Prostaglandin E2 Abrogates Endothelin-induced Vasoconstriction in Renal Outer Medullary Descending Vasa Recta of the Rat

Transcription

1 Prostaglandin E2 Abrogates Endothelin-indued Vasoonstrition in Renal Outer Medullary Desending Vasa Reta of the Rat Enk P. Silldorff, Sai Yang, and homas L. Pallone Department of Mediine, Division of Nephrology, Pennsylvania State University, Hershey Medial Center, Hershey, Pennsylvania 1733 Abstrat Endothelins (E) and prostaglandin E2 are synthesized in the inner medulla by olleting dut epithelium and interstitial ells, respetively. All asending vasa reta (AVR) blood returns from the inner medulla to the ortex in outer medullary vasular bundles. We reasoned that hormones might influene medullary blood flow by diffusing aross AVR fenestrations to modulate vasoonstrition of outer medullary desending vasa reta (OMDVR). o investigate this possibility, OMDVR disseted from vasular bundles were exposed to E-1, 2, or 3. Eah endothelin isoform indued stable vasoonstrition with poteny, E-1 > E-2 > E-3 (EC5, 1.8 x 1-15, 5.9 x 1-12, and 8.8 x 1-1 M, respetively). he EA reeptor antagonists BQ-123 and BQ-61 (1-6 M), as well as an EA and EB reeptor antagonist ombination, attenuated vasoonstrition due to E-1 (1-12 M). BQ-123 had no effet on the response to E-3 (1 - M). he EB reeptor antagonist BQ-788 (1-6 M) attenuated the response to E-3 (11 M), but not that to E-1 (1-2 M). Finally, PGE2 (1-6 M) reversibly dilated OMDVR preonstrited with E-1 (112 M) or E-3 (1-8 M) but not E-1 (11 M). We onlude that E-1, 2, and 3 are potent onstritors of OMDVR and the response to E-1 is mainly EA reeptor subtype mediated, while E- 3 ats via the EB. PGE2 modulates E indued onstrition. hese findings are onsistent with interative feedbak and ontrol of medullary perfusion by loally synthesized hormones. (J. Clin. Invest : ) Key words: miroirulation * miroperfusion * videomirosopy * vasoonstrition * vasodilation Introdution Endothelins (E)' are potent vasoative peptides whose role in the ontrol of the systemi and renal miroirulations ontinues to be eluidated (1, 2). In their investigations of the sites of endothelin prodution, Kitamura et al. found the renal medulla to have the highest onentrations of endothelins (3). Address orrespondene to Erik P. Silldorff, Dept. of Mediine, Division of Nephrology, Hershey Medial Center, P.O. Box 85, 5 University Dr., Hershey, PA Phone: ; FAX: Reeivedfor publiation 22 July 1994 and in revisedform I February Abbreviations used in this paper: AVR, asending vasa reta; DVR, desending vasa reta; E, endothelins; OMDVR, outer medullary desending vasa reta. J. Clin. Invest. he Amerian Soiety for Clinial Investigation, In /95/6/2734/7 $2. Volume 95, June 1995, Substantial evidene has shown that endothelins are synthesized in the renal medulla. Prodution of E-1 and, to a lesser extent, E-3 has been demonstrated in the olleting tubule (4-6). All blood flow to the renal medulla is derived from the efferent flow of juxtamedullary glomeruli (7) so that synthesis of endothelins by glomerular endothelial, mesangial (8), or epithelial ells (9) might also supply hormone to the medulla. Endothelins are known to exert a variety of effets inluding mitogenesis ( 1), modulation of sodium transport ( 11 ) and vasoonstrition (12). Additionally, E stimulation of renal medullary interstitial ells leads to prodution of the abundant medullary eiosanoid, PGE2 ( 13). PGE2 is a renal vasodilator the blokade of whih leads to diminution of medullary blood flow (14-16) and potentiation of E-1-indued renal vasoonstrition (1). he unique anatomial arrangement of the vasa reta suggests the existene of a feedbak system for ontrol of the total and regional distribution of medullary blood flow. Outer medullary desending vasa reta (OMDVR) are ontratile mirovessel segments (17, 18) that arise from efferent arterioles ofjuxtamedullary glomeruli and subsequently traverse the inner stripe of the outer medulla in tightly paked vasular bundles (19). Remarkably, the asending vasa reta (AVR) in the bundles are only those whih arise from the inner medulla. AVR derived from reanastomosis of the apillary plexus in the outer medullary interbundle region return to the ortex without rejoining vasular bundles (19). Based on this anatomial arrangement, we hypothesize that hormones (e.g., E and PGE2) produed in the inner medulla might modulate the ontration of periytes on the ablumenal surfae of adjaent vasular bundle desending vasa reta (DVR) by rossing the fenestrated AVR wall (2). Detailed studies of the permseletive properties of AVR have not been performed, however, some experiments have indiated that maromoleules muh larger than endothelins an gain aess to the interstitium aross AVR fenestrations (21-23). hese onsiderations and the reent demonstration by polymerase hain reation of the mrna enoding the EA and EB reeptors in vasular bundles (24), have provided substantial motivation to perform a detailed investigation of the effet of the endothelin isoforms 1, 2, and 3 on OMDVR. hese studies demonstrate that Es are remarkably potent vasoonstritors. Additionally, we have determined that PGE2 an antagonize the effets of E-1 and E-3 in this important regulatory mirovessel segment verifying that loally produed hormones an modulate OMDVR vasomotor tone. Methods In vitro miroperfusion Detailed desriptions of the methods employed to perfuse OMDVR (17) and doument their ontratility (18) have been provided. he method is a minor variation of that originally desribed by Burg (25) for the study of renal tubules. Vasa reta harvested from the outer 2734 Silldorif et al.

2 medullary vasular bundles and perfused in vitro are exlusively OMDVR and not AVR whih are also present in vasular bundles ( 17 ). In brief, young female Sprague-Dawley rats (Harlan Sprague Dawley, In., Indianapolis, IN) were anesthetized by intraperitoneal injetion of thiopental (5 mg/kg) after whih the kidneys were harvested, slied and plaed into old (4C) Hepes buffer (5 mm Hepes, 14 mm NaCl, 1 mm NaAetate, 5 mm KC1, 1.2 mm MgCl2, 1.71 mm Na2HPO4,.29 mm NaH2HPO4, 1 mm CaCl2, 5 mm Alanine, 5 mm Gluose, and.5 g/dl Albumin, adjusted to ph 7.4 and bubbled with 1% 2). Vasa reta were disseted from vasular bundles. he bundles were loated by progressively peeling medullary tissue from the ortiomedullary juntion until vasa reta were enountered. Vasa reta are readily identifiable by their small diameter and irregular wall struture. One or two vessels were taken from eah rat within 2 h of death. he onstrution of pipettes has been desribed (17, 18). Holding pipettes have outside diameters of jum and inner onstritions < 9 jim. Perfusion pipettes have outside diameters of 6 jim. Colletion pipettes were maintained in the range of 6-12 jim and interhanged to ahieve an aeptable fit to the OMDVR segment under study. Harvested vessels were transferred to the stage of an inverted mirosope (Nikon diaphot) equipped with Differential Interferene Contrast (DIC, Nomarski) optis. After annulation, vessels were perfused at 37 C with the same Hepes buffer used for dissetion. he bath solution was idential to the perfusate exept for addition of hormones. he bath flow rate was 3 1l/min, delivered by a syringe pump. he addition of hormonal agents during experiments was ahieved by three rapid exhanges of the bath from a prewarmed syringe. Miromanipulators, perfusion and olletion apparatus and perfusion hamber were purhased from Instruments ehnology and Mahinery (San Antonio, X). emperature of the perfusion hamber was maintained at 37 C with a feedbak system employing a CN911 1A ontroller (Omega Engineering In., Stamford, C). Following a 3-min equilibration period, perfusions were standardized by the adjustment of olletion rate to the desired value (5 nl/min in most experiments). imed olletions of vessel effluent were obtained with volumetri onstrition pipettes after whih olletion rate was adjusted by hanging the pressure in the perfusion pipette. Driving pressure, one adjusted to obtain 5 nl/min olletion rate was maintained onstant throughout eah experiment. Continuous measurement of lumenal pressure in the miroperfused vessel has not been feasible, however, we have performed separate experiments that show lumenal pressures to be less than 15 mmhg (18). Videomirosopy and measurement of vessel diameters o evaluate the effets of vasoative agents on OMDVR diameters, miroperfusion experiments were reorded on video tape. he inverted mirosope was equipped with a 2/8% beam splitter and a side port for attahment of a video amera (CCD model 72; Dage-MI). During experimentation, OMDVR were observed with a 4x objetive to yield a final magnifiation of 1,3 on the video sreen. Experiments were reorded on a Panasoni model AG 196 VCR with a mirophone for audio reording of experimental events. During playbak, vessel internal diameters were measured by alipers at the point of greatest onstrition (following E administration). his method has been demonstrated to yield data indistinguishable from that obtained by more umbersome image analysis (18). Beause of the sensitivity and intensity of the vasopressor response to endothelins, the edge detetion algorithm of omputer imaging software was, in many ases, unable to distinguish lumenal edges during near 1% onstrition. he fous of the mirosope and adjustment of the differential interferene ontrast optis were optimized to yield maximal ontrast to the vessel wall. In this study, hanges in vessel inner diameter are expressed as perent onstrition, defined in terms of the basal diameter in the absene of hormones (Do) and the experimental diameter (D) by the expression, %Constrition = (1 - D/Do) x 1. Reagents Endothelins 1, 2, and 3, BQ-123, BQ-61 and IRL-138 were purhased from Peninsula Laboratories. RES-71-1 was obtained from Alexis and BQ-788 from Amerian Peptide Company (Sunnyvale, CA). Indomethain was purhased from Sigma. Peptides (exept IRL-138, RES-71-1, and BQ-788) were initially dissolved in deionized water. All pharmaologial agents were solubilized and frozen at -2'C in small aliquots of 1-'-1-5 M. Aliquots were thawed and diluted at least 1,-fold on the day of the experiment. Unused portions were disarded. PGE2 (Sigma) was dissolved in 1% ethanol and stored at -2'C at a onentration of 1-3M. IRL-138 was solubilized in 4-.1% trifluoroaeti aid and 6% aetonitrile; RES-71-1 was dissolved in 15% DMSO and BQ-788 was dissolved in 1% methanol. Following 2 1,-fold dilution of aliquots with buffer the solvents used did not produe vasoative effets themselves nor did they affet the ations of added hormones. Experimental protool Stability of OMDVR vasoonstrition by E-J, 2, and 3. After annulation, OMDVR were perfused for 3 min for equilibration. Subsequently E-1, 2, or 3 (1 1 M) was added to the bath. Lumenal diameter measurements were taken every minute for the first 5 min and every 5 min thereafter for an additional 3 min. At 3 min E was removed from the bath to observe reovery. After removal of hormone, reovery diameter measurements were taken every minute for 5 min and again at 1 min. Conentration dependene of vasoonstrition by E-J, 2, and 3. Following equilibration, E- 1, 2, or 3 was added to the bath in inremental onentrations. Due to their high poteny E- 1 and E-2 onentrations were inreased from 1-2 M to 1-8 M by 1-fold inrements. E-3 was inreased in the bath from 1-12 M to l-7 M by 1-fold inrements. Based upon the results of experiment 1 (see Results), 1 min were permitted to elapse following onentration hanges before videoreording for diameter measurement. he effets of transmural pressure on E-1 poteny in OMDVR were examined by generating trunated dose-response urves (1-16 M to 1-1 M) at differing olleting pressures ( and 1 mmhg). Similarly, the effet of bath and perfusate albumin onentration (.5 or 4. g/dl) on E-1 poteny was examined. 1 min were allowed between E-1 onentration hanges and measurements as previously desribed. Albumin at.5 g/dl was used in all other experimental proedures. Attenuation ofe-j and E-3 indued vasoonstrition by EA and EB reeptor antagonists. o determine the ontribution of the EA reeptor subtype to the responses of E-1 and E-3 we utilized the speifi EA reeptor antagonists, BQ-123 and BQ-61. Following ontrol measurements, either BQ-123 or BQ-61 (1-6 M) was added to the bath for ten minutes to allow reeptor binding and to determine possible agonisti ativity. Next, E-l (1-12 M) or E-3 (1-8 M) was added to the bath in the presene of antagonist. Finally, the antagonist was removed from the bath leaving E alone. As a time ontrol, hanges in lumenal diameter in the absene of antagonists were also doumented. Similar experiments were onduted to determine the ontribution of the EB reeptor to the vasoonstritor responses of E-1 and E-3 (a seletive EB agonist). he EB reeptor antagonists IRL-138 (1-5 M), RES-71-1 (1-6 M), and BQ-788 (1-6 M) were added separately to the bath before and during the presene of E-3 (1-8 M). In addition, the ability of BQ-788 (1-6 M) to attenuate E-l (1-1 M) indued onstrition was assessed alone and in onjuntion with BQ- 123 (an EA antagonist) in a manner similar to that used for other antagonists. Control experiments were performed in the absene of antagonist. Modulating effets of PGE2 on E-J- and E-3-indued vasoonstrition. 1 min after hormone addition, OMDVR onstrited with either E-1 (1 1 M) or E-3 (1-8 M) were videotaped for diameter measurement. Subsequently,.1% ethanol (the vehile for PGE2) was added to the bath and reording was repeated after five minutes. Next, PGE2 (1-6 M) was added in the presene of E-l or E-3. Diameters were again measured. Finally, PGE2 and vehile were removed from the bath leaving E alone. Diameters were reorded after 1 min. Several laboratories have doumented a lak of effet of PGE2 to Endothelin Ativity in Outer Medullary Desending Vasa Reta of the Rat 2735

3 C - L so,224 1 E- 1 a 4 o : Qo=2nl/min 2 i **-: Qo=5nl/mln 2 l so Fj to I i go W E E-3 61 L ~~~~~~ : I., ime (min) Figure 1. ime ourse of ation for E-l, 2, and 3. Perent onstrition is shown as a funtion of time following addition of E-l, 2, or 3 at 1`o M (n = 12, 1, and 15, respetively). he vertial dotted line at 3 min represents the removal of E from the bath. Initial olletion rates of 5 (solid irles) or 2 (open irles) nl/min were established before addition of hormone. Data are means±se o 8 * *-*: E-1 * * U-U: E-2 A-A: E-3 I AI /i/ E (log i) Figure 2. Conentration dependene of OMDVR vasoonstrition by E-1, 2, and 3. Perent onstrition is shown as a funtion of log-molar onentration of E-1, E-2, and E-3 (n = 7, 9, and 7, respetively). *Signifiant hange from OMDVR inner diameters prior to hormone addition (P <.5). (1-1 M, n = 12, 1, or 15, respetively) onstrited from a mean internal diameter of 9.9±2.2 gm to reah a minimum in - 1 min (Fig. 1). he onstrition was stable for the duration of exposure to E (3 min) and only moderately reversed 1 min following removal of E from the bath. E indued vasoonstrition was independent of initial perfusion rate (5 or 2 nl/min). Conentration dependene of E-1-, 2-, and 3-indued vasoonstrition. Signifiant onentration dependent vasoonstrition was observed with graded inrease in onentration of all three E isopeptides (P <.5, Fig. 2). By interpolation of the data, EC5 values for E-1, 2, and 3 are, 1.8 x 1-15 M (n = 7), 5.9 X 1-'2 M (n = 9), and 8.8 X 1` M (n = 7), respetively. he graded response to E-1 is reprodued in a representative photomirograph (Fig. 3). he vessel shown in Fig. 3 demonstrates diffuse onstrition in response to E-1, A abrogate severe renal vasoonstrition indued by high onentrations of E-l. PGE2 has, however, been effetive at submaximal onentrations of E-l (1). Based on those observations, we tested the effet of PGE2 on OMDVR preonstrited with 11o M E-l and 1-12 M E-1. o address the possible role of endogenous basal release of prostaglandins and/or E-l -stimulated prostaglandins in our preparation we attempted to potentiate preexisting E onstrition (E-l, 1-12 M) by adding indomethain (1-6 M) to the bath solution. o assure that timeourse of antagonist addition and prostaglandin halflife did not affet the results of aute indomethain addition to OMDVR, a separate group of rats was injeted with 5 mg/kg indomethain (intraperitoneal) 3 min before i.p. sarifie and subsequent exposure to E-l (1-12 M). Indomethain (1-6 M) was also added to both the bath and perfusate. hese results were ompared to those obtained from sham injeted rats. Statistial analysis Experimental results are reported as mean±se. Statistial omparisons employ a paired t-test or repeated measures analysis of variane (AN- OVA) as appropriate. For ANOVA, signifiane was determined by the Student-Newman-Keuls test. P values of less than.5 are onsidered signifiant. Results Stability of E-J-, 2-, and 3-indued vasoonstrition. OMDVR exposed to ablumenal appliation of E-1, 2, or 3 Figure 3. Representative photomirograph of graded OMDVR vasoonstrition in response to E-l. he same vessel photographed during ontrol (A), 1-16 M (B), l-'4 M (C), and 1-12 M (D) E-l. Diffuse onstrition ourred in this vessel. Foal onstrition was equally ommon in other OMDVR Silldorff et al.

4 1 * so8 i I E-1 (log i) Figure 4. Effet of varying olletion pressure (P) and albumin onentration on E-1 vasoativity. runated dose-response urves for E-1 are shown with P = mmhg and albumin =.5 g/dl (squares, n = 7), P = 1 mmhg and albumin =.5 g/dl (irles, n = 6), and P = mmhg and albumin = 4. g/dl (triangles, n = 7). No signifiant differenes are observed. however, foal onstrition at only a few loations along the vessel axis was equally ommon. Of the three isopeptides only E-3 was unable to ompletely obliterate the lumenal spae at maximally effetive onentration. Inreasing olletion pressure from mmhg (n = 7) to 1 mmhg (n = 6) had no effet on the response of OMDVR to E-1 ( 1-16 M to 1 1 M). Inreasing albumin onentration from.5 to 4. g/dl (n = 7) was also without effet (Fig. 4). Attenuation of E-J- and E-3-indued vasoonstrition by EA and EB reeptor antagonists. he EA reeptor antagonists BQ-123 and BQ-61 (1-6 M) signifiantly attenuated the response to E-1 ( 11` M, p <.5, Fig. 5, n = 5 and 8, respetively), however, BQ-123 failed to attenuate the effet of E-3 (1-8 M, Fig. 6, n = 6, BQ-61 not tested). he EB reeptor antagonist BQ-788 (1-6 M) was tested alone (n = 8) and in onjuntion (n = 1) with an EA reeptor antagonist (BQ-123, 1-6 M) to determine the possible ontribution of EB reeptor stimulation to the E-1 response. In both ases 12 -: BQ-75 (1pM)+BQ-123 (1MM) 1 A-A: SQ-123 (1MM) -:Q-61(1MM) 4 -:8-785(1MM) + o 8 *-:CONROL * -2 *-: Po OmmHg. ALB.5g/dl -: P=1OmmHg, ALJO.5g/dl A-A: P=OmmHg. ALB4.Og/dI ;S< 1_~ Control Antagonist r-i (u.1) E-i (.1mM) E-i (1mM) or SHAM w/wo antogonli Figure 5. Funtional antagonism of E- 1 by endothelin reeptor antagonists. Antagonism of E-1 mediated vasoonstrition by BQ-123 (open triangles, n = 5), BQ-61 (open irles, n = 8), BQ-788 (open diamonds, n = 8), and BQ-788 and BQ-123 (open squares, n = 1) is shown along with time ontrols in whih no antagonist was added (losed irles, n = 4). Blokade of E-l -indued vasoonstrition ourred with BQ-123, BQ-61, and BQ BQ-788 (*P <.5). Signifiant onstrition by E-1 ourred following removal of these same antagonists (+ P <.5). 4-._ U) Co o Control Antagonlst or E-3 (IOnM) E-3 (IOnM) SHAM w/wo antogonist Figure 6. Funtional antagonism of E-3 by endothelin reeptor antagonists. Perent onstrition indued by E-3 alone (solid irles, n = 6) and with BQ-123 (open diamonds, n = 6), IRL-138 (open squares, n = 5), RES-71-1 (open triangles, n = 6) or BQ-788 (open irles, n = 7). Blokade of onstrition ourring with BQ-788 (**P <.5) was followed by reovery (+P <.5). BQ-788 tended to attenuate E-1 -indued onstrition but the derease was not statistially signifiant (Fig. 5). o verify the ability of the vessels to respond to E-1 following effetive blokade, antagonist(s) were subsequently removed leaving E-1 (1-1 M) in the bath. Following the removal of antagonist(s), further E-1 indued onstrition ourred (P <.5) but the reovery was not omplete, suggesting a high affinity of the antagonist (s) for the reeptor(s) (Fig. 5). he EB reeptor antagonists IRL-138 ( 1-5 M) and RES (1-6 M) as well as the EA reeptor antagonist, BQ- 123 (1-6 M) failed to attenuate the response to E-3 (1-8 M, Fig. 6, n = 5, 6, and 6, respetively). In ontrast, the EB reeptor antagonist BQ-788 (1-6 M) signifiantly diminished the E-3 response (Fig. 6, n = 7). he antagonists did not alter baseline diameters signifiantly when administered alone (Figs. 5 and 6). Modulating effets of PGE2 on E-J- and E-3-indued vasoonstrition. Ablumenal appliation of PGE2 (1-6 M) to vessels preonstrited with E-1 (1-1 M, n = 6) had no vasodilatory effet (Fig. 7). However, vessels preonstrited by a lesser onentration of E-1 ( 1-12 M, n = 8) relaxed by 65% within 5 min of PGE2 appliation (P <.5, Fig. 7). Vessels preonstrited with E-3 (1-8 M, n = 9) were similarly dilated (63%) in the presene of PGE2. Removal of PGE2 from the bath leaving E-1 or E-3 alone resulted in signifiant reovery of onstrition. PGE2 was added to the bath by 1,- fold dilution of 1-3 M stok dissolved in ethanol. Vehile alone (ethanol) had no effet on vessels preonstrited with either E-1 or E-3 (Fig. 7). Endothelin-1 (1-12 M) indued onstrition was not altered by the addition of indomethain ( 1-6 M) to the bath (n = 8, Fig. 8). Similarly, onstrition was unaffeted in OMDVR of rats pretreated with indomethain (5 mg/kg, n = 8) when ompared to diluent injeted rats (n = 7, data not shown). Disussion Blood flow to the renal inner and outer medulla is derived exlusively from the efferent arterioles of juxtamedullary glo- Endothelin Ativity in Outer Medullary Desending Vasa Reta of the Rat 2737

5 ._ 1 1 -: E-1 (O.lnM) -: E-1 (1pM) 8 8 *-E*: E-3 (I(nM) 6 II -.-._ 6 -: E-1 sham exhanges (1 pm) -: experimental (E-1, 1pM) -.- U) 4 F U) 4 2 F + U *~~~~~ C. 2 I -2 Control E E + VEHICLE E + PGE2 E Figure 7. Modulating effets of PGE2 on E-1- and E-3-indued vasoonstrition. E-1 at 1-1 M (open irles, n = 6) and 1-12 M (losed irles, n = 8) and E-3 (losed squares, n = 9) before, during, and after addition of 1-6 M PGE2 to the bath. Signifiant vasodilation (*P <.5) and reovery of onstrition (+P <.5) were observed with E-3 and E-1 (1-12 M). -2 Control E E/I El/POE1 E/I Figure 8. Lak of endogenous prostaglandin effets in E-1 treated OMDVR. Indomethain (I) effets on E-l-indued onstrition and PGE2 dilation were examined (open irles, n = 8) and ompared to untreated OMDVR (n = 7). Signifiant vasodilation (*P <.5) ouffed with PGE2, followed by reovery of onstrition following removal (+P <.5). meruli. Aglomerular routes have largely been disounted (19). In the outer stripe of the outer medulla, juxtamedullary efferent arterioles give rise to OMDVR all of whih traverse the inner stripe in vasular bundles. Vasular bundles ontain OMDVR, AVR returning from the inner medulla and, depending upon the speies, the thin desending limbs of short looped nephrons (19). hus, OMDVR perfuse the entire medulla and are exposed to AVR blood returning from the inner but not the outer medulla of the kidney. OMDVR on the bundle periphery give rise to the apillary plexus that perfuses nephron segments in the outer medullary interbundle region of the inner stripe while those in the bundle enter ross the inner-outer medullary juntion to supply the inner medulla (19, 26). In the present ontext, two features of this miroanatomial arrangement are partiularly relevant. First, endothelin produed in the inner medulla an reah periytes on the ablumenal surfae of OMDVR by diffusing aross the walls of AVR in vasular bundles. Seond, if radial variation of OMDVR responsiveness to endothelin exists within the bundles, endothelins might play a role to determine the fration of juxtamedullary efferent flow that perfuses the outer vs inner medulla of the kidney. An integral assumption is that the AVR wall is permeable to endothelin (mol wt = 2492). Detailed investigations of AVR permseletivity have not been performed, however, histohemial traers as large as atalase and ferritin have been observed to ross AVR fenestrations (22, 23) and the refletion oeffiient of AVR to albumin is less than 1. (21). E-1 and E-3 produed by the inner medullary olleting dut probably diffuses into adjaent inner medullary AVR plasma. Reently, we found signifiant permeation of OMDVR by inulin (Pallone,. L., S. Nielsen, E. P. Silldorff, and S. Yang, manusript submitted for publiation) so that endothelins might also be expeted to enter DVR plasma. Based upon these onsiderations it seems reasonable to suggest that ounterurrent exhange and trapping of endothelin in the renal medulla ould our. he generation of endothelins in the inner medulla may exert diureti and natriureti effets through diret stimulation of tubular ells ( 1). Additionally, endothelins stimulate medullary interstitial ells to produe PGE2 (13), an autaoid whih has been shown to derease sodium reabsorption by the thik asending limb of Henle (27) and the olleting tubule (28). he ability of endothelins to onstrit OMDVR might at to abrogate these primary and seondary salureti stimuli. Inreases in renal perfusion pressure are known to indue natriuresis (termed pressure natriuresis). he detailed mehanisms responsible for pressure natriuresis have not been fully delineated. However, based on results showing that inreases in medullary interstitial pressure and pressure natriuresis are jointly eliminated by renal deapsulation (29), inreases in vasa reta hydrauli pressure and medullary interstitial pressure appear to be primary events. hrough their ability to onstrit OMDVR, endothelins might blunt inreases in medullary miroirulatory and interstitial hydrauli pressure thus reduing the ability of the kidney to undergo pressure natriuresis. he net effet on renal sodium handling of the tubular and vasular ations of endothelins is diffiult to predit. his is the first study to demonstrate that E-1, 2, and 3 vasoonstrit outer medullary desending vasa reta (OMDVR). In view of the ritial loation of these vessels and the high prodution (1) and onentration of endothelins in the renal medulla (3), pathophysiologial impliations may exist. Brezis and olleagues have demonstrated that the highly metaboli thik asending limb of Henles loop is partiularly vulnerable to ishemi insult (3). he medullary portion of the thik asending limb of Henles loop resides in the interbundle region of the outer medulla and is therefore supplied with nutrients and oxygen by OMDVR on the vasular bundle periphery. Marked vasoonstrition of those vessels by endothelins or other hormones might play a role in the indution or exaerbation of ishemi aute renal failure. In support of this hypothesis, Chan et al. and Gellai et al. demonstrated abrogation of ishemi aute renal failure in rats treated with the EA reeptor bloker, BQ-123 (31, 32). Reognizing this is the first study of the ations of endothelins on OMDVR, we metiulously haraterized the effet of eah isopeptide and the reeptor subtypes whih mediate vasoonstrition. Maximal vasoonstrition ours within minutes after appliation of endothelins (Fig. 1) and these hormones have a remarkable poteny (threshold effet 118 M, Fig. 2). It should be reognized, however, that our observations with the in vitro perfused OMDVR provides only a system for bioassay of onentration dependent hormonal effets on vasoon Silldorif et al.

6 strition. As suh, the intense vasoonstrition observed in vitro might not exist in vivo even if loal endothelin onentrations are as high as those employed in our experiments. he absene of modulating vasodilators probably aentuates the apparent response in the in vitro perfused mirovessel preparation. he threshold onentrations apable of eliiting OMDVR vasoonstrition in our preparation are well below typial plasma levels of immunoreative endothelins. his implies, first, that endothelins ould ontribute signifiantly to basal OMDVR vasomotor tone in vivo and, seond, that modulating endogenous vasodilators may play important roles to maintain and modulate renal medullary perfusion. In support of this onept, several authors (14-16) have shown that ylooxygenase inhibition diminishes medullary blood flow in vivo. We demonstrated that E-1, at least in part, ats on OMDVR through stimulation of an EA reeptor subtype while E-3 appears to at through the EB reeptor (Figs. 5 and 6). It is generally aepted that E-1 ativates EA or EB while E-3 ation ours by stimulation of EB alone. he EB reeptor subtype has been loalized to the vasular endothelium where it mediates the seretion of nitri oxide (33) and to smooth musle where it mediates vasoonstrition (12). he inability to attenuate the response of E-3 with either IRL-138 or RES may indiate a lak of affinity or speifiity of these antagonists for the EB reeptor subtype in the rat (34). Vasoonstrition of the renal mirovasulature by endothelins has previously been demonstrated by a number of authors. Edwards et al. showed that rabbit efferent and afferent arterioles - onstrit in response to E-1 and E-2 (ECm 1 nm) while E-3 is signifiantly less potent (35). Similarly, in the hydronephroti rat kidney both efferent and afferent arterioles onstrited in response to topial or lumenal administration of Es, however, the afferent arteriole may be more sensitive (36-38). Other investigators have demonstrated the ability of E to alter glomerular filtration rate and renal vasular resistane in vivo (1). Wilkes et al. demonstrated stimulation of PGE2 prodution in renal medullary interstitial ells by E-1, but not by E-3 (13). his suggests the existene of a mehanism to modulate the ations of E through opposition by PGE2. In our studies, PGE2 reversibly dilated OMDVR preonstrited with 112 M E-1 or E-3, but not 1 1 M E-1. hese results are onsistent with the finding that ylooxygenase inhibition potentiates mild to moderate, but not severe renal vasoonstrition indued by E-1 (1). he differing sensitivities of E-1 - and E-3- indued onstrition to PGE2 might imply that the mehanism of interation between PGE2 and the ellular signal generated by E-1 and E-3 are different. Our experiments using indomethain in isolated OMDVR demonstrate that endogenous ylooxygenase produts do not signifiantly affet the poteny of endothelin or that adequate substrate is laking in this in vitro perfused preparation. In summary, this study provides the first desription of vasoonstrition of OMDVR by endothelins. All E isopeptides are extremely potent and indue stable vasoonstritive responses. he reeptor mediating the response to E-l is primarily EA, while the E-3 response is generated via the EB subtype. he known stimulation of PGE2 prodution in the renal medulla by E-1 (13), oupled with the ability of PGE2 to modulate OMDVR vasoonstrition by E-1 or E-3 suggests that a system may exist for feedbak regulation of medullary perfusion by these autaoids. hat feedbak system might well be influened by other vasoative agents that are present in the renal medulla. Aknowledgments his work was supported by National Institutes of Diabetes and Digestive and Kidney Diseases Grant DK42495 and a Grant-in-Aid from the Amerian Heart Assoiation. hese studies were performed during the tenure of an Established Investigatorship of the Amerian Heart Assoiation (. L. Pallone). Referenes 1. Kohan, D. E Endothelins in the kidney: physiology and pathophysiology. Am. J. Kid. Dis. 22: Simonson, M. S Endothelins: Multifuntional renal peptides. Physiol. Rev. 73: Kitamura, K.,. anaka, J. Kato,. Eto, and K. anaka Regional distribution of immunoreative endothelin in porine tissue: abundane in inner medulla of kidney. BiohenL Biophys. Res. Commun. 161: Kohan, D. E., and F.. J. Fiedorek Endothelin synthesis by rat inner medullary olleting dut ells. J. Amn So. Nephrol. 2: Ujiie, K., Y. erada, H. Nonoguhi, M. Shinohara, K. omita, and F. Marumo Messenger RNA expression and synthesis of endothelin-i along rat nephron segments. J. Clin. Invest. 9: Uhida, S., F. akemoto, E. Ogata, and K. Kurokawa Detetion of endothelin-1 mrna by R-PCR in isolated rat renal tubules. Biohem. Biophys. Res. Commun. 188: Moffat, D. B he Mammalian Kidney. Cambridge Univ. Press, London Kohan, D. E Prodution of endothelin-1 by rat mesangial ells: regulation by tumor nerosis fator. J. Lab. Clin. Med. 119: Kasinath, B. S.,. A. Fried, S. Davalath, and P. A. Marsden Glomerular epithelial ells synthesize endothelin peptides. AnL J. Physiol. 141: Simonsou, M. S., S. Wann, P. Mene, G. R. Dubyak, M. Kester, Y. Nakazato, J. R. Sedor, and M. J. Dunn Endothelin stimulates phospholipase C, Na/H exhange, -fos expression, and mitogenesis in rat mesangial ells. J. Clin. Invest. 83: Perio, N., R. P. Comejo, A. Benigni, B. Malanhini, J. R. Ladny, and G. Remuzzi Endothelin indues diuresis and natriuresis in the rat by ating on proximal tubular ells through a mehanism mediated by lipoxygenase produts. J. AnL So. Nephrol. 2: Summer, M. J.,. R. Cannon, J. W. Mundin, D. G. White, and I. A. Watts Endothelin EA and EB reeptors mediate vasular smooth musle ontration. Br. J. Pharmaol. 17: Wilkes, B. M., A. S. Ruston, P. Mento, E. Girardi, D. Hart, M. Vander Molen, R. Barnett, and E. P. Nord Charaterization of endothelin-l reeptor and signal transdution mehanisms in rat medullary interstitial ells. AnL J. Physiol. 29:F579-F Solez, K., J. A. Fox, M. Miller, and R. H. Heptinstall Effets of indomethain on renal inner medullary plasma flow. Prostaglandins. 7: Cupples, W. A.,. Sakai, and D. J. Marsh Angiotensin II and prostaglandins in ontrol of vasa reta blood flow. Am. J. Physiol. 254:F779- F Lemley, K. V., S. L. Shmitt, C. Holliger, M. J. Dunn, C. R. Robertson, and R. L. Jamison Prostaglandin synthesis inhibitors and vasa reta erythroyte veloities in the rat. AnL J. Physiol. 247:F562-F Pallone,. L., J. Work, R. L. Myers, and R. L. Jamison ransport of sodium and urea in outer medullary desending vasa reta. J. Clin. Invest. 93: Pallone,. L Vasoonstrition of outer medullary vasa reta by angiotensin H is modulated by prostaglandin E2. An. J. Physiol. 266:F85-F Lemley, K. V., and W. Kriz Cyles and separations: the histotopography of the urinary onentrating proess. Kidney Int. 31: Shwartz, M. M., M. J. Karnovsky, and M. A. Venkatahalam Ultrastrutural differenes between rat inner medullary desending and asending vasa reta. Lab. Invest. 35: Pallone,. L Moleular sieving of albumin by the asending vasa reta wall. J. Clin. Invest. 9: Venkatahalam, M. A., and M. J. Karnovsky Extravasular protein in the renal medulla. Lab. Invest. 27: Shimamura,., and A. B. Morrison Vasular permeability of the renal medullary vessels in the mouse and rat. Am. J. Pathol. 71: erada, Y., K. omita, H. Nonoguhi, and F. Marumo Different loalization of two types of endothelin reeptor mrna in mirodisseted rat nephron segments using reverse transription and polymerase hain reation assay. J. Clin. Invest. 9: Endothelin Ativity in Outer Medullary Desending Vasa Reta of the Rat 2739

7 25. Burg, M. B Perfusion of isolated renal tubules. Yale J. Biol. Med. 45: Pallone,. L., C. R. Robertson, and R. L. Jamison Renal medullary miroirulation. Physiol. Rev. 7: Stokes, J. B Effet of prostaglandin E2 on hloride transport aross the thik asending limb of Henle. Seletive inhibition of the medullary portion. J. Clin. Invest. 64: Stokes, J. B., and J. P. Kokko Inhibition of sodium transport by prostaglandin E2 aross the isolated perfused olleting tubule. J. Clin. Invest. 25: Roman, R. J., and A.-P. Zou Influene of the renal medullary irulation on the ontrol of sodium exretion. Am. J. Physiol. 265:R963-R Brezis, M., S. Rosen, and P. Silva Seletive vulnerability of the medullary thik asending limb to anoxia in the isolated perfused rat kidney. J. Clin. Invest. 73: Chan, L., A. Chittinandana, J. I. Shapiro, P. F. Shanley, and R. W. Shrier Effet of an endothelin-reeptor antagonist on ishemi aute renal failure. Am. J. Physiol. 266:F135-F Gellai, M., M. Jungus,. Flether, R. DeWolf, and P. Nambi Reversal of postishemi aute renal failure with a seletive endothelina reeptor antagonist in the rat. J. Clin. Invest. 93: De Nui, G., R. homas, P. D'Orleans-Juste, E. Antunes, C. Walder,. D. Warner, and J. R. Vane Pressor effets of irulating endothelin are limited by its removal in the pulmonary irulation and by the release of prostaylin and endothelium-derived relaxing fator. Pro. Natl. Aad. Si. USA. 85: Urade, Y., Y. Fujitani, K. Oda,. Watakabe, I. Umemura, M. akei,. Okada, K. Sakata, and H. Karaki Retration onerning an endothelin B reeptor-seletive antagonist: IRL 138, [Cysl"-Cys"5]-endothelin-1 (11-21). FEBS (Fed. Eur. Biohem. So.) letters 342: Edwards, R. M., W. rizna, and E. H. Ohlstein Renal mirovasular effets of endothelin. Am. J. Physiol. 259:F217-F Loutzenhiser, R., M. Epstein, K. Hayashi, and C. Horton Diret visualization of effets of endothelin on the renal mirovasulature. Am. J. Physiol. 258:F61 -F Bloom, I.. M., F. R. Bentley, M. A. Wilson, and R. N. Garrison, In vivo effets of endothelin on the renal miroirulation. J. Surg. Res. 54: Fretshner, M., K. Endlih, E. Gulbins, R. E. Lang, K. Shlottmann, and M. Steinhausen Effets of endothelin on the renal miroirulation of the split hydronephroti rat kidney. Renal Phys. Biohem. 14: Silldorif et al.