- Email: [email protected]
Thromboxane receptor blockade improves cyclosporine nephrotoxicity in rats
PROSTAGLANDINS THROMBOXANE RECEPTOR BLOCKADE IXPROVES CYCLOSPORINE NEPHROTOXICITY IN RATS R.F. Spurney, S.D. Ma~ros, D. Collins, P. Ruiz, P.E. Klotman, and T. Coffman Departments of Medicine and Pathology, Duke University a n d Durhtm VA Med/cal Centers, Durham, North Carolina 27710 and *Molecular Medicine Section, L.D.B.A./N. I.D.R., National Institutes of Health Bethesda, Maryland 20892
ABSTRACT Cyclosporine A (CyA) nephrotoxicity is associated with impaired renal hemodynamic function and increased production of the vasoconstrictor eicosanoid thromboxane A2 (TxA2). In CyA toxic rats, lenal dysfunction can be parüally reversed by inhibitors of thromboxane synthase. However, interpr¢taüon of these results is complicated since inlübiüon of thromboxane synthase may cause accumulafion of prostaglandin endoperoxides that can act as parüal agonists at the TxA2 receptor and may blunt the efficacy of treatrnent. Furthermore, these endoperoxides may be used as substrate for producüon of vasodilator prostaglandins causing beneficial effects on hemodynamics which am independent of thromboxane inhibition. To more specifically examine the role of TxA2 in CyA toxicity, we investigated the effects of the thromboxane receptor antagonist GR32191 on renal hemodynamics in a rat model of CyA nephrotoxicity. In this model, administration of CyA resulted in a significant decrease in glomer-!ar f'fllration rate (GFR) (2.85"~0.26 [CyA] vs 6.82-2-2-2~.96tal/min/kg [vehicle]; p<0.0005) and renal blood flow (RBF) (21.65.+.2.31 [CyA] vs 31.87_+3.60 n~/min/kg [vehicle]; p<0.025). Renal vascular resistance (RVR) was significanfly higher in rats given CyA compared to animals treated with CyA vehicle (5.32.%0.55 [CyA] rs. 3.54:£-0.24 mm Hg/min/tal/kg [vehicle]; p<0.05). These renal hemodynamic alteraüons were associated with a significant increase in urinary excreüon of unmetabolized, "nativ¢" thromboxane B2 (TxB2) (103+18 [CyA] vs 60:1:16 pg/hour [vehicle]; p<0.05). Only minimal histomorphologic chang¢s were apparent by light microscopic examinafion of kidneys from both CyA and vehicle ffeated animals. However, with immunoperoxidase staining, a significanfly greater number of cells expressing the rat cornmon lenkocyte antigen was found in the renal interstiüum of rats given CyA*. There was no detectable increase in monocytes/macrophages in the kidneys of CyA toxic anirnals. In rats treated with CyA, intraarwrial infusion of GR32191 at maximally tolerated doses significantly increased GFR and RBF, and decreased RVR. Although both RBF and RVR were restored to levels not different from controls, GFR remained significantly reduced following administration of GR32191. These data suggest that the potent vasoconstrictor TxA2 plays an important role in mediating renal dysfuncüon in CyA nephrotoxicity. Howevet, other factors may be important in producing nephrotoxicity associated with CyA. INTRODUCTION CyA is a potent immunosuppressive agent which has been extremely effective in prolonging aUograft survival (1,2) and suppressing autoimmune disease (3,4). Unfortunately, the clinical use of CyA has been complicated by a significant incidence of drug-related nephrotoxicity (5). The nephrotoxicity is characterized by a dosedependent decrease in GFR that may limit the c"lmical utility of the drug (6,7).
FEBRUARY 1990 VOL. 39 NO. 2
135
PROSTAGLANDINS The specific mechanisms responsible for CyA nephrotoxicity are not known. How¢ver, in both humans and animals given CyA, renal vasoconstriction has been demonstrated (7,8,9). Furthermore, in a rat model, EngIish et al. (I0) have shown that renal hemodynamic alterations rather than direct tubular injury are important in the pathogenesis of nephrotoxicity. Although several potential mediators have been proposed (11,12,13), recent observations suggest that the vasoconstrictor eicosanoid TxA2 may be important in producing the renal hemodynamic alterations associated with CyA administration (14,15). For exampte, in animal models of CyA nephrotoxicity, utinary thromboxane excretion is increased (16,17) and a significant negative correlation between urinary TxB2 excretion and OFR has be.en reported (14). In addition, thromboxane production by the kidney is increased (15,17,18) and this enhancement of renal thromboxane production is associated with reduced OFR and RBF (15,17,18). Evidence for a functional role for thromboxane was suggested in studies by Elzinga and coworkers (15) who demonstrated that fish oft administration improved renal function and reduced renal cortical thromboxane production in rats given CyA. In addition, Perico et aL (14) showed that selective inhibition of thromboxane production with a specific thromboxane synthase inhibitor significantly improved GFR in rats with CyA nephrotoxicity. However, following treatment, GFR remained below normal Two other groups have also reported beneficial effects of thromboxane synthase inhibitors in rats with CyA toxicity (19,20). However, physiologic interpretation of studies using thromboxane synthase inhibitors is complicated. Inhibition of thromboxane synthase may result in accumulaüon of the pt'ostaglandin endoperoxides which can act as agonists at the thromboxane receptor (21) and may blunt the efficacy of treatment. Also, these accumulated endoperoxides may be metabolized to form vasodilator prostaglandins (22,23). Thus, some of the beneficial effects of thromboxane synthase inhibitors may relate to increased production of vasodilator prostaglandins rather than specific inhibiüon of the acüons of TxA2. The alm of the present s~iies was to more specifically examine the functional role of ~omboxane in CyA toxicity. To address this issue, we evaluated the effects of the specific thromboxane receptor antagonist GR32191 ([1R.[l(Z),2,3,5]]-(+)-7-[5[[(1, l'-biphenyl)-4-yl] methoxy]-3-hydroxy-2(1-piperidinyl).cyclopentyl]-4-beptenoic acid, hydrocMoride) (24,25) on renal hemodynanücs m a rat model of CyA nephrotoxicity. This compoand offers the advantage of specifically interfering with the interaction of TxA2 and its rec¢ptor without directly affecting production of arachidonic acid metabolites. METHODS
CyA Nephroroxicity: CyA nephrotoxicity was produced in male ACI rats weighing 150-250 grams as previously described (17). Animals wem anesthetized with halothane and the right kidney was removed by tank incision. To simulate condiüons associated with r¢nal transplantaüon and to eliminate effects of i n ~ sympathctic tone (8,13), all animals underwent denervation of the left kidney followed by cross-clamping of left renal artery for 30 minutes. Denervation was accomplished by disruption of visible nerve bundles followed by application of a 10% phenol solution to the mnal pediceL Following surgery, rats were given CyA (50mg/kg) or its vehicle (olive oft) by daily intraperitoneal injection. Cyclosporine treated rats and vehicle treated animals were studied 12-14 days after op¢raüon. On the day before the renal hemodynamic study, al~imals were placed in metabolic cages and urine was collected for 24 hours. The urine was chiUed throughout the collection period and then stored at minus 70°C and later
136
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS
assayed for TxB2 and 6-keto-prostaglandin F l a (6-keto-PGFla) by radioimmunoassay (RIA) following HPLC separation as described below.
RenalHemodynamicStudies:
Hemodynamic studies were performed as described previously (17,26). On the day of study, animals were anesthetized with 0.04mg/gram pentobarbital, and a polyethylene catheter (PE 240) was inserted into the trachea to facilitate spontaneous venfilation. The left caroti_d artery and leit jugular vein were cannulated with polyethylone catheters ~E-50) for intravenous infusions, to monitor mean arterial pressure (MAP) (Gould-Statham strain gauge), and to allow intermittent sampling of arterial blood. Following surgery, normal saline (2.0% of the body weight) was infused intravenously over 20 nünutes to replace surgical losses. A priming dose of carboxyl-14C-inulin and glycyl-3H-PAH was given, followed by infusion of pentobarbital, carboxyl-14C-inulin, and glycyl-3H-PAH in normal saline at a rate of 25 gl/minute/100gram body weight. The leftrenalvein was catheterizedwith a curved 0.965-mm OD teflon catheter inserted through the left femoral vein to collect samples of renal venous blood. The left meter was cannulated with a PE-10 catheter to facilitate collecüon of urine. A polyethylene catheter (PE-10) was inserted throngh the femoral artery and positioned in the aorm just above the origin of the ler renal artery for in_fusion of the thromboxane receptor antagonist GR32191 or its vehicle (0.45% normal saline) at 25 pJ/min. After 30 minutes of equilibration, inulin clearance and RBF were measured during two consecutive 30 minute periods. Foilowing these baseline renal hemodynamic measurements, the thromboxane receptor antagonist GR32191 was infused into the renal attery. After 30 minutes of equilibrafion,renal function was measured during two additional30 minute pcriods. C y A treatedanimals received either Il.ig/kg/min,10gg/kg/min, or 500gg/kg/min of GR32191. In addition,as con~Is, a group of animals treatcdwith C y A vehiclewere given GR32191 at a dose of 10~tg/kg/min. Clearances of inulin and P A H were calculatedusing standardformulas. Renal P A H extracüon (E-PAH) was determined and R B F was calculatedusingthe formula: R B F = C P A H / [ I ~ A H x (1 - Hct)]. R V R was cstimatedby dividingmean arterialpressureby RBF. After the hemodynamic studies, the rare were sacrificed and the l e r kidney was excised. A portion of the kidney was fLxed in f o r m a ~ for histologic examinaüon. The rcmaining kidney tissuewas frozenin liquidnitrogenand savcd for immunohistology as outlincdbelow.
Eicosanoid Assays: Using prcviously described m¢thods (17,26), eicosanoids wem exwacteclfrom urine using Sep-Pak C18 cartridges(Waters Associates,Milford,Mass.) and then separated using a Waters Model 840 system and Pccosphcre HS3-C18 colunm. Eicosanoids were eluted with a linear gradient from 100% 0.017 M orthophosphoric acid to 100% acetonitrile over 10 minutes at 3 tal/minute. Appropriate fractions were collectedbased on retenfionümes of known standards. The eluate was dried under nitrogen and resuspended in RIA buffer. All samples were correctedfor recovery of tritiatedstandardsadded to the originalsamples. RIA rcagents wem purchased from AMI, Boston, M A and IC'N,Lysle,IL.
FEBRUARY 1990 VOL. 39 NO. 2
137
PROSTAGLANDINS For 6-keto-PGFla, sample and standards wem incubated with a mixture of anüsera and a known amount of tritiated standard at 4°C for 20 hours. For TxB2, the procedure was similar except the incubation time was 1 hour. After incubafion, free cicosanoid was adsorbcd with dextran-coatedcharcoal,and the Iritiumremaining in the supcmatant was measured with a liquidscinüllationcounter. The unknowns, corrected for recovery, were compared with a standard curve in which the logarithm of the concentraüon was plottedagainstthe logitof the B/Bo value. The resultsarc expresscd as pg/minutc of eicosanoid.
Histologic Studie.s:
Formalln fixed tissue stained with hematoxylin and eosin was examined by light microscopy. In addition, to bettet evaluate inflammatory cell infütrates, immunohistologic studies were performed on sections comparing rats given CyA to animals treated with CyA vehicle. Since several reports have suggested that macrophages may be an important cellular source of enhanced thromboxane production in CyA n¢phrotoxicity (27,28), phenotypes of infiltrating cells were evaiuated using monoclonaI antibodies against rat leukocyte-common antigen (OX-I), and rat monocytes and macrophages (OX-42). Immunoperoxidase staining of sections of fresh frozen üssue was performed as previously reported (29). Light microscopic and immunoperoxidase studies were reviewed by a pathologist (PR) blinded to the trcatment groups. Light microscopic scctions w e m examined for tubulointerstitial,vascular,and glomerular lesions. Abnormalities w e m graded using a semiquan•taüve scale where 0 was no abnormality,and 1+, 2+, 3+, and 4+ represented mild, mod¢rate, moderatcly-s¢vere, and scvere abnormalities, respectively. Immunoperoxidase stained sections wem examined for the presence of OX-1 or OX-42 positive cells in glomeruli, mbtdes, vessels, or renal intersütium. The severity of inflammatory cell infillrates was graded semiquantitatively where 0 indicated the absence of cells staining with the OX-42 or OX-1 monoclonal antibody, and 1+ to 4+ represented minimal to marked infiltration by OX-42 or OX-1 positive cells.
Statistical Analysis:
Data am presented as the mean + standard error of the mean. For the hemodynamic smdies, data points for each animal represent the mean of the values measured during the two clearance periods. For comparisons between two groups, staüsücal significance was assessed using an unpaired t-test. For comparisons within the same group a paired t-test was used. For comparisons between more than two groups, statistical analysis was performed by one-way analysis of variance followed by Bonferroni's procedure for multiple pairwise comparisons (30).
RESULTS The effectof C y A on baselineG F R and R B F is shown in Figure i. Both G F R (2.85+0.26 [CyA] vs. 6.82.~:0.96 tal/min/kg [vehicle]; p<0.0005) and R B F (21.65-+-2.31[CyA] vs. 31.87:1:3.60tal/min/kg [vehicle];p<0.025) w e m significantly reduced in rats given CyA compared to animals which rec¢ived CyA v¢hicle. In addition, calculated RVR was significantly increased in CyA treated rats compared to vehicle treated animals (5.32_+0.55 [CyA] vs. 3_54~0.24 mm Hg/tal/min/kg [vehicle]; p<0.05). Baseline nw,an arterial pressures measured at the time of smdy wem sinülar in CyA and vehicle treated rats (94.33+4.01 [CyA] vs 104+_5.32 mm Hg [vehicle]; p=NS).
138
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAG LAN Dl NS
[ ] CyA 7.0
- -
~ . 0 --
GFR (tal/min/kg)
3 . 0 --
l GFR
T
i
--
35
[] --
25
Vehicle
RBF
(tal/min/kg) --
15
RBF
Figure 1: Both GFR and RBF wem significantly reduced in rats given CyA compared with animals which received CyA vehicle. ([*]p<0.025 vs vehicle treated rats, [**]p
I=
~= ùu .: ~"
200~
[] CyA I-I V e h i c l e
lO0
o TxB2
6-keto-PGFlalpha
Figure 2: Urinary excretion of TxB2 and 6-keto-PGFla wem sign/ficanfly increased in rats g/ven CyA compared to an/mals treated with CyA vehicl¢. ([**]p<0.05 vs vehicle treated rats). Histomorphologic examinafion of formalJn t-lxed lädneys from rats treated with either CyA or CyA vehicle revealed minimal changes by üght microscopy. There were no detectable differences in light microscopic fmdings between the CyA and vehicle treated groups. In contrast, there wem some subtle but significant differences detected
FEBRUARY 1990 VOL. 39 NO. 2
139
PROSTAGLANDINS with immunoperoxidase staining. Cells bearing the OX-i (rat common leukocyte) or OX-42 (rat macrophage/monocyte) anfigens were found in focal aggregates or diffusely in the renal interstiüum of both CyA and vehicle treated animals. OX-1 or OX-42 positive cells were found only rarely in glomeruli of either group. There was a significantly greater number of OX-1 positive cells present in the renal intersfitium of CyA treated rats compared to vehicle treated animals (histologic score: 1.75_4"0.16 [CyA] vs 1.25_+.21 [vehicle]; p<0.05). However, there was no difference in the preponderance of interstitial OX-42 positive cells in rats given CyA compared to vehicle treated rats (histologic score: 1.0(~-4-0.15 [CyA] vs 1.10i-0.17 [vehicle]; p>0.40). The effect of the specific thromboxane receptor antagordst GR32191 on systemJc and renal hemodynamics in rats with CyA nephrotoxicity is shown in figure 3 and table 1. As secn in figure 3A, during administration of ll.tg/kg/min of GR32191, GFR increased by 22.7% from 3.04_+0.42 to 3.73_+0.54 tal/min/kg (p<0.05). Similarly, with 10 ~tg/kg/min of GR32191, GFR incrcased by 33.3% from 2.43"!-0.38 to 3.24:L'0.44 tal/min/kg (p<0.01). Following treatment with either 1 or i0 ~tg/kg/min, GFR in CyA toxic rats remained significantly lower (p<0.01) than GR32191 treated controls (5.27-Z-0.17 ml/min/kg). When a higher dose (500gg/kg/min) of receptor antagonist was given, a fall in GFR from 3.44+_0.60 to 1.78+_0.97 tal/min/kg (p<0.05) was observed. This decrcase in GFR was associatcd with a significant reduction in MAP from 89.0"~.7 to 73.3+13.5 mm Hg (1)<0.05) (Table 1). In addiüon, these animals developed evidence of intravascular hemolysis with reddish-brown discoloration of the plasma and urine associated with a falling hematocnt. Orte or 10~tg/kg/min GR32191 did not significanüy affect MAP in CyA toxic rats (Table 1). In control animals, GR32191 had no significant ¢ffects on systemic or renal hemodynamics (not shown). TABLE 1: Effect of GR32191 on MAP and RVR in CvA toxic rats MAP RVR (mm H«) (mm He/ml/min/k~) Control period 95.227.4 5.23_+0.72 GR32191 ( 1gg/kg/min) 104.7+8.6 4.62._+0.74 Control period GR32191 (10ttg/kg/min) Control period GR32191 (500tt~/k~/min~ *P<0.05 vs contr0fpeziod
96.4~.9 94.4+_2.7
5.73+1.04 4.27_+0.67
89.0B'9.7 73.3+13.5"
4.64:L-0.80 4.05_+1.49
The effects of the thtomboxane receptor antagonist on RBF in CyA toxic ardmals are shown in figure 3B. As seen in the figure, administration of lgg/kg/min of GR32191 caused renal blood flow to increase by 27.8% from 20.76d:3.05 to 26.54-+4.50 nil/min/kg (p<0.05). 10ttg/kg/min ofGR32191 caused renal blood flow to increase by a similar degree from 22.06"2.3.89 to 26.01+~.69 nil/min/kg (p<0.05). Following treatm¢nt with eithe~ 1 or 10 gg/kg/min of GR32191, renal blood flow in rats given CyA was restored to levels not sigräficantly different from GR32191 treated controls (25.99-2.2.30 nil/kg/min). In addition, infusion of the thromboxane receptor antagonist, reduced RVR in all rats given CyA (table 1) to levels not significantly diffexent from control animals (4.41:£-0.65 mm Hg/tal/min/kg).
140
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS
A.
.{ i
z-
O
2,
g
i •
~
3-
2
i
control
B.
.
i
1
GR32191 (1gg/kg/min)
45-
45-
35,
35,
*
25 ¸
E
25
n-
15
i
i
control
GR32191 (10p.g/kg/min)
control
GR32191 (101~g/kg/min)
.= m
=
15
control
GR32191 (11~oJkg/min)
Figure 3: Tzeatment with the specific thromboxane receptor antagonist GR32191 at l~tg/kg/min or 101~g/kg/min significantJy increased GFR (Fig 3A) and RBF (Fig 3B)in CyA treated animals. Closed cizcles represent data Ix)ints for individual animals. Open circles represent the mean values. ([**]p<0.01 vs baseline renal hemodynamics, [*]p<0.05 vs baseline renal hemodynamics)
DISCUSSION Although early observations in animal models suggested that cyclosporine might be a tubular toxin (31), recent animal studies and clinical observaüons have demonstrated that rcnal dysfunction associatcd with cyclosporine administrationis
mcdiated primar~y by h¢modynamic factors (8,9,10). Thus, CyA nephrotoxicityis
FEBRUARY 1990 VOL. 39 NO. 2
141
PROSTAG LAN Dl NS
characterized clinically by rapid reversal of renal dysfunction even after prolonged exposure to the drug (6,32). Although the specific mechanisms responsible for CyA nephrotoxicity are not known, recent evidence suggests that the vasoconstrietor eicosanoid TxA2 may be an important mediator of the renal hemodynamic alterations associated with administration of CyA (14,15). For example, urinary thromboxane excretion and renal thromboxane production am stimulated in animal models of CyA toxicity (16,17). In addition, fish oil administration, which reduces renal TxA2 production, improves function in CyA toxic rats (15). Finally, thromboxane synthase inhibition ameliorates CyA nephrotoxicity (14,19,20). To more specifically examine the role of thromboxane in CyA toxicity, we evaluated the effects of a thromboxane receptor antagonist in a rats #ven CyA. Thus, we hoped to bettet define the hemodynamic role of thromboxane in this process. In the present study, administration of CyA to rats reduced GFR and RBF and significanfly incmased RVR. These renal hemodynamic edterations wem associated with increased urinary excretion of unmetabolized, "native" TxB2. In rats given CyA, acute intraartefial infusion of the thromboxane receptor antagonist GR32191 at maximally tolerated doses significanfly incmased GFR, and restored RBF and RVR to levels not different from controls. These data provide additional evidence for a specific functional role of TxA2 in CyA nephrotoxicity. However, as in the studies using thromboxane synthase inkibitors (14,19,20), GFR was not normalized following acute treatment with the TxA2 receptor antagonisL The inability to restore GFR to levels comparable with control animals may be mlated to several factors. Since litfle is known about the pharmacology of mnal thromboxane receptors, it is possible that thromboxane receptor blockaäe by GR32191 was incomplete. Altemaüvely, more prolonged administration of the thromboxane receptor antagonist may be necessary to achieve the maximal benefits of therapy. In addition, other factors may be involved in the pathogenesis of CyA nephrotoxicity. For example, sfimulaüon of the renin-angiotensin system (12) or mnal sympatheüc nervous system (8,13) and alterations in the synthesis of the vasodilator prostaglandins (11,33,34) may be important in producing renal vasoconstricUon and contributing to the decrease in GFR associated with administration of CyA. In intact animals, cyclosporine administraüon has been generally reported to stimulate producüon of thromboxane as weg as other eicosanoids (15,17,18). However, the mechanism of this sümulatory effect is unclear. Furthermore, in a number of in v/tro systems, CyA appears to reduce producüon of arachidonic acid metabolites (11,33,35,36). The reasons for these apparent discmpant effects of CyA in whole animals compared to in vitro systems is not lmown. However, the results of this study along with pmvious reports (15,18,19,20) provides compelling evidence that stimulation of renal thromboxane producüon contributes, in part, to nephrotoxicity in animals given cyclosporine. The cellular source of increased thromboxane production in cyclosporine toxicity is not known. In other models of thromboxane dependent renal disease, inf'fltraüng macrophages appear to be important contributors to renal eicosanoid production in the lddney (37,38). In support of a potential role for infiltrating cells in CyA toxicity, several previous studies have suggested that infiltraüon of inflammatory cells into the kidney may occur in animals given cyclosporine. For example, Humes et al. (39) reported an increase in mononuclear cells expressing common leukocyte anügen in the renal intersüüum of cyclostx)rine treated rats. Benigni and coworkers (27) also reported
142
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS an increasednumbcr of mononuclear inflammatory cellsin glomeruli from ratswith C y A ncphrotoxicity. W h e n examined by electron microscopy, these ccUs had the morphologic characterisücsof macrophages. In addition,Rogcrs et al. (28) reccntly rcportcd thatmacrophages from C y A treatedanimals exhibitenhanced thromboxane production. However, in both humans and animals receiving CyA, an increase in inflammatory cells within the kidney have not been observed consismntly (5,31,40). In the prescnt study, we found a significant increase in cells expressing the rat leukocyte c o m m o n antig¢n in the rcnal interstitiumof animals givcn C y A compared to vchiclc ~eatcd conu'ols. However, there was no increasein ceUs labcllingwith the OX-42 macrophagelmonocyte marker in the renalinterstiüum or in glomcruliof C y A toxic rats. In summary, in a rat model, treatmentwith C y A reduced G F R and RBF, and incrcascdRVR. These renalhemodynamic abnormalitieswcre associatedwith elevatcd urinarycicosanoidexcretionand an increascdnumber of mononuclcar ccUs in the rcnal interstitiumexpressingrat c o m m o n leukocyte antigcn. Administrationof a specific t~omboxane r¢ceptor antagonistimproved renal hemodynamics as evidcnced by a significant~but incomplete,recovery of GFR, and restorationof R B F and R V R to levcls comparablc to controls. These dataprovide futter evidencc supportinga rolefor TxA2 in rcnal dysfunction associatedwith C y A toxicity. Furthcrmorc, they suggcst that thromboxane antagortismmay be usefulin the trcatmcntof patientswith cyclosporinc ncphrotoxicity.
Acknowledgments These studies wem supported by grams from the Rcsearch Service of the Vetcrans Admim'stration and the NIH (P01 DK38108-01A1). Portions of this work wem pr¢sented at the 21st Annual M¢cting of the American Society of Nephrology, and wcrc published in abstract form in Kidney Int 35:505A, 1989. The authors wish to thank Chris Best, Pat Rannery, Julie Full¢r, and Joseph Swanson for their expert technical assistance and Ms. Norma Turner for superb secretarial assistance. Cyclosporine was provided by Sandoz, Inc. and GR32191 was provid¢d by Glaxo, Inc.
References I) European Multic¢ntre Trial Group. Cyclosporin in cadavedc r¢nal transplantation: one-ye,ar follow-up of a multicentm trial. Lanc¢t 2:986-989, 1983.
2) Canadian Mulücentrc Transplant Study Group. A randomized clinical trial of cyclosporine in cadaveric renal transplantation. N Engl J Med 314:1219-1225, 1986.
3) Isenberg DA, Snaith ML, Morrow WJ, AI-Khader AA, Cohen SL, Fish¢r CC, Mowbray J. Cyclosporin A for treatment of systerräc lupus erythematosus. Int J Immunopharmacol 3:163-169, 1981.
4) Palesüne AG, Austin HA, Balow JE, Antonovych TT, Sabnis SG, Preuss HG, Nussenblatt RB. R¢nal histologic aJterafions in paüents treated with cyclosporinc for uveiüs. N Engl J Med 314:1293-1298, 1986. 5) Myers BD, Ross J, Newton L, Luetscher J, Perlroth M. Cyclosporine-associated chronic nephropathy. N Engl J Med 311".699-705, 1984.
FEBRUARY 1990 VOL. 39 NO. 2
143
PROSTAGLANDINS 6) Klintmalm GB, Iwatasulfi S, Starzel TW. Nephrotoxicity of cyclosporin A in liver and kidney u'ansplantrecipients. Lancet 1:470-471, 1981.
7) Myers BD, Sibley R, Newton L, Tomlanovich SJ, Boshkos C, Stinson E, Luetscher JA, Whitney DJ, Krasny D, Coplon NS, Perlroth MG. The long-term course of cyclosporine-associatedchronic nephropathy. Kidney Int 33:590-600, 1988.
8) Murray BM, Paller MS, Ferris TF. Effect of cyclosporineadministrationon renal hemodynamicsin conscious rat.s. Kidney Int 28:76%774, 1985. 9) Curüs JJ, Luke RG, Dubovsky E, Diethelm AG, Welchel JD, Jones P. Cyclosporin-A in therapeutic doses increases renal allograft vascular resistance.Lancet 1:477-479, 1986.
1o) English J, Andrew E, Houghton DC, B¢nnett WM. Cyclosporine-inducedacute renal dysfunctionin the rat. Transplantation44:135-141, 1987.
11) Neild GH, Rocchi G, Imab¢rü L, Fumagalli F, Brown Z, Remuzzi G, Williams DG. Effect of cyclosporineon prostacyclinsynthesis by vascular tissue in rabbits. Transplant Proc 4(suppl 1):2398-2400, 1983. 12) Siegl H, Ryffel B, Shoemaker P, Muller A. Conatsch P, Mihatsch M. Cyclosporine, the renin-angiotensin-aldosterone system and renal adverse reacüons. Transplant Proc ~(suppl 1):2719-2725, 1983. 13) Moss NG, Powell SL, Falk RJ. Inwavenouscyclosporine acüvates afferent and efferent renal nerves and causes sodium retention in innervated kidneys in rats. Proc Natl Acad Sci USA 82: 8222-8226, 1985.
14) Perico N, Benigni A, Zoja C, Delaini F, Remuzzi G. Funcüonal significance of exaggerated renal thromboxane A2 synthesis induced by cyclosporin A. Am J Physiol 251:F581-587, 1986.
15) Elzinga L, Kelley VE, HoughtonDC, Bennett W1VI. Modificationof experimental nephrotoxicity with fish oil as the vehicle for cyclosporine. Transplantation 43:271-274, 1987. 16) Kawaguchi A, Golman MH, Shapiro R, Foegh ML, Ramwell PW, Lower RR. Increas¢ in urinary thromboxane B2 in rats caused by cyclosporine. Transplantaüon 40:214-216, 1985. 17) Coffman TM, Carr DR, Yarg«r WE, Klotman PE. Evidenc« that renal prostaglandin and thromboxane production is sümulated in chronic cyclosporine nephrotoxicity.Transplantation43:282-285, 1987. 18) Perico N, Zoja C, Benigni A, Ghilardi F, Gualandris L, Remuzzi G. Effect of short-term cyclosporine administration in rats on renin-angiotensin and thromboxaneA2: Possiblerelevanceto the reductionin glomeralarfiltration rate. J Pharmacol Exp Ther. 239:229-235, 1986.
144
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS 19) Petric R, Freeman D, Wallace A, McDonald J, Stiller C, Keown P. Protective effect of thromboxane synthetase inhibition in the rat with eyclosporine nephrotoxicity. The second international eongress on cyclosporine, Washington, D.C., USA, November,1987, p. 38.
20) Smeesters C, Chaland P, Giroux L, Moutquin J, Etienne P, Douglas F, Corman J, St Louis GS, Daloze P. Prevention of acute cyclosporine A nephrotoxicity by a thromboxane synthetase inhibitor. Transplant Proc (suppl 3):658-664, 1988. 21) Oates JA, Fitzgerald GA, Branch RA, Jackson EK, Knapp HR, Roberts I_3. Clinical implicaüons of prostaglandin and thromboxane A2 formation (first of two patts). N Engl J Med 319:689-698, 1988. 22) FitzGerald GA, Oates JA. Selecfive and nonselective inhibifion of thromboxane formation. Clin Pharmaeol Ther 35:633-640, 1984. 23) FitzGerald GA, Reilly LAG, Pedersen AK. The biochemical pharmacology of thromboxane synthase inhibition in man. Circ 72:1194-1201, 1985. 24) Homby EJ, Foster MR, McCabe PJ, Stratton LE. The inhibitory effect of GR32191, a thromboxane receptor blocking drug, on human platelet aggregaüon, adhesion and secretion. Thromb Haemostas 61:429-436, 1989. 25) Lumley P, White BP, Humphrey PPA. GR32191, a highly potent and specific thromboxane A2 receptor blocking drug on platelets and vascular and airways smooth muscle in vitro. Br J Pharmaco197:783-794, 1989. 26) Coffman TM, Ruiz P, Sanfilippo F, Klotman PE. Chronic thromboxane inhibition preserves funcüon of rejecUon rat renal a11ografts. Kidney Int 35:24-30, 1989. 27) Benigni A, Chibrando C, PiccineUi A, Perico N, Gavinelli M, Furci L, Patino O, Abbate M, Bertani T, Remuzzi G. Incmased urinary excretion of thromboxane B2 and 2,3-dinor-TxB2 in cyclosporin A nephrotoxicity. Kidney Int 34:164-174, 1988. 28) Rogers TS, Elzinga L, Bennett W, Kelley VE. Selective enhancement of thromboxane in macrophages and kidneys in cyclosporine-inducednephrotoxicity. Transplantation, 45:153-156, 1988. 29) Ruiz P, Coffman TM, HoweU DN, Straznickas J, Scroggs MW, Baldwin WM, Klotman PE, Sanfdippo F. Evidence that pretransplant donor blood transfusion prevents rat renal aUograft dysfuncüon but not the in situ cellular alloimmune or morphologic manifestations of rejeeüon. Transplantation 45:1-7, 1988.
30) Wallenstein S, Zucker CL, Fleiss JL. Some statistical methods useful in circulaüon research. Circ Res 47:1-9, 1980. 31) Whiting PH, Thomson AW, Blair JT, Simpson JG. Exp¢rim¢ntal cyclosporin A nephrotoxicity. Br J Exp Path. 63: 88-94, 1982.
FEBRUARY 1990 VOL. 39 NO. 2
145
PROSTAGLANDINS
32) Chapman ]'R, Gfiffiths D, Harding NG, Morris PH. Reversibility of cyclosporine nephrotoxicity after three months' tmatment. Lancet 1:128-130, 1985. 33) Stahl RA, Kudelka S. Chronic cyclosporine A treatment reduces prostaglandin E2 formaüon in isolated glomenfli and papilla of rats kidneys. Clin Nephro125:(suppl 1):$78-$82, 1986. 34) Paller MS. The prostaglandin E1 analog misoprostol reverses acute cyclosporine nephrotoxicity. Transplant Proc (suppl 3): 634-637, 1988. 35) Bunke L, Wilder L. Cyclosporine inhibits mesangial cell prostaglandin producüon.Kidney Int 35:402, 1989. 36) Stahl RAK, Adler S, Baker PJ, Johnson RJ, Chen YP, Pritzl P, Couser WG. Cyclosporin A inhibits prostaglandin E2 formaüon by rat mesangial eells in culture. Kädney Int 35:1161-1167, 1989. 37) Okegawa T, Ionas PE, DeSchryver K, Kawasaki A, Needleman P. Metabolic and cellular alterations underlying the exaggerated renal prostaglandin and thromboxane synthesis in ur¢ter obstrucüon in rabbits. J Clin Invest 71:81-90, 1983. 38) Kelley VE, Altboum I, Boswell I Thromboxane producüon by acüvated renal cortical macrophages. Fed Proc 46:1330, 1987. 39) Humes HD, Jackson NM. Cyclosporine effects on isolated membranes, proximal mbule cells, and interstiüum of the kiduey. Transplant Proc 3(suppl 3):748-758, 1988.
40) Verani R. Cyclosporine nephrotoxiäty in the Fischer raL Clin Nephrol 25(suppl 1): $9-S13, 1986.
Editcr: I~~ Pamwell
146
I~ceived: 10-26-89
Acoepted: 12-8-89
FEBRUARY 1990 VOL. 39 NO. 2
ABSTRACT Cyclosporine A (CyA) nephrotoxicity is associated with impaired renal hemodynamic function and increased production of the vasoconstrictor eicosanoid thromboxane A2 (TxA2). In CyA toxic rats, lenal dysfunction can be parüally reversed by inhibitors of thromboxane synthase. However, interpr¢taüon of these results is complicated since inlübiüon of thromboxane synthase may cause accumulafion of prostaglandin endoperoxides that can act as parüal agonists at the TxA2 receptor and may blunt the efficacy of treatrnent. Furthermore, these endoperoxides may be used as substrate for producüon of vasodilator prostaglandins causing beneficial effects on hemodynamics which am independent of thromboxane inhibition. To more specifically examine the role of TxA2 in CyA toxicity, we investigated the effects of the thromboxane receptor antagonist GR32191 on renal hemodynamics in a rat model of CyA nephrotoxicity. In this model, administration of CyA resulted in a significant decrease in glomer-!ar f'fllration rate (GFR) (2.85"~0.26 [CyA] vs 6.82-2-2-2~.96tal/min/kg [vehicle]; p<0.0005) and renal blood flow (RBF) (21.65.+.2.31 [CyA] vs 31.87_+3.60 n~/min/kg [vehicle]; p<0.025). Renal vascular resistance (RVR) was significanfly higher in rats given CyA compared to animals treated with CyA vehicle (5.32.%0.55 [CyA] rs. 3.54:£-0.24 mm Hg/min/tal/kg [vehicle]; p<0.05). These renal hemodynamic alteraüons were associated with a significant increase in urinary excreüon of unmetabolized, "nativ¢" thromboxane B2 (TxB2) (103+18 [CyA] vs 60:1:16 pg/hour [vehicle]; p<0.05). Only minimal histomorphologic chang¢s were apparent by light microscopic examinafion of kidneys from both CyA and vehicle ffeated animals. However, with immunoperoxidase staining, a significanfly greater number of cells expressing the rat cornmon lenkocyte antigen was found in the renal interstiüum of rats given CyA*. There was no detectable increase in monocytes/macrophages in the kidneys of CyA toxic anirnals. In rats treated with CyA, intraarwrial infusion of GR32191 at maximally tolerated doses significantly increased GFR and RBF, and decreased RVR. Although both RBF and RVR were restored to levels not different from controls, GFR remained significantly reduced following administration of GR32191. These data suggest that the potent vasoconstrictor TxA2 plays an important role in mediating renal dysfuncüon in CyA nephrotoxicity. Howevet, other factors may be important in producing nephrotoxicity associated with CyA. INTRODUCTION CyA is a potent immunosuppressive agent which has been extremely effective in prolonging aUograft survival (1,2) and suppressing autoimmune disease (3,4). Unfortunately, the clinical use of CyA has been complicated by a significant incidence of drug-related nephrotoxicity (5). The nephrotoxicity is characterized by a dosedependent decrease in GFR that may limit the c"lmical utility of the drug (6,7).
FEBRUARY 1990 VOL. 39 NO. 2
135
PROSTAGLANDINS The specific mechanisms responsible for CyA nephrotoxicity are not known. How¢ver, in both humans and animals given CyA, renal vasoconstriction has been demonstrated (7,8,9). Furthermore, in a rat model, EngIish et al. (I0) have shown that renal hemodynamic alterations rather than direct tubular injury are important in the pathogenesis of nephrotoxicity. Although several potential mediators have been proposed (11,12,13), recent observations suggest that the vasoconstrictor eicosanoid TxA2 may be important in producing the renal hemodynamic alterations associated with CyA administration (14,15). For exampte, in animal models of CyA nephrotoxicity, utinary thromboxane excretion is increased (16,17) and a significant negative correlation between urinary TxB2 excretion and OFR has be.en reported (14). In addition, thromboxane production by the kidney is increased (15,17,18) and this enhancement of renal thromboxane production is associated with reduced OFR and RBF (15,17,18). Evidence for a functional role for thromboxane was suggested in studies by Elzinga and coworkers (15) who demonstrated that fish oft administration improved renal function and reduced renal cortical thromboxane production in rats given CyA. In addition, Perico et aL (14) showed that selective inhibition of thromboxane production with a specific thromboxane synthase inhibitor significantly improved GFR in rats with CyA nephrotoxicity. However, following treatment, GFR remained below normal Two other groups have also reported beneficial effects of thromboxane synthase inhibitors in rats with CyA toxicity (19,20). However, physiologic interpretation of studies using thromboxane synthase inhibitors is complicated. Inhibition of thromboxane synthase may result in accumulaüon of the pt'ostaglandin endoperoxides which can act as agonists at the thromboxane receptor (21) and may blunt the efficacy of treatment. Also, these accumulated endoperoxides may be metabolized to form vasodilator prostaglandins (22,23). Thus, some of the beneficial effects of thromboxane synthase inhibitors may relate to increased production of vasodilator prostaglandins rather than specific inhibiüon of the acüons of TxA2. The alm of the present s~iies was to more specifically examine the functional role of ~omboxane in CyA toxicity. To address this issue, we evaluated the effects of the specific thromboxane receptor antagonist GR32191 ([1R.[l(Z),2,3,5]]-(+)-7-[5[[(1, l'-biphenyl)-4-yl] methoxy]-3-hydroxy-2(1-piperidinyl).cyclopentyl]-4-beptenoic acid, hydrocMoride) (24,25) on renal hemodynanücs m a rat model of CyA nephrotoxicity. This compoand offers the advantage of specifically interfering with the interaction of TxA2 and its rec¢ptor without directly affecting production of arachidonic acid metabolites. METHODS
CyA Nephroroxicity: CyA nephrotoxicity was produced in male ACI rats weighing 150-250 grams as previously described (17). Animals wem anesthetized with halothane and the right kidney was removed by tank incision. To simulate condiüons associated with r¢nal transplantaüon and to eliminate effects of i n ~ sympathctic tone (8,13), all animals underwent denervation of the left kidney followed by cross-clamping of left renal artery for 30 minutes. Denervation was accomplished by disruption of visible nerve bundles followed by application of a 10% phenol solution to the mnal pediceL Following surgery, rats were given CyA (50mg/kg) or its vehicle (olive oft) by daily intraperitoneal injection. Cyclosporine treated rats and vehicle treated animals were studied 12-14 days after op¢raüon. On the day before the renal hemodynamic study, al~imals were placed in metabolic cages and urine was collected for 24 hours. The urine was chiUed throughout the collection period and then stored at minus 70°C and later
136
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS
assayed for TxB2 and 6-keto-prostaglandin F l a (6-keto-PGFla) by radioimmunoassay (RIA) following HPLC separation as described below.
RenalHemodynamicStudies:
Hemodynamic studies were performed as described previously (17,26). On the day of study, animals were anesthetized with 0.04mg/gram pentobarbital, and a polyethylene catheter (PE 240) was inserted into the trachea to facilitate spontaneous venfilation. The left caroti_d artery and leit jugular vein were cannulated with polyethylone catheters ~E-50) for intravenous infusions, to monitor mean arterial pressure (MAP) (Gould-Statham strain gauge), and to allow intermittent sampling of arterial blood. Following surgery, normal saline (2.0% of the body weight) was infused intravenously over 20 nünutes to replace surgical losses. A priming dose of carboxyl-14C-inulin and glycyl-3H-PAH was given, followed by infusion of pentobarbital, carboxyl-14C-inulin, and glycyl-3H-PAH in normal saline at a rate of 25 gl/minute/100gram body weight. The leftrenalvein was catheterizedwith a curved 0.965-mm OD teflon catheter inserted through the left femoral vein to collect samples of renal venous blood. The left meter was cannulated with a PE-10 catheter to facilitate collecüon of urine. A polyethylene catheter (PE-10) was inserted throngh the femoral artery and positioned in the aorm just above the origin of the ler renal artery for in_fusion of the thromboxane receptor antagonist GR32191 or its vehicle (0.45% normal saline) at 25 pJ/min. After 30 minutes of equilibration, inulin clearance and RBF were measured during two consecutive 30 minute periods. Foilowing these baseline renal hemodynamic measurements, the thromboxane receptor antagonist GR32191 was infused into the renal attery. After 30 minutes of equilibrafion,renal function was measured during two additional30 minute pcriods. C y A treatedanimals received either Il.ig/kg/min,10gg/kg/min, or 500gg/kg/min of GR32191. In addition,as con~Is, a group of animals treatcdwith C y A vehiclewere given GR32191 at a dose of 10~tg/kg/min. Clearances of inulin and P A H were calculatedusing standardformulas. Renal P A H extracüon (E-PAH) was determined and R B F was calculatedusingthe formula: R B F = C P A H / [ I ~ A H x (1 - Hct)]. R V R was cstimatedby dividingmean arterialpressureby RBF. After the hemodynamic studies, the rare were sacrificed and the l e r kidney was excised. A portion of the kidney was fLxed in f o r m a ~ for histologic examinaüon. The rcmaining kidney tissuewas frozenin liquidnitrogenand savcd for immunohistology as outlincdbelow.
Eicosanoid Assays: Using prcviously described m¢thods (17,26), eicosanoids wem exwacteclfrom urine using Sep-Pak C18 cartridges(Waters Associates,Milford,Mass.) and then separated using a Waters Model 840 system and Pccosphcre HS3-C18 colunm. Eicosanoids were eluted with a linear gradient from 100% 0.017 M orthophosphoric acid to 100% acetonitrile over 10 minutes at 3 tal/minute. Appropriate fractions were collectedbased on retenfionümes of known standards. The eluate was dried under nitrogen and resuspended in RIA buffer. All samples were correctedfor recovery of tritiatedstandardsadded to the originalsamples. RIA rcagents wem purchased from AMI, Boston, M A and IC'N,Lysle,IL.
FEBRUARY 1990 VOL. 39 NO. 2
137
PROSTAGLANDINS For 6-keto-PGFla, sample and standards wem incubated with a mixture of anüsera and a known amount of tritiated standard at 4°C for 20 hours. For TxB2, the procedure was similar except the incubation time was 1 hour. After incubafion, free cicosanoid was adsorbcd with dextran-coatedcharcoal,and the Iritiumremaining in the supcmatant was measured with a liquidscinüllationcounter. The unknowns, corrected for recovery, were compared with a standard curve in which the logarithm of the concentraüon was plottedagainstthe logitof the B/Bo value. The resultsarc expresscd as pg/minutc of eicosanoid.
Histologic Studie.s:
Formalln fixed tissue stained with hematoxylin and eosin was examined by light microscopy. In addition, to bettet evaluate inflammatory cell infütrates, immunohistologic studies were performed on sections comparing rats given CyA to animals treated with CyA vehicle. Since several reports have suggested that macrophages may be an important cellular source of enhanced thromboxane production in CyA n¢phrotoxicity (27,28), phenotypes of infiltrating cells were evaiuated using monoclonaI antibodies against rat leukocyte-common antigen (OX-I), and rat monocytes and macrophages (OX-42). Immunoperoxidase staining of sections of fresh frozen üssue was performed as previously reported (29). Light microscopic and immunoperoxidase studies were reviewed by a pathologist (PR) blinded to the trcatment groups. Light microscopic scctions w e m examined for tubulointerstitial,vascular,and glomerular lesions. Abnormalities w e m graded using a semiquan•taüve scale where 0 was no abnormality,and 1+, 2+, 3+, and 4+ represented mild, mod¢rate, moderatcly-s¢vere, and scvere abnormalities, respectively. Immunoperoxidase stained sections wem examined for the presence of OX-1 or OX-42 positive cells in glomeruli, mbtdes, vessels, or renal intersütium. The severity of inflammatory cell infillrates was graded semiquantitatively where 0 indicated the absence of cells staining with the OX-42 or OX-1 monoclonal antibody, and 1+ to 4+ represented minimal to marked infiltration by OX-42 or OX-1 positive cells.
Statistical Analysis:
Data am presented as the mean + standard error of the mean. For the hemodynamic smdies, data points for each animal represent the mean of the values measured during the two clearance periods. For comparisons between two groups, staüsücal significance was assessed using an unpaired t-test. For comparisons within the same group a paired t-test was used. For comparisons between more than two groups, statistical analysis was performed by one-way analysis of variance followed by Bonferroni's procedure for multiple pairwise comparisons (30).
RESULTS The effectof C y A on baselineG F R and R B F is shown in Figure i. Both G F R (2.85+0.26 [CyA] vs. 6.82.~:0.96 tal/min/kg [vehicle]; p<0.0005) and R B F (21.65-+-2.31[CyA] vs. 31.87:1:3.60tal/min/kg [vehicle];p<0.025) w e m significantly reduced in rats given CyA compared to animals which rec¢ived CyA v¢hicle. In addition, calculated RVR was significantly increased in CyA treated rats compared to vehicle treated animals (5.32_+0.55 [CyA] vs. 3_54~0.24 mm Hg/tal/min/kg [vehicle]; p<0.05). Baseline nw,an arterial pressures measured at the time of smdy wem sinülar in CyA and vehicle treated rats (94.33+4.01 [CyA] vs 104+_5.32 mm Hg [vehicle]; p=NS).
138
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAG LAN Dl NS
[ ] CyA 7.0
- -
~ . 0 --
GFR (tal/min/kg)
3 . 0 --
l GFR
T
i
--
35
[] --
25
Vehicle
RBF
(tal/min/kg) --
15
RBF
Figure 1: Both GFR and RBF wem significantly reduced in rats given CyA compared with animals which received CyA vehicle. ([*]p<0.025 vs vehicle treated rats, [**]p
I=
~= ùu .: ~"
200~
[] CyA I-I V e h i c l e
lO0
o TxB2
6-keto-PGFlalpha
Figure 2: Urinary excretion of TxB2 and 6-keto-PGFla wem sign/ficanfly increased in rats g/ven CyA compared to an/mals treated with CyA vehicl¢. ([**]p<0.05 vs vehicle treated rats). Histomorphologic examinafion of formalJn t-lxed lädneys from rats treated with either CyA or CyA vehicle revealed minimal changes by üght microscopy. There were no detectable differences in light microscopic fmdings between the CyA and vehicle treated groups. In contrast, there wem some subtle but significant differences detected
FEBRUARY 1990 VOL. 39 NO. 2
139
PROSTAGLANDINS with immunoperoxidase staining. Cells bearing the OX-i (rat common leukocyte) or OX-42 (rat macrophage/monocyte) anfigens were found in focal aggregates or diffusely in the renal interstiüum of both CyA and vehicle treated animals. OX-1 or OX-42 positive cells were found only rarely in glomeruli of either group. There was a significantly greater number of OX-1 positive cells present in the renal intersfitium of CyA treated rats compared to vehicle treated animals (histologic score: 1.75_4"0.16 [CyA] vs 1.25_+.21 [vehicle]; p<0.05). However, there was no difference in the preponderance of interstitial OX-42 positive cells in rats given CyA compared to vehicle treated rats (histologic score: 1.0(~-4-0.15 [CyA] vs 1.10i-0.17 [vehicle]; p>0.40). The effect of the specific thromboxane receptor antagordst GR32191 on systemJc and renal hemodynamics in rats with CyA nephrotoxicity is shown in figure 3 and table 1. As secn in figure 3A, during administration of ll.tg/kg/min of GR32191, GFR increased by 22.7% from 3.04_+0.42 to 3.73_+0.54 tal/min/kg (p<0.05). Similarly, with 10 ~tg/kg/min of GR32191, GFR incrcased by 33.3% from 2.43"!-0.38 to 3.24:L'0.44 tal/min/kg (p<0.01). Following treatment with either 1 or i0 ~tg/kg/min, GFR in CyA toxic rats remained significantly lower (p<0.01) than GR32191 treated controls (5.27-Z-0.17 ml/min/kg). When a higher dose (500gg/kg/min) of receptor antagonist was given, a fall in GFR from 3.44+_0.60 to 1.78+_0.97 tal/min/kg (p<0.05) was observed. This decrcase in GFR was associatcd with a significant reduction in MAP from 89.0"~.7 to 73.3+13.5 mm Hg (1)<0.05) (Table 1). In addiüon, these animals developed evidence of intravascular hemolysis with reddish-brown discoloration of the plasma and urine associated with a falling hematocnt. Orte or 10~tg/kg/min GR32191 did not significanüy affect MAP in CyA toxic rats (Table 1). In control animals, GR32191 had no significant ¢ffects on systemic or renal hemodynamics (not shown). TABLE 1: Effect of GR32191 on MAP and RVR in CvA toxic rats MAP RVR (mm H«) (mm He/ml/min/k~) Control period 95.227.4 5.23_+0.72 GR32191 ( 1gg/kg/min) 104.7+8.6 4.62._+0.74 Control period GR32191 (10ttg/kg/min) Control period GR32191 (500tt~/k~/min~ *P<0.05 vs contr0fpeziod
96.4~.9 94.4+_2.7
5.73+1.04 4.27_+0.67
89.0B'9.7 73.3+13.5"
4.64:L-0.80 4.05_+1.49
The effects of the thtomboxane receptor antagonist on RBF in CyA toxic ardmals are shown in figure 3B. As seen in the figure, administration of lgg/kg/min of GR32191 caused renal blood flow to increase by 27.8% from 20.76d:3.05 to 26.54-+4.50 nil/min/kg (p<0.05). 10ttg/kg/min ofGR32191 caused renal blood flow to increase by a similar degree from 22.06"2.3.89 to 26.01+~.69 nil/min/kg (p<0.05). Following treatm¢nt with eithe~ 1 or 10 gg/kg/min of GR32191, renal blood flow in rats given CyA was restored to levels not sigräficantly different from GR32191 treated controls (25.99-2.2.30 nil/kg/min). In addition, infusion of the thromboxane receptor antagonist, reduced RVR in all rats given CyA (table 1) to levels not significantly diffexent from control animals (4.41:£-0.65 mm Hg/tal/min/kg).
140
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS
A.
.{ i
z-
O
2,
g
i •
~
3-
2
i
control
B.
.
i
1
GR32191 (1gg/kg/min)
45-
45-
35,
35,
*
25 ¸
E
25
n-
15
i
i
control
GR32191 (10p.g/kg/min)
control
GR32191 (101~g/kg/min)
.= m
=
15
control
GR32191 (11~oJkg/min)
Figure 3: Tzeatment with the specific thromboxane receptor antagonist GR32191 at l~tg/kg/min or 101~g/kg/min significantJy increased GFR (Fig 3A) and RBF (Fig 3B)in CyA treated animals. Closed cizcles represent data Ix)ints for individual animals. Open circles represent the mean values. ([**]p<0.01 vs baseline renal hemodynamics, [*]p<0.05 vs baseline renal hemodynamics)
DISCUSSION Although early observations in animal models suggested that cyclosporine might be a tubular toxin (31), recent animal studies and clinical observaüons have demonstrated that rcnal dysfunction associatcd with cyclosporine administrationis
mcdiated primar~y by h¢modynamic factors (8,9,10). Thus, CyA nephrotoxicityis
FEBRUARY 1990 VOL. 39 NO. 2
141
PROSTAG LAN Dl NS
characterized clinically by rapid reversal of renal dysfunction even after prolonged exposure to the drug (6,32). Although the specific mechanisms responsible for CyA nephrotoxicity are not known, recent evidence suggests that the vasoconstrietor eicosanoid TxA2 may be an important mediator of the renal hemodynamic alterations associated with administration of CyA (14,15). For example, urinary thromboxane excretion and renal thromboxane production am stimulated in animal models of CyA toxicity (16,17). In addition, fish oil administration, which reduces renal TxA2 production, improves function in CyA toxic rats (15). Finally, thromboxane synthase inhibition ameliorates CyA nephrotoxicity (14,19,20). To more specifically examine the role of thromboxane in CyA toxicity, we evaluated the effects of a thromboxane receptor antagonist in a rats #ven CyA. Thus, we hoped to bettet define the hemodynamic role of thromboxane in this process. In the present study, administration of CyA to rats reduced GFR and RBF and significanfly incmased RVR. These renal hemodynamic edterations wem associated with increased urinary excretion of unmetabolized, "native" TxB2. In rats given CyA, acute intraartefial infusion of the thromboxane receptor antagonist GR32191 at maximally tolerated doses significanfly incmased GFR, and restored RBF and RVR to levels not different from controls. These data provide additional evidence for a specific functional role of TxA2 in CyA nephrotoxicity. However, as in the studies using thromboxane synthase inkibitors (14,19,20), GFR was not normalized following acute treatment with the TxA2 receptor antagonisL The inability to restore GFR to levels comparable with control animals may be mlated to several factors. Since litfle is known about the pharmacology of mnal thromboxane receptors, it is possible that thromboxane receptor blockaäe by GR32191 was incomplete. Altemaüvely, more prolonged administration of the thromboxane receptor antagonist may be necessary to achieve the maximal benefits of therapy. In addition, other factors may be involved in the pathogenesis of CyA nephrotoxicity. For example, sfimulaüon of the renin-angiotensin system (12) or mnal sympatheüc nervous system (8,13) and alterations in the synthesis of the vasodilator prostaglandins (11,33,34) may be important in producing renal vasoconstricUon and contributing to the decrease in GFR associated with administration of CyA. In intact animals, cyclosporine administraüon has been generally reported to stimulate producüon of thromboxane as weg as other eicosanoids (15,17,18). However, the mechanism of this sümulatory effect is unclear. Furthermore, in a number of in v/tro systems, CyA appears to reduce producüon of arachidonic acid metabolites (11,33,35,36). The reasons for these apparent discmpant effects of CyA in whole animals compared to in vitro systems is not lmown. However, the results of this study along with pmvious reports (15,18,19,20) provides compelling evidence that stimulation of renal thromboxane producüon contributes, in part, to nephrotoxicity in animals given cyclosporine. The cellular source of increased thromboxane production in cyclosporine toxicity is not known. In other models of thromboxane dependent renal disease, inf'fltraüng macrophages appear to be important contributors to renal eicosanoid production in the lddney (37,38). In support of a potential role for infiltrating cells in CyA toxicity, several previous studies have suggested that infiltraüon of inflammatory cells into the kidney may occur in animals given cyclosporine. For example, Humes et al. (39) reported an increase in mononuclear cells expressing common leukocyte anügen in the renal intersüüum of cyclostx)rine treated rats. Benigni and coworkers (27) also reported
142
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS an increasednumbcr of mononuclear inflammatory cellsin glomeruli from ratswith C y A ncphrotoxicity. W h e n examined by electron microscopy, these ccUs had the morphologic characterisücsof macrophages. In addition,Rogcrs et al. (28) reccntly rcportcd thatmacrophages from C y A treatedanimals exhibitenhanced thromboxane production. However, in both humans and animals receiving CyA, an increase in inflammatory cells within the kidney have not been observed consismntly (5,31,40). In the prescnt study, we found a significant increase in cells expressing the rat leukocyte c o m m o n antig¢n in the rcnal interstitiumof animals givcn C y A compared to vchiclc ~eatcd conu'ols. However, there was no increasein ceUs labcllingwith the OX-42 macrophagelmonocyte marker in the renalinterstiüum or in glomcruliof C y A toxic rats. In summary, in a rat model, treatmentwith C y A reduced G F R and RBF, and incrcascdRVR. These renalhemodynamic abnormalitieswcre associatedwith elevatcd urinarycicosanoidexcretionand an increascdnumber of mononuclcar ccUs in the rcnal interstitiumexpressingrat c o m m o n leukocyte antigcn. Administrationof a specific t~omboxane r¢ceptor antagonistimproved renal hemodynamics as evidcnced by a significant~but incomplete,recovery of GFR, and restorationof R B F and R V R to levcls comparablc to controls. These dataprovide futter evidencc supportinga rolefor TxA2 in rcnal dysfunction associatedwith C y A toxicity. Furthcrmorc, they suggcst that thromboxane antagortismmay be usefulin the trcatmcntof patientswith cyclosporinc ncphrotoxicity.
Acknowledgments These studies wem supported by grams from the Rcsearch Service of the Vetcrans Admim'stration and the NIH (P01 DK38108-01A1). Portions of this work wem pr¢sented at the 21st Annual M¢cting of the American Society of Nephrology, and wcrc published in abstract form in Kidney Int 35:505A, 1989. The authors wish to thank Chris Best, Pat Rannery, Julie Full¢r, and Joseph Swanson for their expert technical assistance and Ms. Norma Turner for superb secretarial assistance. Cyclosporine was provided by Sandoz, Inc. and GR32191 was provid¢d by Glaxo, Inc.
References I) European Multic¢ntre Trial Group. Cyclosporin in cadavedc r¢nal transplantation: one-ye,ar follow-up of a multicentm trial. Lanc¢t 2:986-989, 1983.
2) Canadian Mulücentrc Transplant Study Group. A randomized clinical trial of cyclosporine in cadaveric renal transplantation. N Engl J Med 314:1219-1225, 1986.
3) Isenberg DA, Snaith ML, Morrow WJ, AI-Khader AA, Cohen SL, Fish¢r CC, Mowbray J. Cyclosporin A for treatment of systerräc lupus erythematosus. Int J Immunopharmacol 3:163-169, 1981.
4) Palesüne AG, Austin HA, Balow JE, Antonovych TT, Sabnis SG, Preuss HG, Nussenblatt RB. R¢nal histologic aJterafions in paüents treated with cyclosporinc for uveiüs. N Engl J Med 314:1293-1298, 1986. 5) Myers BD, Ross J, Newton L, Luetscher J, Perlroth M. Cyclosporine-associated chronic nephropathy. N Engl J Med 311".699-705, 1984.
FEBRUARY 1990 VOL. 39 NO. 2
143
PROSTAGLANDINS 6) Klintmalm GB, Iwatasulfi S, Starzel TW. Nephrotoxicity of cyclosporin A in liver and kidney u'ansplantrecipients. Lancet 1:470-471, 1981.
7) Myers BD, Sibley R, Newton L, Tomlanovich SJ, Boshkos C, Stinson E, Luetscher JA, Whitney DJ, Krasny D, Coplon NS, Perlroth MG. The long-term course of cyclosporine-associatedchronic nephropathy. Kidney Int 33:590-600, 1988.
8) Murray BM, Paller MS, Ferris TF. Effect of cyclosporineadministrationon renal hemodynamicsin conscious rat.s. Kidney Int 28:76%774, 1985. 9) Curüs JJ, Luke RG, Dubovsky E, Diethelm AG, Welchel JD, Jones P. Cyclosporin-A in therapeutic doses increases renal allograft vascular resistance.Lancet 1:477-479, 1986.
1o) English J, Andrew E, Houghton DC, B¢nnett WM. Cyclosporine-inducedacute renal dysfunctionin the rat. Transplantation44:135-141, 1987.
11) Neild GH, Rocchi G, Imab¢rü L, Fumagalli F, Brown Z, Remuzzi G, Williams DG. Effect of cyclosporineon prostacyclinsynthesis by vascular tissue in rabbits. Transplant Proc 4(suppl 1):2398-2400, 1983. 12) Siegl H, Ryffel B, Shoemaker P, Muller A. Conatsch P, Mihatsch M. Cyclosporine, the renin-angiotensin-aldosterone system and renal adverse reacüons. Transplant Proc ~(suppl 1):2719-2725, 1983. 13) Moss NG, Powell SL, Falk RJ. Inwavenouscyclosporine acüvates afferent and efferent renal nerves and causes sodium retention in innervated kidneys in rats. Proc Natl Acad Sci USA 82: 8222-8226, 1985.
14) Perico N, Benigni A, Zoja C, Delaini F, Remuzzi G. Funcüonal significance of exaggerated renal thromboxane A2 synthesis induced by cyclosporin A. Am J Physiol 251:F581-587, 1986.
15) Elzinga L, Kelley VE, HoughtonDC, Bennett W1VI. Modificationof experimental nephrotoxicity with fish oil as the vehicle for cyclosporine. Transplantation 43:271-274, 1987. 16) Kawaguchi A, Golman MH, Shapiro R, Foegh ML, Ramwell PW, Lower RR. Increas¢ in urinary thromboxane B2 in rats caused by cyclosporine. Transplantaüon 40:214-216, 1985. 17) Coffman TM, Carr DR, Yarg«r WE, Klotman PE. Evidenc« that renal prostaglandin and thromboxane production is sümulated in chronic cyclosporine nephrotoxicity.Transplantation43:282-285, 1987. 18) Perico N, Zoja C, Benigni A, Ghilardi F, Gualandris L, Remuzzi G. Effect of short-term cyclosporine administration in rats on renin-angiotensin and thromboxaneA2: Possiblerelevanceto the reductionin glomeralarfiltration rate. J Pharmacol Exp Ther. 239:229-235, 1986.
144
FEBRUARY 1990 VOL. 39 NO. 2
PROSTAGLANDINS 19) Petric R, Freeman D, Wallace A, McDonald J, Stiller C, Keown P. Protective effect of thromboxane synthetase inhibition in the rat with eyclosporine nephrotoxicity. The second international eongress on cyclosporine, Washington, D.C., USA, November,1987, p. 38.
20) Smeesters C, Chaland P, Giroux L, Moutquin J, Etienne P, Douglas F, Corman J, St Louis GS, Daloze P. Prevention of acute cyclosporine A nephrotoxicity by a thromboxane synthetase inhibitor. Transplant Proc (suppl 3):658-664, 1988. 21) Oates JA, Fitzgerald GA, Branch RA, Jackson EK, Knapp HR, Roberts I_3. Clinical implicaüons of prostaglandin and thromboxane A2 formation (first of two patts). N Engl J Med 319:689-698, 1988. 22) FitzGerald GA, Oates JA. Selecfive and nonselective inhibifion of thromboxane formation. Clin Pharmaeol Ther 35:633-640, 1984. 23) FitzGerald GA, Reilly LAG, Pedersen AK. The biochemical pharmacology of thromboxane synthase inhibition in man. Circ 72:1194-1201, 1985. 24) Homby EJ, Foster MR, McCabe PJ, Stratton LE. The inhibitory effect of GR32191, a thromboxane receptor blocking drug, on human platelet aggregaüon, adhesion and secretion. Thromb Haemostas 61:429-436, 1989. 25) Lumley P, White BP, Humphrey PPA. GR32191, a highly potent and specific thromboxane A2 receptor blocking drug on platelets and vascular and airways smooth muscle in vitro. Br J Pharmaco197:783-794, 1989. 26) Coffman TM, Ruiz P, Sanfilippo F, Klotman PE. Chronic thromboxane inhibition preserves funcüon of rejecUon rat renal a11ografts. Kidney Int 35:24-30, 1989. 27) Benigni A, Chibrando C, PiccineUi A, Perico N, Gavinelli M, Furci L, Patino O, Abbate M, Bertani T, Remuzzi G. Incmased urinary excretion of thromboxane B2 and 2,3-dinor-TxB2 in cyclosporin A nephrotoxicity. Kidney Int 34:164-174, 1988. 28) Rogers TS, Elzinga L, Bennett W, Kelley VE. Selective enhancement of thromboxane in macrophages and kidneys in cyclosporine-inducednephrotoxicity. Transplantation, 45:153-156, 1988. 29) Ruiz P, Coffman TM, HoweU DN, Straznickas J, Scroggs MW, Baldwin WM, Klotman PE, Sanfdippo F. Evidence that pretransplant donor blood transfusion prevents rat renal aUograft dysfuncüon but not the in situ cellular alloimmune or morphologic manifestations of rejeeüon. Transplantation 45:1-7, 1988.
30) Wallenstein S, Zucker CL, Fleiss JL. Some statistical methods useful in circulaüon research. Circ Res 47:1-9, 1980. 31) Whiting PH, Thomson AW, Blair JT, Simpson JG. Exp¢rim¢ntal cyclosporin A nephrotoxicity. Br J Exp Path. 63: 88-94, 1982.
FEBRUARY 1990 VOL. 39 NO. 2
145
PROSTAGLANDINS
32) Chapman ]'R, Gfiffiths D, Harding NG, Morris PH. Reversibility of cyclosporine nephrotoxicity after three months' tmatment. Lancet 1:128-130, 1985. 33) Stahl RA, Kudelka S. Chronic cyclosporine A treatment reduces prostaglandin E2 formaüon in isolated glomenfli and papilla of rats kidneys. Clin Nephro125:(suppl 1):$78-$82, 1986. 34) Paller MS. The prostaglandin E1 analog misoprostol reverses acute cyclosporine nephrotoxicity. Transplant Proc (suppl 3): 634-637, 1988. 35) Bunke L, Wilder L. Cyclosporine inhibits mesangial cell prostaglandin producüon.Kidney Int 35:402, 1989. 36) Stahl RAK, Adler S, Baker PJ, Johnson RJ, Chen YP, Pritzl P, Couser WG. Cyclosporin A inhibits prostaglandin E2 formaüon by rat mesangial eells in culture. Kädney Int 35:1161-1167, 1989. 37) Okegawa T, Ionas PE, DeSchryver K, Kawasaki A, Needleman P. Metabolic and cellular alterations underlying the exaggerated renal prostaglandin and thromboxane synthesis in ur¢ter obstrucüon in rabbits. J Clin Invest 71:81-90, 1983. 38) Kelley VE, Altboum I, Boswell I Thromboxane producüon by acüvated renal cortical macrophages. Fed Proc 46:1330, 1987. 39) Humes HD, Jackson NM. Cyclosporine effects on isolated membranes, proximal mbule cells, and interstiüum of the kiduey. Transplant Proc 3(suppl 3):748-758, 1988.
40) Verani R. Cyclosporine nephrotoxiäty in the Fischer raL Clin Nephrol 25(suppl 1): $9-S13, 1986.
Editcr: I~~ Pamwell
146
I~ceived: 10-26-89
Acoepted: 12-8-89
FEBRUARY 1990 VOL. 39 NO. 2