| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
,1
,,2
* Department of Pediatrics, NYU School of Medicine, New York, New York; USA;
Department of Pharmacology, NYU School of Medicine, New York, New York, USA;
Department of Physiology and Neuroscience, NYU School of Medicine, New York, New York, USA;
Department of Medicine, Boston University School of Medicine, Boston, Massachusetts, USA;
|| American Cardiovascular Research Institute, Atlanta, Georgia, USA; and
¶ Department of Medicine, NYU School of Medicine, New York, New York, USA
2Correspondence: Pediatric Cardiology, NYU School of Medicine, 560 First Ave., TCH-521, New York, NY 10016, USA. E-mail: william.coetzee{at}med.nyu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: potassium channels • coronary vasculature • transgenic mouse
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In the coronary vasculature, KATP channels have been implicated in the autoregulation of blood flow, as well as regulating coronary tone in response to pathophysiological insults such as hypoxia and ischemia (2)
. The role of KATP channels in regulating coronary vessel tone is relatively well characterized in smooth muscle, where their opening results in hyperpolarization of the smooth muscle, which in turn leads to vasodilation. Based mainly on pharmacological evidence, there also appears to be a role for endothelial KATP channels in regulating coronary blood flow. For example, vasodilation of coronary arteries by adenosine and hyperosmolarity may be mediated in part by activating endothelial KATP channels (3
, 4)
. Endothelial KATP channels may also contribute to shear stress-induced endothelial release of the vasodilator nitric oxide in rabbit aorta (5)
.
Functional KATP channels can be reconstituted by coexpression of pore-forming Kir6.x (Kir6.1 or Kir6.2) subunits in combination with regulatory SUR (SUR1 or SUR2) subunits (6)
. The Kir6.x subunits determine the biophysical properties and nucleotide sensitivities of KATP channels, whereas SUR subunits fine-tune the channel’s nucleotide sensitivity and confer unique pharmacological specificities. The different properties of KATP channels found in different cells (7)
originate in part by specific combinations of the various Kir6.x and SUR subunits. For example, the cardiac sarcolemmal and the pancreatic ß-cell KATP channel have the same pore-forming subunit (Kir6.2), but differ in their use of the regulatory subunit (respectively SUR2A and SUR1). Smooth muscle KATP channels, in contrast, are composed of Kir6.1 and SUR2B subunits (6)
. We demonstrated that the coronary endothelium contains heteromeric Kir6.1/Kir6.2 pore-forming subunits in combination with regulatory SUR2B subunits (8)
. Although endothelial KATP channels are increasingly well characterized at the molecular level, their physiological role in the coronary vasculature is ill-defined. The creation of Kir6.1 and SUR2 null mice has been invaluable in our understanding of the roles of KATP channels in the vasculature (9
, 10)
. However, gene knockout approaches result in a lack of the target protein in all somatic cells; therefore, it can be difficult to dissociate the role of a channel in a particular cellular compartment (e.g., endothelium) from effects in other somatic cells (e.g., the vascular smooth muscle). Both Kir6.1 and SUR2 null mice have impaired coronary function, but it is not clear whether these defects originate predominantly in the vascular smooth muscle or the endothelium. In support, when SUR2B is transgenically reintroduced into the smooth muscle of SUR2 null mice, vascular problems persist, arguing for a nonsmooth muscle-dependent mechanism (11)
. The defect probably does not arise in the cardiac myocyte, since Kir6.2 null mice (12)
or mice expressing dominant negative Kir6.x subunits in their hearts (13)
do not appear to exhibit vascular complications. We examined the specific role of endothelial KATP channels in coronary function by generating transgenic mice that express Kir6.1 dominant negative subunits only in the endothelium. Our data demonstrate a unique role for endothelial KATP channels in mediating the secretory release of the vasoconstrictor endothelin-1 (ET-1) through an unknown mechanism, thereby regulating coronary function.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Production of the CX1-LEL-Kir6.1-AAA transgenic construct
The transgenic construct consisted of a pore mutant rat Kir6.1 (Gly-Phe-Gly replaced with Ala-Ala-Ala; ref. 13
) with the polyadenylation sequence of the human growth hormone. The construct was subcloned into the pBS-CX1-LEL vector (a gift from Dr Gary Owens, University of Virginia), which contains the chicken ß-actin (CX1) promoter, followed by the eGFP coding region (including a stop codon), flanked by two loxP sites. The Kir6.1-AAA-poly-A cDNA was inserted downstream of the second loxP site to create the pBS-CX1-eGFP-Kir6.1-AAA transgenic construct.
Generation of Tg[CX1-eGFP-Kir6.1-AAA] transgenic mice
The transgene was enzymatically excised, purified, and microinjected into the pronuclei of fertilized FVB/N zygotes (Transgenic/ES cell Core Facility, Skirball Institute, NYU School of Medicine). Surviving embryos were reimplanted into pseudopregnant FVB/N foster mothers. Putative founders were screened for the presence of the transgene by using the polymerase chain reaction (PCR). Primers used were (1331F) 5'-ACGTAAACGGCCACAAGTTC-3' and (1318R) 5'-CTTGTACAGCTCGTCCATGC-3'; (
650 bp). Mice (or their tail tips after snipping) could also be genotyped by their eGFP fluorescence at the excitation wavelength of 488 nm (Fig. 1
). The resulting transgenic mice are referred to as Tg[CX1-eGFP-Kir6.1-AAA] (or GFP+/Cre– genotype). Three lines of transgenic mice were produced. None of the mice had any overt phenotype. They had normal fertility and offspring gender ratios. A single line was selected for further studies.
|
Generation of Tg[
MHC-Kir6.1-AAA] mice
We crossed the Tg[CX1-eGFP-Kir6.1-AAA] mice with transgenic mice that express Cre recombinase under the control of the -MHC promoter (14)
(obtained from Dr Michael Schneider, Baylor College of Medicine, Houston, TX, USA). Genomic DNA from tail clippings was genotyped by PCR for the presence of Kir6.1-AAA (as described above) and/or the presence of Cre [primers used were (1551F) 5'-GCGGTCTGGCAGTAAAAACTATC-3' and (1551R) 5'-GTGAAACAGCATTGCTGTCACTT-3'; (100 bp)].
Generation of Tg[Tek-Kir6.1-AAA] mice
We crossed the Tg[CX1-eGFP-Kir6.1-AAA] mice with B6.Cg-Tg (Tek-cre)12Flv/J mice (Jackson Laboratories, West Grove, PA, USA; 004128), which express Cre recombinase specifically in the endothelium, under the control of the Tek (or Tie-2) promoter. Genotyping was performed as described above.
In vivo electrocardiographic studies
Electrocardiographic parameters were examined as described (13)
. Ergonovine (Sigma, St. Louis, MO, USA; 400 mg/ml in DMSO and diluted with H2O) was applied intravenously at the following cumulative doses (15, 30, 50, and 80 mg/kg applied intraventricular; 10 min each). At the conclusion of the experiment, Trypan blue dye (Life Technologies, Inc., Carlsbad, CA, USA) was infused into the isolated heart by aortic cannulation and the coronary vasculature was visualized for the presence of vasoconstrictions.
RT-PCR
RNA was prepared from mouse hearts and brains and subjected to reverse transcription using oligo-dT primers. Multiplex PCR was performed using the following Kir6.1 primers: (1314F) 5'-GAAAGGCATCACGGAGAAGA-3' and (1315R) 5'-GATGGAAGGTTTCTCGTCCA-3'; (774 bp). Amplification of GAPDH served as a control and the primers used were (1295F) 5'-ACCACAGTCCATGCCATCAC-3' and (1295R) 5'-TCCACCACCCTGTTGCTGTA-3'; (432 bp). PCR was performed using 35 cycles at an annealing temperature of 56°C.
Measurement of coronary perfusion pressure
Adult Sprague Dawley rats or mice were intraperitoneally injected with heparin (0.5 U/g) and Nembutal (120 mg/kg). After deep anesthesia, hearts were excised and rinsed in ice-cold Krebs Henseleit buffer containing (in mmol/L): 118 NaCl, 25 NaHCO3, 1.2 KH2PO4, 4.7 KCl, 1.2 MgSO4, 1.8 CaCl2, and 11 glucose (pH 7.4 when oxygenated with 95% O2/5% CO2). The aorta was rapidly (within 2 min) cannulated and the heart was perfused in Langendorff mode with Krebs Henseleit buffer (at 37°C) using a peristaltic pump (Minipuls 3, Gilson Inc., Middleton, WI, USA). The perfusion rate was 10 ml/min for rat hearts and 2.5 ml/min for mouse hearts; flow was monitored by Doppler (T206 flow meter, Transonic Systems Inc., Ithaca, NY, USA). The coronary perfusion pressure was monitored at a side arm of the aortic cannula (Digi-Med Model 400 Blood Pressure Analyzer, Micro-Med, Louisville, KY, USA). Hearts were electrically paced (220 and 450 bpm, respectively, for rats and mice; S-900 Stimulator and S-910 Stimulus Isolation Unit, Cornerstone by Dagan, Minneapolis, MN, USA). Acquired data were analyzed using OriginPro 7 software (OriginLab Corporation, Northampton, MA, USA) and Excel (Microsoft).
Measurement of endothelin-1 release in the coronary effluent
ET-1 levels were measured in the coronary effluent of mouse or rat hearts by radioimmunoassay (RIA; RAS 6901, Peninsula Labs, San Carlos, CA, USA). The coronary effluent from each heart was spiked with bovine serum albumin (0.02%) before being frozen (–20°C or –80°C) for later analysis. Standards were prepared by adding ET-1 peptide (0.1 to 32 pg; Peninsula Labs) to perfusion buffer. The ET-1 was extracted from the coronary effluent by slow elution (1 drop/10 s) through preconditioned chromatography columns (C18 SEP, Peninsula Labs). After washing with 1% trifluoroacetic acid (twice), ET-1 was eluted with 3 ml buffer B (60% acetonitrile in 1% trifluoroacetic acid). Eluates were frozen, then lyophilized (RC 10.22, Jouan Inc., Winchester, VA, USA). ET-1 measurements were performed using an RIA kit (RAS 6901, Peninsula Labs), following the manufacturer’s guidelines. Measurements were made in duplicate and data were normalized to the heart wet weight and coronary flow rate.
Measurement of endothelin-1 release from human coronary artery endothelial cells
Primary human coronary artery endothelial cells (HCAEC; Cambrex Bio Science Walkersville Inc., Walkersville, MD, USA) were used within two and three passages from purchase, as described (8)
. ET-1 measurements were made in the culturing medium by ELISA (TiterZyme Enzyme Immunometric Assay kit; Assay Designs Inc., Ann Arbor, MI, USA).
Immunostaining
Primary antibodies used were Cy3-conjugated mouse monoclonal -actin smooth muscle antibody (1A4; Sigma), rat monoclonal antibody anti-ICAM-2 (3C4; BD Biosciences PharMingen, San Diego, CA, USA), rabbit polyclonal anti-ET-1 antibody (Peninsula Labs), the CAF-1 anti-Kir6.1 antibody (15)
, and G-16 anti-Kir6.2 antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Secondary antibodies used were Cy5-conjugated F (ab’)2 fragment donkey anti-rat IgG, Cy3-conjugated donkey anti-rabbit IgG, Cy5-conjugated donkey anti-goat, and Cy2-conjugated donkey anti-chicken IgY (Jackson ImmunoResearch Laboratories, Westgrove, PA, USA). Negative controls included experiments using only primary or secondary antibodies. The remaining procedures were essentially the same as described before (15)
.
Reagents
Ergonovine, bradykinin, pinacidil, glibenclamide, tolbutamide, and diazoxide were from Sigma. Cycloheximide was from Calbiochem (San Diego, CA, USA) and bosentan was a generous gift from Actelion Pharmaceuticals Ltd. (Allschwil, Switzerland). In some experiments, mice were pretreated for 3 days with daily intraperitoneal injections of bosentan (30 mg/kg).
Data and statistical analysis
Data are expressed as mean ± SE. Significance between groups was determined using Student’s paired or unpaired t tests, the Mann-Whitney Rank Sum Test, 2-way repeated measures ANOVA, and 2-way ANOVA. Post hoc multiple comparisons were performed with the Student-Newman-Keuls method (SigmaStat, Jandel Corporation). Differences between groups were considered significant at confidence levels >95% (P<0.05).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
,16
,17)
. Using the Cre-Lox system, we generated a generic and versatile mouse model in which the pore mutant Kir6.1-AAA subunit can be expressed specifically in any tissue for which a tissue-specific promoter is available. The Kir6.1-AAA cDNA was subcloned downstream of eGFP (followed by a stop codon), which in turn is flanked by loxP sites (Fig. 1)
. The transgene was placed under the control of the ubiquitous chicken ß-actin (CX1) promoter, which drives expression in all mammalian somatic cells (18)
. The Tg[CX1-eGFP-Kir6.1-AAA] mice, with a GFP+/Cre– genotype, expressed eGFP in all somatic cells (Fig. 1)
. Apart from their green fluorescence, these mice were phenotypically normal in every respect.
Low leakiness of the transgenic construct
For this transgenic mouse model to be useful, the transgenic construct should have low leakiness. In heterologous expression systems, Kir6.1-AAA was not expressed in the absence of Cre recombinase (data not shown). In Tg[CX1-eGFP-Kir6.1-AAA] mice, the low leakiness of the transgene is demonstrated by three independent methods. First, cardiac myocytes from these mice express normal DNP-activated KATP channel currents (data not shown), in contrast to the absence of KATP channel currents when Kir6.1-AAA expression is placed under the control of the -MHC promoter (13)
. Second, Kir6.1 mRNA levels are similar in the hearts of wild-type and Tg[CX1-eGFP-Kir6.1-AAA] mice (Fig. 1)
. Third, by Western blot, Kir6.1 protein is expressed at the same level in the hearts of wild-type mice and transgenics (data not shown).
Cre-mediated excision of eGFP leads to tissue-specific activation of Kir6.1-AAA expression
In heterologous expression systems, transfection with Cre recombinase excises eGFP and activates expression of Kir6.1-AAA (data not shown). To address the question of whether this also occurs in vivo, we crossed Tg[CX1-eGFP-Kir6.1-AAA] mice with -MHC-Cre transgenic mice, which express Cre recombinase in the heart (14)
. The resulting offspring lacked eGFP expression in their hearts but not other somatic tissue (Fig. 1)
. The residual fluorescence in the heart may be due to the fact that the -MHC promoter does not drive expression in all cardiac myocytes or to the possible presence of eGFP-containing nonmuscle cells. These data are indicative of successful Cre-mediated eGFP excision from the transgene in a tissue-specific manner.
The next questions to address were whether Kir6.1-AAA subunits are expressed after Cre-mediated eGFP excision and whether this occurs in a tissue-specific manner. We examined Kir6.1 mRNA expression levels by RT-PCR in hearts and brains of the offspring (Fig. 1)
. GFP+/Cre– mice had similar Kir6.1 mRNA levels compared with their wild-type littermates (GFP–/Cre–). In contrast, GFP+/Cre+ mice had significantly elevated cardiac Kir6.1 mRNA levels (Fig. 1)
. Brains from these animals had normal Kir6.1 mRNA levels. Finally, isolated ventricular myocytes from the hearts of GFP+/Cre+ mice lacked functional KATP channels (data not shown). Thus, the Tg[CX1-eGFP-Kir6.1-AAA] mice have a low basal leakiness of the transgene.
Generation of mice expressing dominant negative Kir6.1 subunits in the endothelium
To target KATP channels in the endothelium, we crossed Tg[CX1-eGFP-Kir6.1-AAA] mice with transgenic mice that express Cre recombinase under the control of the TEK receptor tyrosine kinase (also called Tie-2) promoter (Fig. 1)
. Tek is expressed almost exclusively in endothelial cells in mice, rats, and humans (19)
. We refer to the resulting offspring (GFP+/Cre+ genotype) as Tg[Tek-Kir6.1-AAA] mice. After several generations, we have not observed any gross morphological or behavioral phenotypes.
Phenotype of Tg[Tek-Kir6.1-AAA] mice
No premature mortality was observed. Gross body weight and wet heart weight are similar between these mice and their control (GFP+/Cre–) littermates (supplemental Table 1
). As expected, eGFP expression was absent in the coronary endothelium of the Tg[Tek-Kir6.1-AAA] mice, as evident from the lack of colocalization with the constitutively expressed endothelial cell surface glycoprotein, ICAM-2 (Fig. 2
). We examined whether they had increased susceptibility to coronary vasospasm, as reported for Kir6.1 or SUR2 null mice (9
,10)
, by examining the ECG and visualizing the coronary vasculature after treating mice with ergonovine. Increasing doses of ergonovine caused QRS widening, loss of P-wave activity, and coronary vasospasm; however, differences between the groups were not evident (Fig. 3
).
|
|
|
Impaired coronary function in Tg[Tek-Kir6.1-AAA] mice
We measured the coronary perfusion pressure (CPP) of isolated hearts at a constant flow rate as an index of coronary function. The CPP was significantly higher in Tg[Tek-Kir6.1-AAA] hearts compared with control littermates (Table 1
). Adenosine (7.5 µM) or levcromakalim (30 µM) caused a significant reduction in the CPP in both groups. As expected, glibenclamide caused a significant coronary vasoconstriction in control hearts. In contrast, glibenclamide had no effect on CPP in Tg[Tek-Kir6.1-AAA] hearts.
Elevated ET-1 release from the hearts of Tg[Tek-Kir6.1-AAA] mice
We perfused Tg[Tek-Kir6.1-AAA] hearts with bradykinin (10 µM), which reduced the CPP from 181 ± 8.9 to 167 ± 9.8 mmHg within 3 min (n=6, P<0.05). This observation demonstrates that the inducible nitric oxide pathway is intact in the coronary endothelium from Tg mice. We next investigated whether the increased CPP was due to increased levels of the vasoconstrictor, ET-1. Indeed, we found that ET-1 release (measured by RIA) was elevated in the coronary effluent of isolated Tg[Tek-Kir6.1-AAA] hearts (2.1±0.27 pg · min–1.gww–1, n=6 vs. 1.3±0.19 pg · min–1.gww–1 in control GFP+/Cre– littermates, n=6; P<0.05).
Normalization of the elevated CPP of Tg[Tek-Kir6.1-AAA] hearts by ET-1 antagonism
To examine the relative contribution of ET-1 to the increased CPP of Tg[Tek-Kir6.1-AAA] hearts, mice were pretreated with the ET-1 receptor antagonist, bosentan. In the absence of the drug, the CPP of isolated hearts was substantially higher in Tg[Tek-Kir6.1-AAA] mice (Fig. 4
). In contrast, after bosentan treatment there was no difference in the CPP between the two groups. Over time there was a slight elevation in CPP in both groups. Since this was not observed in the absence of the drug, this is likely due to a nonspecific action of bosentan. We conclude that blockade of ET-1 receptors was sufficient to restore the elevated CPP of Tg[Tek-Kir6.1-AAA] hearts to normal levels.
|
Pharmacological KATP channel blockade increases ET-1 release from isolated hearts
We measured ET-1 release from isolated rat hearts before and after application of KATP channel modulators. The KATP channel blocker tolbutamide (300 µM) caused a significant increase in ET-1 release, whereas the KATP channel opener diazoxide (300 µM) had no effect (Fig. 5
). These data are consistent with the concept that KATP channel closure stimulates ET-1 release from the coronary endothelium. Glibenclamide (10 µM) similarly increased ET-1 release from isolated mouse hearts (as assessed by RIA: from 1.4±0.15 to 2.7±0.35 pg · min–1.gww–1; n=13, P<0.05). In Tg[Tek-Kir6.1-AAA] hearts, the effect of glibenclamide on ET-1 release was not statistically significant (from 2.1±0.27 to 3.2±0.59 pg · min–1.gww–1 in response to glibenclamide application; n=6).
|
Pharmacological KATP channel blockade increases ET-1 release from human coronary artery endothelial cells
We next examined the hypothesis that KATP channel-mediated ET-1 release occurs as a result of an enhanced secretory release from the endothelium. As a model system, we used primary human coronary artery endothelial cells and measured ET-1 in the supernatant with ELISA. As expected, these cells constitutively secreted ET-1. Both glibenclamide (10 µM) and tolbutamide (300 µM) significantly elevated the rate of ET-1 release (Fig. 6
). The half-maximal stimulation of ET-1 release by tolbutamide was 178 µM (slope factor of 1.7), which is consistent with the pharmacological profile and tolbutamide sensitivity of SUR2B-containing KATP channels (20)
. As expected for a KATP channel-mediated response, diazoxide (300 µM) antagonized tolbutamide-induced ET-1 release (Fig. 6)
. By themselves, pinacidil (300 µM) or diazoxide (300 µM) had no effect on ET-1 release (data not shown). Blocking protein synthesis with cycloheximide (10 µM for 24 h) caused a reduction in basal ET-1 release, suggesting that constitutive ET-1 release occurs partly through de novo synthesis (Fig. 6)
. In the presence of cycloheximide, tolbutamide still elevated ET-1 release, consistent with the conclusion that KATP channel blockade stimulates ET-1 release from storage vesicles through a secretory pathway. Indeed, ET-1-containing granules (most likely the Weibel-Palade bodies) are clearly visible when performing immunocytochemistry with ET-1 antibodies (Fig. 7
). Both Kir6.1 and Kir6.2 subunits appear to be expressed in a similar granular pattern (Fig. 7)
, which raises the possibility that KATP channels may be expressed predominantly in subcellular organelles to control ET-1 release independent of effects on endothelial membrane potential.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
KATP channels and regulated exocytosis
It is well established that ion channels participate in regulated exocytosis. For example, by mediating the Ca2+ influx that triggers exocytotic fusion, Ca2+ channels play a central role in a wide range of secretory processes, including the secretory release of neurotransmitters and hormones such as adrenaline and insulin (21)
. An important role for KATP channels in regulated exocytosis is best exemplified by their involvement in triggering and modulating the secretory release of insulin from pancreatic ß-cells (22)
. Recent evidence also implicates KATP channels in the release of glucagon from isolated pancreatic -cells (23)
. KATP channel blockade also affects chorionic gonadotropin-stimulated progesterone release from human ovarian follicular and luteal cells (24)
. Although the release mechanism remains to be clarified, it is possible that exocytotic release from intracellular granules may be responsible (25)
. In the pancreatic ß-cell, glucose-mediated closure of surface-located KATP channels causes membrane depolarization, which in turn leads to spontaneous electrical activity of ß-cells, Ca2+ entry, and regulated exocytosis. Activation of this KATP channel-dependent pathway results in exocytosis of an immediately releasable pool of insulin-containing granules that is responsible for the first phase of glucose-stimulated insulin release. Although KATP channels clearly regulate the first (trigger) phase of release, they may also be involved in the second phase, as evidenced by its attenuation in SUR1 knockout animals (26
; but see ref. 27
). An increasing number of studies suggest that KATP channels may have functions not strictly related to their classic role of merely controlling the membrane potential. For example, some studies have described KATP channel subunits (SUR1) as associating with proteins involved in exocytotic release vesicle dynamics and in regulating granule priming (28)
, including cAMP-GEFs (Epacs) (29
, 30)
and syntaxin 1A (31)
. Thus, there may be a unique role for KATP channels in mediating regulated exocytosis that may be unrelated to controlling excitability, which may also be present in some nonexcitable cells such as endothelial cells.
KATP channels and exocytosis in endothelial cells
Similar to the release of von Willebrand factor and tissue-type plasminogen activator, there are two secretory pathways responsible for ET-1 release from endothelial cells (32)
. Since Tg[Tek-Kir6.1-AAA] hearts have an elevated basal CPP, there may be an involvement of KATP channels in the constitutive pathway, which is implicated in the maintenance of basal vascular tone (32)
. The constitutive pathway involves intracellular trafficking events from the endoplasmic reticulum and Golgi complex, where proteins are continuously shuttled from the trans-Golgi network to the cell surface in secretory vesicles, as well as secretory events of vesicle docking with the plasma membrane. It is not clear how KATP channels could be involved in the constitutive secretory pathway, but reports (in other cell types) of an association of KATP channel subunits with several proteins involved in granule exocytosis, such as syntaxin-1A, Rim2, and Piccolo (29
, 31)
, may eventually provide insight into the subcellular mechanisms involved. Secretion of ET-1 also occurs via a regulated pathway that is under the control of physiological or pathophysiological stimuli (e.g., mechanical stimuli, shear stress, histamine, or thrombin) in which secretion is too rapid to involve de novo protein synthesis (32)
. We found acute application of sulfonylureas to cause rapid ET-1 release from isolated hearts as well as isolated human coronary artery endothelial cells. Furthermore, this effect was not prevented by cycloheximide, a blocker of de novo protein synthesis, suggesting that a component of ET-1 release occurs from immediately releasable storage vesicles (most likely Weibel-Palade bodies). Therefore, it is possible that KATP channels may also be involved in this step.
The mechanism by which KATP channels mediate endothelin-1 release?
Endothelial cells are not excitable and do not depend on voltage-activated Ca2+ channels for Ca2+ entry to occur (33)
. Furthermore, in contrast to pancreatic ß-cells, where membrane depolarization triggers excitability and Ca2+ entry, membrane hyperpolarization is normally thought to be associated with endothelial cell Ca2+ influx (33)
. Finally, although KATP channels have been described as present in the endothelial plasmalemma (34
, 35)
, it is generally agreed they do not contribute significantly to endothelial membrane potential (8
, 33)
. It is therefore unlikely for plasmalemmal KATP channels to have the same role in triggering exocytosis in endothelial cells as their well-described role in the pancreatic ß-cells. At present, the mechanism by which KATP channels may mediate regulated exocytosis of ET-1 is not clear. Our immunocytochemistry data suggest the possibility that KATP channels are not located primarily on the cell surface membrane, but may be present predominantly in intracellular granular organelles. The identity of these granules remains to be established, but it is possible that they are ET-1-containing granules. It is well established that secretory granules contain a variety of ion channels, including KATP channels (36
37
38)
. Their specific role in these granules remains to be investigated, and nonexcitable endothelial cells may provide an excellent role to study this mode of KATP channel function in regulated exocytosis.
A novel role for endothelial KATP channels in regulating the coronary vascular tone
Smooth muscle plasmalemmal KATP channels regulate the coronary flow in response to metabolic demand (2)
and changes in blood pressure (autoregulation) (39)
. By opening, they hyperpolarize the membrane potential, which diminishes Ca2+ influx and causes smooth muscle relaxation (40)
. Genetic targeting of KATP channels (Kir6.1 null or SUR2 null mice) causes an elevated coronary tone and the occurrence of spontaneous coronary vasospasms (9
, 10)
, which may be the reason for premature death in these animals. Transgenic reintroduction of SUR2 in a SUR2 null background does not normalize the elevated CPP or alleviate spontaneous coronary vasoconstriction (11)
, suggesting the defects may arise from a coronary artery vascular smooth muscle-extrinsic process. Some reports suggest that endothelial KATP channels may also be involved in regulating coronary blood flow (see above). By selectively targeting endothelial cells with a dominant negative KATP channel subunit, we demonstrate here a novel role for these channels in regulating coronary function. The basal coronary perfusion pressure was substantially elevated in these animals. Smooth muscle KATP channels were intact in these animals since adenosine and levcromakalim (substances thought to act by modulating smooth muscle KATP channels; refs. 41
42
43
) resulted in coronary vasodilation in Tg[Tek-Kir6.1-AAA] hearts. Ergonovine did not produce coronary vasospasm. Although we did not examine the mice for the occurrence of spontaneous coronary vasospasm, this is unlikely to be a major factor since these animals did not exhibit the early mortality found with Kir6.1 null or SUR2 null mice. Since coronary vasospasm persists in SUR2 null mice after reintroduction of smooth muscle SUR2 (11)
, it is likely that the origin of vasospasm is extravascular. Overall, our data suggest that smooth muscle and endothelial KATP channels are both involved in regulating coronary tone, but they have different roles: smooth muscle KATP channels regulate excitability whereas endothelial KATP channels may be directly involved in regulated exocytosis of the vasoconstrictor peptide, ET-1. It should be noted that endothelial KATP channels may also be involved in other facets of endothelial function. For example, it has been proposed that endothelial KATP channel opening promotes vasodilation by promoting NO release (44)
. This mechanism may be additive to the effect of KATP channels on ET-1 release and may serve to modulate coronary function.
Sulfonylureas mediate coronary vasoconstriction in part via endothelial KATP channels
Sulfonylureas (tolbutamide and glibenclamide) are used therapeutically as anti-diabetic agents. At higher concentrations, these KATP channel blockers also act on the vasculature to cause vasoconstriction (45)
. Although sulfonylureas clearly act on smooth muscle KATP channels (40)
, it is intriguing that glibenclamide has little effect on the resting tone of isolated porcine coronary arteries (46)
, whereas a pronounced coronary vasoconstriction is normally observed in perfused hearts. This apparent discrepancy might be due to the pronounced role of KATP channels in the resistance vessels and may indicate that the contribution of KATP channels to regulating vascular tone increases as the coronary diameter decreases along the vascular bed (47
48
49
50)
. In endothelium-denuded pulmonary resistance vessels, glibenclamide does not cause vasoconstriction, which led to the suggestion that the compound exerts its effect by acting on endothelial KATP channels (44)
. We found that glibenclamide had no effect on the coronary perfusion pressure in hearts isolated from Tg[Tek-Kir6.1-AAA] mice, whereas KATP channel openers caused vasodilation (most likely by acting on smooth muscle KATP channels). These data suggest that the vasoconstrictive effect of sulfonylureas may in large part be due to its effect on endothelial KATP channels. These data may provide new therapeutic opportunities in designing compounds that do not affect endothelial KATP channels, and hence coronary blood flow.
Implications of our findings
KATP channels in the coronary arteries help to maintain constant blood flow to the myocardial tissue (51
, 52)
and regulate coronary flow alterations in response to metabolic demand (2)
. These effects are thought to be mediated by the KATP channels in coronary smooth muscle. Our data suggest that endothelial KATP channels may contribute partly to these responses by mediating the release of the vasoconstrictor endothelin-1 (53)
. It is conceivable that endothelial KATP channels may also contribute to the well-described vasodilatory role of KATP channels in the coronary vasculature in response to hypoxia and ischemia (2)
.
In conclusion, we have found evidence for a novel role of endothelial KATP channels in regulating the coronary vessel tone. The data provide strong evidence that endothelial KATP channels can affect coronary tone by mediating the release of the vasoconstrictor, endothelin-1. This has important implications for the use of sulfonylureas in treating type II diabetes as well as familial mutations of KATP channel subunits. Since endothelial function predicts future development of coronary artery disease (54)
, our results may also have profound implications for understanding and treating patients with myocardial ischemia and infarction, and possibly even hypertension-induced endothelial dysfunction (55)
.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication November 30, 2006. Accepted for publication January 25, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Farzaneh and A. Tinker Differences in the mechanism of metabolic regulation of ATP-sensitive K+ channels containing Kir6.1 and Kir6.2 subunits Cardiovasc Res, September 1, 2008; 79(4): 621 - 631. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Shorter, N. P. Farjo, S. M. Picksley, and V. A. Randall Human hair follicles contain two forms of ATP-sensitive potassium channels, only one of which is sensitive to minoxidil FASEB J, June 1, 2008; 22(6): 1725 - 1736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Elrod, M. Harrell, T. P. Flagg, S. Gundewar, M. A. Magnuson, C. G. Nichols, W. A. Coetzee, and D. J. Lefer Role of Sulfonylurea Receptor Type 1 Subunits of ATP-Sensitive Potassium Channels in Myocardial Ischemia/Reperfusion Injury Circulation, March 18, 2008; 117(11): 1405 - 1413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Roy, D. Patel, S. Khanna, G. M. Gordillo, S. Biswas, A. Friedman, and C. K. Sen Transcriptome-wide analysis of blood vessels laser captured from human skin and chronic wound-edge tissue PNAS, September 4, 2007; 104(36): 14472 - 14477. [Abstract] [Full Text] [PDF]
|
|
| ||||||||
