Increased Glomerular Thromboxane Synthesis as a Possible Cause of Proteinuria in Experimental Nephrosis

Transcription

1 nreased Glomerular Thromboxane Synthesis as a Possible Cause of Proteinuria in xperimental Nephrosis G. Remuzzi, L. mberti, M. Rossini, C. Morelli, C. Carminati, G. M. Cattaneo, and T. Bertani Mario Negri nstitute for Pharmaologial Researh, Via Gavazzeni 11, Bergamo, and Division of Nephrology and Nulear Mediine, Ospedali Riuniti, Bergamo, taly Abstrat Altered glomerular metabolism of arahidoni aid (AA) has already been demonstrated in experimental nephrotoxi nephritis. The enhaned synthesis of thromboxane A2 (TxA2) in isolated glomeruli that has been found may mediate hanges in renal hemodynamis. The objetives of this investigation were: to hek whether glomerular AA metabolism is also altered in a model of glomerulopathy in whih no leukoyte infiltration or platelet deposition ould be demonstrated; to establish a orrelation between the altered AA metabolism and proteinuria; and to explore whether the alteration of the prostaglandin (PG) pathway found in isolated glomeruli is an in vitro artifat or reflets a modifiation in vivo. We used a model of glomerular damage haraterized by heavy and persistent proteinuria, whih was indued in the rat by a single intravenous injetion of adriamyin. At light mirosopy, minimal glomerular abnormalities were found in this model. letron mirosopy showed profound alterations of glomerular epithelial ells with extensive fusion of foot proesses and signs of epithelial ell ativation. letron mirosopy of numerous glomeruli showed no platelet deposition or marophage and leukoyte infiltration in this model. solated glomeruli from nephroti rats studied 14 or 3 d after a single intravenous injetion of adriamyin (7.5 mg/kg) when animals were heavily proteinuri generated signifiantly more TxB2, the stable breakdown produt of TxA2, than normal glomeruli. No signifiant hanges were found in the other major AA metabolites formed through ylooxygenase. Urinary exretion of immunoreative TxB2 was also signifiantly higher in nephroti than in normal animals. Administration of a seletive Tx synthetase inhibitor, UK-38,485, from day 14 to day 18 after adriamyin resulted in a signifiant redution of proteinuria ompared with pretreatment values. Glomerular synthesis and urinary exretion of TxB2 were normal during the UK-38,485 treatment. Additional experiments showed that elevated glomerular synthesis and urinary exretion of TxB2 were not a onsequene of inreased substrate availability. Maximal stimulation of the renin-angiotensin axis with furosemide inreased glomerular TxB2 synthesis in normal rats, whih was signifiantly lower than in nephroti animals. Finally, experiments using a unilateral model of adriamyin nephrosis indiated that the enhanement of glomerular TxB2 This work was presented in part at the Amerian Soiety of Nephrology Annual Meeting in Washington, DC, Reeived for publiation 23 January 1984 and in revised form 23 July J. Clin. nvest. The Amerian Soiety for Clinial nvestigation, n /85/1/94/8 $ 1. Volume 75, January 1985, synthesis is not simply a onsequene of the nephroti syndrome. We onlude that: there is an abnormality of glomerular AA metabolism in nephroti syndrome, whih leads to inreased TxA2 prodution; the inreased Tx generation orrelates with protein exretion and might be responsible for altering the glomerular basement membrane permeability to protein; and the alteration found in isolated glomeruli probably reflets a modifiation in vivo, in that urinary exretion of immunoreative TxB2 is also onsistently inreased in adriamyin nephrosis. ntrodution After the initial observation by Smith and Bell (1) that normal glomerular tissue synthesizes prostaglandins (PG),' several efforts were made to define the role of arahidonate metabolites of glomerular origin in normal onditions and disease. The majority of studies found PGF2,, and PG2 to be the predominant glomerular PG in rats, though signifiant amounts of thromboxane A2 (TxA2) and prostaylin (PG2) ould also be deteted (2-5). Glomeruli that were isolated from rats with glyerol-indued aute renal failure synthesized larger amounts of PG2, PGF2<, and TxB2 than appropriate ontrols (6). t is tempting to speulate that these vasoative substanes have a pathogeneti role in mediating the alterations in regional blood flow typial of aute renal failure. More reently, Lianos et al. (7) and Dunn (8), using the nephrotoxi nephritis model in rats indued by a single injetion of anti-glomerular basement membrane antibody, found that isolated glomeruli from these animals synthesized Tx at 1 times the rate of ontrol glomeruli. These authors made the important observation that the inreased glomerular Tx synthesis orrelates linearly with the extent of proteinuria. Whether this reflets a ause and effet relationship, or both glomerular Tx prodution and proteinuria are two independent markers of glomerular injury, is open to speulation. n order to larify this issue we studied a model of glomerular damage, whih was haraterized by heavy and persistent proteinuria, indued in the rat by a single intravenous injetion of adriamyin (9). We report here a signifiant inrease of both glomerular synthesis and urinary exretion of TxB2 as the obvious alterations of renal arahidoni aid (AA) metabolism in rats affeted by adriamyin nephrosis. The inrease in TxB2 synthesis is onomitant with the appearane of proteinuria and reahes its maximum at the time of peak proteinuria. Administration of a seletive Tx synthetase inhibitor to nephroti animals signifiantly redues the exaggerated renal TxB2 prodution and lowers proteinuria in all animals. 1. Abbreviations used in this paper: AA, arahidoni aid; KRB, Krebs- Ringer phosphate buffer; PG, prostaglandin(s); PGD2, PG2, 6-Keto- PGF1,,, PGF2., PG2, prostaglandins PGD2, PG2, 6-Keto-PGF,, PGF2<, and PG2; TLC, thin-layer hromatography; Tx, thromboxane; TxA2, TxB2, thromboxanes TxA2 and TxB2. 94 Remuzzi, mberti, Rossini, Morelli, Carminati, Cattaneo, and Bertani

2 We suggest that the mehanism of proteinuria, at least in this experimental model, depends on inreased glomerular Tx synthesis, whih probably mediates an alteration in the glomerular barrier to proteins. Methods Materials. Adriamyin (Adriblastina) was a gift from Farmitalia Carlo rba (Milan, taly). UK-38,485, 3-(lH-imidazol-l-yl-methyl)-2-methyl- H-indole-l-propanoi aid) was kindly provided by Dr. H. M. Tyler, (Pfizer Ltd., Sandwih, United Kingdom). UK-38,485 was dissolved in.1 N sodium hydroxide and the ph was adjusted to 8.5 with.1 N hydrohlori aid. Furosemide (Lasix) was purhased from Hoehst AG (Frankfurt, Federal Republi of Germany), sulinda (Clinoril) from Merk Sharp & Dohme (Rome, taly). nulin was purhased from. Merk (Darmstadt, Federal Republi ofgermany). AA, bovine serum albumin (BSA), and PG standards for thin-layer hromatography (TLC) studies were obtained from Sigma Chemial Co. (St. Louis, MO). AA was dissolved in 1 mm sodium arbonate, ph 1, and kept under nitrogen. '4C-AA (57.6 mci/mmol), 3H-TxB2 (15 Ci/ mmol), 3H-6-Keto PGFj,, (13 Ci/mmol), and 3H-PG2 (165 Ci/ mmol) were purhased from New ngland Nulear (Boston, MA). TLC plates (Silia Gel 6 F-254) were a gift from. Merk. Animal experiments. Sprague-Dawley CD-COBS male rats (Charles River Breeding Laboratories n., Wilmington, MA), weighing g at the start of the experiment, were used. Nephroti syndrome was indued by a single intravenous injetion of adriamyin (7.5 mg/ Kg), whih orresponded to 15 mg/square meter through the tail vein of nonanesthetized animals aording to a method previously desribed (9). Controls rats reeived the solvent alone. The Tx synthetase inhibitor, UK-38,485, was injeted (2 mg/kg) intraperitoneally t.i.d. for five onseutive days. 4 h after the last injetion the kidneys were removed. Seleted studies were performed in animals with the left kidney lamped for 5 min just before intravenous adriamyin. Furosemide (12.5 mg/kg) was administered subutaneously twie a day for 1 d. Sulinda (6 mg/kg) was given as a single oral dose daily for three onseutive days. For the determination of inulin learane animals were anesthetized by intraperitoneal injetion of natin (16 mg/1 g of body weight), plaed on a onstant temperature table, and traheotomized. 2 ml of inulin (5% in isotoni saline solution) was injeted as priming dose, followed by onstant infusion at 2.7 ml/h. A 6-min period of onstant infusion was allowed before initiation of learanes. Urine was olleted via a polyethylene tube inserted in the bladder.. 1-ml blood samples were obtained from a polyethylene tube (P-5) inserted in the femoral artery. Plasma and urine inulin were determined using olorimetri assay adapted for miroliter samples. Glomerular isolation. Treated and ontrol rats were anesthetized with ether and a annula was inserted into the lower aorta. Subsequently, the aorta was lamped above the renal arteries, the renal veins were ut, and the kidneys were perfused with 6-8 ml of Krebs-Ringer phosphate buffer (KRB) with the following omposition: mm NaCl, 4.82 mm KC, 1.21 mm MgSO4, 2.57 mm CaCl2, and 15.6 mm phosphate buffer (ph 7.4). When kidney surfaes were ompletely blanhed, the kidneys were removed and immediately immersed in ie-old KRB. All subsequent preparative steps were performed at 4 C. The renal apsules were removed and the ortex was separated from the medulla. The ortial segments were finely mined into 1-2- mm fragments and the glomeruli were separated from residual ortial tissue by passing the homogenate through different size sieves and pressing the preparation through a 16 um sieve to exlude the tubules, and a 75-pm sieve to retain the glomeruli. The sieved tissue, suspended in KRB, ph 7.4, was passed through a 25-gauge needle and entrifuged at 14 g for 12 s. The supernatant was disarded, the pellet resuspended in the same buffer solution, passed again through the needle, and entrifuged. The purity of isolated glomeruli was determined mirosopially by ounting the number of glomerular and nonglomerular partiles suspended in a given volume. The final pellet onsisted of deapsulated glomeruli with <2% tubular ontamination. letron mirosopy of the normal isolated glomeruli showed well-preserved ultrastruture with intat endothelial lining, basement membranes, and distint epithelial foot proesses. The glomeruli of treated and ontrol rats were always prepared in parallel. nubations. Freshly prepared glomeruli from ontrol and treated rats were resuspended in ml of KRB at 37C in room atmosphere in a Dubnoff shaking water bath (12 yles/min). nubation was usually stopped after 6 min by entrifugation at 3, g for 5 min and the supernatant of eah tube was olleted and frozen until PG radioimmunoassay. n seleted experiments, the isolated glomeruli were inubated in presene of AA, 1 and 2 AM, final onentration. Protein onentration was determined aording to the method of Lowry et al. (1) with BSA as a standard. The onentration of glomerular proteins varied in different experiments (2-5 pg/ml) but was kept onstant within the same experiment. For in vitro experiments, isolated glomeruli from normal rats were inubated with adriamyin 2 pg/ml and mg/ml for 5 and 3 min. Sample preparation. Serum for studying platelet TxB2 prodution in response to endogenous thrombin was obtained by leaving multiple 1-ml aliquots of native blood, olleted by intraardia punture from ether-anesthetized animals, at 37C for 3 min, after the method previously reported (11). The prepared sera were frozen and kept at -2'C until assayed. Urines were olleted using metaboli ages over a 24-h period, and their protein ontent was determined by the sulfosaliyli aid method. Serum and urine reatinine were determined by the method of Hare (12) using a Bekman analyzer (Astra 4 model, Bekman nstruments n., Fullerton, California). Creatinine and inulin learanes were alulated in the usual manner. Urinary TxB2, 6-Keto-PGF., and PG2, whih, within limits, reflet the renal synthesis of TxA2, PG12, and PG2, respetively, (13, 14) were measured by radioimmunoassay after extration. Renal tissue speimens were obtained from kidney biopsies. For light mirosopy, fragments of the ortex were fixed in Dubosq-Brazil fluid (8% alohol, 15 ml; formol, 6 ml; aeti aid, 15 ml; piri aid, g) and embedded in paraffin. Setions of 2 pm were stained with Masson's trihrome, hematoxylin and eosin, periodi aid staining (periodi aid-shiff), and Wilder's retiulin. For eletron mirosopy, small piees of the ortex were fixed with phosphate-buffered 2.5% glutaraldehyde (ph 7.2) for 6 h, then rinsed in.2 M aodylate buffer (ph 7.4). Subsequently the samples were postfixed in 1% osmium tetroxide at 4 C for h, washed in buffer, and immered in.5% Veronal uranyl aetate. Samples were then dehydrated through graded alohol and embedded in Spurr resin. Ultrathin setions (6-8 nm) were ut on a Reihert Om A-2 ultramirotome and examined with a Zeiss M 19. Assays. Radioimmunoassay: unextrated supernatants of glomerular preparations and extrated urinary samples were assayed for TxB2, 6- Keto-PGF., the stable breakdown produt of PG12, and PG2. Glomerular supernatants were diluted in 5 mm phosphate buffer (ph 7.4) and assayed in a volume of 1.5 ml, at a final dilution of 1:12-1:3 for TxB2, 1:3 for 6-Keto-PGF., and 1:3 for PG2. PG were measured in extrated urinary samples diluted 1:3-1:12. 5, dpm of 3H-TxB2 or 3H-6-Keto-PGF1. or 3H-PG2 and appropriately diluted anti-txb2 (final dilution 1:1,,) or anti-6-keto-pgf,, (final dilution 1:35,) or anti-pg2 sera (final dilution 1:5,) were mixed and added in a volume of 1.25 ml to eah assay tube. The smallest onentrations that ould be measured with 95% onfidene were 5 pg/ml for TxB2, 1 pg/ml for 6-Keto-PGFa, and 12.5 pg/ ml for PG2. The detetion limits were.6 ng for TxB2, 3 ng for 6- Keto-PGF1., and 375 pg for PG2 per milliliter of glomerular supernatant, and 3 ng for TxB2 and 6-Keto-PGF. and 375 pg for PG2 per milliliter of extrated urinary sample. Radioativity of samples was ounted in a liquid sintillation ounter (Berthold model Betaszint b F 5/3). Results were expressed as nanograms of TxB2, 6-Keto- nreased Glomerular Thromboxane Synthesis 95

3 PGF., and PG2 per milligram of protein per hour for glomerular preparations and nanograms per day for urine. TLC: isolated glomeruli (-1 mg protein ontent per tube of glomerular inubation) were inubated in KRB ontaining.8 MCi '4C-AA in a final volume of 2 ml. The tubes were plaed in a shaking water bath (1 yles/min) at 37C for h. nubation was terminated by entrifugation at 4VC for 5 min at 3, g and the supernatant was removed. The pellet was washed thrie with KRB ontaining 1 mg/ml of essentially fatty aid-free albumin. PG synthesis was examined by inubating the washed, prelabeled glomeruli resuspended in 2 ml of KRB ontaining 4 mm CaC12 at 37C for 15 min. nubation was terminated by entrifugation as above. The supernatant was removed, immediately aidified to ph with 1 N HC, and extrated twie with 3 vol of ethyl aetate. Reovery of "4C-AA was 9-95%. The extrats were evaporated to dryness under nitrogen, resuspended in hloroform-ethanol (2:1), spotted on silia gel TLC plates together with 2-5 jug of PG and Tx standards, and developed twie in the organi phase of ethyl aetate/iso-otane/aeti aid/water (11:5:2:1, vol/vol). The standards were visualized by exposing the thin-layer plates to iodine vapor. The plates were ut into 3 segments and the silia gel from eah segment was transferred to a sintillation vial; radioativity was determined by sintillation ounting in 8 ml of a ounting mixture (Ready-Solv MP, Bekman Analytial S.R.L., Milan, taly). Counting effiieny, determined with a radioative external standard, was 4-5%. Urinary TxB2, 6-Keto-PGF,, and PG2 were Urine extration. measured by radioimmunoassay after extration and silii aid olumn hromatography. 2, pm of 3H-PG2 (New ngland Nulear, 165 Ci/mM) in 2 ml of 5 mm phosphate buffer solution, ph 7.4, were added to 6-ml aliquots of urine to serve as traer during extration and purifiation. Urine was aidified to ph 3.5 with formi aid (99%) and extrated with hloroform 1:3. The hloroform was removed by evaporation at 37 C..5 g of silii aid was dissolved in 4 ml of benzene/ethyl aetate (6:4) (solvent 1) and added to a small glass olumn (6 X.7 m). The olumn was washed with 5 ml of benzene/ ethyl aetate/methanol (6:4:2) (solvent 2) and 1.5 ml of solvent 1. The extrated residue was dissolved in.2 ml of benzene/ethyl aetate/ methanol (6:4:1) (solvent 3), and.8 ml of solvent was added. The mixture was applied to the olumn, and PG frations were eluted with 5 ml of solvent (fration : POA and PGB) and 5 ml of solvent 2 (fration : PG and PGF). Fration was taken to dryness at 37 C and dissolved with 6 ml of 5 mm phosphate buffer solution, ph 7.4. A portion of fration was radioimmunoassayed for TxB2, 6-Keto-PGFa, and PG2. Perentage reovery was alulated using 2, pm of 3H-PG2. PG2 was hosen as it was extrated with the same effiieny and eluted from the silii aid olumn in the same fration as TxB2 and 6-Keto-PGF.. Overall reovery for all the ompounds ranged from 6 to 75%. Statistial analysis. Results were analyzed by analysis of variane, unpaired t, and Dunan's multiple range tests. Results are expressed as mean± SD. Results Time-ourse of proteinuria and ultrastrutural findings in adriamyin nephrosis. After a single intravenous injetion of adriamyin (7.5 mg/kg), all treated animals developed a nephroti syndrome. Proteinuria started 5-7 d after adriamyin, reahed high values after 14 d (887±144 mg/d), and persisted during the following weeks. Light mirosopy showed only moderate swelling of glomerular viseral ells and some hyaline asts in the lumens of distal tubules, but under eletron mirosopy glomerular viseral epithelial ells displayed very pronouned hanges, whih onsisted of extensive 'fusion' of foot proesses replaed by expanses of epithelial ytoplasm, and many protein droplets were found within the ytoplasm. No signifiant hanges were found in the glomerular basement membrane, endothelium, or mesangium. Careful eletron mirosopy examination of a large number of glomeruli failed to reveal platelets or inflammatory ells. PG and Tx prodution by isolated glomeruli of rats with adriamyin nephrosis. solated glomeruli were studied from 21 rats given a single intravenous injetion of adriamyin and 2 normal rats. Two groups of adriamyin-treated rats were onsidered. The first group of animals were killed 3 d after adriamyin when no proteinuria ould be deteted. The pattern of Tx and PG prodution from isolated glomeruli of these animals, measured by a radioimmunoassay, was omparable to ontrols. n ontrast, isolated glomeruli from rats treated with adriamyin 14 and 3 d before, and proteinuri at the moment of the study, synthesized Tx at 4-5 times the rates of ontrol glomeruli (Fig. 1). Synthesis of 6-Keto-PGFa and PG2 inreased transiently. However, 14 d after the single intravenous injetion of adriamyin, Tx was the only AA metabolite that was persistently synthesized at inreased rates, ontrary to 6-Keto-PGF, and PG2 (Table ). Radio TLC onfirmed the onsistent inrease in Tx synthesis by isolated glomeruli of adriamyin-treated animals in omparison with O 1, 8-, 6- x. 2 - a- -1.C 1 C o a' n o~~~o o 3rdd 14thd 3othd Control - Adriamyin nephrosis - Figure 1. Protein exretion and isolated glomeruli Tx synthesis in rats at various intervals after a single intravenous adriamyin injetion. TxB2 synthesis (ng/mg protein/h, n = 1) was respetively 2.32±.94 and 2.86±.53 at 14 and 3 d when the animals were heavily proteinuri. Both values were signifiantly (P <.1) different from those in ontrol animals (.51±.26) and in animals treated with adriamyin 3 d before (.3±.8) when proteinuria was within the normal range.. *- 96 Remuzzi, mberti, Rossini, Morelli, Carminati, Cattaneo, and Bertani

4 Table. Glomerular Synthesis and Urinary xretion ofpg by Nephroti Animals* solated glomeruli (ng/mg protein/h) Urine (ng/day) Adriamyin Adriamyin Controls nephrosis Controls nephrosis 6-keto-PGF,. 1.95± ± ± ±25.2 PG2 3.48± ± ± ±29.32 Unpaired t test showed no signifiant differene for eah omparison between the two groups. * 14 d after a single intravenous injetion of adriamyin. ontrols. Synthesis of TxB2 aounted for 9.4% of the total reovered radioativity in adriamyin-treated rats as against 4.7% in ontrol rats. Synthesis of PGF2a, 5.7%, PG2, 5.2%, 6-Keto-PGFa, 4.2%, and PGD2, 2.8% in adriamyin-treated animals were all similar to ontrols (respetively, 6.2, 4.8, 3.9, and 3.3%). 45.8% of the total radioativity in adriamyintreated animals and 49% in ontrol animals was reovered unhanged as C2:4. Total radioativity reovered from the TLC plate was 5, dpm. Two other experiments gave similar results. Urinary exretion of Tx in adriamyin nephrosis. n order to establish whether the inreased glomerular Tx synthesis in nephroti animals reflets a modifiation operating in vivo or is influened by the in vitro manipulation of the kidneys, urinary exretion of Tx was measured in ontrols and nephroti animals. Urinary exretion of TxB2 was signifiantly inreased in experimental nephrosis, whih was in parallel with the data from isolated glomeruli (Fig. 2). These findings are onsistent with reent evidene that urinary exretion of Tx mostly reflets renal synthesis of the vasoative parent ompound in health and disease (15). Urinary exretion of 6-Keto-PGFia and PG2 was not signifiantly modified in nephroti animals (Table ). ffet of a seletive Tx synthetase inhibitor on glomerular Tx synthesis and proteinuria. n order to understand better the relationship between the inreased Tx synthesis and proteinuria we studied two additional groups of animals injeted with adriamyin (7.5 mg/kg); eight rats reeived three daily intraperitoneal injetions of the Tx inhibitor, UK-38,485 (16), at doses of 2 mg/kg during the period of heavy proteinuria (days after adriamyin), and eight rats reeived buffer alone for the same period. The dose of UK-38,485 used, when given to normal rats, aused 98% (from 142.8±76.7 to 2.45±1.27 ng/ml) inhibition of platelet Tx generation, as measured by the amount of Tx deteted in serum after blood oagulation in vitro (11). All rats were killed on day 18, 4 h after the last UK-38,485 or buffer injetion. As shown in Fig. 3, proteinuria during the treatment with UK-38,485 was signifiantly lower than both pretreatment values and the values in a omparable group of rats, not given the inhibitor, at the same interval from adriamyin injetion. Urinary exretion of TxB2 was also signifiantly redued during UK-38,485 treatment (Fig. 3). Sham-treated nephroti ontrols presented no signifiant differenes in urinary exretion of TxB2 before and after the period of buffer injetion (18.34±6.52 vs ±8.24 ng/d). We also found that Tx synthesis from isolated nephroti glomeruli beame normal 1-% Cm CD a, 4 x D Con trol Adriamyin nephrosis ( 14th d ) Figure 2. Urinary exretion of immunoreative TxB2 in ontrol and adriamyin treated rats. TxB2 exretion (ng/d, n = 8) was 4.5±1.2 in ontrol vs ±5.1 in adriamyin (7.5 mg/kg) treated animals 14 d after the single intravenous injetion (P <.1). after the treatment with UK-38,485 (Fig. 3). On the ontrary, 6-Keto-PGF and PG2 syntheses in nephroti animals were not signifiantly modified by the treatment with UK-38,485 in isolated glomeruli (2.26±.76 vs ± 1.7 ng of 6-Keto- PGF,1/mg protein per h and 3.61±.93 vs. 4.±1.18 ng of PG2/mg protein per h) as well as in urine (58.14±21.98 vs ±23.1 ng of 6-Keto-PGF/d and 68.14±2.1 vs ±25.85 ng of PG2/d). Moreover, the treatment with UK-38,485 inhibited glomerular Tx synthesis in ontrol rats (3.±.5 vs..45±.3 ng/mg protein per h). ffet of the seletive Tx inhibitor UK-38,485 on renal funtion in normal and nephroti animals. During treatment with UK-38,485 at a dose inhibiting serum Tx formation by 98%, no signifiant hanges in renal funtion, measured by endogenous reatinine and inulin learanes, ould be deteted in normal or nephroti animals (reatinine learanes: 1.63±.68 vs. 1.3±.19 in normals and 1.36±.35 vs. 1.7±.11 ml/min in nephrotis, inulin learane rate: 1.2±.12 vs. 1.14±.32 in normals and 1.25±.25 vs. 1.18±.28 ml/min per 1g weight in nephrotis). These results are onsistent with the findings of Lianos et al. (7), who showed in a model of nephrotoxi nephritis that inulin and para-aminohippurate learanes and filtration fration were not worsened by UK-38,485. ffets of sulinda on platelet and renal Tx generation. Sulinda (6 mg/kg) was given to 1 nephroti animals from day 14 to day 16 after a single intravenous injetion of adriamyin, and serum TxB2 onentration and urinary exretion of TxB2 were measured before and 2 h after the last dose. Sulinda inhibited serum TxB2 formation by >7% in nreased Glomerular Thromboxane Synthesis 97

5 z Go n - e. 2 a, 1 x ^ 1, O- 81 '-- 6C)- 4 x 4 1- C. u O - Adriamyin Adriamyin nephrosis nephrosis (14th d) +UK (18th d) Figure 3. The effet of a seletive Tx synthetase inhibitor (UK- 38,485) on isolated glomeruli Tx synthesis (top) and urinary exretion (middle) of TxB2. Rats (n = 8) were treated with the Tx inhibitor for 5 d when they were heavily proteinuri, as desribed in Methods. 4 h after the last UK-38,485 dose, glomerular TxB2 syntheti rates after death were signifiantly (P <.1) redued (.41±.64) ompared with pretreatment values (2.86±.53 ng/mg protein per h). Before death, 24-h urinary exretion of TxB2 was also signifiantly (P <.1) redued (6.9±3.6) during UK-38,485 treatment, ompared with pretreatment values (21.96±5.1 ng/d). The 24-h urinary protein exretion (bottom) were signifiantly (P <.5) lower (4.5±65) during the UK-38,485 treatment than before treatment (887±144 mg/d). all animals, but had no effet on urinary exretion of TxB2 (22.72±6.66 vs. baseline 24.47±7.25 ng/d). Urinary protein exretion as well was not signifiantly modified by sulinda in respet to the pretreatment values (1,2±12 vs. 1,5 ±9 mg/d). ffet of AA added in vitro on PG and Tx synthesis by isolated glomeruli. n order to establish whether the inreased Tx synthesis of isolated glomeruli in nephroti animals was due to inreased substrate availability, Tx prodution from isolated glomeruli was studied in vitro with and without the addition of 1 and 2 AM exogenous AA. n the presene of AA, TxB2 and 6-Keto-PGF1. syntheses from normal and nephroti glomeruli (1 animals for eah group) was not signifiantly modified ompared to basal onditions (for TxB2: ontrols =.53±.32 and.48±.18 vs..42±.15; nephrotis = 2.36±1.24 and 2.51±1.65 vs. 2.1±.95 ng/mg protein per h; for 6-Keto-PGFa: ontrols = 2.51±.61 and 2.66±.59 vs. 2.33±.53; nephrotis = 2.64±.65 and 2.75±.67 vs. 2.43±.55 ng/mg protein per h). At variane, PG2 prodution of ontrol and nephroti isolated glomeruli in the presene of 1 and 2 gm AA was higher, although not signifiantly different, from the values of unstimulated glomeruli (ontrols = 4.19±1.2 and 4.47±1.15 vs. 3.89±.83; nephrotis = 4.±1.13 and 4.45±1.5 vs. 3.63±.68 ng/mg protein per h). ffet of hroni furosemide on Tx synthesis by isolated glomeruli. 2 normal rats were injeted with furosemide or buffer for 1 d in order to ahieve maximum stimulation of the renin-angiotensin axis. Although this was an aute experiment and may not be fully representative, we deided to kill the animals at 14 d, when proteinuria usually reahed top values. TxB2 synthesis by glomeruli isolated from furosemidetreated animals was signifiantly elevated (.91±.8 vs..59±.1, ng/mg protein per h in ontrols, P <.1). However, Tx generation by isolated glomeruli from these normal animals was still signifiantly (P <.1) lower than in the glomeruli of nephroti rats studied simultaneously (2.5±1.3 ng/mg protein per h). These results suggest that adriamyin enhanes glomerular Tx synthesis with a mehanism independent from stimulation of the renin-angiotensin system. ffet of the nephroti syndrome per se on glomerular PG and Tx prodution. n order further to exlude that the inreased glomerular Tx prodution in adriamyin nephrosis is simply a onsequene of the nephroti syndrome, we performed some additional experiments. 1 rats were given adriamyin in a single intravenous injetion (7.5 mg/kg), but the left renal artery was lamped just before the injetion. The lamp was removed 15 min later. 2 wk later all the animals were heavily proteinuri, but the left kidney was proteted from adriamyin toxiity and on eletron mirosopy examination appeared normal, whereas adriamyin-indued hanges were more severe than usual in the right kidney. Thus, it appears that diret exposure of kidney to adriamyin in vivo is neessary for the demonstration of glomerular damage. This model serves to study PG prodution by isolated glomeruli in normal or diseased kidneys in animals with heavy proteinuria and nephroti syndrome. Failure to detet any abnormality in the generation of AA metabolites by the left kidney indiates that the nephroti syndrome per se has no effet on glomerular synthesis of PG and Tx (Table ). Urinary exretion of TxB2 from animals affeted by unilateral disease was slightly, but not signifiantly, dereased in omparison to values obtained in the onventional two kidneys model (16.8±3.62 vs ±4.51 ng/d). Aute in vitro and in vivo effet ofadriamyin on glomerular Table Synthesis of TXVB2 by solated Glomeruli* Adriamyin nephrosis Left renal artery lamp Controls Left kidney ±.21 Right kidney 4.26±.8.5±.1t * Nanogram per milligram protein per hour. * P <.1 (unpaired t test). 98 Remuzzi, mberti, Rossini, Morelli, Carminati, Cattaneo, and Bertani

6 PG and Tx synthesis. We investigated whether the altered AA metabolism was the immediate onsequene of exposure of the glomerular struture to adriamyin by inubating the glomerular preparation with adriamyin (2 lsg/ml and 1 mg/ml) in vitro for 5 or 3 min. This signifiantly depresses the isolated glomeruli's apaity to synthesize Tx, ompared to ontrols, i.e., 2 sg/ml adriamyin:.26±.15 vs..81±.11 ng/mg protein per h for 5 min inubation (P <.1), and.9±.4 vs..75±.6 ng/mg protein per h for 3 min inubation (P <.1); 1 mg/ml adriamyin:.23±.1 vs..63±.11 ng/mg protein per h for 5 min (P <.1), and.12±.4 vs..77±.5 ng/mg protein per h for 3 min inubation (P <.1). Similarly, isolated glomeruli of rats injeted with 7.5 mg/kg adriamyin and killed 5 min and 12 h after the treatment generated signifiantly less arahidonate metabolites than ontrols (.11±.5 vs..67±.6 (P <.1) and.14±.7 vs..7±.9 (P <.1) ng/mg protein per h, respetively). These experiments indiate that adriamyin is not diretly responsible for the exaggerated glomerular synthesis of Tx observed in the experimental model desribed so far. Disussion The finding that several experimental models of glomerular damage show ertain ommon quantitative and qualitative alterations in AA metabolism suggests the importane of this modulatory pathway in glomerular pathophysiology (17-19). Glomerular prodution of TxB2 might be one of the fators responsible for the altered glomerular filtration rate in disease (8). ndeed, in experimental nephrotoxi nephritis, a marked initial inrement of glomerular Tx has been demonstrated, whih parallels the derements in renal blood flow and glomerular filtration rate seen a few hours after the indution of the disease. n ontrast, the glomerular synthesis of vasodilatory PG2 inreases progressively in the following days and appears to aount for the spontaneous reovery in renal blood flow and glomerular filtration rate observed in these animals after 14 d (2). n this ontext, Lianos et al. (7) also found in a rat model of nephrotoxi nephritis that Tx was the only arahidonate metabolite that was persistently synthesized at an inreased level during prolonged inubation of glomerular preparations. More important, they found a linear orrelation between proteinuria and the inrease in glomerular Tx synthesis. As the authors ommented, this orrelation ould represent a ause-effet relationship, or both glomerular Tx and proteinuria ould be independent markers of the extent of glomerular injury. The results of the present study show that a model of glomerular injury quite different from those studied previously was also haraterized by a seletive inrease in urinary exretion and glomerular prodution of TxB2. We also found transient differenes between normal and nephroti animals in other produts of AA metabolism, but these differenes never reahed statistial signifiane. Together with previous observations in nephrotoxi nephritis and in a model of aute renal failure due to bilateral ureteral obstrution, these results might be taken to suggest that glomerular funtion in normal onditions might be ritially under the ontrol of Tx synthesis (21). Our study also investigated the relationship between altered Tx synthesis and proteinuria. The adriamyin model of nephroti syndrome in rats is partiularly suitable, in that the time ourse and the amount of protein exretion are well defined (9). A single adriamyin injetion in rats indues a nephroti syndrome with proteinuria appearing 5 d after injetion, and reahing high values after 14 d. At variane with aminonuleoside nephrosis, where animals are free of proteinuria after 4 wk, in adriamyin nephrosis proteinuria persists several months (22). Under light mirosopy, pathologial hanges are almost unremarkable, onsisting in moderate swelling of glomerular viseral epithelial ells and in some tubular asts. n ontrast, ultrastrutural findings omprise marked hanges of podoytes with extensive 'fusion' of foot proesses and pronouned 'ativation' of glomerular viseral epithelial ells. Our results indiate that Tx synthesis by isolated glomeruli signifiantly inreased in onomitane with the onset of proteinuria, and reahed a plateau as onstant values for protein exretion were attained. The signifiant redution in proteinuria noted after administration of UK-38,485, a seletive Tx synthetase inhibitor, indiates that in our model inreased glomerular Tx synthesis might be one of the fators responsible for persistent proteinuria. An interesting feature of our study is also that the Tx inhibitor, at a dose inhibiting platelet Tx prodution by 98%, normalized but did not abolish the exaggerated Tx prodution by nephroti glomeruli. This has also been observed by Patrignani et al. (23) who reported failure of dazoxiben to redue urinary TxB2 exretion by >3% in man. As these authors speulated, this ould be attributed to more than one fator, i.e., rapidly reversible inhibition and different sensitivity of ortex Tx-synthase ompared to the platelet enzyme. Appropriate measurements of reatinine and inulin learanes in normal and nephroti animals exposed to UK-38,485 indiated that our results were not influened by modifiations of glomerular filtration rate indued by the drug. We also found that the inreased Tx synthesis in nephroti glomeruli was not the immediate result of glomerular apillaries being exposed to adriamyin, sine isolated glomeruli exposed to the drug in vitro showed a marked redution in their apaity to generate TxB2. Similar results were obtained when isolated glomeruli from rats killed 5 min and 12 h after adriamyin injetion were examined. Thus, while the latter findings appear to reflet a nonspeifi, diret, aute toxi effet of adriamyin on glomerular apillaries, the marked inrease in urinary exretion and glomerular prodution of TxB2 appears peuliar to the nephroti syndrome. As for the mehanism underlying the alteration of glomerular AA metabolism in experimental nephrosis, an inrease in preursor availability might be a oneivable explanation of the enhaned glomerular TxB2 synthesis. However, failure to neutralize the differene between normal and nephroti rats in the presene of an exess of AA added in vitro makes suh an event unlikely. An alternative possibility is that the altered glomerular AA metabolism reflets ativation of the reninangiotensin system whih haraterizes the nephroti ondition (24). f this is the ase, aptopril, an inhibitor of angiotensin -onverting enzyme, should eliminate the exaggerated Tx output from nephroti glomeruli. However, it has reently been proven that aptopril, besides inhibiting the renin-angio- nreased Glomerular Thromboxane Synthesis 99

7 tensin axis, also diretly stimulates the renal AA metabolism (25). To overome this problem we studied a group of normal rats under hroni furosemide treatment, whih had limited aess to water, to ensure maximal stimulation of the reninangiotensin system (26). The glomerular Tx synthesis was signifiantly higher in adriamyin than in furosemide-treated animals, so it is unlikely that the altered AA metabolism observed in nephroti syndrome is simply a onsequene of stimulation of the renin-angiotensin system. Finally, the inreased prodution of Tx by isolated glomeruli in adriamyin nephrosis might be a onsequene of the hypoalbuminemia whih develops with the nephroti syndrome. The rationale for this possibility derives from studies by Yoshida and Aoki (27) and by Shieppati et al. (28), who demonstrated that hypoalbuminemia ofthe nephroti syndrome raises the apaity of irulating platelets to form Tx. However, the present study shows that blood modifiations indued by the nephroti syndrome do not alter the AA metabolism, in that isolated glomeruli from kidneys unilaterally proteted from adriamyin toxiity, but taken from animals with heavy proteinuria, generated a normal amount of TxB2. As regards the ellular origin of TxB2, it is reasonable to believe that glomerular epithelial ells are the major soure of Tx in our experimental ondition. n our model, at variane with previous ones, no monoyte or leukoyte infiltration and no platelet deposition ould be deteted, and the only ultrastrutural abnormalities were onfined to viseral epithelial ells (9). Sine morphologial evidene does not neessarily rule out the possibility of ontamination of our glomerular preparation with entrapped platelets and leukoytes, we studied the effet of sulinda in adriamyin nephrosis. To larify better the issue of the urinary TxB2 origin we used the pharmaologial approah proposed by Ciabattoni et al. (29). These authors provided evidene in humans that sulinda substantially inhibits Tx synthesis by irulating ells but does not affet renal Tx synthesis as measured by urinary exretion of TxB2. These studies give useful information on the origin of urinary TxB2, distinguishing between irulating ells and intrarenal Tx synthesis. We repeated the same experiments in nephroti rats. Despite the fat that platelet ount in these animals rises to twie basal values 14 d after the treatment with adriamyin (3), we found that the dose of sulinda used inhibited platelet TxB2 by >7% in all animals but did not affet urinary exretion of TxB2. This would suggest that the inreased urinary exretion of Tx in experimental nephrosis is unlikely to be of platelet origin, but possibly reflets an inreased intrarenal synthesis. nreased Tx prodution an alter glomerular funtion in a omplex way. t was suggested that basal renal vasular tone is in some way under the influene of TxA2, sine imidazole (31) inreases renal blood flow in animals by reduing renal resistane. Moreover, evidene is aumulating that this potent vasoative substane may ontribute to the intense afferent arteriolar vasoonstrition that ours with ureteral obstrution. The results of the present study suggest that TxA2 ould have an additional effet on glomerular funtion by altering the basement membrane's permeability to proteins. n this ontext, a possible role of PG in modulating seletivity and permeability of the glomerular barrier an be inferred from old studies in whih indomethain (32) was reported onviningly, though inonsistently (33), to redue proteinuria in animals and humans. However, the onfliting data obtained with indomethain might be interpreted as a onsequene of nonseletive inhibition of PG with different, and possibly opposite, ativities in the glomeruli, thus making it diffiult to assess the net result. Appropriate studies are warranted to evaluate the potential role of drugs seletively inhibiting Tx synthesis in the treatment of proteinuria. Aknowledaments The authors are grateful to Drs. Manuela Livio and Ariela Benigni for fruitful ooperation, and to Professor Carlo Patrono for disussion and onstrutive ritiism. Cristina Signorelli and Judith Baggott helped prepare the manusript. This study was supported by a grant from the National Researh Counil (CNR) (Program 'Tenihe Sostitutive di Funzioni d'organo'; Progetto Finalizzato 'Mediina Preventiva' No ). Referenes 1. Smith, W. L., and T. G. Bell mmunohistohemial loalization of the prostaglandin-forming ylo-oxygenase in renal ortex. Am. J. Physiol. 235: Hassid, A., M. Koniezkowsky, and M. J. Dunn Prostaglandin synthesis in isolated rat kidney glomeruli. Pro. Nati. Aad. Si. USA. 76: Shlondorff, D., S. Rozniak, J. A. Satriano, and V. W. Folkert Prostaglandin synthesis by isolated rat glomeruli: effet of angiotensin. Am. J. Physiol. 238: Folkert, V. W., and D. Shlondorff Prostaglandin synthesis in isolated glomeruli. Prostaglandins. 17: Sraer, J., N. Ardaillou, J. D. Sraer, and R. Ardaillou n vitro prostaglandin synthesis by human glomeruli and papillae. Prostaglandins. 23: Sraer, J. D., L. Moulonguet-Doleris, F. Delarue, J. Sraer, and R. Ardaillou Prostaglandin synthesis by glomeruli isolated from rats with glyerol-indued aute renal failure. Cir. Res. 49: Lianos,. A., G. A. Andres, and M. J. Dunn Glomerular prostaglandin and thromboxane synthesis in rat nephrotoxi serum nephritis. J. Clin. nvest. 72: Dunn, M. J The role of arahidoni aid metabolites in glomerulonephritis. n Glomerular njury: 3 Years After Morgagni. T. Bertani and G. Remuzzi, editors. Wihtig ditore, Milan Bertani, T., A. Poggi, R. Pozzoni, F. Delaini, G. Sahi, Y. Thoua, G. Mea, G. Remuzzi, and M. B. Donati Adriamyinindued nephroti syndrome in rats. Sequene of pathologi events. Lab. nvest. 46: Lowry,. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall Protein measurement with the folin phenol reagent. J. Biol. Chem. 193: Remuzzi, G., A. Benigni, P. Dodesini, A. Shieppati, M. Livio, J. S. Day, W. L. Smith,. Pina, P. Patrignani, and C. Patrono Redued platelet thromboxane formation in uremia. videne for a funtional ylooxygenase defet. J. Clin. nvest. 71: Hare, R. S ndogenous reatinine in serum and urine. Pro. So. xp. Biol. Med. 74: Frolih, J. C., T. W. Wilson, B. J. Sweetman, M. Smigel, A. S. Nies, K. Carr, J. T. Watson, and J. A. Oates Urinary prostaglandins. dentifiation and origin. J. Clin. nvest. 55: Pugliese, F., and G. Ciabattoni nvestigation of renal arahidoni aid metabolites by radioimmunoassay. n Prostaglandins and the Kidney. M. J. Dunn, C. Patrono, and G. A. Cinotti, editors. Plenum Publishing Corp., New York Remuzzi, mberti, Rossini, Morelli, Carminati, Cattaneo, and Bertani

8 15. Fitzgerald, G. A., A. K. Pedersen, and C. Patrono Analysis of prostaylin and thromboxane biosynthesis in ardiovasular disease. Cirulation. 67: Parry, M. J., M. J. Randall,. Hawkeswood, P.. Cross, and R. P. Dikinson nhaned prodution of prostaylin in blood after treatment with seletive thromboxane synthetase inhibitor, UK- 38,485. Br. J. Pharmaol. 77:547. (Abstr.) 17. Zipser, R., S. Myers, and P. Needleman xaggerated prostaglandin and thromboxane synthesis in the rabbit with renal vein onstrition. Cir. Res. 47: Benabe, J.., S. Klahr, M. K. Hoffman, and A. R. Morrison Prodution of thromboxane A2 by the kidney in glyerolindued aute renal failure in the rabbit. Prostaglandins. 19: Dighe, K. K., G. W. Smith, A. Ungar, and P. H. Whelpdale Renal prostaglandins in renal hypertensive dogs. Clin. Si. Mol. Med. 54: Stork, J.., and M. J. Dunn mmuneglomerulonephritis in the rat: renal funtion, thromboxane A2 (TxA2) and PG2. Clin. Res. 32:565A. 21. Morrison, A. R., K. Nishikawa, and P. Needleman Thromboxane A2 biosynthesis in the ureter obstruted isolated perfused kidney of the rabbit. J. Pharmaol. xp. Ther. 25: Bertani, T., and G. Remuzzi pithelial ell damage. n Glomerular njury: 3 Years After Morgagni. T. Bertani and G. Remuzzi, editors. Wihtig ditore, Milan Patrignani, P., P. Filabozzi, F. Catella, F. Pugliese, and C. Patrono Differential effets of dazoxiben, a seletive thromboxane-synthase inhibitor, on platelet and renal prostaglandin-endoperoxide metabolism. J. Pharmaol. xp. Ther. 228: Lifshitz, M. D., and J. H. Stein Renal vasoative hormones. n The Kidney. B. M. Brenner and F. C. Retor, editors. W. B. Saunders Co., Philadelphia, London, and Toronto Swartz, S. L., G. H. Williams, N. K. Hollenberg, L. Levine, R. G. Dluhy, and T. J. Moore Captopril-indued hanges in prostaglandin prodution. Relationship to vasular responses in normal man. J. Clin. nvest. 65: Weber, P. C., B. Sherer, and C. Larsson nrease of free arahidoni aid by furosemide in man as the ause of prostaglandin and renin release. ur. J. Pharmaol. 41: Yoshida, N., and N. Aoki Release of arahidoni aid from human platelets. A key role for the potentiation of platelet aggregability in normal subjets as well as in those with nephroti syndrome. Blood. 52: Shieppati, A., P. Dodesini, A. Benigni, M. Massazza, G. Mea, G. Remuzzi, M. Livio, G. de Gaetano, and. C. Rossi The metabolism of arahidoni aid by platelets in nephroti syndrome. Kidney nt. 25: Ciabattoni, G., G. A. Cinotti, and C. Patrono Renal effets of antiinflammatory drugs. ur. J. Rheumatol. nflammation. 3: Poggi, A., L. Kornblihtt, F. Delaini, T. Colombo, L. Mussoni,. Reyers, and M. B. Donati Delayed hyperoagulability after a single dose of Adriamyin to normal rats. Thromb Res. 16: Brenner, B. M., and N. Shor Studies of prostaglandin ation on the glomerular miroirulation. n Prostaglandins and the Kidney. M. J. Dunn, C. Patrono, and G. A. Cinotti, editors. Plenum Publishing Corp., New York Mihielsen, P., R. Verberkmoes, V. Desmet, and W. Hermerijkx volution histologique des glomerulonephrites proliferatives diffuses trait6es par l'indomethaine. J. Urol. Nephrol. 75: Glassok, R. J., A. H. Cohen, C. M. Bennett, and M. Martinez- Maldonado Primary glomerular diseases. n The Kidney. B. M. Brenner and F. C. Retor, editors. W. B. Saunders Co., Philadelphia, London, and Toronto nreased Glomerular Thromboxane Synthesis 11