sponsiveness. Acetybcholine increased preglomerular vascu- icant difference from control value (P < 0.05). lan diameter 22 ± 2% before and - 1 ± 1% after CHAPS; nitroprusside increased vessel diameter 19 ± 4% before and 21 ± 2% after CHAPS. After endothelial removal, the perfus- inhibition as well as the response to addition of the ate was replaced with the reconstituted blood solution, and corresponding diob, 1 1 , 1 2-DIHETE, are depicted in the response to 5,6-EET-Me or 1 1 , 1 2-EET was redetermined. Figure 2. Interbobular and afferent arteniobar diame- Statistics tens increased by 14 ± 2% and 15 ± 2%, respectively, in response to 1 1 , 1 2-EET ( 1 00 nM). 1 1 , 1 2-DIHETE Data are presented as mean ± SE. Significance of differ- (0. 1 to 100 nM) did not significantly affect the diame- ences in mean values for dose-response effect was evaluated tens of the intenbobubar and afferent arteriole. Acetyb- by analysis of variance for repeated measures followed by Duncan’s multiple range test. One-way analysis of variance choline ( 1 M) dilated interbobular and affenent ante- followed by the Bonferroni test was used to determine signif- nioles by 20 ± 2% and 24 ± 3%, respectively. icance of differences between control EET responses com- Endothebial removal by CHAPS did not alter the pre- pared with DIHETE responses and responses after endothe- gbomerular vascular response to 1 1 , 1 2-EET. After en- hal removal and cyclooxygenase inhibition. A value of P < dothelial removal, the responses to 1 1 , 1 2-EET were 0.05 was considered statistically significant. maintained and the diameter of the interbobubar and afferent arterioles increased by 15 ± 2% and 19 ± 3%, RESULTS respectively. In addition, the preglomerubar vasodila- Effect of Epoxygenase Metabolites on the tion to 1 1 , 1 2-EET was not altered by cycbooxygenase Preglomerular Vasculature inhibition indicating that this effect is not dependent Interlobular and afferent arteniolar responses to increasing concentrations of EET are presented in AFFF.RENT A.TEk1OLE INTERLORULAR ARTERIOLE

Figure 1 . 1 1 , 1 2-EET vasodibated the interbobubar and afferent arterioles in a dose-dependent manner. At a concentration of 1 00 nM 1 1 , 1 2-EET, the diameters of the interlobubar and afferent arterioles increased from 22 ± 1 to27 ± 1 mand 14 ± 1 to 17 ± 1 m, respectively. 14, 15-EET significantly increased diameters ofboth the interbobular and afferent arterioles by

OI1.I2UT Oia lEST 8 ± 3% and 1 1 ± 3%, respectively. Although not Scs*rs. in rrr OIAPS+I1.12 ttr _cvcLo. IIJZLET cyo.o+,ulrxT reaching statistical significance, the pregbomerubar I.I2D1IETZ #{149}II.I2DIHETT RIAcITYLOIOLU4S arteniolar diameter increased slightly to 8,9-EET. In contrast, 5,6-EET-Me resulted in a dose-dependent Figure 2. Effect of endothellal removal and cyclooxygenase constriction of interbobular and afferent arterioles. At inhibition on the preglomerular vascular response to 11,12- a concentration of 100 nM 5,6-EET-Me, interbobuban EET. Interlobular arteriolar diameter averaged 25 ± 1 Mm (N = 21) and afferent arteriolar diameter averaged 17 ± 1 and afferent arteniolar diameters decreased from 2 1 ± 2 to 17 ± 1 j.m and 16 ± 1 to 13 ± 1 m, respectively. Mm (N = 19) In the presence of norepinephrine (0.5 MM). After endothelial removal, interlobular and afferent arterioles

averaged 26 ± 2 Mm (N = 6) and 16 ± 2 Mm (N = 6), 1 1 , 12-EET Actions Are Endothelial and Cyclooxygenase Independent respectively. Interlobular diameter averaged 25 ± 2 Mm (N = 7) and afferent arteriolar diameter averaged 18 ± 2 Mm (N =

The pregbomerular vascular response to 1 1 , 12-EET 6) in the presence of indomethacin (10 MM). * indicates after endotheliab removal and during cycbooxygenase significant difference from control response (P < 0.05).

2366 Volume 7 - Number 1 1 ‘ 1996 Imig et al

AFFERENT ARTERIOLE INTERLOBULAR ARTERIOLE AFFERENT ARTERIOLE 1NTERLOBULAR ARTERIOLE 5 2C u.ns,arr D MEET-M, 0 54 EET-M. CHAPS+ S4EET-M, CRAPS + 5,6 LET-Ms EE11.12(R,S)-EET #{149}CYCLO+ 5.SEET-M. #{149}CYCLO + 84 LET-Ms . S4DIEETE U LUDEIIETE I- LU - * : -5 0 * 0 I- -1 z ______0 0 U U -15 -1

-lC. S 10” 10.10 i#{176} 1O hr7 IOU ir’5 10’ 10-’ -25- -2 CONCENTRATION (Ml CONCENTRATiON IM1 Figure 4. Effect of endothellal removal and cyclooxygenase

Figure 3. Effect of 1 1 , 12(R,S)-EET and 1 1 , 12(S,R)-EET on the inhibition on the preglomerular vascular response to 5,6-EET- afferent (left panel) and interlobular arteriole (right panel). Me. lnterlobular arteriolar diameter averaged 28 ± 1 Mm

Interlobular arteriolar diameter averaged 26 ± 1 Mm (N = 12) (N = 17) and afferent arteriolar diameter averaged 18 ± 1 and afferent arteriolar diameter averaged 16 ± 1 Mm (N = Mm (N = 20) in the presence of noreplnephrine (0.5 MM).

15) in the presence of norepinephrine (0.5 MM). * indicates After endothelial removal, interlobular and afferent arterioles significant difference from control value (P < 0.05). averaged 27 ± 2 Mm (N = 5) and 18 ± 1 Mm (N = 6),

respectively. Interlobular diameter averaged 25 ± 2 Mm (N =

6) and afferent arteriolar diameter averaged 18 ± 1Mm (N = on conversion and/or generation of cycbooxygenase 8) in the presence of indomethacin (10 MM). indicates metabobites. In the presence of indomethacin, 1 1 , 12- significant difference from control response (P < 0.05). EET increased the diameter of the intenbobular ante- niobe by 16 ± 2% and the afferent arteriole by 16 ± 3%. Interbobular and affenent arterioles were not signifi-

Vascular Response to 1 1, 12-EET is cantly decreased (3 ± 1 % and 3 ± 1 %) during super- Stereoselective fusion with 100 nM 5,6-DIHETE.

Figure 3 presents the effects of 1 1 , 1 2(R,S)-EET and Afferent Arteriolar Vasoconstriction to 1 1 , 1 2(S,R)-EET on the interbobular and afferent ante- 5,6-EET-Me Involves TP Receptor Activation nioles. 1 1 , 1 2(R,S)-EET resulted in dose-dependent vasodibation of the intenbobular and affenent arterioles Figure 5 presents the affenent arteriolar responses that was similar to that elicited by the nacemic mixture to 5,6-EET and U-46619 before and during TP recep- of 1 1 , 1 2-EET. The interbobular arteriobar diameter ton blockade. Afferent arteniobar diameter decreased increased from 25 ± 2 to 28 ± 2 m and affenent by 13 ± 3% and 9 ± 2% during superfusion of arteniolar diameter increased from 1 6 ± 1 to 1 9 ± 1 5,6-EET-Me (100 nM) and U-46619 (100 nM), nespec- m in response to superfusion of 1 1 , 1 2(R,S)-EET. In tively. The afferent arteniolar vasoconstriction to contrast, 1 1 , 1 2(S,R)-EET had a small vasoconstrictor U-466 1 9 was eliminated by the TP receptor blockade. effect which reached statistical significance at a con- During TP receptor blockade the afferent arteriolar centration of 100 nM. vasoconstniction to 5,6-EET was attenuated by 60% and the diameter decreased by 5 ± 2%. 5,6-EET-Me Actions are Endothelial and Cyclooxygenase Dependent DISCUSSION The pregbomerular vascular responses to 5,6-EET- A wide array of biologic actions have been ascribed Me after endothelial removal, during cycbooxygenase to the cytochrome P450 epoxygenase metabolites of inhibition and the vascular responses to the corre- arachidonic acid. Epoxygenase metabolites are pro- sponding diob, 5,6-DIHETE, are depicted in Figure 4. duced by renal microvessels (29) and tubular cells 5,6-EET-Me ( 100 nM) decreased the diameter of the (30), and the renal production ofEET appears to be via interbobubar and affenent arterioles by 16 ± 3% and the cytochnome P450 2C gene family (7,3 1 ). However, 1 7 ± 3% , respectively. Endothelial removal or cyclo- the isoforms responsible for endothelial production of oxygenase inhibition prevented or greatly attenuated EET have yet to be identified or cloned. Renal enzy- the preglomerulan vasoconstniction to 5 , 6-EET-Me. matic epoxygenase activity has been reported to in- After removal of the endothelium with CHAPS, the crease in some strains of rats on a high-salt diet (7,8) intenbobular and afferent arteriolar diameters de- but not in others (32), and it has been shown to be creased by 5 ± 1 % and 6 ± 1 % , respectively, In altered during the development of hypertension ( 1,6). response to 100 nM 5,6-EET-Me. In the presence of Furthermore, Its production in proximal tubules is indomethacin, 5,6-EET-Me decreased the interbobular increased by angiotensin II ( 1 4). Renal epoxygenase and affenent anteniolar diameters by 2 ± 4% and 2 ± metabolites have been implicated in the renal mainte- 2%, respectively. The pregbomerulan vasculature was nance of water and electrolyte balance (7,9, 12-14). In not significantly affected by the corresponding diol, this regard, EET have been reported to affect both 5,6-DIHETE (0. 1 to 100 nM), and diameters of the renal hemodynamics (3, 1 1 ,20,2 1 ) and tubular so-

Journal of the American Society of Nephrology 2367 Afferent Arteriolar Responses to EEl

AFFERENT ARTERIOLE tion and the decrease in gbomerular filtration rate were blocked by indomethacin (3). In the isolated perfused rabbit kidney preconstnicted with phenylephrine, the EJ 5,6 EET-Me negioisomens 5,6-EET, 8,9-EET, and 1 1 , 1 2-EET de- SQ29,548 + 5,6 EET-Me . U-46619 creased renal vascular resistance with 5,6-EET being SQ29,548 + U-466b9 the most potent ( 1 1). In this study 14, 15-EET elicited vasoconstriction, and the vasodibation to 5,6-EET was abolished by cyclooxygenase inhibition ( 1 1 ). Vaniabib- * ity in the reported responses to epoxygenase metabobites previously observed may depend on the species and vasculature studied, route of administration, or subsequent metabolism of the parent compound. Ep- C * oxygenase metabobites are produced by renal epithe- hal and endothebial cells and are released into the F- z- penicellular fluid to interact with the preglomerular C vasculature. In considering such a paracnine robe for C-) these vasoactive agents, the actions of epoxygenase .,0 0 metabobites were investigated during adventitial ad- -.1 ministration to the renal microvascubature. In this study, adventitiab administration of 1 1 , 1 2-EET and 14, 15-EET vasodibated and 5,6-EET resulted in vasoconstriction of the renal pregbomenular vasculatune. To determine whether the preglomerubar vascular Figure 5. Effect of TP receptor blockade on the afferent responses to 1 1 , 1 2-EET were dependent on an intact arteriolar response to 5,6-EET-Me. Afferent arteriolar diameter endothebium or conversion by cycbooxygenase, diam-

averaged 18 ± 1 Mm (N = 6) in the presence of norepineph- eter changes to 1 1 , 1 2-EET before and after indometh- rine (0.5 MM). In the presence of SQ 29,548 (5 MM) afferent acm or CHAPS administration were observed. Inter- arterioles averaged 18 1 Mm (N = 6), respectively. lobular and afferent artenlobar vasodibatory responses indicates significant difference from control response to the to 1 1 , 1 2-EET were not affected by the cyclooxygenase same agonist (P < 0.05). inhibitor, indomethacin. Endothebial removal with CHAPS had no effect on the preglomerular vasodiba- dium reabsorption ( 12, 14). This study directly exam- tony response to 1 1 , 1 2-EET. In addition, the predom- med the responses of the pregbomerular vasculature inant metabolite, 1 1 , 1 2-DIHETE, did not have a sig- to all four negioisomeric EET by using the in vitro nificant effect on the pregbomerular vasculature. blood-perfused juxtamedullary nephron preparation. These results suggest that 1 1 , 1 2-EET acts directly on Addition of 1 1 , 1 2-EET or 1 4, 1 5-EET to the superfu- vascular smooth muscle cells and does not act via sate vasodilated the interbobular and afferent ante- stimulation of cycbooxygenase metabolites nor is it noles. In contrast, 5,6-EET resulted in a vasoconstnic- metabolized to a vasoactive product via the cycbooxy- tion of the pregbomerular vascubature and 8,9-EET did genase pathway. In the isolated perfused rabbit kid- not significantly affect the renal microvascubature. ney, the renal vasodibation to 1 1 , 1 2-EET was also Vasoactivity of epoxygenase metabobites has been reported to be cycbooxygenase independent ( 1 1 ). Inhi- demonstrated in several species and vascular beds bition of cycbooxygenase with indomethacin or endo- ( 1 8,33-37). In the intestinal circulation 5,6-EET, 8,9- thebial removal did not alter the coronary artery vaso- EET, and 1 1 , 1 2-EET increased blood flow, whereas dilation to any of the regioisomeric EET (15). 14-15-EET did not alter blood flow (34). 5,6-EET was Additionally, production of EET by the coronary artery the only negioisomer to exhibit vasoactivity in the was increased in vessels with an injured endothebium isolated perfused rat tail artery and it resulted in suggesting that these metabobites may act to counter- vasodilation (33). Epoxygenase metabolites have also act the vascular effects of thromboxane ( 15). Consis- been demonstrated to be vasodibatory in the cerebral tent with the current observations of a direct effect of (36,37) and coronary circulations (15,35). In the kid- 1 1 , 1 2-EET on the vascular smooth muscle are recent ney, epoxygenase metabolites have been found to be findings demonstrating the presence of Ca2 -acti- either renal vasoconstrictive or vasodibatory depend- vated K channels in the preglomeruban vascubature ing on the experimental conditions and species stud- (22,29,38) and that 1 1 , 1 2-EET activates these chan- ied (3, 1 1 , 1 8). Takahasi et al. (3) infused 5,6-EET and nebs (22,35). Furthermore, the action of EET on yes- 8,9-EET into the renal artery of the rat and observed sels and Ca2-activated K channels are consistent an increase in renal vascular resistance, a decrease in with the possibility that these metabobites are endo- gbomerubar filtration rate, and no change in sodium thelium-denived hyperpobarizing factors (EDHF). In- excretion. The renal vascular effects of 5,6-EET and deed, epoxygenase metabobites have recently been 8,9-EET may be related to the formation of a cycboox- suggested to be EDHF (39,40,4 1 ), and EDHF appears ygenase compound, because the renal vasoconstric- to mediate a barge portion of the vasodilaton response

2368 Volume 7 . Number 1 1 ‘ 1996 Imig et al

to bradykinin in the coronary arteries (40) and iso- 5,6-EET, 5,6-epoxy-PGE1, and 5-hydnoxy-PGI1 were lated perfused kidneys (42). without affect ( 1 2). Differences in the nature of the Steneoselective production of epoxygenase metabo- cycbooxygenase dependency of 5,6-EET responses lites of arachidonic acid has been demonstrated in the may depend on the species studied, the cyclooxygen- kidney. In man, the predominant epoxygenase metab- ase metabolites produced by a cell type, and the action olite formed in the kidney is 14, 15-EET and 66% of of those metabolites. Because the processing of the this regioisomer is in the (R,S) configuration (5). Like- blood removed 95% of the platelets (23), it is unlikely wise 1 1 , 1 2(R,S)-EET is the major epoxide produced in that platelets are metabolizing 5,6-EET to a vasocon- the nat kidney (4). In this study the interbobular and stnictor metabolite in this study. Elimination of the afferent arteniolan vasodilation to 1 1 , 1 2-EET demon- renal vasoconstriction to 5,6-EET by endothebiab re- stnated steneosebectivity. The major epoxide produced moval provides evidence that the endothelium metab- in the rat kidney, 1 1 , 1 2(R,S)-EET, but not 1 1 , 1 2(S,R)- olizes 5,6-EET via the cycbooxygenase pathway EET, vasodibated the interbobulan and affenent ante- and/or is stimulated to produce a cyclooxygenase niobes. This is in agreement with recent data demon- vasoconstnictor metabobite. strating that 1 1 , 1 2(R,S)-EET vasodilated rat renal In summary, 1 1 , 1 2-EET and 1 4, 1 5-EET dilated arteries preconstnicted with phenylephnine and acti- whereas 5,6-EET constricted the preglomerulan ante- vated the large conductance Ca2-activated K chan- nioles of the juxtamedullany nephrons. The vasocon- nels in vascular smooth muscle cells isolated from stniction of the interbobular and affenent arterioles to these vessels (22). 5,6-EET was both cyclooxygenase and endotheliab Although the vasodibatory effects of 1 1 , 1 2-EET ap- dependent. In contrast, the renal vasodilation caused peared to be caused by a direct action on the vascular by 1 1 , 1 2-EET was not dependent on either cyclooxy- smooth muscle cells, the vasoconstnicton effects of genase metabolism on the endothelium. The direct 5,6-EET demonstrated a different pattern. The pneglo- vascular smooth muscle effect of 1 1 , 1 2-EET demon- menubar vasoconstriction to 5,6-EET was greatly at- strated steneoselectivity because the predominant rat tenuated by cycbooxygenase inhibition and after ne- kidney metabobite 1 1 , 1 2(R,S)-EET but not 1 1 , 1 2(S,R)- moval of the endothelium. The epoxide hydrolase EET was found to increase diameters of the intenbob- product 5,6-DIHETE did not significantly alter inter- ubar and afferent arterioles. This observation of a lobular or afferent anteniobar diameter. Previous stud- direct action of 1 1 , 1 2-EET on vascular smooth muscle ies have demonstrated that both the renal vascular cells and observations of previous studies demon-

(3, 1 1 ,2 1 ) and tubular ( 1 2) actions of 5,6-EET are strating that 1 1 , 1 2-EET increases activity of Ca2 - cyclooxygenase dependent. Carroll et al. ( 1 1 ) demon- activated K channels (22,36) provide evidence to stnated that the vasodilation to 5,6-EET in the isolated support the concept that an epoxygenase metabolite perfused rabbit kidney was abolished by cycbooxygen- of arachidonic acid may be EDHF. ase inhibition but was unaffected by the oxygen nadi- cal inhibitors, catabase and superoxide dismutase. ACKNOWLEDGMENTS Analysis of renal venous effluent material revealed The authors thank Anthony Cook and Paul Deichmann for excellent that the cyclooxygenase-dependent vasodilator activ- technical assistance with these experiments and Dr. Ed Inscho for ity of 5,6-EET is caused by release of prostaglandins helpful comments concerning portions of this manuscript. This work was supported by Grants DK-08676, GM-31278, and HL-18426 from (PG) E2 and 12 and metabolism of 5,6-EET to 5,6- the National Institutes of Health and a Grant-in-Aid from the Amen- xy1 (2 1 ). In the rat, cyclooxygenase inhibition can Heart Association. converted the 5,6-EET decrease in renal blood flow and gbomerulan filtration rate to a vasodibatory re- REFERENCES sponse and an increase in glomerulan filtration rate 1 . Gebremedhin D, Ma YH, Imig JD, Harder DR. Roman RI: (3). The vasoconstniction to 5,6-EET could have been Role of cytochrome P-450 in elevating renal vascular tone the result of the ability of cycbooxygenase to transform in spontaneously hypertensive rats. J Vase Res 1993;30: 53-60. 5,6-EET to a vasoconstnicton analog ofPGH2 or throm- 2. Knickle LC, Webb CD, House AA, Bend JR: Mechanism- boxane A2 (43). Indeed, results of this study demon- based inactivation of cytochrome P-450 1A1 by N-anal- strate that the affenent arteniolan vasoconstriction to kyl- 1 -aminobenzotriazoles in guinea pig kidney in vivo and in vitro: minimal effects on metabolism of arachi- 5,6-EET does require activation of the TP receptor. In donic acid by renal P450-dependent monooxygenases. a recent study, Fulton et a!. (44) demonstrated that J Pharmacol Exp Ther 1993;267:758-764. incubation of 5,6-EET with washed rat platelets yields 3. Takahashi K, Capdevila J, Karara A, Faick JR. Jacobson HR, Badr KF: Cytochnome P-450 arachidonate metabo- cycbooxygenase-dependent products which vasocon- bites in rat kidney: characterization and hemodynamic strict the isolated perfused rat kidney. The actions of responses. Am J Physiol 1990;258:F781-F789. 5,6-EET to inhibit sodium reabsorption in the cortical 4. Karara A, Makita K, Jacobson HR, Falck JR, collecting duct are also the result of cyclooxygenase Guengerich FP, DuBois RN, Capdevila JH: Molecular cloning, expression, and enzymatic characterization of activity ( 1 2). In this renal cell type, 5,6-EET stimu- the rat kidney cytochrome P-450 anachidonic acid epoxy- bated PGE2 synthesis and it was demonstrated that genase. J Biol Chem 1993;268: 13565-13570. PGE2 had actions similar to 5,6-EET on cortical col- 5. Karara A, Dishman E, Jacobson H, Faick JR. Capdevila J: Stereochemical analysis of endogenous epoxyeicosa- becting duct transepithebial voltage and intracellular tnienoic acids of human kidney cortex. J Biol Chem calcium levels, but the cyclooxygenase metabolites of 1 993;268:24543-24546.

Journal of the American Society of Nephrology 2369 Afferent Arteriolar Responses to EEl

6. Sacerdoti D, Abraham NG, McGiffJC, Schwartzman ML: cosatnienoic acids and hetenoatom analogs. Methods Renal cytochrome P450-dependent metabolism of ara- Enzymol 1990; 187:357-364. chidonic acid in spontaneously hypertensive rats. Bio- 26. Capdevila JH, Dishman E, Karara A, Falck JR: Cyto- chem Pharmacol 1988;37:521-527. chrome P450 anachidonic acid epoxygenase: Stereo- 7. Capdevila J, Wei S. Yan J, et at.: Cytochnome P-450 chemical characterization of epoxyeicosatnienoic acids. arachidonic acid epoxygenase: regulatory control of the Methods Enzymol 1 99 1 ;206:44 1-453. renal epoxygenase by dietary salt loading. J Biol Chem 27. Ma Y H, Gebremedhin D, Schwartzman ML, et aL: 1 992;267:2 1720-21726. 20-Hydnoxyeicosatetraenoic acid is an endogenous vaso- 8. Makita K, Takahashi K, Karara A, Jacobson HR, Faick constrictor of canine renal ancuate arteries. Cinc Res JR. Capdevila JH: Experimental and/on genetically con- 1993;72: 126-136. trolled alterations of the renal micnosomal cytochnome 28. Imig JD, Gebremedhin D, Zou AP, et aL: Formation and P450 epoxygenase Induce hypertension in rats fed a high actions of 20-hydroxyeicosatetnaenoic acid in the renal salt diet. J Clin Invest 1994;94:2414-2420. micnocincubation. Am J Physiol 1 996;270:R2 1 7-R227. 9. Takahashi K, Harris RC, Capdevila JH, et at.: Induction 29. Ogletree ML, Harris DN, Greenberg R, Haslanger MF, of renal anachidonate cytochrome P-450 epoxygenase Nakane M: Pharmacological actions of SQ 29,548, a after uninephrectomy: Counterregulation of hyperfiltna- novel selective thnomboxane antagonist. J Phanmacob tion. J Am Soc Nephrol 1993;3: 1496-1500. Exp Then 1985;234:435-441. 10. Imakoka S. Naka.mura M, Ishizaki T, et at.: Regulation 30. Schwartzman ML, Abraham NG, Carroll MA, Levere RD, of renal cytochnome P450s by thyroid hormone in dia- McGIffJC: Regulation ofarachidonic acid metabolism by betic rats. Biochem Pharmacol 1 993;46:2 197-2200. cytochnome P-450 in rabbit kidney. Biochem J 1986; 1 1 . Carroll MA, Garcia MP, Faick JR. McGiff JC: Cycbooxy- 238:283-290. genase dependency of the renovasculan actions of cyto- 3 1 . Capdevila 1m, Jacobson H, Zeldin D: Molecular cloning, chrome P450-derived arachidonate metabolites. J Phan- expression and enzymatic characterization of a human macol Exp Then 1992;260: 104-109. kidney cytochnome P450 anachidonlc acid epoxygenase 12. Sakairi Y, Jacobson HR. Noland TD, Capdevila JH, lAbstractl. J Am Soc Nephnol 1994;5:677. Fabck JR, Breyer MD: 5.6-EET inhibits ion transport in 32. Ma YH, Schwartzman ML, Roman RI: Altered nenal collecting duct by stimulating endogenous prostaglan- P-450 metabolism of arachidonic acid in DahI salt- din synthesis. Am J Physiol 1995;268:F931-F939. sensitive rats. Am J Physiol 1994;267:R579-R589. 1 3. Schbondorff D, Petty E, Oates JA, Jacoby M, Levine SD: 33. Carroll MA, Schwartzman M, Capdevila J, Falck JR, Epoxygenase metabobites of anachidonic acid inhibit va- McGiff JC: Vasoactivity of anachidonic acid epoxides. sopressin response In toad bladder. Am J Physiol 1987; EunJ Phanmacol 1987;138:281-283. 253:F464-F470. 34. Proctor KG, Faick JR, Capdevila J: Intestinal vasodila- 14. Madhun ZT, Goldthwait DA, McKay D, Hopfer U, Doug- tion by epoxyeicosatrienoic acids: Arachidonic acid me- las JG: An epoxygenase metabolite of arachidonic acid tabobites produced by cytochnome P-450 monooxygen- mediates angiotensin Il-Induced rises in cytosolic cab- ase. Cinc Res 1987;60:50-59. cium in rabbit proximal tubule epithebial cells. J Clin 35. Rosobowsky M, Campbell WB: Role of PGI2 and epoxyei- Invest l991;88:456-461. 15. Rosobowsky M, Campbell WB: Synthesis of hydroxyeico- cosatrienoic acids in relaxation of bovine coronary arteries to anachidonic acid. Am J Physiol 1993;264:H327- satnaenolc (HETEs) and epoxyeicosatrienoic acids (EETs) H335. by cultured bovine coronary artery endothelial cells. Biochim Biophys Acta 1996;1299:267-277. 36. Gebremedhin D, Ma YH, Roman RI, VanRollins M, 16. Pfister SL, Faick JR. Campbell WB: Enhanced synthesis Harder DR: Cellular mechanisms of cerebral epoxyeico- of epoxyeicosatrienoic acids by cholesterol-fed rabbit satnienoic acids on cerebral arterial smooth muscle. aorta. Am J Physiol 199 l;261:H843-H852. Am J Physiol 1992;263:H519-H525. 1 7. Moore SA, Spector AA, Hart MN: Eicosanoid metabolism 37. Ellis EF, Police RI, Yancey L, McKinney JS, Amruthesh in cenbromicnovascular endothelium. Am J Physiob SC: Dilation of cerebral arterioles by cytochrome P-450 1 998;254:C37-C44. metabobites ofanachidonic acid. Am J Physiol 1990;259: 18. Harder DR. Campbell WB, Roman RI: Role of cyto- Hi 171-Hi 172. chrome P-450 enzymes and metabolites of anachidonic 38. Gebremedhin D, Kaldunski M, Jacobs ER, Harder DR, acid in the control ofvascuban tone. J Vase Res 1995;32: Roman RI: Coexistence of two types of Ca2-activated 79-92. K channels in rat renal arterioles. Am J Physiob 1996; 19. Imig JD, Navar LG: Afferent arteniobar response to ara- 270:F69-F8 1. chidonic acid: Involvement of metabolic pathways. Am J 39. Hu 5, Kim HS: Activation of K channel in vascular Physiol 1996;271 :F87-F93. smooth muscles by cytochnome P450 metabolites of 20. Katoh T, Takahashi K, Capdevila J, et aL: Glomerular anachidonic acid. Eur J Pharmacol 1993;230:215-22i. steneospecific synthesis and hemodynamic actions of 40. Hecker M, Bara AT, Bauersachs J, Busse R: Character- 8,9-epoxyeicosatrienoic acid in the nat kidney. Am J ization of endothelium-denived hyperpobanizing factor as Physiol 1991;261:F578-F586. a cytochnome p450-derived anachidonic acid metabolite

2 1 . Carroll MA, Balazy M, Margiotta P, Faick JR, McGiffJC: in mammals. J Physiol 1 994;48 1:407-414. Renal vasodilaton activity of 5,6-epoxyeicosatnienoic acid 4 1 . Campbell WB, Gebremedhin D, Pratt PF, Harder DR: depends upon conversion by cycbooxygenase and release Identification of epoxyeicosatrienoic acids as endotheli- ofprostaglandins. J Biol Chem 1993;268: 12260-12266. um-denived hyperpolanizing factors. Cinc Res 1996;78: 22. Zou AP, Fleming JT, Falck JR. et at.: Stereospecific 415-423. effects of eoxyeicosatnienoic acids on renal vascular 42. Fulton D, McGiff JC, Quilley J: Contribution of nitric tone and K -channel activity. Am J Physiob 1996;270: oxide and cytochnome P450 to the vasodilaton effect of F822-F832. bradykinin in the rat kidney. Br J Pharmacol 1992; 107: 23. Imig JD, Gebremedhin D, Harder DR. Roman RI: Mod- 722-725. ulation of vascular tone in renal microcinculation by 43. Balazy M: Metabolism of 5,6-epoxyeicosatnienoic acid by erythnocytes: role of EDRF. Am J Physiob 1993;264: the human platelet: Formation of novel thromboxane H 190-H 195. analogs. J Biob Chem 199i;266:23561-23567. 24. Casellas D, Navar LG: In vitro perfusion of juxtamedul- 44. Fulton D, Balazy M, McGiff JC, Quilley J: Possible lary nephrons from rat. Am J Physiob 1984;246:F349- contribution of platelet cycbooxygenase to the renal vas- F358. cuban action of5,6 epoxyeicosatrienoic acid. J Phanmacol 25. Fabck JR, Yadagiri P, Capdevila J: Synthesis of epoxyei- Exp Then 1996;277: 1 195-i 199.

2370 Volume 7 ‘ Number 1 1 ‘ 1996