点击显示 收起
【摘要】 The Na + /K + /2Cl - cotransporter (NKCC) plays diverse roles in the kidney, contributing sodium reabsorption and tubuloglomerular feedback (TGF). However, NKCC is also expressed in smooth muscle and inhibitors of this transporter affect contractility in both vascular and nonvascular smooth muscle. In the present study, we investigated the effects of NKCC inhibitors on vasoconstrictor responses of the renal afferent arteriole using the in vitro perfused hydronephrotic rat kidney. This preparation has no tubules and no TGF, eliminating this potential complication. Furosemide and bumetanide inhibited myogenic responses in a concentration-dependent manner. Bumetanide was 20-fold more potent (IC 50 1.0 vs. 20 µmol/l). At 100 and 10 µmol/l, furosemide and bumetanide inhibited myogenic responses by 72 ± 4 and 68 ± 5%, respectively. The maximal level of inhibition by bumetanide was not affected by nitric oxide synthase inhibition (100 µmol/l N G -nitro- L -arginine methyl ester). However, the time course for the dilation was slowed (from t 1/2 = 4.0 ± 0.5 to 8.3 ± 1.7 min, P = 0.04), suggesting either a partial involvement of NO or a permissive effect of NO on relaxation kinetics. Bumetanide also inhibited ANG II-induced afferent arteriolar vasconstriction at similar concentrations. Finally, NKCC1, but not NKCC2, expression was demonstrated in the afferent arteriole by RT-PCR and the presence of NKCC1 in afferent arteriolar myocytes was confirmed by immunohistochemistry. In concert, these results indicate that NKCC modulation is capable of altering myogenic responses by a mechanism that does not involve TGF and suggest a potential role of NKCC1 in the regulation of vasomotor function in the renal microvasculature.
【关键词】 renal autoregulation tubuloglomerular feedback vascular smooth muscle chloride
THE N a/ K /2 Cl COTRANSPORTER (NKCC) is involved in multiple aspects of renal function. NKCC2 (also known as BSC1) is selectively expressed in the apical membrane of the thick ascending limb (TAL) of the loop of Henle and in the macula densa cells of the terminal portion of this nephron segment ( 17, 43 ). This transporter plays a critical role in Na reabsorption in the TAL and in the sensing of flow-dependent signals by the macula densa. A second isoform of this transporter, NKCC1 (also known as BSC2), exhibits a wider pattern of expression ( 17, 43 ). In the kidney, NKCC1 is expressed in basalateral epithelial membranes ( 43 ), mesangial cells, and renin-producing juxtaglomerular (JG) cells of the afferent arteriole ( 8, 26 ). Indeed, recent studies suggest that the effects of NKCC inhibitors, such as furosemide, on renin secretion are mediated, in part, by a direct effect on JG cells ( 8 ).
In regard to the regulation of the renal microvasculature, the role of NKCC2 is firmly established. This transporter is essential in the tubuloglomerular feedback (TGF) response, which senses and regulates the delivery of filtrate to the early distal tubule. As the filtrate passes through the TAL, the reabsorption of sodium and retention of water results in a flow-dependent dilution of and reduction in osmolarity in the fluid presented to the luminal surface of the macula densa. Elevated TAL flow increases , stimulating NKCC2 activity at the macula densa and triggering the release of ATP ( 4 ), which is then converted to adenosine by ectonucleotidases ( 4, 7 ). Adenosine and ATP both elicit a selective constriction of the afferent arteriole ( 24, 45 ) completing the feedback loop by reducing glomerular filtration rate (GFR) and TAL flow. Pharmacological inhibitors of NKCC interfere with this signaling pathway and effectively block TGF responses. Accordingly, tools such as furosemide have been used extensively to indirectly assess the physiological role of TGF and, for example, to determine the effects of eliminating TGF on renal autoregulation.
NKCC1 is also expressed in many smooth muscle types and is thought to play an important role in establishing the outward chloride gradient (reviewed in Ref. 10 ). The outward movement of Cl - represents an important "inward" or depolarizing current in smooth muscle, and a modulation of the Cl - gradient could impact basal membrane potential and/or cell signaling. Indeed, NKCC inhibitors have been shown to exert direct vasodilatory effects in renal and nonrenal vascular beds (e.g., Ref. 3 ) and have been demonstrated to elicit relaxant effects in both vascular and nonvascular smooth muscle (e.g., Refs. 2, 15, 28, 30, 38, 46, 47, 50 ). To determine whether inhibiting NKCC exerts a similar vasodilatory effect on the renal afferent arteriole through mechanisms that are independent of TGF signaling, we examined the effects of furosemide and bumetanide on afferent arteriolar vasoconstrictor responses using the in vitro perfused hydronephrotic rat kidney model. In this preparation, chronic ureteral ligation produces a complete atrophy of all tubular elements, facilitating direct visualization of the microvasculature, but also eliminating TGF and any possible effect of acute alterations in TGF on the afferent arteriole.
METHODS
The use of the in vitro perfused hydronephrotic kidney model to study the renal myogenic response has been described in detail in previous publications (e.g., Refs. 9, 21, 33 ). The protocol for this study was approved by the University of Calgary Animal Care Committee, and the procedures comply with the regulations of the Canadian Council on Animal Care. Chronic unilateral hydronephosis is induced to facilitate visualization of the renal microvasculature and results in a complete atrophy of the tubular elements. The left ureters of 6- to 7-wk-old (80-100 g) male Sprague-Dawley rats are ligated under halothane-induced anesthesia. The left kidney is then harvested after 6-8 wk. The left renal artery is cannulated in situ and the kidney is excised and transferred to a heated chamber on the stage of an inverted microscope, without disrupting perfusion. The renal perfusate is DMEM (Sigma, St. Louis, MO) containing 30 mmol/l bicarbonate, 5 mmol/l glucose, and 5 mmol/l HEPES, and equilibrated with 95% air-5% CO 2 (pH 7.40). Temperature is maintained at 37°C. Medium is pumped on demand through a heat exchanger to a pressurized reservoir feeding the renal arterial cannula. The effluent drains into the bath, providing access to both abluminal and adluminal sites. A fiber optic probe is used to stabilize and transilluminate a portion of the membranous renal cortex for observations of renal microvascular responses. Afferent arteriolar diameters are measured by on-line image processing. Diameter measurements are obtained at each pixel and averaged over the entire segment length (20-40 µm). Data are collected at a scanning rate of 5-8 Hz and diameter measurements are averaged over the plateau of the each response ( 1-min period).
Kidneys were allowed to equilibrate for at least 1 h following the establishment of in vitro perfusion. Basal renal perfusion pressure was held at 80 mmHg (measured at the level of the renal artery) during the equilibration period. All experiments were performed in the presence of 10 µmol/l ibuprofen. Myogenic responses were evoked by raising renal arterial pressure from 80 to 160 mmHg in 20-mmHg steps, holding each pressure for at least 1 min to attain steady-state responses (see Ref. 33 ). Kidneys were then pretreated with either furosemide or bumetanide for 10 min and myogenic responses were reevaluated (illustrated in Fig. 1 ). In some experiments (e.g., Fig. 1 ), the inhibitor was washed out and responses were reassessed after 1 h. The myogenic responses of this preparation do not spontaneously abate over the time course of these studies. For example, we showed previously that, under these experimental conditions, identical myogenic responses are attained with repeated pressure ramps over periods in excess of 7 h (see Ref. 9 ). To evaluate the possible involvement of endothelial factors, a separate series of kidneys were pretreated with 100 µmol/l N G -nitro- L -arginine methyl ester ( L -NAME) for 15 min and responses to 10 µmol/l bumetanide were followed in a time-dependent manner. We also evaluated the effects of bumetanide on the afferent arteriolar responses to 0.1 nmol/l ANG II in an additional series of kidneys. Ibuprofen (10 µmol/l) was included in these experiments and perfusion pressure was maintained constant at 80 mmHg. Bumetanide was added at increasing concentrations (0.1-10 µmol/l) to kidneys that had been preconstricted with ANG II.
Fig. 1. Original tracing illustrating the effects of bumetanide on myogenic responses of the afferent arteriole. Note concentration-dependent inhibition of response and full recovery of myogenic reactivity following washout of bumetanide.
RT-PCR studies were performed using single afferent arterioles isolated from normal rat kidneys using the gel perfusion technique developed in our laboratory ( 32 ). Briefly, kidneys were flushed of blood in situ with warm medium and then perfused with 2% agarose solution at 37°C. The kidneys were excised and chilled ( 4°C) to solidify the agarose. Thin cortical slices were treated with collagenase IV, dispase II, and DNase to dissociate vessels. The isolated arterioles were then washed three times. Twenty vessels were pooled for the assay and whole kidney homogenate and distilled water aliquots were included as positive and negative controls. Samples were lysed in Igepal 630 (nonyl phenol 9 mole ethoxylate, Sigma), stored at -80°C, and processed using the OneStep RT-PCR kit (Qiagen, Mississauga, ON). The forward and reverse primers used to assay for NKCC1 (gtcacctggtaccaaggatgtg and gctgacgatccagtcactctg) amplify positions 2929 to 3809 of the mRNA sequence (predicted size 880 bases) and span seven intron sites. The primers used to assay for NKCC2 (gagttaccgccaagttcgactg and gaatcaggtgggatagtctccag) amply positions 3298 to 3626 of the mRNA sequence (predicted size 328 bases) and span two intron sites. PCR was performed at 35 cycles. The products were separated on 2% agarose, visualized by ethidium bromide, and sequenced (University of Calgary DNA Sequencing Facilities, www.med.ucalgary.ca/dnaservices ), confirming their identification as the predicted NKCC1 and NKCC2 mRNA amplicons.
Immunocytochemistry was performed using myocytes that were freshly dispersed from isolated single afferent arterioles (technique described in Ref. 32 ). The isolated myocytes were fixed in 1% formalin (Sigma) and made permeable to the antibodies by treatment with 0.1% Triton/PBS. The cells were then treated with 3% donkey serum to block nonspecific binding sites and exposed to a 1/50 dilution of the primary antibody, anti-NKCC1 (sc-21547, Santa Cruz Biotechnology, Santa Cruz, CA) overnight. The primary antibody was washed in 3% donkey serum/PBS and exposed to the secondary antibody, Alexa Fluor 488 donkey anti-goat IgG (A11055 , Molecular Probes, Eugene, OR) and then washed in PBS. For controls (nonspecific binding), myocytes were treated identically, but exposure to the primary antibody was omitted. Images were obtained using an Olympus Fluoview 300 laser-scanning confocal microscope. All images were obtained under the same exposure and conditions.
Furosemide, bumetanide, ibuprofen, and L -NAME were obtained from Sigma. Fresh stock solutions of furosemide and bumetanide were prepared in dimethylsulfoxide immediately before use. All data are presented as means followed by the standard error of the mean as an index of dispersion. Differences between treatment groups were assessed by one-way ANOVA and Student?s t -test. Differences between means exhibiting P values <0.05 were considered to be statistically significant.
RESULTS
The tracing depicted in Fig. 1 illustrates the protocol used to examine the effects of NKCC inhibition on myogenic responses of the renal afferent arteriole. Mean data obtained from such studies are presented in Fig. 2. The mean data for bumetanide are shown in Fig. 2 A. In this group increasing renal arterial pressure from 80 to 100, 120, 140, and 160 mmHg reduced afferent arteriolar diameters in the controls from 18.5 ± 1.0 to 14.2 ± 1.1, 10.2 ± 0.9, 7.5 ± 0.6, and 6.2 ± 0.8 µm, respectively. The administration of bumetanide at concentrations ranging from 0.01 to 10 µmol/l had no effect on basal (80 mmHg) diameters ( P 0.7). This finding is consistent with previous observations that this preparation exhibits little basal tone at low perfusion pressures. For example, calcium antagonists similarly do not affect basal diameters ( 21, 34 ). Bumetanide caused a concentration-dependent inhibition of myogenic vasoconstriction. At the lowest level (0.01 µmol/l), bumetanide significantly attenuated the response at 140 mmHg ( P = 0.04). At concentrations of 0.1 to 10 µmol/l, bumetanide significantly reduced the contractile responses to each pressure step (100-160 mmHg). At the highest renal arterial pressure (160 mmHg), diameters were 6.2 ± 0.8 µm for controls and 8.1 ± 0.5 ( P = 0.068), 10.7 ± 0.7 ( P = 0.0017), 12.4 ± 0.8 ( P = 0.0017), and 14.7 ± 1.1 µm ( P < 0.0001) following treatment with 0.01, 0.1, 1.0, and 10 µmol/l bumetanide, respectively.
Fig. 2. Mean data illustrating concentration-dependent inhibition of myogenic responses by bumetanide ( A, n = 6) and furosemide ( B, n = 6). Control responses indicated by.
Similar effects were seen with furosemide ( Fig. 2 B ). In the controls for this series, increasing renal perfusion pressure from 80 to 100, 120, 140, and 160 mmHg reduced afferent arteriolar diameters from 19.4 ± 0.9 to 16.7 ± 0.6, 12.0 ± 0.6, 9.4 ± 0.4, and 7.5 ± 0.2 µm, respectively. As seen with bumetanide, furosemide had no effect on basal diameters at 80 mmHg, but attenuated responses elicited by elevated renal arterial pressures. At a concentration of 0.1 µmol/l, furosemide caused a significant inhibition of the response seen at 160 mmHg ( P = 0.033), whereas higher concentrations (1.0, 10, and 100 µmol/l) significantly attenuated responses at all elevated pressures (100 to 160 mmHg). At 160 mmHg, diameters were 7.5 ± 0.2 in controls and 9.3 ± 0.8 ( P = 0.03), 11.3 ± 0.3( P = 0.02), 12.9 ± 0.5 ( P = 0.02), and 16.5 ± 0.9 µm ( P < 0.0001) in the presence of 0.1, 1.0, 10, and 100 µmol/l furosemide, respectively.
These same data are depicted in dose-response form in Fig. 3, to facilitate a comparison of the concentration dependency of the actions of the two NKCC inhibitors. Bumetanide is a more potent inhibitor of NKCC than furosemide and its diuretic effects are elicited at 20- to 40-fold lower concentrations. As shown in Fig. 3, a similar relationship is seen in regard to the effects of these two agents on the afferent arteriolar myogenic response. Thus the IC 50 for bumetanide-induced inhibition of myogenic response was 20-fold lower than that for furosemide, consistent with an effect mediated by NKCC blockade.
Fig. 3. Comparison of the concentration dependency of the inhibitory effects of bumetanide ( ) and furosemide ( ) on the afferent arteriolar response to pressure (160 mmHg). Note that IC 50 for actions of bumetanide (1.0 µmol/l) was 20-fold less than that of furosemide (20 µmol/l).
The RT-PCR and immunocytochemistry studies demonstrated the expression of NKCC1 in the afferent arteriole and in afferent arteriolar myocytes ( Figs. 4 and 5 ). As shown in Fig. 4, RT-PCR of pooled, intact isolated afferent arterioles resulted in a single product of the predicted size (880 bases) and corresponding to the product obtained with the whole kidney homogenate. The identification of this product as the NKCC1 amplicon was confirmed by sequencing. The RT-PCR studies assaying for NKCC2 demonstrated expression in whole kidney but failed to reveal any expression in afferent arterioles (data not shown). The paired images presented in Fig. 5, A and B, illustrate the presence of immunoreactive protein in afferent arteriolar myocytes treated with anti-NKCC1 primary antibody and fluorescent secondary antibody. As shown by the images of Fig. 5, C and D, no significant fluorescence was seen in myocytes treated with the secondary antibody alone and subjected to the same imaging conditions.
Fig. 4. Results of RT-PCR revealed message for NKCC1 in isolated afferent arterioles. Lane 1 : size standards. Lane 2 : (H 2 O) distilled water (negative control). Lane 3 : (KID) whole kidney homogenate (positive control). Lanes 4 and 5 : (AA) isolated afferent arteroles (20 vessels in each sample); 2% agarose gel. NKCC2 expression was seen in the whole kidney homogenate, but not in the AA (data not shown).
Fig. 5. NKCC1 expression in myocytes freshly dispersed from single afferent arterioles. A and B : light (DIC) and fluorescent images of myocytes treated with NKCC1 primary antibody and Alexa Fluor secondary antibody. C and D : images of cells from same preparation treated with the secondary (Alexa Fluor) antibody alone and taken at the same exposure.
The literature suggests NKCC blockers may evoke vasodilation, in part, through endothelial- or nitric oxide (NO)-dependent mechanisms. To examine this possibility, studies were conducted in the presence of the NO synthase blocker L -NAME (100 µmol/l). Renal perfusion pressure was increased from 80 to 160 mmHg and then bumetanide (10 µmol/l) was administered as pressure was maintained at 160 mmHg. As illustrated in Fig. 6 A, elevating renal perfusion pressure from 80 to 160 mmHg reduced afferent arteriolar diameters from 20.2 ± 1.2 to 9.2 ± 0.6 µm in controls ( n = 4), and subsequent addition of 10 µmol/l bumetanide increased diameters to 20.8 ± 0.8 µm. In the presence of L -NAME ( n = 4), increasing renal perfusion pressure from 80 to 160 mmHg reduced diameters from 20.4 ± 1.0 to 8.9 ± 1.3 µm, and the addition of bumetanide returned diameters to 19.0 ± 1.2 µm. Thus the presence of L -NAME did not affect basal diameters ( P = 0.88), diameters at 160 mmHg ( P = 0.82), or final diameter responses to bumetanide ( P = 0.25). These data, expressed as percent inhibition, yielded bumetanide-induced vasodilatory responses of 107 ± 5% for control and 90 ± 6% for the L -NAME group ( P = 0.26). Although the maximal responses were thus not altered by NOS blockade, the time course for the vasodilation was significantly impaired by L -NAME treatment. Mean data, expressed as percent vasodilation, and half time to maximal responses ( t 1/2 ) are depicted in Fig. 6 B. The t 1/2 values were 4.0 ± 0.5 min in controls vs. 8.3 ± 1.7 min in L -NAME ( P = 0.04).
Fig. 6. Effects of nitric oxide (NO) synthase inhibition [100 µmol/l N G -nitro- L -arginine methyl ester ( L -NAME)] on 10 µmol/l bumetanide-induced inhibition of myogenic tone. Note that L -NAME had no significant effect on basal diameters, pressure-induced contraction, or the steady-state dilation induced by bumetanide ( A ). However, as shown in B, the rate of vasodilation induced by bumetanide was significantly slower in the presence of L -NAME vs. control ( n = 4).
Finally, to ascertain whether the effects of NKCC inhibition were specific to the myogenic response, we determined the effects of bumetanide on afferent arteriolar vasoconstriction induced by ANG II. The results of these studies are shown in Fig. 7. As depicted, 0.1 nmol/l ANG II reduced afferent arteriolar diameters from 15.8 ± 1.0 to 4.5 ± 0.6 µm ( n = 5). Bumetanide at concentrations of 0.1, 1.0, and 10 µmol/l returned diameters to 7.3 ± 1.1 ( P = 0.055), 10.4 ± 1.1 ( P = 0.001), and 13.7 ± 0.8 ( P = 0.00002 vs. ANG II alone), corresponding to 25 ± 6, 52 ± 5, and 82 ± 5% inhibition of the ANG II response.
Fig. 7. Effects of bumetanide on afferent arteriolar responses to ANG II ( n = 5). Note that the concentration dependency for the bumetanide-induced inhibition of the vasoconstriction elicited by ANG II was similar to that subtending the inhibition of myogenic vasoconstriction ( Fig. 3 ).
DISCUSSION
The present study demonstrates that NKCC inhibitors exert marked effects on the reactivity of the afferent arteriole, reducing vasoconstrictor responses to elevated pressure and ANG II. Such actions are independent of secondary modulating effects via TGF and do not appear to require endothelium-derived NO or cyclooxygenase products. These findings suggest a potentially important role of NKCC in the afferent arteriole. The Cl - concentration within the smooth muscle myocyte is maintained above equilibrium levels and outward Cl - movements represent an important depolarizing current in most vascular smooth muscles. The NKCC1 transporter is suggested to play a critical role in establishing this Cl - gradient ( 10 ) and could thus be important in maintaining normal physiological reactivity of the afferent arteriole.
We interpreted the fact that we observed these effects in the in vitro perfused hydronephrotic rat kidney as providing unequivocal evidence that TGF is not involved in the modulating actions of NKCC inhibitors on afferent arteriolar reactivity. This preparation lacks tubules and does not have a functioning JGA, eliminating any possibility of acute TGF-dependent modulation. Our model has been used extensively to study the renal myogenic response, but it is perhaps appropriate to comment on its characteristics in regard to the present study. As shown in Figs. 1 and 2 and illustrated in past publications (e.g., Refs. 9, 33 ), robust and graded myogenic responses are evoked over the same range in renal perfusion pressures as that subtending renal autoregulation in the normal kidney. Moreover, the kinetic attributes of the myogenic response seen in this preparation have been shown to be identical to the myogenic component of autoregulation in the normal intact kidney ( 11, 33 ). Pathophysiological models exhibiting altered renal autoregulation in the intact kidney demonstrate similar alterations in myogenic reactivity when studied using this in vitro preparation ( 21, 23, 48, 53 ). Similarly, the profile of the effects of pharmacological treatments on the myogenic response of this preparation parallels the effects seen on autoregulation in the normal intact kidney ( 22, 27, 34 ). Accordingly, while one cannot rule out the possibility that the effects of NKCC we observed reflect an abnormal property of the myogenic response of this preparation, this argument is not supported by the existing literature.
We suggest that the vasorelaxant effects of furosemide and bumetanide that we observed on the afferent arteriole reflect the consequence of blocking NKCC. Previous reports clearly document direct smooth muscle-relaxing effects of NKCC inhibitors on isolated blood vessels ( 2, 5, 15, 46, 47, 50 ) and in nonvascular smooth muscle ( 28, 30, 38 ). The specific involvement of NKCC1 in the vasodilator effects of NKCC inhibitors is supported by the observation that elimination of this isoform by gene deletion eliminates the depressor effects of bumetanide on the intrinsic spontaneous vasomotor activity of the isolated portal vein ( 5, 36 ). The NKCC1 knockout is also associated with a hypotensive phenotype, which is suggested to reflect, at least in part, reduced vasomotor tone ( 36, 55 ). Most cell types are thought to express some form of this transporter ( 17, 43 ) and in the present study we provide evidence that NKCC1 is expressed in the smooth muscle myocytes of the afferent arteriole. This finding is consistent with previous reports of NKCC1 expression in the mouse afferent arteriole and in renin-producing JG cells of the mouse ( 8, 26 ). The concentration dependencies of the vasodepressor effects we observed for bumetanide and furosemide are also consistent with an action involving NKCC1. Bumetanide inhibited myogenic responses with an IC 50 that was 20-fold less than that of furosemide ( Fig. 3 ). Rubidium flux assays demonstrate a similar 20-fold difference in the inhibitory actions of these agents on HEK cells expressing hNKCC1 ( 19 ) and similar differences in potencies are noted for the diuretic actions of these agents ( 41 ).
Several mechanisms have been suggested to be involved in the vasodilator effects of NKCC inhibition, but the most widely supported mechanism relates to a critical role of NKCC in establishing the Cl - gradient in smooth muscle myocytes. There are three potential mechanisms for the maintenance of the normal Cl - gradient in smooth muscle. These include the chloride/bicarbonate exchanger, a partially characterized acetazolamide-sensitive transporter ("pump III") and NKCC (reviewed in Ref. 10 ). Early studies clearly demonstrated that NKCC inhibitors reduce smooth muscle Cl content ( 6 ). Since Cl - efflux represents a depolarizing current, a reduction in intracellular [Cl - ] and the outward Cl - gradient would be anticipated to elicit hyperpolarization and attenuate depolarizing signals. Indeed, NKCC inhibitors elicit hyperpolarization in the rabbit pulmonary artery ( 29 ), in rat femoral and saphenous arteries ( 12, 13 ), and in human umbilical and placental arteries ( 14 ). These effects mirror changes in intracellular [Cl - ] ( 12 - 14, 29 ). We previously demonstrated that the afferent arteriolar myogenic response and the response to ANG II are attenuated by manipulations known to elicit hyperpolarization ( 9, 35, 42 ).
NKCC is also expressed in the vascular endothelium, and the Cl - gradient plays an important role in Ca 2+ signaling and NO synthase activation in this tissue ( 39, 43 ). The vasorelaxant effects of furosemide in the blood perfused rat tail artery were reported to be endothelium dependent ( 18 ), whereas endothelial dependence has not been observed in other preparations ( 50 ). In the present study, the level of bumetanide-induced dilation was not significantly affected by L -NAME, consistent with a direct smooth muscle action. However, NOS inhibition was observed to significantly slow the rate of relaxation. This effect may suggest either an impact of basal NO on relaxation kinetics and/or a concomitant effect of bumetanide on endothelial NO production. A reduction in intracellular [Cl - ] could cause endothelial cell hyperpolarization, stimulating Ca 2+ entry and NO formation. However, the mechanisms underlying this effect of L -NAME were not pursued and its biological significance is not clear at the present time. Nevertheless, it can be concluded that while NO may contribute to the kinetics of the response, NO formation is not an obligate component of the vasodilatory actions of bumetanide on the afferent arteriole, as the magnitude of the steady-state dilation was not significantly affected. Similarly, it is important to note that the effects of NKCC inhibition seen in the present study were observed in the presence of cyclooxygenase inhibition (see METHODS ), whereas some of the in vivo vascular effects of NKCC inhibitors have been reported to depend on prostaglandin formation (reviewed in Ref. 16 ).
The lack of a vasoconstrictor effect of L -NAME alone in this model merits comment, as L -NAME generally elicits renal vasoconstriction in vivo. The effectiveness of this treatment in eliminating endothelium-derived NO induced vasodilation this preparation has been established by our previous work ( 51, 52, 56 - 58 ). However, it is quite typical that neither blockade of NO formation by L -NAME ( 51, 52, 56 - 58 ) nor inhibition of NO?s downstream target, soluble guanylyl cyclase ( 52 ), alters basal afferent arteriolar diameter in this model. We do not believe that this reflects an altered reactivity to NO. For example, L -NAME also has no effect on basal renal perfusate flow in normal rat kidneys studied under the same in vitro conditions ( 20 ) and we previously showed that the afferent arteriole of the hydronephrotic kidney model exhibits normal reactivity to authentic NO ( 52 ). We have not measured NO levels in our system and it is conceivable that basal NO concentrations are lower, due to a lack of exogenous secretagogues, dilution by high perfusate flows, or a lack of macula densa-derived NO. However, we believe that the most likely explanation may be that the vasoconstriction seen with NO removal in vivo is due to the presence of unopposed vasoconstrictor stimuli, which are lacking in our in vitro preparations.
Furosemide is often used as a tool to inhibit TGF ( 4 ) and its direct effects on the afferent arteriole could complicate such an approach. However, it is difficult to extrapolate to what extent, if any, this occurs or to estimate the magnitude of such potential effects from the findings of the present study. Important considerations include drug concentrations, the routes of administration, and the peculiar pharmacology of furosemide. Under our experimental conditions, 100 µmol/l furosemide inhibited myogenic responses by 70%. Such marked effects on autoregulation are certainly not seen in vivo. However, we found that as little as 1 µmol/l significantly attenuated myogenic responses over all pressures. In micropuncture studies, furosemide is typically administered to the tubular lumen at concentrations of 100-200 µmol/l ( 31, 37 ). This route of administration would limit direct exposure of the smooth muscle to this agent, unless it leaked into the surrounding tissues. When used systemically to block TGF, typical intravenous doses of 5-10 mg/kg are used in the rat ( 25, 40, 44 ). Assuming a volume of distribution in the rat of 0.3 l/kg ( 1 ), this approach would result in peak plasma concentrations (bound plus free) of 50-100 µmol/l, which could potentially produce direct effects, at least in the initial periods following injection. However, circulating levels would rapidly decline as the drug is eliminated by the kidney. Moreover, furosemide is avidly bound to plasma proteins 95%), reducing the fraction of free drug available to interact with NKCC1. This protein binding does not impede the tubular actions of furosemide, as the drug is actively secreted into the tubular lumen and concentrated in the TAL where it interacts with luminal NKCC2. Thus the tubular actions of furosemide dominate. The extent to which the direct actions of furosemide might contribute to the effects attributed to TGF in such in vivo experiments is unknown. However, the effects of furosemide on the myogenic component of autoregulation are very modest. Studies of TGF conducted using in vitro preparations often involve adding furosemide directly to an artificial perfusate. Concentrations in the range of 50 µmol/l are generally maintained throughout the study ( 24, 49, 54 ). The amount of free drug available to directly affect the renal vasculature would depend on the drug-binding capacity of the perfusate. It is not known whether direct effects are seen in this setting. Studies addressing this possibility would be of interest.
The present study points to an important role of NKCC1 in the renal afferent arteriole. A resolution of the cellular mechanisms underlying the vasodilator effects of NKCC inhibitors on this vessel would be of considerable interest. For example, are alterations in the Cl - gradient and/or membrane potential involved, as studies in other vascular preparations have suggested ( 10 )? Another interesting issue awaiting further study concerns the possibility that afferent arteriolar reactivity might be modulated by physiological or pathophysiological events that have been shown to modulate NKCC1 activity in other smooth muscle types. In this regard, it is of interest that the activity of this transporter is modulated by intracellular signaling (reviewed in Ref. 17 ) and is reported to be altered in disease models such as hypertension (see Ref. 10 ). Whether NKCC1 activity in the afferent arteriole is similarly affected in pathophysiological settings and whether such alterations result in altered afferent arteriolar reactivity are interesting questions for future investigations.
GRANTS
These studies were supported by a grant from the Heart and Stroke Foundation of Alberta, NWT, and Nunavut. Dr. Loutzenhiser is an Alberta Heritage Foundation for Medical Research (AHFMR) Scientist.
ACKNOWLEDGMENTS
The authors thank Dr. B. Walcott for advice on NKCC antibodies.
【参考文献】
Andreasen F, Tietze I, Hansen FA, Petersen JS, Christensen S, Steiness E. Furosemide kinetics and dynamics in rats and humans. Comp Biochem Physiol C 100: 635-641, 1991.
Anfinogenova YJ, Baskakov MB, Kovalev IV, Kilin AA, Dulin NO, Orlov SN. Cell-volume-dependent vascular smooth muscle contraction: role of Na +, K +, 2Cl - cotransport, intracellular Cl- and L-type Ca 2+ channels. Pflügers Arch 449: 42-55, 2004.
Barthelmebs M, Stephan D, Fontaine C, Grima M, Imbs JL. Vascular effects of loop diuretics: an in vivo and in vitro study in the rat. Naunyn Schmiedebergs Arch Pharmacol 349: 209-216, 1994.
Bell PD, Lapointe JY, Peti-Peterdi J. Macula densa cell signaling. Annu Rev Physiol 65: 481-500, 2003.
Blair-West JR, McKinley MJ, McKenzie JS. Effect of furosemide on the reactivity of rat portal vein. J Pharm Pharmacol 24: 442-446, 1972.
Cantiello H, Copello J, Muller A, Mikulic L, Villamil MF. Effect of bumetanide on potassium transport and ionic composition of the arterial wall. Am J Physiol Renal Fluid Electrolyte Physiol 251: F537-F546, 1986.
Castrop H, Huang Y, Hashimoto S, Mizel D, Hansen P, Theilig F, Bachmann S, Deng C, Briggs J, Schnermann J. Impairment of tubuloglomerular feedback regulation of GFR in ecto-5'-nucleotidase/CD73-deficient mice. J Clin Invest 114: 634-642, 2004.
Castrop H, Lorenz JN, Hansen PB, Friis U, Mizel D, Oppermann M, Jensen BL, Briggs J, Skott O, Schnermann J. Contribution of the basolateral isoform of the Na-K-2Cl- cotransporter (NKCC1/BSC2) to renin secretion. Am J Physiol Renal Physiol 289: F1185-F1192, 2005.
Chilton L, Loutzenhiser R. Functional evidence of an inward rectifier potassium current in the renal afferent arteriole. Circ Res 88: 152-158, 2001.
Chipperfield AR, Harper AA. Chloride in smooth muscle. Prog Biophys Mol Biol 74: 175-221, 2000.
Cupples WA, Loutzenhiser R. Dynamic autoregulation in the in vitro perfused hydronephrotic kidney. Am J Physiol Renal Physiol 275: F126-F130, 1998.
Davis JP. The effects of Na + -K + -Cl - cotransport and Cl - -HCO 3 - exchange blockade on the membrane potential and intracellular chloride levels of rat arterial smooth muscle, in vitro. Exp Physiol 77: 857-862, 1992.
Davis JP. Evidence against a contribution by Na + -Cl - cotransport to chloride accumulation in rat arterial smooth muscle. J Physiol 491: 61-68, 1996.
Davis JP, Chien PF, Chipperfield AR, Gordon A, Harper AA. The three mechanisms of intracellular chloride accumulation in vascular smooth muscle of human umbilical and placental arteries. Pflügers Arch 441: 150-154, 2000.
Deth RC, Payne RA, Peecher DM. Influence of furosemide on rubidium-86 uptake and alpha-adrenergic responsiveness of arterial smooth muscle. Blood Vessels 24: 321-333, 1987.
Dormans TP, Pickkers P, Russel FG, Smits P. Vascular effects of loop diuretics. Cardiovasc Res 32: 988-997, 1996.
Gamba G. Molecular physiology and pathophysiology of electroneutral cation-chloride cotransporters. Physiol Rev 85: 423-493, 2005.
Gerkens JF, Armsworth SJ, Dosen PJ, Smith AJ. Endothelium-dependent inhibition of sympathetic vasoconstriction by frusemide administration to rats. Clin Exp Pharmacol Physiol 15: 449-455, 1988.
Gillen CM, Brill S, Payne JA, Forbush B 3rd. Molecular cloning and functional expression of the K-Cl cotransporter from rabbit, rat, and human. A new member of the cation-chloride cotransporter family. J Biol Chem 271: 16237-16244, 1996.
Gui Y, Loutzenhiser R, Hollenberg MD. Bidirectional regulation of renal hemodynamics by activation of PAR1 and PAR2 in isolated perfused rat kidney. Am J Physiol Renal Physiol 285: F95-F104, 2003.
Hayashi K, Epstein M, Loutzenhiser R. Pressure-induced vasoconstriction of renal microvessels in normotensive and hypertensive rats: studies in the isolated perfused hydronephrotic kidney. Circ Res 65: 1475-1484, 1989.
Hayashi K, Epstein M, Loutzenhiser R. Determinants of the renal actions of atrial natriuretic peptide (ANP): lack of effect on ANP on pressure-induced vasoconstriction. Circ Res 67: 1-10, 1990.
Hayashi K, Epstein M, Loutzenhiser R, Forster H. Impaired myogenic responsiveness of the renal afferent arteriole in streptozotocin-induced diabetic rats: role of eicosanoid derangements. J Am Soc Nephrol 2: 1578-1586, 1992.
Inscho EW, Cook AK, Imig JD, Vial C, Evans RJ. Physiological role for P2X1 receptors in renal microvascular autoregulatory behavior. J Clin Invest 112: 1895-1905, 2003.
Just A, Arendshorst WJ. Nitric oxide blunts myogenic autoregulation in rat renal but not skeletal muscle circulation via tubuloglomerular feedback. J Physiol 569: 959-974, 2005.
Kaplan MR, Plotkin MD, Brown D, Hebert SC, Delpire E. Expression of the mouse Na-K-2Cl cotransporter, mBSC2, in the terminal inner medullary collecting duct, the glomerular and extraglomerular mesangium, and the glomerular afferent arteriole. J Clin Invest 98: 723-730, 1996.
Kirton CA, Loutzenhiser R. Alterations in basal PKC activity modulate renal afferent arteriolar myogenic reactivity. Am J Physiol Heart Circ Physiol 275: H467-H475, 1998.
Kovalev IV, Baskakov MB, Anfinogenova YD, Borodin YL, Kilin AA, Minochenko IL, Popov AG, Kapilevich LV, Medvedev MA, Orlov SN. Effect of Na +,K +,2Cl - cotransport inhibitor bumetanide on electrical and contractile activity of smooth muscle cells in guinea pig ureter. Bull Exp Biol Med 136: 145-149, 2003.
Kreye VA, Bauer PK, Villhauer I. Evidence for furosemide-sensitive active chloride transport in vascular smooth muscle. Eur J Pharmacol 73: 91-95, 1981.
Lavallee SL, Iwamoto LM, Claybaugh JR, Dressel MV, Sato AK, Nakamura KT. Furosemide-induced airway relaxation in guinea pigs: relation to Na-K-2Cl cotransporter function. Am J Physiol Lung Cell Mol Physiol 273: L211-L216, 1997.
Leyssac PP, Holstein-Rathlou NH. Effects of various transport inhibitors on oscillating TGF pressure responses in the rat. Pflügers Arch 407: 285-291, 1986.
Loutzenhiser K, Loutzenhiser R. Angiotensin II activates differing Ca influx mechanisms in renal afferent and efferent arterioles. Circ Res 87: 551-557, 2000.
Loutzenhiser R, Bidani A, Chilton L. The renal myogenic response: kinetic attributes and physiologic role. Circ Res 90: 1316-1324, 2002.
Loutzenhiser R, Epstein M, Horton C. Inhibition by diltiazem of pressure-induced afferent vasoconstriction of the isolated perfused rat kidney. Am J Cardiol 59: 72A-75A, 1987.
Loutzenhiser R, Parker M. Hypoxia inhibits myogenic reactivity of renal afferent arterioles by activating ATP-sensitive potassium channels. Circ Res 74: 861-869, 1994.
Meyer JW, Flagella M, Sutliff RL, Lorenz JN, Nieman ML, Weber CS, Paul RJ, Shull GE. Decreased blood pressure and vascular smooth muscle tone in mice lacking basolateral Na + -K + -2Cl - cotransporter. Am J Physiol Heart Circ Physiol 283: H1846-H1855, 2002.
Moore LC, Casellas D. Tubuloglomerular feedback dependence of autoregulation in rat juxtamedullary afferent arterioles. Kidney Int 37: 1402-1408, 1990.
Mozhayeva MG, Bagrov YaYu. The inhibitory effects of frusemide on Ca 2+ influx pathways associated with oxytocin-induced contractions of rat myometrium. Gen Physiol Biophys 14: 427-436, 1995.
Nilius B, Droogmans G. Ion channels and their functional role in vascular endothelium. Physiol Rev 81: 1415-1459, 2001.
Pires SL, Julien C, Chapuis B, Sassard J, Barres C. Spontaneous renal blood flow autoregulation curves in conscious sinoaortic baroreceptor-denervated rats. Am J Physiol Renal Physiol 282: F51-F58, 2002.
Ramsay LE, McInnes GT, Hettiarachchi J, Shelton J, Scott P. Bumetanide and frusemide: a comparison of dose-response curves in healthy men. Br J Clin Pharmacol 5: 243-247, 1978.
Reslerova M, Loutzenhiser R. Divergent mechanisms of ATP-sensitive K + channel-induced vasodilation in renal afferent and efferent arterioles. Evidence of L-type Ca 2+ channel-dependent and -independent actions of pinacidil. Circ Res 77: 1114-1120, 1995.
Russell JM. Sodium-potassium-chloride cotransport. Physiol Rev 80: 211-276, 2000.
Shi Y, Wang X, Chon KH, Cupples WA. Tubuloglomerular feedback-dependent modulation of renal myogenic autoregulation by nitric oxide. Am J Physiol Regul Integr Comp Physiol 290: R982-R991, 2006.
Schnermann J, Levine DZ. Paracrine factors in tubuloglomerular feedback: adenosine, ATP, and nitric oxide. Annu Rev Physiol 65: 501-529, 2003.
Stanke F, Devillier P, Breant D, Chavanon O, Sessa C, Bricca G, Bessard G. Frusemide inhibits angiotensin II-induced contraction on human vascular smooth muscle. Br J Clin Pharmacol 46: 571-575, 1998.
Stanke-Labesque F, Cracowski JL, Bedouch P, Chavanon O, Magne JL, Bessard G, Devillier P. Furosemide inhibits thromboxane A2-induced contraction in isolated human internal mammary artery and saphenous vein. J Cardiovasc Pharmacol 35: 531-537, 2000.
Takenaka T, Forster H, De Micheli A, Epstein M. Impaired myogenic responsiveness of renal microvessels in Dahl salt-sensitive rats. Circ Res 71: 471-480, 1992.
Takenaka T, Harrison-Bernard LM, Inscho EW, Carmines PK, Navar LG. Autoregulation of afferent arteriolar blood flow in juxtamedullary nephrons. Am J Physiol Renal Fluid Electrolyte Physiol 267: F879-F887, 1994.
Tian R, Aalkjaer C, Andreasen F. Mechanisms behind the relaxing effect of furosemide on the isolated rabbit ear artery. Pharmacol Toxicol 67: 406-410, 1990.
Trottier G, Hollenberg M, Wang X, Gui Y, Loutzenhiser K, Loutzenhiser R. PAR-2 elicits afferent arteriolar vasodilation by NO-dependent and NO-independent actions. Am J Physiol Renal Physiol 282: F891-F897, 2002.
Trottier G, Triggle CR, O'Neill SK, Loutzenhiser R. Cyclic GMP-dependent and cyclic GMP-independent actions of nitric oxide on the renal afferent arteriole. Br J Pharmacol 125: 563-569, 1998.
Van Rodijnen WF, van Lambalgen TA, Tangelder GJ, van Dokkum RP, Provoost AP, ter Wee PM. Reduced reactivity of renal microvessels to pressure and angiotensin II in fawn-hooded rats. Hypertension 39: 111-115, 2002.
Walker M 3rd, Harrison-Bernard LM, Cook AK, Navar LG. Dynamic interaction between myogenic and TGF mechanisms in afferent arteriolar blood flow autoregulation. Am J Physiol Renal Physiol 279: F858-F865, 2000.
Wall SM, Knepper MA, Hassell KA, Fischer MP, Shodeinde A, Shin W, Pham TD, Meyer JW, Lorenz JN, Beierwaltes WH, Dietz JR, Shull GE, Kim YH. Hypotension in NKCC1 null mice: role of the kidneys. Am J Physiol Renal Physiol 290: F409-F416, 2006.
Wang X, Hollenberg MD, Loutzenhiser R. Redundant signaling mechanisms contribute to the vasodilatory response of the afferent arteriole to proteinase-activated receptor-2. Am J Physiol Renal Physiol 288: F65-F75, 2005.
Wang X, Loutzenhiser R. Determinants of renal microvascular response to ACh: afferent and efferent arteriolar actions of EDHF. Am J Physiol Renal Physiol 282: F124-F132, 2002.
Wang X, Trottier G, Loutzenhiser R. Determinants of renal afferent arteriolar actions of bradykinin: evidence that multiple pathways mediate responses attributed to EDHF. Am J Physiol Renal Physiol 285: F540-F549, 2003.
作者单位：Smooth Muscle Research Group, Department of Pharmacology and Therapeutics, University of Calgary Faculty of Medicine, Calgary, Alberta, Canada