INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

BLOOD FLOW TO THE MEDULLA of the kidney is maintained through the tonic and stimulated release of nitric oxide (NO). Intrarenal infusion of vasoconstrictors such as norepinephrine and angiotensin II (ANG II) cause vasoconstriction and a reduction of renal blood flow. In response to vasoconstrictor infusions, compensatory release of NO helps to preserve the small fraction of the total renal blood flow that reaches the medulla of the kidney (38, 43, 44). An important physiological role for the production of NO in the control of medullary blood flow has also been proposed based on the observation that NO synthase (NOS) inhibition leads to reduction of medullary blood flow, sodium retention, and hypertension (16, 17, 23). Calcium sensitive isoforms of NOS are expressed in a variety of structures within the renal medulla including the medullary thick ascending limb (mTAL), outer medullary collecting duct (OMCD), and descending vasa recta (DVR) endothelia (17, 23). ANG II causes a compensatory increase in medullary NO generation (44) but reduces DVR endothelial intracellular calcium concentration ([Ca²⁺]_i) and inhibits increases in [Ca²⁺]_i by bradykinin (BK) (25, 31).

Previous studies have shown that ANG II type 1 (AT₁) and 2 (AT₂) receptors are expressed in outer medullary vascular bundles (19, 39). Motivated by this, we sought to determine which receptor subtype mediates the unusual suppressive effect of ANG II on DVR endothelial calcium signaling. We found that the AT₁ receptor blocker losartan inhibits the ANG II effect on DVR endothelia, but a high concentration of losartan is required. Interestingly, the ability of ANG II to reduce endothelial [Ca²⁺]_i elevation by BK and acetylcholine (ACh) is markedly enhanced when AT₂ receptors are blocked with PD-123319, implying an important role for AT₂ receptor activation to facilitate endothelial [Ca²⁺]_i signaling by vasodilators. We also examined the [Ca²⁺]_i transients induced by ANG II in the cytoplasm of isolated DVR smooth muscle/pericytes. In contrast to effects on endothelia, ANG II-induced pericyte [Ca²⁺]_i transients exhibited classical peak and plateau responses that were blocked by losartan but not PD-123319.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vitro isolation of DVR. Studies were performed in 267 vessels from Sprague-Dawley rats (80-150 g body wt, Harlan) anesthetized before nephrectomy by an intraperitoneal bolus of thiopental (50 mg/kg body wt). DVR were microdissected as previously described from outer medullary vascular bundles and mounted on glass pipettes using a standard apparatus (ITM; San Antonio, TX) (26). The following buffer was used for the dissection, bath, and loading of fura 2 (in mM): 5 HEPES, 140 NaCl, 10 Na-acetate, 5 KCl, 1.2 MgCl₂, 1.71 Na₂HPO₄, 0.29 NaH₂PO₄, 1 CaCl₂, 5 alanine, and 5 glucose with 0.5 g/dl albumin; pH = 7.4.

Measurement of vasoreactivity in microperfused DVR. To evaluate the effects of vasoactive agents on DVR diameters, microperfusion experiments were recorded on videotape using a Panasonic model AG 1980 videocassette recorder with a microphone for audio recording of experimental events. The inverted microscope was equipped with a beam splitter and a side port for the attachment of a videocamera (model 72 CCD, Dage-MTI). DVR were observed with a ×40 objective to yield a final magnification of approximately ×1,300 on the video screen. During playback, vessel diameters were measured with calipers at the point of greatest constriction. Changes are expressed as follows: %constriction = (1  D/D₀) × 100, where D is the internal diameter and D₀ is the diameter before the introduction of hormones and reagents into the bath.

Measurement of endothelial intracellular calcium. DVR were loaded with fura 2 by exposing them to bath containing 2 µM fura 2-AM (Molecular Probes; Eugene, OR). At the time the bath was exchanged to contain fura 2-AM, the feedback controller was turned on, gradually warming the vessel to 37°C over ~5 min. Total loading time was 20 min. We (24) have previously shown that fura 2 preferentially loads into endothelial cells rather than pericytes. For measurement of [Ca²⁺]_i, fura 2-loaded DVR were excited using 350/380-nm dual-wavelength combinations. The background-subtracted ratio of the fluorescent emission was calculated for conversion to the equivalent [Ca²⁺]_i assuming a dissociation constant for fura 2 at 37°C of 224 nM. The respective Ca²⁺-free and Ca²⁺-saturated 350-to-380-nm fluorescence ratios were measured as previously described by exposing vessels to buffer containing 5 or 0 mM CaCl₂ and 0.5 mM EGTA, respectively, along with 10 µM ionomycin and converted to [Ca²⁺]_i as previously described (24, 25). A photon-counting photomultiplier assembly (PMT) was employed to measure fluorescent emission at 510 nm from fura 2-loaded DVR endothelia. The light for the excitation of fura 2 was provided by a 75-W xenon arc lamp directed through a computer-controlled shutter. The excitation wavelengths were isolated using a computer-controlled monochrometer (PTI; Lawrenceville, NJ). DVR were observed through a 1.3 numerical aperture, Nikon CF fluor ×40 oil immersion objective. Fluorescent emission was isolated with a 510WB40 filter (Omega Optical; Brattleboro, VT) and directed to the PMT. PMT output was monitored with commercial hardware and software (PTI).

Measurement of Mn²+ influx into DVR endothelia using fura 2. Mn²⁺ quench of fura 2 fluorescence was used to measure Mn²⁺ influx into DVR endothelia as previously described (25). For these protocols, DVR were loaded with fura 2 and excited at the isosbestic (calcium insensitive) wavelength (360 nm) while the fluorescent emission (F₃₆₀) was monitored at 510 nm with a PMT. After MnCl₂ (500 µM) was added to the bath, the rate of decline of F₃₆₀ provided a measure of plasmalemmal Mn²⁺ influx.

Isolation of DVR pericytes and loading of fura 2 into pericyte cytoplasm. To isolate pericytes from individual DVR, small wedges of the renal medulla were separated from kidney slices by dissection and transferred to CaCl₂-free physiological saline solution (PSS) containing collagenase 1A (0.45 mg/ml, Sigma), protease XIV (0.4 mg/ml, Sigma), and albumin (1.0 mg/ml) as previously described (31). These were incubated at 37°C for 22 min and then transferred back to CaCl₂ (1 mM)-containing PSS and held at 4°C in a petri dish. At intervals, vessels were isolated from the digested renal tissue by microdissection and transferred to the perfusion chamber. We have previously shown that this enzymatic digestion can be used to enable gigaohm seal formation on pericytes for electrophysiological recording (22).

Reagents. BK, ACh, ANG II, cyclopiazonic acid (CPA), PD-123319, CGP-42112A, and ionomycin were purchased from Sigma. Losartan was a gift from Merck. These agents were dissolved in water at 10¹-10⁵ M and stored frozen at 20°C. Aliquots were thawed on the day of the experiment, and excess was discarded at the end of each day. Calcium ionophore was dissolved in ethanol at 10 mM. CPA and fura 2-AM (Molecular Probes) were stored frozen in anhydrous DMSO.

Statistical analysis. Experimental results are reported as means ± SE. In the figures, most error bars have been suppressed for clarity of presentation. Plateau [Ca²⁺]_i values quoted in the text are means ± SE at the end of periods in designated protocols. Statistical comparisons employed a paired t-test, an unpaired t-test, ANOVA, or repeated-measures ANOVA as appropriate. For ANOVA, significance was determined by the Student-Newman-Keuls test. P < 0.05 was used for significance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inhibition of CPA- and ACh-stimulated [Ca²+]_i by ANG II. As previously described, ANG II reduces basal endothelial [Ca²⁺]_i, but the effect is difficult to study because resting [Ca²⁺]_i is low and only small changes occur after ANG II application. When [Ca²⁺]_i is increased by treatment with an endothelium-dependent vasodilator or a sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) pump inhibitor, the ability of ANG II to lower endothelial [Ca²⁺]_i is dramatic (25). In this study, we used CPA and ACh to stimulate increases in endothelial [Ca²⁺]_i. The effect of ANG II (10⁸ M) on endothelial [Ca²⁺]_i in vessels treated with CPA or ACh is shown in Figs. 1 and 2. After application of CPA (10⁵ M), endothelial [Ca²⁺]_i increased from basal values of <100 to ~600 nM. After 5 min, the bath was exchanged to contain either vehicle or ANG II (10⁸ M). In the presence of ANG II, endothelial [Ca²⁺]_i fell from 592 ± 125 to 328 ± 105 nM (Fig. 1; P < 0.05). The effect of ACh on DVR endothelial [Ca²⁺]_i has not been previously described. For that reason, we examined its concentration dependence. Between 10⁸ and 10⁵ M, successively increasing concentrations of ACh led to additional [Ca²⁺]_i elevation (Fig. 2A). Threshold effects were observed at 10⁸ M; endothelial [Ca²⁺]_i increased from 45 ± 9 to 67 ± 13 nM (P < 0.05). Significant increases were observed throughout the concentration range, and endothelial [Ca²⁺]_i reached 331 ± 34 and 305 ± 37 nM at 10⁵ and 10⁴ M, respectively. As shown in Fig. 2B, a peak and plateau response to ACh was observed. The plateau value of 306 ± 61 nM fell to 154 ± 20 nM (P < 0.05) after ANG II (10⁸ M) was introduced into the bath.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   ANG II inhibition of cyclopenzoic acid (CPA)-stimulated endothelial intracellular calcium concentration ([Ca2+]i). Basal endothelial [Ca2+]i was monitored for 1 min, after which CPA (105 M) was introduced into the bath for 5 min to elevate [Ca2+]i. Subsequent addition of ANG II (108 M) reversed the CPA-induced increase in [Ca2+]i (n = 7 in each group).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   ANG II inhibition of ACh-stimulated endothelial [Ca2+]i. A: descending vasa recta (DVR) endothelial [Ca2+]i was measured at baseline and at 5-min intervals as ACh was introduced into the bath at log molar increments from 108 to 104 M (n = 7). B: DVR endothelial [Ca2+]i was measured at baseline and for 15 min after exposure to ACh at 104 M. After 10 min of ACh, ANG II (108 M) was also introduced into the bath along with ACh, resulting in a reduction of the plateau [Ca2+]i response (n = 7).

Effect of losartan and PD-123319 on ANG II suppression of DVR endothelial [Ca²+]_i. We next examined whether the specific ANG II AT₁ and AT₂ receptor antagonists losartan and PD-123319 could block the ability of ANG II to reduce DVR endothelial [Ca²⁺]_i. In the protocol illustrated in Fig. 3, [Ca²⁺]_i was measured for 1 min (baseline) and then during 5 min of treatment with CPA. Subsequently, either losartan (10⁶ M; Fig. 3A) or PD-123319 (10⁶ M; Fig. 3B) was introduced for 5 min, followed by a concomitant addition of ANG II (10⁸ M) for another 5 min. In a final period, the blocker was removed from the bath so that the vessels were exposed to ANG II alone. When applied alone, neither losartan nor PD-123319 affected CPA-stimulated endothelial [Ca²⁺]_i. At 10⁶ M, neither agent prevented the suppression of endothelial [Ca²⁺]_i by ANG II. As shown in Fig. 3A, CPA-stimulated [Ca²⁺]_i fell from 633 ± 113 to 304 ± 54 nM in the presence of ANG II and losartan (P < 0.01). As shown in Fig. 3B, it fell from 498 ± 137 to 226 ± 33 nM (P < 0.01). Given the surprising inability of either the AT₁ or AT₂ receptor antagonists to block the effects of ANG II, we repeated the examination but with a higher concentration of losartan and PD-123312 (10⁵ M) and a lower concentration of ANG II (10⁹ M). Under these conditions, blockade by losartan but not PD-123312 was obtained (Fig. 4). As shown in Fig. 4A, CPA-stimulated [Ca²⁺]_i changed from 701 ± 82 to 707 ± 71 nM in the presence of ANG II and losartan [P = not significant (NS)]. As shown in Fig. 4B, it fell from 797 ± 193 to 326 ± 160 nM (P < 0.01). The need for such a high concentration of losartan to achieve blockade is atypical of the AT₁ receptor (3, 28). For this reason, we verified the result using an intragroup comparison (Fig. 5). In this experiment, losartan was introduced into the bath of CPA-treated vessels at 10⁵ M, after which ANG II was added at 10⁹ M. The losartan concentration was subsequently lowered in log molar increments. Consistent with the findings of Figs. 3 and 4, loss of losartan blockade of endothelial [Ca²⁺]_i inhibition occurred between 10⁵ and 10⁶ M as CPA-stimulated endothelial [Ca²⁺]_i fell from 481 ± 97 to 214 ± 29 nM ( P < 0.01).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Effect of losartan (Los; 106 M) and PD-123319 (PD; 106 M) on ANG II (108 M) endothelial [Ca2+]i suppression. A: DVR endothelial [Ca2+]i was measured at baseline, after which CPA (105 M) was added to the bath to elevate [Ca2+]i. At 5-min intervals after CPA, the bath was exchanged to contain CPA + Los (106 M), Los (106 M) + ANG II (108 M), and then ANG II alone. At these concentrations, Los failed to block the effect of ANG II to reduce DVR endothelial [Ca2+]i (n = 6). B: protocol in A was repeated, substituting PD (106 M) for Los. PD also failed to block the effect of ANG II to lower DVR endothelial [Ca2+]i (n = 7).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Effect of Los (105 M) and PD (105 M) on ANG II (109 M) endothelial [Ca2+]i suppression. A: DVR endothelial [Ca2+]i was measured at baseline, after which CPA (105 M) was added to the bath to elevate [Ca2+]i. At 5-min intervals after CPA, the bath was exchanged to contain CPA + Los (105 M), Los (105 M) + ANG II (109 M), and then ANG II alone. At these concentrations, Los successfully blocked the effect of ANG II to inhibit DVR endothelial [Ca2+]i (n = 10). B: protocol in A was repeated, substituting PD (105 M) for Los. PD failed to block the effect of ANG II to lower DVR endothelial [Ca2+]i (n = 6).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Intragroup comparison of the concentration dependence of Los blockade of ANG II (109 M) endothelial [Ca2+]i suppression. DVR endothelial [Ca2+]i was measured at baseline, after which CPA (105 M) was added to the bath to elevate [Ca2+]i. Los was added to the bath at 104 M, followed by ANG II (109 M). At 5-min intervals, the concentration of Los was reduced in log molar increments. Below 105 M, Los failed to block the effect of ANG II to lower DVR endothelial [Ca2+]i (n = 7).

5

5

9

4

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Effect of Los and PD (105 M) on ANG II (109 M) endothelial [Ca2+]i suppression. Endothelial [Ca2+]i was increased by preexposing vessels to ACh (104 M). Subsequently, Los (105 M), PD (105 M), or vehicle (n = 6 in each group) was exchanged into the bath, followed by ANG II (109 M). Los, but not PD, blocked ANG II suppression of endothelial [Ca2+]i. After the removal of Los, the blockade was reversible.

Modulation of endothelial [Ca²+]_i signaling by AT₂ receptors. When ANG II receptors are stimulated before the induction of endothelial [Ca²⁺]_i transients, we observed that AT₂ receptor blockade with PD-123319 led to an augmentation of ANG II inhibition of endothelial responses to BK (10⁷ M) and ACh (10⁴ M). To increase our confidence that PD-123319 was selectively blocking AT₂ receptors, we reduced its concentration to 10⁸ M for these studies. Figure 7 shows a comparison of BK-induced [Ca²⁺]_i signaling in DVR endothelia under four conditions. BK was introduced into the bath by itself or with ANG II (10⁸ M), ANG II + PD-123319 (10⁸ M each), or PD-123319 (10⁸ M). As shown in Fig. 7A, blockade of the AT₂ receptor enhanced the ability of ANG II to inhibit BK-induced [Ca²⁺]_i signaling. Upon stimulation with BK, BK + ANG II, or BK + ANG II + PD-123319, the peak [Ca²⁺]_i achieved was 842 ± 224, 482 ± 114 (P = NS vs. BK), and 204 ± 51 nM (P < 0.05 vs. BK), respectively. BK-induced [Ca²⁺]_i responses tended to wane markedly, and we could not demonstrate a significant difference in plateau [Ca²⁺]_i at 15 min. By itself, PD-123319 did not inhibit BK-induced [Ca²⁺]_i signaling (Fig. 7B). BK + PD-123319 achieved peak and plateau [Ca²⁺]_i of 687 ± 126 and 171 ± 42 nM, respectively. The peak but not the plateau was significantly higher (P < 0.01) than that achieved by the combined action of BK + ANG II + PD-123319.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of angiotensin type 2 (AT2) receptor blockade on ANG II inhibition of bradykinin (BK)-induced [Ca2+]i signaling. A: after 1 min of baseline recording, BK (107 M, n = 9), BK + ANG II (108 M, n = 9), or BK + ANG II + PD (108 M, n = 8) was added to the bath for 15 min and then washed out. When AT2 receptors were blocked, the ability of ANG II to inhibit BK-induced [Ca2+]i signaling was markedly enhanced. B: PD, when applied without ANG II, had no effect on BK [Ca2+]i signaling (n = 8).

View larger version (29K): [in this window] [in a new window]   Fig. 8.   Effect of AT2 receptor blockade on ANG II inhibition of ACh-induced [Ca2+]i signaling. A: after 1 min of baseline recording, PD (105 M, n = 7) or Los (105 M, n = 6) was added to the bath. Subsequently, ANG II (109 M) was added (at 3 min), followed by ACh (105 M, at 8 min). The ACh-induced [Ca2+]i transient was markedly inhibited by preexposure to ANG II/PD. B: ACh (105 M, n = 6), ACh + ANG II (108 M, n = 6), or ACh + ANG II + PD (108 M, n = 5) was added to the bath from 2 to 12 min and then washed out. When AT2 receptors were blocked, ANG II inhibition of the ACh-induced [Ca2+]i transient was markedly enhanced. C: endothelial [Ca2+]i transients were measured after simultaneous introduction of ACh (105 M), ANG II (108 M), and PD (108 M) to the bath. This was examined in the absence (n = 9) or presence of AT2 receptor preactivation with CGP-42112A (CGP; 108 M, n = 10). AT2 receptor preactivation prevented the inhibition of ACh-induced [Ca2+]i responses.

8

Blockade of ANG II suppression of Mn²+ entry into DVR endothelia. We have previously shown that the ability of ANG II to suppress endothelial [Ca²⁺]_i is at least partially explained by an inhibition of divalent cation influx (25). When Mn²⁺ enters fura 2-loaded cells, it quenches fura 2 fluorescence. The rate of Mn²⁺ entry can therefore be gauged from the rate of decline of fura 2 fluorescence when fura 2 is excited at its Ca²⁺-insensitive isosbestic point, 360 nm. As shown in Fig. 9, ANG II inhibits Mn²⁺ quench, an effect that is blocked by losartan but not PD-123319.

View larger version (17K): [in this window] [in a new window]   Fig. 9.   Inhibition of Mn2+ influx into DVR endothelia by ANG II. A: fluorescence of fura 2 was measured during excitation at 360 nm (F360) and normalized to its value at time 0 (F0; ordinate). The addition of 500 µM MnCl2 to the bath resulted in an ~20% reduction of fluorescence over 120 s. The addition of ANG II (109 M) blocked the rate of fura 2 quench attributable to influx of Mn2+ into the endothelia. B: for purposes of statistical analysis, the average of the last 5 s of fluorescence was computed for each vessel from recordings such as those depicted in A. Bars show means ± SE of that averaged final fluorescence for five groups in which Mn2+ was not added (vehicle, n = 6), Mn2+ was added alone (n = 8), Mn2+ was added along with ANG II (109 M, n = 8), or Mn2+ was added with ANG II plus either Los (105 M, n = 9) or PD (105 M, n = 7). Los, but not PD, blocked the ability of ANG II to inhibit Mn2+ of fura 2. *P < 0.05.

Blockade of the ANG II-induced increase in pericyte [Ca²+]_i. The reduction of DVR endothelial [Ca²⁺]_i by ANG II contrasts with the expected effect of stimulation of the AT₁ receptor to increase [Ca²⁺]_i via inositol (1,4,5)-trisphosphate [Ins(1,4,5)P₃]-mediated signaling (2, 40). Because ANG II is a potent vasoconstrictor of DVR, it is anticipated that it would increase smooth muscle/pericyte [Ca²⁺]_i through AT₁ receptor stimulation. Pericytes were isolated and loaded with fura 2 in the presence of probenicid to prevent fura 2 leakage from the cytoplasm (31). Pericyte [Ca²⁺]_i was measured during exposure to either ANG II (10⁸ M), ANG II + losartan (10⁵ M), or ANG II + PD-123319 (10⁵ M) (Fig. 10). As expected for smooth muscle, ANG II increased pericyte [Ca²⁺]_i from 56 ± 46 nM to a peak of 639 ± 145 nM. A plateau of 146 ± 30 nM was achieved 10 min after ANG II stimulation. In the presence of losartan, contemporary values for pericyte [Ca²⁺]_i were 54 ± 21 and 74 ± 26 nM, respectively (P < 0.01, peak and plateau phases). AT₁ receptor blockade completely eliminated the ANG II response. In contrast, PD-123319 had no significant effect to alter either the peak (473 ± 100 nM) or plateau (144 ± 55 nM) phase pericyte [Ca²⁺]_i. PD-123319 alone had no effect on pericyte [Ca²⁺]_i (Fig. 10).

View larger version (26K): [in this window] [in a new window]   Fig. 10.   Los inhibits ANG II-induced increases in pericyte [Ca2+]i. Pericytes were isolated by stripping them from the abluminal surface of microdissected DVR and loading them with fura 2 in the presence of probenecid (1 mM) (31). Baseline [Ca2+]i was recorded for 1 min, after which ANG II (108 M), ANG II + Los (105 M), or ANG II + PD (105 M) was added to the bath for 10 min (n = 6 in each group). ANG II induced a peak and plateau [Ca2+]i response that was blocked by Los but not PD. In a separate group, the bath was exchanged to include PD alone (105 M, n = 4). In the absence of ANG II, PD did not increase pericyte [Ca2+]i.

Role of AT₁ and AT₂ receptor activation in DVR vasoconstriction. We examined the ability of AT₁ and AT₂ receptor blockade to influence vasoconstriction (Fig. 11). In one series, losartan (10⁶ M), PD-123319 (10⁶ M), or vehicle was introduced into the bath of microperfused DVR followed by ANG II (10⁸ M). As shown in Fig. 11A, losartan blocked and PD-123319 augmented constriction. These results demonstrate that AT₁ receptor activation mediates and AT₂ receptor activation modulates vasoconstriction of DVR. In separate experiments, microperfused DVR were sequentially exposed to PD-123319 at 10⁸ M and then 10⁶ M concentrations (5-min intervals each). At the end of each period, %constriction reached 0.7 ± 0.9 and 2.4 ± 3.3, respectively (not significantly different from zero, data not shown). We also tested whether AT₂ receptor preactivation with CGP-42112A (10⁸ M) could modulate subsequent ANG II-induced constriction. During a 5-min preactivation period, CGP-42112A did not affect basal vasomotor tone. Compared with controls, vessels preactivated with CGP-42112A exhibited a delayed response to ANG II (Fig. 11B).

View larger version (23K): [in this window] [in a new window]   Fig. 11.   AT2 receptor activation modulates vasoconstriction. A: microperfused DVR were exposed to Los (106 M, n = 8), PD (106 M, n = 10), or vehicle (n = 9) at time 0. After the baseline was recorded for 1 min, ANG II (108 M) was introduced into the bath. After 10 min, the bath was exchanged to remove the blocker. AT1 receptor blockade with Los inhibited and AT2 receptor blockade augmented ANG II-induced vasoconstriction. B: vessels were preexposed to the AT2 receptor agonist CGP (108 M, n = 8) or vehicle (n = 6) for 5 min, after which ANG II (108 M) was introduced into the bath. Preactivation of AT2 receptors delayed ANG II-induced vasoconstriction. (*P < 0.05 vs. control).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We (25) previously demonstrated that ANG II reduces both basal DVR endothelial [Ca²⁺]_i and BK- or thapsigargin-stimulated [Ca²⁺]_i responses of DVR endothelium. That effect occurred when ANG II was applied to either the luminal or abluminal surface of the vessel and when pericytes were separated from the endothelium (31). Taken together, those findings implicate both an endothelial receptor and an endothelial signal transduction cascade through which ANG II modulates [Ca²⁺]_i. In this study, we extended the prior work by investigating the ability of AT₁ and AT₂ receptor blockade to mediate and modulate the endothelial [Ca²⁺]_i response to ANG II. Two principal findings emerged. First, the AT₁ receptor blocker losartan inhibits ANG II-induced suppression, but a high concentration of losartan (10⁵ M) is required (Figs. 3-5). A second major finding is that AT₂ receptor blockade with PD-123319 enhances the ability of ANG II to suppress endothelial [Ca²⁺]_i responses. This is true provided that AT₂ receptors are activated at the same time as, or prior to, the [Ca²⁺]_i-stimulating agent (Figs. 7 and 8). The latter effect is likely to be AT₂ receptor specific because it is achieved at a low PD-123319 concentration (10⁸ M).

As previously discussed (25), based on the known signal transduction pathways of the G protein-coupled AT₁ receptor, the finding that DVR endothelial [Ca²⁺]_i is suppressed rather than stimulated by ANG II is unexpected. ANG II stimulation of cultured aortic endothelial cells by Pueyo and colleagues (29, 30) resulted in a rise in [Ca²⁺]_i, phospholipase C activation, phospholipase A₂ activation, and peroxynitrite production. The opposite response that we observed in acutely isolated DVR might be explained by phenotypic alteration of cultured aortic cells or an effect of acute isolation on DVR endothelia. It seems more likely that the difference in [Ca²⁺]_i signaling is related to a true regional difference between macrovascular aortic and microvascular DVR endothelia. We hypothesized that ANG II inhibits [Ca²⁺]_i signaling in DVR endothelia to reduce vasodilator production by these cells, thereby enabling the adjacent thick ascending limbs of Henle to control vasomotion in vascular bundles (25, 31).

Measurement of endothelial [Ca²+]_i in isolated DVR. The DVR wall is comprised of smooth muscle/pericytes and endothelial cells. In this study, we loaded fura 2 into DVR freshly isolated by microdissection from the renal medulla. We (24) have previously shown that this yields a strong endothelial fluorescent signal with near exclusion of fura 2 from the pericyte cytoplasm. Recently, we devised a method for removing pericytes from the abluminal surface of collagenase-treated DVR and used this to show that selective endothelial loading of fura 2 is partially explained by export of deesterified fura 2 from pericytes by a pathway that is blocked by probenecid. We also found that the ability of ANG II to suppress basal and CPA-stimulated endothelial [Ca²⁺]_i persists even when pericytes have been removed. In contrast, isolated, fura 2/probenecid-loaded pericytes respond to ANG II with a classical peak and plateau [Ca²⁺]_i increase (Fig. 10) (31). Because ANG II can exert its opposite effects to raise and lower [Ca²⁺]_i in DVR pericytes and endothelia, respectively, even after these cells have been separated from one another, we conclude that ANG II receptors exist on both cell types.

11

10

Inhibition of DVR endothelial Ca²+ influx. Although the signaling within DVR endothelia that is responsible for ANG II-induced [Ca²⁺]_i inhibition remains unknown, a role for inhibition of Ca²⁺ influx has been established. We have previously shown that the rate of refilling of Ca²⁺ into Ca²⁺-depleted endothelia is reduced by pretreatment with ANG II. Similarly, Mn²⁺ entry, as gauged by the quench of fura 2 in Ca²⁺-replete cells, is blocked by ANG II (25). With the use of the latter technique, we have now shown that losartan but not PD-123319 can reverse the ANG II inhibition of Mn²⁺ entry (Fig. 9). The pathway through which Ca²⁺ enters DVR endothelia is unknown. Endothelia are generally thought to lack voltage-operated Ca²⁺ channels and instead to rely on nonselective cation channels and changes in membrane potential to augment or reduce the driving force for divalent cation influx by hyperpolarization or depolarization, respectively (1, 12). According to that hypothesis, ANG II depolarization of pericytes would be expected to favor a [Ca²⁺]_i increase (Fig. 10), whereas ANG II depolarization of DVR endothelium would favor a [Ca²⁺]_i decrease (Figs. 1-8). In recent studies, we verified that ANG II depolarizes both cell types and showed that BK hyperpolarizes DVR endothelia (22, 31).

Modulation of ANG II [Ca²+]_i suppression by losartan and PD-123319. ANG II exerts its effects through AT₁ (AT_1A and AT_1B) and AT₂ receptors (13). AT_1A and AT_1B receptors are 7 membrane-spanning G protein-coupled receptors with 95% amino acid sequence homology and are widely localized to vascular and tubular elements of the kidney (10). AT₁ receptor activation signals largely via phospholipase C-mediated Ins(1,4,5)P₃ generation and [Ca²⁺]_i responses. Those actions have been observed in a variety of cell types including smooth muscle and endothelia (2, 30, 31, 40). Losartan is a nonpeptide, highly selective antagonist of angiotensin AT_1A and AT_1B receptors. This compound inhibits ANG II binding to AT₁ receptors at concentrations between 10⁹ and 10⁸ M and has a >10,000-fold selectivity for AT₁ over AT₂ receptors (3, 28). Given this sensitivity and the high degree of selectivity, it is somewhat surprising that [Ca²⁺]_i suppression by ANG II was only blocked when losartan was used at 10⁵ M (Figs. 3-5). Possibly, isolation of vessels alters receptor affinity or a separate population of low-affinity receptors exists on DVR endothelia. In any case, it cannot be firmly concluded that AT₁ receptors mediate the suppressive DVR endothelial [Ca²⁺]_i response to ANG II. Studies with ANG II receptor blockers alone will probably not suffice to confidently identify the receptor that mediates DVR endothelial ANG II [Ca²⁺]_i suppression.

8

Interplay between AT₁ and AT₂ receptors. It is generally accepted that AT₂ receptor activation exerts effects that oppose and modulate the activation of AT₁ receptors (5, 8, 11, 14, 35-37). AT₂ receptor expression is markedly reduced after fetal development but remains in the kidney into adult life (18). AT₂ receptors are inducible by ANG II infusion and a low-salt diet and are localized to vascular structures and glomeruli (2, 40). AT₁ receptor activation induces cell proliferation, vasoconstriction, and activation of protein kinases, whereas AT₂ receptor stimulation is vasodilatory, antiproliferative, and generally increases phosphatase activity. The findings of this study are entirely compatible with those notions.

Renal effects of AT₂ receptor stimulation. In recent years, a number of studies have been performed to examine the physiological role that AT₂ receptor stimulation serves in the kidney (9, 15, 19, 35, 36, 39, 42). The data support a role for AT₂ receptor stimulation to function in a counterregulatory manner to oppose AT₁ receptor-mediated vasoconstriction. With the use of intrarenal tissue microdialysis, Siragy and colleagues (5, 35-37) found that ANG II tonically stimulates BK production leading to cGMP formation and NO release. Those observations were corroborated in AT₂ receptor null mice. Those animals exhibit sodium retention, a failure to generate intrarenal BK, and exaggerated hypertension in response to chronic ANG II infusion (36). Conversely, overexpression of AT₂ receptors leads to vasodilation through kinin activation (42). Gross et al. (9) examined the relationship of AT₂ receptor activation with medullary blood flow and sodium excretion. Natriuresis and changes in medullary blood flow that accompany increased renal perfusion pressure were blunted in AT₂ receptor knockout mice (9).

Effect of ANG II on renal blood flow and NO production. Perfusion of the renal medulla is sensitive to NO inhibition. NOS blockade reduces medullary blood flow, favors sodium retention, and induces hypertension (17, 20). Ca²⁺-sensitive isoforms of NOS (NOS1 and NOS3) are expressed in thick ascending limb, vasa recta, and collecting ducts (17, 23). Endothelial NOS (NOS3) is stimulated by elevation of [Ca²⁺]_i. Thus ANG II suppression of endothelial [Ca²⁺]_i should favor rather than oppose vasoconstriction and tend to reduce blood flow through DVR. In contrast to this reasoning, past studies have pointed to a reciprocal relationship between ANG II stimulation and intrarenal NO production (41, 44). Infusion of ANG II into the medullary interstitium leads to an increase in intrarenal NO generation, implying a role for NO to prevent ANG II constriction of medullary resistance vessels (44). In contrast to this, we found that ANG II reduces DVR endothelial [Ca²⁺]_i and inhibits Ca²⁺ signaling by the endothelium-dependent vasodilator BK. To explain this apparent paradox, we (25) have noted that ANG II raises [Ca²⁺]_i in OMCD and mTAL. DVR on the periphery of outer medullary vascular bundles lie adjacent to these structures. On the basis of this proximity, we speculated that ANG II-induced Ca²⁺ activation of NOS1 and NOS3 within the cytoplasm of mTAL and OMCD stimulates NO that diffuses to the vascular bundle periphery to preferentially dilate the vessels that supply them with oxygen and nutrients. Suppression of endothelial [Ca²⁺]_i would facilitate that process to provide a feedback mechanism that transfers control of vasomotion away from endothelium to transporting epithelia. The latter may be of importance in the relatively hypoxic renal medulla to prevent ischemic insult to metabolic active NaCl-transporting nephron segments (6).