To investigate whether bradykinin (BK) participates in the inhibition of renal effects of exogenous angiotensin II (AngII) by AngII type 1 receptor (AT1R) blockade, eight salt-repleted volunteers underwent four p-aminohippurate- and inulin-based renal studies of AngII infusion at increasing rates of 0.625, 1.25, and 2.5 ng·kg·min−1 for 30 min. Studies 1 and 2 were preceded by 3 days of placebo, whereas studies 3 and 4 used 240 to 320 mg·day−1 valsartan. Bradykinin B2-type receptor (BKB2R) antagonist icatibant (50 μg·kg−1) was coinfused in studies 2 and 4. Mean blood pressure (MBP), glomerular filtration rate (GFR), renal blood flow (RBF), and renal sodium excretion (UNaV) were measured. In study 1, MBP rose by 12.8%, UNaV decreased by 68%, and GFR and RBF also fell (p < 0.001 for all). In study 2, GFR and RBF fell as in study 1, but the rise in MBP and the fall in UNaV were accentuated [+20.0%, analysis of variance (ANOVA), p < 0.02 versus study 1 and −80.0%, p < 0.05, respectively]. In study 3, AngII had no effects, and in study 4, renal hemodynamics remained unaffected, but MBP still rose and UNaV fell (ANOVA, p < 0.02 and 0.005 versus study 3, respectively). Icatibant accentuated AngII-induced changes in MBP and UNaV. Previous AT1R blockade prevented any systemic and renal effects of AngII, but significant changes in MBP and UNaV still followed AngII plus icatibant even after AT1R blockade. BK, through BKB2Rs, participates in the inhibitory action of AT1R blockers toward actions of exogenous AngII on MBP and UNaV in healthy humans.
Many physiological relationships exist between the vasoconstrictor-antinatriuretic RAS and the vasodilator-natriuretic KKS (Carretero and Scicli, 1995; Siragy et al., 1996; Madeddu et al., 2007). Therefore, attention has been paid to the participation of the main KKS-derived peptide BK in the beneficial effects of RAS blockade with either ACEi or ARBs in hypertension, CHF, and renal disease (Weir, 2007).
BK, through BKB2R, induces systemic and renal vasodilatation and inhibits tubular sodium reabsorption, thus opposing the actions of RAS (Siragy et al., 1994, 1996; Carretero and Scicli, 1995; Madeddu et al., 2007; Sivritas et al., 2008). Because in the systemic vasculature KKS, unlike RAS, is activated by high sodium intake, elevated BK contributes to salt repletion to maintain normal BP by counteracting endogenous vasoconstrictors (Murphey et al., 2004; Madeddu et al., 2007; Sivritas et al., 2008). In contrast, in the kidney both KKS and RAS are stimulated by salt restriction and suppressed by salt repletion, with a potent activation of KKS by RAS itself on low sodium intake (Siragy et al., 1994, 1996; Murphey et al., 2004; Madeddu et al., 2007; Sivritas et al., 2008; Siragy, 2010). Therefore, BK should contribute significantly to UNaV in both salt depletion, where its elevated intrarenal levels oppose RAS-mediated antinatriuresis, and salt repletion, despite much lower intrarenal levels (Murphey et al., 2004; Sivritas et al., 2008; Siragy, 2010). Because ACE, in addition to converting AngI to AngII, degrades kinins (Carretero and Scicli, 1995; Madeddu et al., 2007), the vascular actions of ACEi include predictably the participation of slow BK inactivation. Accordingly, BKB2R blockade in humans blunted the ACEi-induced increase in flow-dependent, endothelium-mediated vasodilation (Hornig et al., 1997) and lowered BP (Gainer et al., 1998; Squire et al., 2000), also leading to vasoconstriction in ACEi-treated CHF (Cruden et al., 2004).
In contrast, the contribution of BK to the effects of ARBs in humans is much less clarified. Studies in AT1R-blocked animals showed an augmented production of BK, NO, and cGMP in systemic vasculature and kidney mediated by AT2R (Siragy et al., 1996; Tsutsumi et al., 1999; Carey, 2005; Jones et al., 2008; Yayama and Okamoto, 2008; Siragy, 2010). AT2R, activated by elevated AngII, either endogenously produced on sodium restriction or exogenously infused on sodium repletion (Siragy, 2010), should stimulate BK–NO–cGMP cascade, thus opposing systemic and renal AT1R-mediated vasoconstriction. Because natriuresis also follows intrarenal AT1R blockade mediated by AT2R (Tsutsumi et al., 1999; Carey and Padia, 2008), intrarenal production of BK–NO–cGMP stimulated by AT2R could be implicated in increased UNaV after AT1R blockade (Yayama and Okamoto, 2008; Siragy, 2010).
Although some plausible mechanisms exist whereby endogenous BK contributes to the systemic and renal effects of AT1R blockade, human studies have been addressed only for the systemic vasculature, providing conflicting results. No effect of BKB2R blockade was shown on the forearm vasodilatation after losartan in CHF patients (Davie et al., 1999) and the BP-lowering action of losartan and valsartan (VAL) in sodium-restricted humans (Gainer et al., 1998; LeFebvre et al., 2007). In contrast, the favorable effect of candesartan on flow-dependent, endothelium-mediated vasodilation in coronary patients was blunted by BKB2R blockade (Hornig et al., 2003), and the AngII-induced constriction of ex vivo human coronary arteries was inhibited by ARBs and accentuated by AT2R blockade (Batenburg et al., 2004), indicating an AT2R-mediated vasodilatatory action of AngII. The latter was abolished by either NO inhibition or BKB2R blockade, thus suggesting activation of the BK–NO–cGMP cascade (Batenburg et al., 2004).
Because ARBs are known to counteract effectively any action of exogenous AngII on systemic vasculature, adrenal tissues, and kidney (Gandhi et al., 1996; Morgan et al., 1997; Schmitt et al., 1998), low-dose AngII infusion may be a suitable experimental condition for determining the role of BK in both the renal actions of AngII in humans and their inhibition exerted by AT1R blockade. The purpose of the present work was to investigate whether BKB2R blockade affects BP and renal changes during AngII infusion in salt-repleted, healthy humans with and without a previous short-term AT1R blockade.
Materials and Methods
Eight young healthy volunteers, four men and four women, after written informed consent, participated in the study, according to the ethical protocols of our institution. None had evidence or history of heart, liver, kidney, or endocrine diseases or alcohol or drug abuse, and none were under medical treatment. Before the study, all participants had a clinical examination, repeated blood pressure measurements, an electrocardiogram, and routine laboratory screening (Table 1).
Each participant underwent in a randomized order four separate 90-min p-aminohippurate (PAH) and inulin (INU)-based renal hemodynamic studies of AngII infusion (Montanari et al., 2003; Biggi et al., 2007). Subjects were trained by a dietitian to maintain before each study a detailed written diet providing daily 250 ± 10 mmol sodium, 80 ± 6 mmol K, and 2450 ± 70 kcal (55% carbohydrates, 15% protein, and 30% lipids) and a fixed intake of antioxidant fruit and vegetables for 5 days (Biggi et al., 2007). At each infusion, adherence to diet was estimated based on the food record for the last 3 days, and baseline urinary urea nitrogen excretion (UUN) was used as a marker of protein intake (Table 2). Washout period between infusion studies was at least 2 weeks. Females were examined during the follicular phase of their menstrual cycle.
AngII Infusion Studies.
After an overnight fast, experiments were initiated at 8:00 AM with the participant in a sitting position (Montanari et al., 2003). A plastic indwelling catheter was placed into a cubital vein, a priming dose of 3000 mg × 1.73 m2 body surface area of INU and 600 mg × 1.73 m2 of PAH was injected, and an infusion of PAH and INU was initiated and continued throughout the entire study by using a 50-ml syringe precision pump (Perfusion Secura; Braun, Melsungen, Germany) to obtain plasma levels of approximately 15 mg/liter for PAH and 200 mg/liter for INU. A second indwelling catheter for blood sampling was placed immediately at the controlateral arm. After 60 min of equilibration, participants emptied their bladders, then a 30-min baseline clearance period was performed. Then, after voiding, a pump infusion of human AngII was initiated at stepwise increasing rates of 0.625, 1.25, and 2.5 ng·kg·min−1, each for 30 min (infusion periods 0–30, 30–60 and 60–90 min, respectively). Tap water (300 ml) was administered hourly to ensure an appropriate urine flow. BP and heart rate were measured every 3 min with an automated oscillometric monitoring device (TM 2421; A and D Co., Tokyo, Japan). Samples from urine of each clearance period were taken for sodium excretion rate (UNaV). Blood samples also were drawn for plasma renin activity (PRA) at 60 min of baseline period, for plasma PAH and INU every 10 min, and for plasma sodium at 0 and 30 min of each period during the entire study.
The four studies were performed as follows: In study 1, AngII was infused after a 3-day placebo (PL) treatment. In study 2, after PL, BKB2R antagonist icatibant (d-Arg-Arg-Pro-Hyp-Gly-Thi-Ser-d-Tic-Oic-Arg, HOE140) (50 μg·kg−1) was coinfused in the first 10 min of the first step of AngII infusion with 0.625 ng·kg·min−1 (period 0–30). In study 3, AngII infusion followed a 3-day pretreatment with VAL (240 to 320 mg·day−1 according to body weight) taken orally at 10 PM. In study 4, with VAL pretreatment, HOE140 was coinfused as in study 2. Doses of HOE140 of the same order of magnitude have been previously shown to attenuate the forearm vasodilator effect of intra-arterial BK (Cockcroft et al., 1994; Brown et al., 2000), without altering BP or baseline UNaV (Gainer et al., 1998). Accordingly, in a pilot study of PAH and INU infusion on the habitual sodium intake of four additional volunteers, HOE140 (50 μg·kg−1) did not affect mean blood pressure (MBP), UNaV, effective renal plasma flow (ERPF), and glomerular filtration rate (GFR).
The responses of renal hemodynamics and UNaV to either infused AngII or AT1R blockade are known to be different according to salt intake, because of their dependence on the levels of endogenous AngII in the kidney. Actually, exogenous AngII is followed by more pronounced renal effects on sodium repletion than on salt restriction (Gandhi et al., 1996; Visser et al., 2008), whereas renal vasodilatory and natriuretic responses to acute AT1R blockade are accentuated by sodium restriction (Burnier et al., 1996; Price et al., 1997; Gainer et al., 1998). Therefore, the present study was performed on sodium intake as elevated as 250 mmol·day−1 to inhibit endogenous AngII production as much as possible. To prevent any further confounding interference resulting potentially from the effect of an acute AT1R blockade on endogenous AngII (Gandhi et al., 1996; Burnier et al., 1996; Schmitt et al., 1998), a short-term, high-dose VAL pretreatment timed to provide the last dose in the evening before AngII infusion was preferred instead of administering an ARB single dose close to the beginning of AngII infusion. An effective AT1R blockade was assumed to be maintained in the morning after the last VAL administration on the basis of previous data indicating that the pressor effect of exogenous AngII was blunted even 12 h after a single dose of 80 to 160 mg of VAL (Morgan et al., 1997; Latif et al., 2001).
MBP was calculated as: systolic BP + (2×diastolic BP)/3. A satisfactory steady state for PAH and INU in plasma was obtained with our infusion technique, as indicated by an interassay coefficient of variability for plasma PAH and INU within the baseline period (1.9 and 2.4%, respectively), which was very close to the intra-assay coefficient for duplicate PAH and INU measurements. Thus, ERPF and GFR were estimated by using a constant-infusion technique to avoid an unethical bladder catheterization as required in standard urinary clearance studies. PAH and INU measured in the infusate were multiplied for the volume of infused solution per minute. The resulting infusion rate of PAH or INU (mg·min−1) was divided for each measured plasma concentration, thus obtaining four clearance values at baseline and three in each infusion period. The mean values were used in the expression of data for each period. Filtration fraction (FF) was calculated as the ratio between GFR and ERPF, RBF was calculated from ERPF and hematocrit, renal vascular resistance (RVR) was calculated from MBP and RBF, and fractional excretion rate of sodium (%FENa) was calculated from UNaV, mean plasma sodium, and GFR (Montanari et al., 2003; Biggi et al., 2007).
PAH (20% solution) and INU (10% solution) were purchased from J. Monico, Venice, Italy. Pharmaceutical grade human AngII and HOE 140 were obtained from Clinalfa (Laufelfingen, Switzerland). Commercially available 80- and 160-mg capsules of VAL were used.
Data are expressed as mean ± S.E.M. The values at baseline of various infusions were compared with Student's t test for paired values. One-way ANOVA was used to analyze time-dependent effects in each infusion. The differences among various infusions were analyzed by two-way ANOVA followed by post hoc multiple comparisons with the Student-Newman-Keuls test. Differences at the 5% level or less were considered to be statistically significant.
The results of each AngII infusion study are summarized in Table 2.
Stepwise infusion of AngII after PL (study 1) was followed by a 12.8% increase in MBP (one-way ANOVA, p < 0.001). Both RBF and GFR showed a progressive decrease (−23.9 and −16.5%, respectively; p < 0.001), with increased FF (+8.1%; p < 0.05) and RVR (+46.7%; p < 0.001). Both absolute (UNaV) and fractional to GFR (%FENa) sodium excretion rates declined substantially (−68.3 and −61.6%, respectively; p < 0.001).
In study 2, with PL + AngII + HOE140, the increase in MBP was larger than that after PL + AngII (+20% from baseline, two-way ANOVA; p < 0.02 versus study 1). UNaV and %FENa also decreased significantly more than with PL + AngII (p < 0.05). Changes in ERPF, GFR, RBF, FF, and RVR were essentially the same as those after PL + AngII (two-way ANOVA; p not significant).
At baseline of studies 3 and 4 (VAL + AngII and VAL + AngII + HOE140, respectively), GFR, UNaV, and %FENa were almost identical to those measured after PL (studies 1 and 2). In contrast, MBP as an average was 4.8% reduced, ERPF and RBF were 13.5% increased, and RVR and FF were lowered by 17.7 and 11.7%, respectively, (paired t test; p < 0.05 for all). Baseline PRA was approximately four times augmented (p < 0.001).
In study 3 (VAL + AngII), no significant time-dependent change took place for MBP, ERPF, GFR, RBF, RVR, FF, UNaV, and %FENa. Their time courses also were significantly different from those observed in studies 1 (PL + AngII) and 2 (PL + AngII + HOE140) (two-way ANOVA; p < 0.001 for all, except p < 0.05 for FF).
In study 4 (VAL+ AngII + HOE-140), as in study 3, renal hemodynamics remained essentially unaffected. MBP, however, still rose significantly (+6.8% from baseline; one-way ANOVA, p < 0.005), with time-dependent changes larger than those observed in study 3 (VAL + AngII) (two-way ANOVA, p < 0.02) and smaller than those observed in both study 1 (PL + AngII; p < 0.05) and study 2 (PL + AngII + HOE140; p < 0.02). Both UNaV and %FENa also fell by 34% (p < 0.005), with time-dependent changes larger than those shown in study 3 (VAL + AngII; p < 0.005) and smaller than those shown in study 1 (PL + AngII; p < 0.02) and study 2 (PL + AngII + HOE140; p < 0.01).
As the main new finding in this study, BKB2R blockade accentuated in humans the rise in MBP and the fall in UNaV caused by exogenous AngII, with or without a previous AT1R blockade.
The expected decrease in ERPF, RBF, and GFR and elevation in FF and RVR after AngII (study 1) (Gandhi et al., 1996; Morgan et al., 1997; Schmitt et al., 1998; Montanari et al., 2003) were not affected (study 2) by HOE140, coinfused at a rate devoid of significant actions at baseline (Gainer et al., 1998) but known to counteract the BK-induced vasodilatation (Cockcroft et al., 1994). In contrast, HOE140 accentuated the elevation in MBP and the fall in UNaV, indicating that endogenous BK opposes in humans the effects of infused AngII on BP and UNaV. This agrees with the potentiated hypertension and sodium retention with BKB2R blockade in desoxycorticosterone acetate-salt or AngII-infused rats (Madeddu et al., 1992, 2007) and human data showing a slightly increased baseline BP with high-dose HOE140 (Squire et al., 2000) and a higher BP with BKB2R blockade after RAS stimulation with furosemide (Murphey et al., 2000).
We also hypothesized involvement of BK in the inhibition by AT1R blockade of the effects of exogenous AngII. ARBs may cause on their own acutely lowered BP, renal vasodilatation, and natriuresis (Gandhi et al., 1996; Burnier et al., 1996; Price et al., 1997; Gainer et al., 1998; Schmitt et al., 1998). Thus, when timed to be administered close to an AngII infusion (Gandhi et al., 1996; Schmitt et al., 1998), their action toward infused AngII could not be distinguished from that exerted toward endogenous AngII, which may account for 50% of the apparent inhibitory effect on renal changes from exogenous AngII (Schmitt et al., 1998).
Although endogenous AngII was suppressed by sodium repletion and the last dose of VAL had been given 11 h before the study, baseline MBP and RVR were lowered in studies 3 and 4 and ERPF and RBF were increased, indicating a contribution of endogenous AngII to the baseline systemic and renal hemodynamics even on salt repletion. In addition, AngII infusion after VAL was superimposed to a higher baseline endogenous AngII, as indicated by a much higher PRA (Maillard et al., 2002; Azizi et al., 2004), thus reaching higher AngII levels than in studies 1 and 2. Notwithstanding this limitation, VAL prevented almost completely systemic and renal actions of infused AngII (study 3) (Gandhi et al., 1996; Morgan et al., 1997; Schmitt et al., 1998; Maillard et al., 2002).
A distinct rise in MBP with decreased UNaV and %FENa still followed AngII when combined with VAL and coinfused HOE140 (study 4), showing that BKB2R also “buffered” changes in MBP and UNaV to exogenous AngII in the presence of an effective AT1R blockade. This suggests that BK, through BKB2 receptors, participates in humans in the inhibitory action of ARBs on MBP and UNaV changes after exogenous AngII. Conflicting results have been reported on the relationship of endogenous BK to BP regulation, with or without AT1R blockade. Resting BP (Cockcroft et al., 1994; Gainer et al., 1998; Brown et al., 2000) or the acute and chronic BP lowering actions of ARBs (Gainer et al., 1998; LeFebvre et al., 2007) were not affected in humans by BK–B2R blockade.
However, in the latter studies, an extreme sodium restriction (10 mM × day 1) could have suppressed the systemic KKS (Murphey et al., 2004; Sivritas et al., 2008), thus offsetting any contribution of endogenous BK to the ARB action, in spite of elevated endogenous AngII. In our salt-repleted individuals, although AngII infusion in each study precluded any investigation on the effect of BKB2R blockade on baseline BP, HOE140 blunted the action of VAL in preventing the AngII-induced BP changes. In addition, BKB2R blockade significantly increased BP in humans just on sodium repletion, either at baseline with high-dose inhibitor (Squire et al., 2000) or after acute RAS stimulation with furosemide (Murphey et al., 2000). Therefore, a high systemic BK, related to an unrestricted salt intake, should be required to demonstrate an action of endogenous BK on BP, either at baseline, during elevation of endogenous AngII with furosemide, or during an exogenous AngII infusion, such as in the present study. This agrees with the notion that on sodium repletion BK modulates at baseline the AT1R-mediated systemic vasoconstriction (Murphey et al., 2004; Madeddu et al., 2007; Sivritas et al., 2008), also indicating that, without VAL (study 2), the potentiated BP rise by HOE140 reflected essentially the abolished modulation by BK of AT1R-mediated effects. In contrast, with VAL, which on its own prevented almost completely any AT1R-mediated action of AngII on BP (study 3), participation may be suggested of AT2R, activated by infused AngII as known from animal studies (Tsutsumi et al., 1999; Carey, 2005; Jones et al., 2008; Yayama and Okamoto, 2008; Siragy, 2010). AT2R should stimulate BK–NO–cGMP cascade, leading to a BK-mediated vasodilatatory component of AngII action, which in turn should be offset by HOE140. This explanation remains uncertain, because additional studies with a specific AT2R blocker were not feasible.
Similar to the further decreased UNaV with AngII plus HOE140, UNaV still fell by 34% when AngII and HOE140 were coinfused after VAL (study 4). Because the 68% decrease in UNaV in study 1 was prevented in study 3 (VAL + AngII), approximately half the inhibiting power of a prior ARB on sodium retention from exogenous AngII seemed to depend on BKB2R.
In both studies with HOE140, UNaV fell without differences in renal hemodynamics compared with paired studies without BKB2R blockade, suggesting that AngII and BK interacted in modulating UNaV directly at the tubular level, without a contribution of hemodynamically mediated changes. However, BK is known to redistribute RBF intrarenally from cortex to medulla (Carretero and Scicli, 1995; Omoro et al., 2000), rather than to augment the whole RBF. Because such changes could not be reflected by variations in PAH-estimated RBF, a participation in UNaV modulation may not be excluded of a BKB2R-dependent, vasodilatating component in kidney medulla, possibly stimulated by infused AngII and prevented by HOE140.
Significant natriuresis, abolished by intrarenal AT2R antagonism, follows selective AT1R blockade in the rat, indicating that AT2R contributes to natriuresis in response to ARBs, with the potential mediation of increased renal BK, NO, and cGMP (Carey, 2005; Carey and Padia, 2008; Jones et al., 2008; Yayama and Okamoto, 2008; Siragy, 2010). The experimental conditions in our human study were quite close to those used in rats. After an effective AT1R blockade by VAL, the infused AngII should have led to the activation of AT2R, thus stimulating the natriuretic BK–NO–cGMP cascade. This view, as for BP changes, remains speculative, because a specific AT2R blocker could not be tested.
AT2Rs activated by a chronically elevated AngII also are postulated to participate, besides AT1R blockade, in the clinical benefit of ARB treatment (Savoia et al., 2007; Weir, 2007; Jones et al., 2008). It is obvious that our data in short-term AT1R-blocked, healthy humans infused with AngII cannot be extended to a chronic ARB treatment in patients with cardiovascular or renal disease. However, in the present study we show an interaction between AngII and BKB2R, possibly via AT2R activated by infused AngII, in young, AT1R-blocked healthy individuals, i.e., just in a condition associated with the lowest level of AT2R expression (Jones et al., 2008). On the other hand, in animal studies (Tsutsumi et al., 1999; Jones et al., 2008; Yayama and Okamoto, 2008), up-regulation of AT2R may result from aging, vascular disease, and chronic AT1R blockade and in human clinical conditions as well, such as diabetes or insulin resistance (Savoia et al., 2007; Brillante et al., 2008; Jones et al., 2008). Thus, participation of BKB2R in the control of BP and UNaV under AT1R blockade, as shown by the limits of our model, may pertain to long-term conditions of elevated AngII, up-regulated AT2R, and stimulated BK–NO–cAMP cascade, such as in ARB-treated patients.
To summarize, BKB2R blockade accentuated in sodium-repleted humans elevation in BP and decrease in UNaV after AngII infusion, without altering renal vasoconstriction and decrease in GFR. Previous high-dose VAL prevented any effect of AngII almost completely.
With combined AT1R and BKB2R blockade, AngII was followed by significant changes in MBP and UNaV, without affecting renal hemodynamics. Therefore, endogenous BK, possibly stimulated by AT2R activation, participates, through BKB2R, in the inhibitory action by AT1R blockade of actions of exogenous AngII on both BP and UNaV. Caution should be taken in extending such results beyond the limits of a human model of short-term AT1R blockade and exogenous AngII infusion. Nonetheless, these data might also be relevant for clinical conditions of chronically stimulated AngII, with up-regulation of AT2R and BK–NO–cAMP cascade, such as in long-term, ARB-treated cardiovascular and renal patients.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- angiotensin II
- angiotensin II type 1 receptor
- angiotensin II type 2 receptor
- renin-angiotensin system
- kinin-kallikrein system
- bradykinin B2-type receptor
- angiotensin-converting enzyme
- ACE inhibitor
- angiotensin II type 1 receptor antagonist
- congestive heart failure
- renal sodium excretion
- urinary urea nitrogen excretion
- analysis of variance
- blood pressure
- mean BP
- glomerular filtration rate
- renal blood flow
- plasma renin activity
- effective renal plasma flow
- filtration fraction
- renal vascular resistance
- fractional excretion rate of sodium.
- Received February 5, 2010.
- Accepted May 25, 2010.
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics