Role of Cyclooxygenase-2–Derived Metabolites and NO in Renal Response to Bradykinin
Abstract—It has been reported that bradykinin (BK) can induce or activate both cyclooxygenase (COX) isoforms and that the renal effects of BK seem to be mediated by prostaglandins and NO. The first objective of this study was to evaluate the relative contribution of both COX isoforms in mediating the renal response to BK in anesthetized dogs. The second objective was to examine whether COX-2 inhibition potentiates the renal effects induced by NO reduction during BK administration. Intrarenal BK infusion (8 ng · kg−1 · min−1, n=6) elicited a significant increment in renal blood flow, sodium excretion, urine volume, and the fractional excretion of lithium. COX-2 inhibition (nimesulide, 5 μg · kg−1 · min−1, n=6) reduced the renal vasodilatation but did not significantly modify the natriuresis or diuresis secondary to BK. Administration of a nonspecific isozyme COX inhibitor (meclofenamate, 5 μg · kg−1 · min−1; n=6) did not induce greater effects than those produced by nimesulide. NO synthesis reduction (NG-nitro-l-arginine methyl ester [L-NAME], 3 μg · kg−1 · min−1) prevented the renal vasodilatation and the increment in the fractional excretion of lithium induced by BK but did not affect the natriuretic or diuretic response. Simultaneous nimesulide infusion did not modify the renal effects of L-NAME during BK infusion (n=6). Finally, inhibition of both COX isoforms with meclofenamate, in dogs treated with L-NAME (n=6), completely prevented the vasodilator and excretory actions of BK. The results of this study suggest that (1) NO and prostanoids dependent on COX-2 seem to be involved in the renal vasodilatation induced by BK, and (2) there is an interaction between NO and COX-1–derived metabolites in mediating the natriuretic and diuretic response to BK.
The renal vasodilatation and natriuresis induced by bradykinin (BK) seem to be secondary to several mediators, such as prostaglandins (PGs), NO, and cytochrome P-450.1 2 3 4 5 6 The role of endogenous PGs in mediating the renal effects of BK has been suggested by studies showing that BK stimulates vascular and tubular production of PGs7 and that the renal effects of BK are reduced by PG synthesis inhibition.1 3 5 It has also been reported in “in vitro” studies that BK elevates the activity and/or expression of both cyclooxygenase (COX) isoforms.8 9 However, the relative contribution of the PGs derived from the action of each COX isoform in mediating BK-induced renal effects is unknown. The first objective of the present study was to evaluate the relative contribution of metabolites derived from both COX isoforms in the decrease of renal vascular resistance (RVR) and natriuresis elicited by BK. This objective was accomplished by evaluating the renal changes induced by BK during infusion of a non–isozyme-specific COX inhibitor (meclofenamate) or a selective COX-2 inhibitor (nimesulide).
The importance of NO in the renal changes induced by BK was suggested in one study showing that infusion of an NO synthesis inhibitor to dogs pretreated with meclofenamate completely prevents the renal effects of BK.4 These results suggest that NO plays a major role in mediating the renal effects of BK when both COX isoforms are inhibited. However, it remains to be demonstrated whether NO synthesis inhibition alone modifies the renal vasodilatation and natriuresis elicited by BK. It is also unknown whether NO synthesis inhibition completely prevents the renal effects of BK when only COX-2 activity is reduced. The second objective of the present study was to evaluate whether COX-2 inhibition potentiates the renal effects induced by NO synthesis blockade during BK infusion. Based on the results obtained in previous studies,10 11 the hypothesis is that NO mediates the decrease in RVR induced by BK by a direct effect and by increasing the production of COX-2–derived metabolites.
Experiments were performed in mongrel dogs of either gender (15 to 24 kg) with free access to tap water and a normal sodium intake. Protocols were designed according to the Guiding Principles approved by the Council of the American Physiological Society. The evening before the experiment, LiCO3 (800 mg) was given orally to evaluate changes in fractional lithium excretion (FeLi) in groups 1 and 4. Surgical preparation was performed in dogs anesthetized with sodium pentobarbital (30 mg/kg IV), as previously described.12 13 14 Catheters were placed in the femoral artery for measurement of mean arterial pressure (MAP) and in the femoral vein for infusion of inulin and additional anesthetic (1.2 mL/min). The right renal artery was fitted with a noncannulating electromagnetic flow probe and was connected to a flowmeter. Distal to the flow probe, a curved 23-gauge needle attached to polyethylene tubing was inserted into the right renal artery and was connected to a peristaltic pump for the infusion of saline or BK (0.6 mL/min). Finally, a 45-minute stabilization period was allowed before experimental maneuvers were begun.
Group 1 (n=6)
After two 15-minute control clearances, BK (8 ng · kg−1 · min−1) was infused into the right renal artery for the duration of the experiment. Ten minutes after initiating BK infusion, 3 more 15-minute clearances were obtained. In preliminary experiments, it was demonstrated that the dose of BK used elicited a significant increase in renal blood flow (RBF), natriuresis, and diuresis without inducing changes in MAP and in the renal hemodynamic and excretory function of the contralateral kidney. It was also found in preliminary experiments (n=6) that urinary prostaglandin E2 (PGE2) excretion increased (P<0.05) from 23.5±3.9 to 48.2±9.9 ng/min in response to BK infusion. Urinary concentration of PGE2 was measured by use of a commercial enzyme-linked immunoassay (Neogen Corp).
Group 2 (n=6)
Nimesulide was infused as a bolus (0.75 mg/kg) and then continuously (5 μg · kg−1 · min−1) for the duration of the experiment. Forty-five minutes after initiating this continuous infusion, two 15-minute control clearances were obtained. Ten minutes after initiating BK (8 ng · kg−1 · min−1) infusion, three 15-minute clearances were taken. The dose of nimesulide used does not modify the arachidonic acid–induced platelet aggregation.12 In preliminary experiments (n=3), it was found that nimesulide reduced urinary PGE2 excretion (P<0.05) before BK infusion (23.5±3.9 to 13.2±1.0 ng/min). During BK infusion, urinary PGE2 excretion was lower in dogs pretreated with nimesulide (31.4±5.0 ng/min) than in those pretreated with saline (48.2±9.9 ng/min).
Group 3 (n=6)
Forty-five minutes after initiating an infusion of meclofenamate (5 μg · kg−1 · min−1), two 15-minute control clearances were taken. Ten minutes after starting a continuous BK infusion (8 ng · kg−1 · min−1), three 15-minute clearances were obtained. The dose of meclofenamate used blocked the arachidonic acid–induced platelet aggregation.12 In preliminary experiments (n=3), it was found that meclofenamate reduced urinary PGE2 excretion (P<0.05) before BK infusion (23.5±3.9 to 5.2±0.1 ng/min). During BK infusion, PGE2 excretion was lower in dogs pretreated with meclofenamate (12.7±4.8 ng/min) than in those pretreated with nimesulide (31.4±5.0 ng/min) or pretreated with saline (48.2±9.9 ng/min).
Group 4 (n=6)
The protocol in this group was similar to that used in group 1, with the only difference being that the 2 control clearances were obtained 45 minutes after starting a continuous intrarenal infusion of NG-nitro-l-arginine methyl ester (L-NAME, 3 μg · kg−1 · min−1).
Group 5 (n=6)
The experimental protocol performed was similar to that used in group 4, with the difference being that L-NAME (3 μg · kg−1 · min−1) was simultaneously infused with nimesulide for 45 minutes before starting the 2 control clearance periods. Nimesulide was administered at the same dose used in group 2.
Group 6 (n=6)
The experimental protocol performed was similar to that used in group 4, with the difference being that L-NAME (3 μg · kg−1 · min−1) was simultaneously infused with meclofenamate (5 μg · kg−1 · min−1) for 45 minutes before starting the 2 control clearance periods.
Renal clearances were taken during each experimental period to determine the glomerular filtration rate (GFR), sodium, potassium, and lithium excretion, and urine flow rate (UV). Blood samples for plasma sodium, potassium, lithium, and inulin concentrations were also obtained. GFR was measured by the clearance of inulin. Inulin concentrations were analyzed by the anthrone method. Concentrations of sodium and potassium were measured by flame photometry (Corning 435). Proximal tubule sodium reabsorption was estimated by the lithium clearance technique. Lithium concentrations (μmol/L) were measured only in groups 1 and 4 by flame emission spectrophotometry (Perkin-Elmer, model 5500). That lithium is a marker for changes in proximal tubule sodium reabsorption is suggested by results reported by Thomsen et al.15
The data for the 2 control clearance periods were averaged for statistical comparisons, because the fluid and solute excretions were in steady-state conditions. Data are expressed as mean±SE. The significance of differences between values of each period in the same group was evaluated by using a 1-way ANOVA and the Fisher test. The significance of differences between the values obtained in different groups was calculated by using a 2-way ANOVA and the Duncan test. A value of P<0.05 was considered significant.
As occurred in the other 5 groups, MAP and GFR did not change significantly through the experiment. With respect to RBF, it increased from a basal value of 157±20 to 217±20 mL/min during BK infusion (P<0.05) (Figure 1⇓). BK also elicited a significant increment in urinary sodium excretion (UNaV) and UV (Figure 2⇓). The greater increments in UNaV (209±57 μmol/min) and UV (1.38±0.35 mL/min), with respect to the basal period (74±20 μmol/min and 0.36±0.07 mL/min, respectively), were found during the first clearance obtained during BK infusion. BK-induced natriuresis and diuresis were smaller (P<0.05) during the last clearance taken (151±36 μmol/min and 1.00±0.21 mL/min, respectively). Figure 3⇓ shows the percent change in FeLi during BK infusion in each dog with respect to the values found during the basal period. It can be observed that FeLi increased in response to BK and that this increment was significant (P<0.05) only during the first clearance obtained after BK infusion had begun.
Pretreatment with nimesulide reduced the BK-induced renal vasodilatation (Figure 1⇑). The elevation in RBF (from 147±24 to 169±25 mL/min, P<0.05) was smaller than that found in the control group (Figure 1⇑). BK also induced an increase (P<0.05) in UNaV (from 31±6 to 129±26 μmol/min) and UV (from 0.20±0.04 to 1.16±0.18 mL/min) that remained significant until the end of the experiment (Figure 4⇓). This excretory response to BK was not significantly different than that found in the control group.
BK infusion elicited an increase in RBF (from 155±13 to 183±20 mL/min, P<0.05). This increment was similar to that found in nimesulide-pretreated dogs (group 2) and smaller (P<0.05) than that found in group 1, in which PG synthesis was not inhibited (Figure 1⇑). Meclofenamate pretreatment did not inhibit the excretory response to BK (Figure 4⇑). UNaV and UV increased (P<0.05) from 24±5 μmol/min and 0.15±0.04 mL/min to 103±14 μmol/min and 0.78±0.20 mL/min, respectively, during the first clearance obtained during BK infusion. These increments remained (P<0.05) until the end of the experiment (Figure 4⇑). No significant differences were found between the increments in UNaV and UV in this group, in which a non–isozyme-specific COX inhibitor was administered, and in the control group.
Previous inhibition of NO synthesis completely prevented the BK-induced renal vasodilatation observed in the control group (Figure 1⇑). As can be observed in Figure 2⇑, BK elicited an increase (P<0.05) in UNaV (from 32±5 to 81±9 μmol/min) and UV (from 0.26±0.02 to 0.94±0.21 mL/min) in dogs pretreated with L-NAME. These increments remained significant throughout the experiment (Figure 2⇑). Although the BK-induced increments in UNaV and UV tended to be smaller in this group than in the vehicle-treated animals (group 1), the increase in renal excretory ability elicited by BK was not significantly modified by previous NO synthesis inhibition. As shown in Figure 3⇑, BK infusion did not induce a significant change in FeLi when NO synthesis was inhibited, because FeLi was similar before and after BK infusion.
As occurred in group 4, in which only L-NAME was administered, intrarenal BK infusion did not induce changes of RBF in dogs pretreated with nimesulide and L-NAME (130±16 versus 129±16 mL/min in the basal period). However, in contrast to what was found in group 6, in which both COX isoforms were inhibited, selective COX-2 inhibition did not prevent the increase in renal excretory ability induced by BK (Figure 5⇓). UNaV and UV increased (P<0.05) from 14±2 μmol/min and 0.14±0.03 mL/min to 77±13 μmol/min and 0.64±0.11 mL/min, respectively, during the first clearance obtained during BK infusion. Both excretory parameters remained elevated (P<0.05) throughout the experiment during BK infusion.
The rise in RBF elicited by BK was also prevented by the simultaneous infusion of L-NAME and the nonselective COX inhibitor (meclofenamate) (122±20 versus 118±17 mL/min in the basal period). Figure 5⇑ shows that BK infusion was not able to modify significantly UNaV and UV when endogenous synthesis of NO and PG were inhibited with L-NAME and meclofenamate.
The results of the present study provide new evidence suggesting that BK-induced renal vasodilatation seems to be mediated by NO and COX-2–derived metabolites. It is also proposed that the natriuresis and diuresis elicited by BK are mainly secondary to the tubular effects of NO and metabolites derived from COX-1 activity.
Several studies have proposed that the renal effects induced by endothelium-dependent vasodilators are mediated by different vasoactive substances, such as NO, COX-derived metabolites, and cytochrome P-450 monooxygenase.1 2 3 4 5 6 An interaction between NO and PG in mediating the renal effects elicited by acetylcholine has been reported in studies performed in anesthetized dogs.16 It has been shown that administration of either a COX or NO synthesis inhibitor does not modify the renal effects of acetylcholine. However, simultaneous inhibition of COX and NO synthesis completely prevents the renal effects of acetylcholine.16 With respect to BK, it is unknown whether there is also an interaction between NO and PG in mediating its renal effects.
It has been demonstrated that BK increases the PG synthesis derived from both COX isoforms,8 but the relative contribution of each isoform in producing the PGs involved in mediating the renal vasodilatation and excretory response to BK has not been elucidated. To evaluate the role of COX-2–derived metabolites in the renal effects elicited by BK, we have used an inhibitor (nimesulide) with a high COX-2 selectivity.17 18 Nevertheless, the possibility cannot be completely ruled out that the effects of nimesulide are partly due to COX-1 inhibition. Supporting the possibility that the renal effects elicited by the dose of nimesulide used are not secondary to an inhibition of COX-1, we have reported that platelet aggregation elicited by arachidonic acid is absent in plasma from dogs treated with meclofenamate and not significantly altered in plasma from nimesulide-treated dogs.12 These results are relevant because the thromboxane A2 involved in platelet aggregation is COX-1 dependent.19 We have also found that (1) nimesulide reduces urinary PGE2 excretion to a lower extent than does meclofenamate,12 and (2) meclofenamate, but not nimesulide, potentiates the effects induced by NO synthesis inhibition on renal hemodynamic and excretory function.20
The decrease in RVR induced by BK in the present study occurred without changes in GFR and MAP. As previously proposed by Edwards,21 our results suggest that BK has a greater effect on the efferent than on the afferent arteriole. As far as we know, to date, the role of NO in mediating the BK-induced renal vasodilatation in the intact animal has been examined in only one study, in which this hemodynamic effect was prevented by reducing NO synthesis in meclofenamate-pretreated dogs.4 Our results suggest that NO is mainly responsible for the BK-induced renal vasodilatation because, without PG inhibition, the elevation in RBF was completely prevented by reducing NO synthesis. The fact that the administration of a selective COX-2 inhibitor or a non–isozyme-specific COX inhibitor reduces to the same extent the renal vasodilatation elicited by BK suggests that COX-2–derived metabolites are also involved in the renal hemodynamic response to BK. Taken together with the results obtained during L-NAME administration, it can be proposed that BK induces an increase in NO that elevates the production of COX-2–derived metabolites. Then, NO and COX-2 derived metabolites are finally responsible for the decrease in RVR secondary to BK infusion (Figure 6⇓). The existence of an interaction between NO and COX-2 has been proposed previously by several studies,10 11 22 in which it has been suggested that NO seems to be an endogenous regulator of the COX-2 activity. It also has been shown that BK induces the expression of COX-2.8 9
The BK-induced natriuretic and diuretic response decreased progressively in the control group, inasmuch as the increments in UNaV and UV at the end of the experiments were lower than those found during the first clearance obtained after the initiation of BK infusion (Figure 2⇑). In 5 dogs, it was observed that UNaV and UV decreased even more significantly when BK infusion was maintained for 120 minutes. Supporting the idea that BK infusion induces a natriuretic and diuretic response that decreases and even disappears with time, Granger and Hall23 have found that UNaV and UV are not elevated when BK infusion is maintained for several hours or days. The renal excretory changes elicited by BK have been proposed to be secondary to an indirect hemodynamic effect24 and a direct action on sodium reabsorption in several tubular segments.3 In the present study, the renal excretory response to BK seems to be mediated mainly by changes in tubular reabsorption, inasmuch as the renal vasodilatation, but not the natriuresis and diuresis elicited by BK, was prevented by NO synthesis inhibition. The results of the present study suggest that proximal sodium reabsorption decreases only transitorily during BK infusion and that this change is mediated by NO, because it was prevented by the previous L-NAME administration. This notion is supported by a recent study proposing that BK and acetylcholine, through an NO-mediated mechanism, decrease sodium transport in proximal tubule epithelial cells.25
As previously mentioned, the role of endogenous PGs in mediating the natriuretic and diuretic responses to BK has been suggested by studies showing that BK infusion elevates PG synthesis7 and that the COX inhibition reduces the renal excretory changes elicited by BK.1 Based on studies demonstrating the presence of COX-2 in the renal tubules26 and in results showing that BK elevates COX-2 expression,8 9 it could be proposed that COX-2–derived metabolites are involved in mediating the renal excretory response to BK. The results of the present study show that the mean increments in UNaV and UV elicited by BK tended to be lower in dogs pretreated with nimesulide or meclofenamate than in the control group. However, there were not significant differences between the BK-induced natriuresis and diuresis in these experimental groups. In support of the notion that PGs are involved in the BK-induced natriuresis,4 we have confirmed that meclofenamate prevents the changes in renal excretory function elicited BK in L-NAME–treated dogs (Figure 5⇑). A new finding of the present study is that the administration of a COX-2 inhibitor does not modify the renal excretory response to BK in dogs in which NO synthesis is reduced (Figure 5⇑). These results support the concept that the PGs mediating the BK-induced natriuresis and diuresis are mainly derived from COX-1 activity (Figure 6⇑).
In summary, from the results obtained in the present study, the following can be proposed: (1) NO is primary responsible of the renal vasodilatation elicited by BK. (2) COX-2–derived metabolites are also involved in the renal hemodynamic response to BK infusion. However, this involvement is evident only when NO synthesis is not modified. (3) NO and COX-1–derived metabolites are mainly responsible for the renal excretory response to BK administration.
This study was supported by grants from the Fondo de Investigaciones Sanitarias (FIS 98/1309 and FIS 99/1024) of Spain. F.R. was supported by a grant from the Fondo de Investigaciones Sanitarias (FIS 98/1309). M.T.L. was supported by a grant from University of Murcia. C.M. was supported by a grant from the Spanish Ministerio de Educación y Ciencia (PN-95). Meclofenamate and nimesulide were kindly provided by Parke-Davis Laboratories and Roche Laboratories, respectively.
- Received May 18, 2000.
- Revision received June 19, 2000.
- Accepted July 3, 2000.
Blasingham MC, Nasjletti A. Contribution of renal prostaglandins to the natriuretic actions of bradykinin in the dog. Am J Physiol. 1979;237:F182–F187.
Kauker M. Kallidin effect on renal tubular function in meclofenamate- and vehicle-pretreated rats. Proc Soc Exp Biol Med. 1990;193:60–64.
Lahera V, Salom MG, Fiksen-Olsen M, Romero JC. Mediatory role of endothelium derived nitric oxide in renal vasodilatory and excretory effects of bradykinin. Am J Hypertens. 1991;4(pt 3):260–262.
Navar LG, Inscho W, Majid SA, Imig JD, Harrison-Bernard LM, Mitchell KD. Paracrine regulation of the renal microcirculation. Physiol Rev. 1996;76:425–536.
Ferreri NR, Mc Giff JC, Miller MJS, Schwartzman M, Spokas EG, Wong PY. The kidney and arachidonic acid metabolism. In: Samuelson B, Paoletti R, Ranwell R, eds. The Kidney and Arachidonic Acid Metabolism. New York, NY: Raven Press; 1983;11:481.
Pang L, Knox AJ. PGE2 release by bradykinin in human airway smooth muscle cells: involvement of cycooxygenase-2 induction. Am J Physiol. 1997;273:L1132–L1140.
Ichihara A, Imig JD, Navar LG. Cyclooxygenase-2 modulates afferent arteriolar responses to increases in pressure. Hypertension. 1999;34:843–847.
Rodríguez F, Llinás MT, González JD, Salazar FJ. Renal changes induced by a cyclooxygenase-2 inhibitor during normal and low sodium intake. Hypertension.. 2000;36:276–281.
Llinás MT, González JD, Nava E, Salazar FJ. Role of angiotensin II in the renal effects induced by nitric oxide and prostaglandin synthesis inhibition. J Am Soc Nephrol.. 1997;8:543–550.
Salazar FJ, Llinás MT, González JD, Quesada T, Pinilla JM. Role of prostaglandins and nitric oxide in mediating renal response to volume expansion. Am J Physiol. 1995;268:R1442–R1448.
Thomsen K, Holstein-Rathlou NH, Leissac PP. Comparison of the three measures of proximal tubular reabsorption: lithium clearance, occlusion time, and micropuncture. Am J Physiol.. 1981;241:F348–F355.
Salom M, Lahera V, Romero JC. Role of prostaglandin and endothelium-derived relaxing factor on the renal response to acetylcholine. Am J Physiol. 1991;260:F145–F149.
Taniguchi Y, Ikesu A, Yokohama K, Noda K, Nakamura T, Toda A, Shimeno H. Selective inhibition by nimesulide, a novel non-steroidal anti-inflammatory drug, with prostaglandin endoperoxide synthase-2 activity in vitro. Pharmacol Sci. 1995;1:173–175.
Cullen L, Kelly L, Connor S, Fitzgerald DJ. Selective cyclooxygenase-2 inhibition by nimesulide in man. J Pharmacol Exp Ther. 1998;287:578–582.
Llinás MT, Rodríguez F, Moreno C, Salazar FJ. Role of cyclooxygenase-2 derived metabolites and nitric oxide in regulating renal function. Am J Physiol.. 2000;279:R1641–R1646.
Edwards RM. Response of isolated renal arterioles to acetylcholine, dopamine, and bradykinin. Am J Physiol.. 1985;248:F183–F189.
Salvemini D, Misko TP, Mansferrer JL, Seibert K, Currie MG, Needleman P. Nitric oxide activates cyclooxygenase enzymes. Proc Natl Acad Sci U S A. 1993;90:7240–7244.
Granger JP, Hall JE. Acute and chronic actions of bradykinin on renal function and arterial pressure. Am J Physiol. 1985;248:F87–F92.
Brenner BM, Falchuk KH, Keimowitz RI, Berliner RW. The relationship between peritubular capillary and fluid reabsorption by renal proximal tubule. J Clin Invest. 1969;48:1519–1530.
Linas SL, Repine JE. Endothelial cells regulate proximal tubule epithelial cell sodium transport. Kidney Int. 1999;558:1251–1258.
Vio CP, Cespedes C, Gallardo P, Mansferrer JL. Renal identification of cyclooxygenase-2 in a subset of thick ascending limb cells. Hypertension. 1997;30:687–692.