INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

BRADYKININ (BK) and Lys-BK (kallidin) are peptide hormones released on cleavage of their high (HMW)- or low (LMW)-molecular-weight kininogen precursors. HMW and LMW kininogens are enzymatically degraded by the action of kallikreins, serine proteases that exist either in the circulation as plasma kallikreins or in various organs as tissue kallikreins. HMW kininogen is readily cleaved by plasma kallikrein, whereas LMW kininogen is a poorer substrate for this enzyme. In contrast, tissue kallikreins release bioactive products from either precursor (4, 11).

The kidney is a site for LMW kininogen synthesis, principally in epithelial cells of the distal nephron including the cortical and medullary collecting ducts (9). A recent study using sensitive reverse transcription-polymerase chain reaction (RT-PCR) and immunohistochemical staining with antibodies specific for kininogens demonstrated expression of the HMW as well as the LMW kininogen in these locations (11). Interestingly, collecting duct cells expressing kininogens lie adjacent to those producing tissue kallikrein, suggesting that local production of BK and kallidin might occur in the renal medulla to enable regulation of various processes by these hormones (9, 12, 13, 21, 27, 30, 31).

Although BK and kallidin are often thought of as inflammatory mediators (4, 11, 18), there is substantial evidence that they act as paracrine regulators of renal medullary sodium reabsorption and blood flow (15, 24, 25). Outer medullary descending vasa recta (OMDVR) are vasoactive branches of juxtamedullary efferent arterioles through which the majority of blood flow to the medulla occurs. We hypothesize that BK partially exerts its actions in the renal medulla through receptor-operated calcium signaling in descending vasa recta endothelia. With the use of microdissected, in vitro perfused OMDVR from rats, we show that kinins increase endothelial intracellular calcium concentration ([Ca²⁺]_i) and that BK dilates vessels preconstricted with angiotensin II (ANG II).

	METHODS

Top Abstract Introduction Methods Results Discussion References

In vitro microperfusion. Microperfusion of OMDVR dissected from vascular bundles in the inner stripe of the outer medulla has been described in detail (19, 20). Vessels were perfused on the stage of an inverted microscope (Nikon Diaphot) using apparatus from Instruments Technology and Machinery (San Antonio, TX). In these studies, an oil immersion objective with short working distance was required to obtain fluorescent videoimages, making temperature near the specimen difficult to control. To overcome this obstacle, we custom built perfusion chambers that permit rapid exchange of the bath without deviation of temperature from 37°C. Bathing buffer is warmed immediately upstream of the chamber under feedback control (CN9111A, Omega Engineering). The thermocouple probe of a second, identical controller is located in the narrow inlet region of the chamber next to the perfused vessel. Heating from the controllers is separately applied by passing current through nickel-chromium wires, one surrounding the inlet tubing upstream of the chamber and another embedded in the acrylic around the inlet region.

Videomicroscopy and measurement of vessel diameters. To measure the effects of vasoactive substances on OMDVR diameters, microperfusion experiments were captured on videotape. The inverted microscope is equipped with a 0/100% beam splitter and a side port with a C mount for attachment of a video camera (Panasonic WV-BL90). Experiments were recorded on a Panasonic model AG 1960 videocassette recorder equipped with a microphone for voice recording. Experiments were played back, and diameters were measured from the video screen with calipers (20, 26).

(1)

Fluorescent imaging system. A digital fluorescent imaging system was employed to measure [Ca²⁺]_i of fura 2-loaded OMDVR. Light for excitation of fura 2 was provided from a 75-W xenon arc lamp (PTI, South Brunswick, NJ) directed through a computer-controlled shutter (Uniblitz, Rochester, NY). The excitation wavelength was isolated using 340HT15 and 380HT15 bandpass filters (Omega Optical, Brattleboro, VT) mounted on a solenoid and computer syncronized for image acquisition. OMDVR were observed through a 1.3 NA Nikon CF fluor ×40 oil immersion objective. Fluorescence was monitored at 510 nm and directed to a charge-coupled device (CCD) camera (Loral, Milpitas, CA) equipped with an 18-mm image intensifier (Varo, Garland, TX) via a 0/100% beam splitter. The automatic gain control of the CCD camera was disabled. The intensifier provides a constant gain of 20,000 but has an automatic cutoff to protect against excessive illumination. Output of the CCD camera was captured with a 4-MB GRB silicon video MUX frame grabber (Epix, Northbrook, IL) with a resolution of 8 bits/pixel. Digitized frames were displayed in real time on a color monitor as a pseudocolor image.

Measurement of [Ca²⁺]_i in fura 2-loaded OMDVR. OMDVR were loaded with the Ca²⁺-sensitive fluorescent indicator fura 2 by exposure to bath containing 2 µM fura 2-acetoxymethyl ester (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. A total fura 2-AM exposure time of 15 min was consistently used to load OMDVR. Stock fura 2-AM was stored frozen in anhydrous dimethyl sulfoxide (DMSO) at a concentration of 1 mM.

(2)

(3)

(4)

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Linearity of fura 2 response. Top: mean value of fluorescence ratio image (R) was measured at room temperature as Ca-ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA) buffers containing 2 µM fura 2 pentapotassium salt, 10 mM potassium MOPS, 100 mM KCl and 0 mM Mg2+ (Calcium Calibration Buffer Kit no. 1, Molecular Probes) were sequentially exchanged through the lumen of a 15-µm (outer diameter) glass pipette. [CaEGTA]-to-[EGTA] ratio is proportional to free calcium concentration ([Ca2+]f) (Eq. 4). Values are means ± SE; N = 4 experiments. Sf2, Sb2, free and bound fluorescence intensity at 380 nm, respectively; Rmin, Rmax, minimum and maximum R, respectively. Bottom: same protocol as top (N = 4) except that free Mg2+ concentration was 1 mM and chamber temperature was maintained at 37°C. Buffers were prepared by pH titrimetric method and [Ca2+]f was calculated from dissociation constant of MgEGTA and CaEGTA as described by Tsien and Pozzan (28).

Reagents. BK, kallidin, and related receptor subtype-specific agonists and antagonists (Table 1) and ANG II were purchased from Sigma. The V₁-receptor agonist [Phe²,Ile³,Orn⁸]vasopressin (V1AG) was obtained from Peninsula (Belmont, CA). These agents were dissolved in water at 10³-10⁵ M and stored frozen at 20°C. Aliquots of 25-100 µl were thawed on the day of the experiment. The excess was discarded at the end of each day. 4-Bromo-A-23187 calcium ionophore was dissolved in ethanol at 10 mM, and fura 2-AM was stored frozen in anhydrous DMSO at a concentration of 1 mM.

Statistical analysis. Experimental results are reported as means ± SE. Statistical comparisons used paired t-test, unpaired t-test, or repeated-measures analysis of variance (ANOVA) as appropriate. For ANOVA, significance was determined by the Student-Newman-Keuls test.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Localization of fura 2 signal. The descending vasa recta wall is made up of two cell types, endothelial cells, which line the lumen, and smooth muscle remnants called pericytes, the cell bodies of which protrude from the abluminal side (19, 20). White light and fluorescent micrographs were obtained with the Spectrasource cooled CCD camera in fura 2-loaded OMDVR (N = 8). Three representative reproductions are shown in Fig. 2. Endothelial cells and pericytes can be readily discerned from the high-magnification white light images in the left panels of Fig. 2. For comparison, corresponding fluorescent images captured at 510 nm using 340-nm excitation are shown in the right panels of Fig. 2. The pattern of fluorescence indicates an endothelial origin with little or no contribution from pericytes. Pseudocolor 340-nm, 380-nm, and 340/380 images of fura 2-loaded OMDVR obtained with the intensified CCD camera before and after application of BK (10⁸ M) are shown in Fig. 3. Again, a pattern consistent with an endothelial origin is apparent, although the definition of the images is inherently limited by the relatively poor resolution of the image intensifier.

View larger version (103K): [in this window] [in a new window]   View larger version (63K): [in this window] [in a new window]   Fig. 2.   Comparison of high-power white light and fluorescent images of microperfused outer medullary descending vasa recta (OMDVR) loaded with fura 2. Left: white light micrographs of OMDVR. Arrowheads point to pericyte cell bodies. Bar, ~8 µm. Right: fluorescent micrographs of same OMDVR as left panel. Excitation at 340 nm and images captured at 510 nm using time exposure of charge-coupled device (CCD) sensor cooled to 30°C. Images using 380-nm excitation were similar. A-C: 3 representative vessels of N = 8 experiments.

View larger version (81K): [in this window] [in a new window]   Fig. 3.   Pseudocolor images of fura 2-loaded OMDVR. Images (340 nm, 380 nm, and background-subtracted ratio) obtained with intensified CCD camera before and after exposure to 108 M bradykinin (BK). Color bar indicates value corresponding to individual pixel locations of ratio image. Pattern of distribution of fluorescence and [Ca2+]i change are consistent with endothelial cell origin of fura 2 fluorescence.

Response of fura 2-loaded OMDVR to BK and kallidin. The typical response of OMDVR to abluminal BK exposure is shown in the top panel of Fig. 4. Immediately after application of BK, [Ca²⁺]_i increased to yield peak and plateau phases. After removal of BK from the bath, [Ca²⁺]_i promptly returned to control levels. The concentration dependence of the [Ca²⁺]_i response to abluminal BK is shown in the bottom panel of Fig. 4. Baseline [Ca²⁺]_i was consistently in the range from 50 to 100 nM. BK, at concentrations >10⁷ M, gave a maximal response with mean peak values between 600 and 800 nM and a plateau [Ca²⁺]_i of ~200 nM. The response of OMDVR to Lys-BK (kallidin) was similar to that of BK (Fig. 5).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Global [Ca2+]i response of OMDVR to BK. Top: [Ca2+]i as a function of time in fura 2-loaded OMDVR as BK is added to and removed from bath. Bottom: [Ca2+]i response as a function of log-molar BK concentration ([BK]). Data are shown as means ± SE of baseline, peak, and plateau phases. Peak and plateau responses were >baseline (P < 0.05) for all [BK] >1011 M. Separate groups of vessels were run at each [BK]; N = 5 or 6 experiments each.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Comparison of global [Ca2+]i response of OMDVR to BK and kallidin (Lys-BK): baseline, peak, and plateau [Ca2+]i in OMDVR treated with 108 M BK (N = 6 experiments) or Lys-BK (N = 6). BK data are same as Fig. 4, bottom.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effect of [Phe3,Ile3,Orn3]vasopressin (V1AG) and BK on OMDVR [Ca2+]i response. Top: V1AG constricts OMDVR (N = 7 experiments) compared with time controls (N = 7). * P < 0.05 vs. control. Bottom: V1AG alone fails to elicit a [Ca2+]i change in fura 2-loaded OMDVR, but addition of BK with V1AG produces a typical peak and plateau response (N = 5).

BK receptor subtype mediating calcium response. To determine whether the calcium response to BK is mediated by B₁ or B₂ receptors, we performed a series of experiments using subtype-specific inhibitors. For convenience, the inhibitors, their actions, and the abbreviations we use to define them are summarized in Table 1. The ability of the inhibitors to raise [Ca²⁺]_i was examined (Table 2). When the inhibitors were applied by themselves, no discernible peak responses were seen but very small plateau increases in [Ca²⁺]_i were found with [Thi^5,8,D-Phe⁷]BK (TPBK) and D-Arg-[Hyp³,Thi^5,8,D-Phe⁷]BK (HTPBK) {but not D-Arg-[Hyp³,D-Phe⁷,Leu⁸]BK (HPLBK)}, indicating some slight agonist activity for these analogs. This is consistent with the actions of these inhibitors at the B₂ receptor as summarized by Regoli et al. (22) and other authors (4, 11). The ability of the inhibitors (10⁶ M) to block the [Ca²⁺]_i response generated by 10⁸ M BK was examined in a separate series of experiments. [Ca²⁺]_i was measured in fura 2-loaded, microperfused OMDVR, after which BK along with vehicle or antagonist was added to the bath. The B₁ blocker des-Arg⁹-[Leu⁸]BK (DALBK) had no significant effect on either the peak (Fig. 7, top) or plateau (Fig. 7, bottom) phase of the BK response. HTPBK and HPLBK are thought to be among the most specific B₂ blockers, with HPLBK devoid of any agonist activity (22). Both HTPBK and HPLBK completely blocked the [Ca²⁺]_i response. The mixed agonist/antagonist TPBK reduced the peak response but, on average, did not eliminate the plateau.

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Inhibition of OMDVR [Ca2+]i response by receptor subtype-specific antagonists. Mean ± SE peak (top) and plateau (bottom) responses of OMDVR exposed to BK (108 M, N = 6 experiments) with vehicle (Veh) or inhibitor (106 M). N = 8, 5, 5, and 6 for des-Arg9-[Leu8]BK (DALBK), [Thi5,8,D-Phe7]BK (TPBK), D-Arg-[Hyp3,Thi5,8,D-Phe7]BK (HTPBK), and D-Arg-[Hyp3,D-Phe7,Leu8]BK (HPLBK), respectively. * P < 0.05 vs. baseline.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Effect of BK B1-receptor agonist des-Arg9-BK (DBK) on fura 2-loaded OMDVR. Top: small [Ca2+]i response shown as function of time during DBK exposure. Compare ordinate scale to that of bottom panel and to Fig. 4. Bottom: comparison of baseline, peak, and plateau [Ca2+]i at BK or DBK concentrations of 109, 108, and 106 M (N for DBK = 5, 5, and 4 and N for BK = 6, 6, and 5 experiments, respectively). BK data are same as in Fig. 4, bottom. Compared with BK, DBK yielded a much smaller or negligible [Ca2+]i response.

Vasomotor response of OMDVR to BK. Anticipating that BK might behave as an endothelium-dependent vasodilator, we tested its effects in vessels preconstricted with ANG II. OMDVR were exposed to ANG II (10⁸ M) for 10 min after which BK (10⁷ M) was added and then removed. As previously observed, ANG II constricted OMDVR (19). A tendency for BK to reversibly dilate the vessels was consistently observed but achieved statistical significance only when 1 or 100 µM L-arginine was present in the bath throughout the experiment (Fig. 9).

View larger version (17K): [in this window] [in a new window]   Fig. 9.   Effect of BK on preconstricted OMDVR. OMDVR were preconstricted with abluminal ANG II (108 M) for 10 min after which BK (107 M) was added to bath for 10 min and then removed. ANG II constricted OMDVR [P < 0.05 vs. control period, all L-arginine (L-Arg) concentrations]. BK dilated preconstricted OMDVR in presence of 1 (N = 10 experiments) or 100 (N = 9) µM L-arginine, but vasodilation did not achieve significance without L-Arg (N = 15).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Molecular machinery for generation of kinins is present in the renal medulla, suggesting local paracrine roles for these peptide hormones. Kallikreins, enzymes that release kinins from kininogen precursors, have been localized to connecting tubule of the rat, portions of the outer medulla, and collecting ducts of the papilla (9, 21, 30, 31). Recently, it has been established that both HMW kininogen and LMW kininogen are expressed in the distal nephron of the mammalian kidney. In the human, LMW kininogen has been demonstrated in the distal nephron in close proximity to cells that express tissue kallikrein (9). With the use of RT-PCR and specific antibodies, Hermann and colleagues (12) also verified that HMW kininogen exists in a similar distribution.

The presence of receptors on tubular and vascular elements of the kidney also favors a paracrine role for BK and kallidin (8). Kinins exert their actions via at least two receptor subtypes, B₁ and B₂. With the use of anti-peptide and anti-ligand antibodies, Figueroa and colleagues (8) showed that B₂ receptors are expressed throughout the outer and inner medulla with cellular localization to a variety of nephron segments as well as afferent arteriolar smooth muscle. Information on the distribution of B₁ receptors is limited; however, it is notable that mesangial cells, which have smooth muscle-like functions, express both B₁ and B₂ receptors, each of which mediates signaling through changes in [Ca²⁺]_i (1, 2). Interestingly, activation of B₂ receptors leads to afferent dilation but mesangial cell contraction (1).

Evidence exists to support a role for kinins to modulate blood flow to and sodium reabsorption by the renal medulla in normal and pathophysiological states. A 20% reduction of papillary blood flow during infusion of a kinin antagonist was found by Roman and colleagues (24). Enhancement of kinin activity through inhibition of kininases with enalaprilat or phosphoramidon increased both papillary blood flow and sodium excretion. Seino et al. (25) showed that injection of a competitive BK antagonist into Sprague-Dawley, Wistar-Kyoto, and spontaneously hypertensive rats (SHR) caused a reduction of blood pressure and renal blood flow. The blood pressure reduction in the SHR was less than that for the other groups, suggesting that some deficiencies of the kinin system can contribute to hypertension (25). A later study by Mattson and Cowley (15) supported a role for kinins as an endothelium-dependent vasodilator. Interstitial infusion of BK increased sodium and water excretion and enhanced papillary blood flow, an effect blocked by nitric oxide synthase inhibition.

In this study, we performed the first characterization of the role of kinins in modulation of the functions of descending vasa recta by measuring the [Ca²⁺]_i response of fura 2-loaded, in vitro perfused OMDVR and the effect of BK on OMDVR vasomotor tone. This is also the first study to report [Ca²⁺]_i responses in OMDVR. For this reason we sought to carefully determine which distribution of cell type(s) load fura 2 by cleavage of the acetoxymethyl ester. Surprisingly, comparison of white light and fluorescent video images of OMDVR indicates virtually no loading into pericytes (Figs. 2, 3). The reason for this is uncertain, but fura 2 is known to vary with respect to its loading by deesterification in different cell types (17), and some cells extrude fura 2 via the action of organic anion transporters (7). A second line of evidence favoring an endothelial cell origin for the fura 2 fluorescent signal is provided by the lack of a response of fura 2-loaded OMDVR to vasopressin analogs. Vasopressin contracts smooth muscle by binding to the V₁ receptor, leading to activation of phospholipase C and the phosphatidylinositol pathway and an increase in [Ca²⁺]_i (16). Application of V1AG to OMDVR leads to vasoconstriction but does not induce a fura 2 response (Fig. 6).

Blockade of B₂ but not B₁ receptors with subtype-specific antagonists eliminated the rise in [Ca²⁺]_i induced by BK (Fig. 7), verifying the presence of a B₂-type receptor on OMDVR endothelia. Because of the limitations of these methods, however, we cannot rule out the existence of B₁ or B₂ receptors on the pericytes that surround the OMDVR wall.

The observation that BK increases [Ca²⁺]_i in endothelial cells led us to hypothesize that it would act as a vasodilator through release of nitric oxide, an effect previously observed in response to acetylcholine (32). BK-induced vasodilation of ANG II-preconstricted vessels was confirmed (Fig. 9). To achieve maximal vasodilation, it was necessary to have L-arginine in the bath because the tendency toward vasodilation did not achieve significance in the absence of L-arginine. A dependence of endothelial cells on external L-arginine for maximal nitric oxide generation has been observed in cell culture. Buckley et al. (5) increased [Ca²⁺]_i in bovine aortic endothelial cells using BK or thapsigargin, an inhibitor of the intracellular calcium adenosinetriphosphatase responsible for intracellular calcium uptake and storage. Nitric oxide generation increased markedly when medium L-arginine concentration was increased from 0 to 1 µM, with no additional effect of increasing L-arginine further to 100 µM (5). This sensitivity of OMDVR vasodilation to external L-arginine might have physiological significance. Larson and Lockhart (14) found that L-arginine infusion restored the sensitivity of vasa recta blood flow to increases in perfusion pressure in SHR. The possibility that medullary interstitial L-arginine concentrations might be modulated through countercurrent exchange or multiplication is suggested by the work of Dantzler and Silbernagl (6), who showed carrier-mediated reabsorption of L-arginine from the loops of Henle.

In summary, we examined the [Ca²⁺]_i in isolated, in vitro perfused OMDVR by fluorescent ratio imaging using fura 2. BK and kallidin, which are mixed B₂- and B₁-receptor agonists, each elicited an increase in [Ca²⁺]_i from baseline values in the range from 50 to 100 nM to peak values of 600-800 nM followed by a sustained plateau of 150-250 nM. The BK [Ca²⁺]_i response was blocked by selective B₂-receptor antagonists but not a B₁ antagonist. BK vasodilated microperfused OMDVR that had been preconstricted with 10⁸ M ANG II. We conclude that BK is a vasodilator of preconstricted OMDVR and that its actions are probably mediated through release of endothelium-dependent vasodilators in response to increases in [Ca²⁺]_i. It is likely that kinins modulate renal medullary blood flow, at least partially, through these effects on OMDVR.