K_V1 channel expression and vasomotor function in human coronary resistance arteries

Abstract

Objectives

K_V channels are important regulators of vascular tone, but the identity of specific K_V channels involved and their regulation in disease remain less well understood. We determined the expression of K_V1 channel subunits and their role in cAMP-mediated dilation in coronary resistance arteries from subjects with and without coronary artery disease (CAD).

Methods

Human coronary arteries and arterioles (HCA) from non-CAD and CAD patients were assessed for mRNA and protein expression of K_V1 channel subunits with molecular techniques and for vasodilator response with isolated arterial myography.

Results

Assays of mRNA transcripts, membrane protein expression, and vascular cell-specific localization revealed abundant expression of K_V1.5 in vascular smooth muscle cells of non-CAD HCAs. Isoproterenol and forskolin, two distinct cAMP-mediated vasodilators, induced potent dilation of non-CAD arterioles, which was inhibited by both the general K_V blocker 4-AP and the selective K_V1.5 blocker DPO-1. The cAMP-mediated dilation was reduced in CAD and was accompanied by a loss of or reduced contribution of 4-AP-sensitive K_V channels.

Conclusions

K_V1.5, as a major 4-AP-sensitive K_V1 channel expressed in coronary VSMCs, mediates cAMP-mediated dilation in non-CAD arterioles. The cAMP-mediated dilation is reduced in CAD coronary arterioles, which is associated with impaired 4-AP-sensitive K_V channel function.

1 INTRODUCTION

Potassium (K⁺) channels in the vascular smooth muscle cell (VSMC) of resistance arteries and arterioles play a central role in the regulation of vascular tone.^1–4 Current evidence indicates that VSMCs express at least four distinct classes of K⁺ channels: voltage-gated (K_V), calcium-activated (K_Ca), inwardly rectifying (K_ir) including ATP-sensitive (K_ATP), and two pore domain (K_2P) K⁺ channels.^1,5 Although the relative importance of each class may vary depending on the species and vascular beds, accumulating studies support the role for K_V, particularly shaker-family K_V (K_V1) channels in the regulation of the resting vascular tone and vasodilation of the microvasculature in response to various endothelial, neurohumoral, and metabolic factors.^6–9 For instance, H₂O₂, an endothelium-derived hyperpolarization factor^10–12 as well as a metabolic dilator released from beating cardiac myocytes^13,14 activates K_V channels to induce dilation in the canine, rat, and porcine coronary circulation.^7,15,16 Similarly, stimulation of β₂-adrenoceptors (e.g., with isoproterenol), an important mechanism of coronary blood flow increase during exercise,¹⁷ activates 4-AP sensitive K_V channel currents in rabbit portal vein and rat coronary artery.^18,19 Alterations in K_V channel function have been found in animal models of hypertension, metabolic syndrome, and diabetes.^4,6,20–22. Although K_V channels have been studied in various animal models, the functional roles of these channels in the human microvasculature are less well understood.²³

Recently, our laboratory reported that K_V1.5 is the major type of K_V1 channels expressed and active in VSMCs of human adipose arterioles.²⁴ We also found that hydrogen peroxide (H₂O₂) induces potent smooth muscle-dependent vasodilation that is largely mediated by K_V1.5 and large-conductance Ca²⁺-activated K⁺ (BK_Ca) channels in human adipose arterioles from subjects without coronary artery disease (non-CAD).²⁴ This H₂O₂-induced dilation is reduced in adipose arterioles from subjects with coronary artery disease (CAD) and is accompanied by a loss of K_V1.5- but not BK_Ca-dependent dilation.²⁴ The expression of specific K_V channel subunit gene products and proteins varies among species, vascular beds, or vessel size.²⁵ In addition, there is evidence that certain vascular beds such as the coronary circulation may adapt to limit the deleterious effects on vascular function during disease.^26,27 It remains unknown whether results obtained in adipose arterioles regarding the expression and function of K_V1 channels, as well as their potential alterations in disease, are also applicable to other vascular beds of humans. In the present study, we identified major K_V1 channel subtypes in human coronary arteries/arterioles (HCA) using molecular and immunohistochemical approaches. We also examined the functional role of K_V1 channels in cAMP-mediated vasodilation induced by the β₂-adrenergic receptor agonist isoproterenol and the adenylate cyclase activator forskolin in isolated human coronary arterioles from non-CAD and CAD subjects.

2 MATERIALS AND METHODS

2.1 Tissue acquisition

Fresh human atrial tissues were obtained as discarded surgical specimens (n=21) from patients undergoing cardiopulmonary by-pass surgeries. Unused whole hearts (n=22) were acquired from the Wisconsin Organ Donor Network. De-identified patient demographic data were collected using the Generic Clinical Research Database at the Medical College of Wisconsin. All procedures were approved by the Institutional Review Board of the Medical College of Wisconsin/Froedtert Hospital. Patient demographic information is summarized in Table 1.

Table 1.

Patient demographics (n=43)

	Atria (n=21)		Donor heart (n=22)
	non-CAD (n=7)	CAD (n=14)	non-CAD (n=13)	CAD (n=9)
Gender, Male/Female	4(57%)/3(43%)	6(43%)/8(57%)	5(38%)/8(62%)	6(67%)/3(33%)
Age, yr (mean ± SEM)	62±5	63±3	43±5	60±3
Body mass index (mean ± SEM)	32±3	27±2†	27±2	31±2
Underlying diseases/Risk factors
Coronary artery disease (CAD)	0	14(100%)	0	9(100%)
Hypertension	4(57%)	9(64%)	3(23%)	6(67%)
Hyperlipidemia	3(43%)	7(50%)	1(8%)	2(33%)
Diabetes mellitus	0	0	0	1(11%)
Atrial fibrillation	0	4(29%)	1(8%)	0
Congestive heart failure	0	3(21%)	0	0
Myocardial infarction	0	3(21%)	0	4(44%)
None of the above	0	0	8(62%)	0

^†Total of 1 patient with no information on height and weight.

2.2 Vascular Reactivity

Human coronary arterioles (internal diameter, 100 to 200 μm) were dissected from the atrial tissue and mounted in a wire myograph (Danish Myo Technology A/S) for isometric tension recording (Figs. 4–5) as previously described.²⁸ In some experiments (Fig. 6), human coronary arterioles (50 to 150 μm) were cannulated and pressurized (60 mmHg) with two glass micropipettes, and the internal diameter of arterioles was measured with a video system (Boeckeler VIA-100).²⁹ Arterioles were preconstricted with endothelin-1 to approximately 30–50% of the tension developed during exposure to 80 mM high K⁺ in physiological saline solution (wire myograph) or the baseline internal diameter (cannulated preparation). Relaxation responses to cumulative concentrations of the mixed β₂-adrenergic receptor agonist isoproterenol (10⁻⁹ to 10⁻⁵ M), the adenylate cyclase activator forskolin (10⁻⁹ to 10⁻⁵ M), or the nitric oxide donor sodium nitroprusside (10⁻⁹ to 10⁻⁴ M) were determined in the absence and/or presence of 30 minutes preincubation with one of K⁺ channel modulators, including the general K_V1 channel blocker 4-aminopyridine (4-AP; 5 × 10⁻³ M) and selective K_V1.5 channel blocker diphenyl phosphine oxide-1 (DPO-1; 10⁻⁶ M). Unless otherwise indicated, experiments were performed on endothelium-intact arterioles and in the presence of the nitric oxide synthase inhibitor N^G-nitro-L-arginine methyl ester (10⁻⁴ M) and the cyclooxygenase inhibitor indomethacin (10⁻⁵ M). At the end of each experiment, papaverine (10⁻⁴ M), an endothelium-independent vasodilator, was added to the vessel bath to determine the maximal dilation for normalization of dilator responses. Vasodilator responses are expressed as a percentage of maximal relaxation relative to endothelin-1 constriction, with 100% representing full relaxation to basal tension or the maximal diameter. The average baseline (cannulated preparation) or normalized (wire myograph) internal diameter of arterioles used in vessel reactivity studies is 168±29 μm (n=31).

Figure 4. Role of K_V channels in isoproterenol-induced dilation of human coronary arterioles from non-CAD and CAD subjects.

The β₂-adrenergic receptor agonist isoproterenol induced dose-dependent dilation in non-CAD arterioles, and this dilation was significantly reduced in CAD arterioles (A). The isoproterenol-induced dilation was markedly reduced by the general K_V1 channel blocker 4-AP (5 × 10⁻³ M) in non-CAD (B) but not in CAD vessels (C). Vasodilation to the NO donor sodium nitroprusside (SNP) was similar in non-CAD and CAD arterioles (D). *P<0.05 versus non-CAD or control; n=5–9 vessels/group.

Figure 5. Role of K_V channels in forskolin-induced dilation of human coronary arterioles from non-CAD and CAD subjects.

The adenylate cyclase activator forskolin-induced dilation was markedly reduced in CAD arterioles as compared with non-CAD arterioles (A). The dilation was reduced by the general K_V1 channel blocker 4-AP (5 × 10⁻³ M) in non-CAD (B) and to a lesser extent in CAD (C) arterioles. *P<0.05 versus non-CAD or control; n=4–8 vessels/group.

Figure 6. Role of K_V1.5 in isoproterenol-induced dilation of human coronary arterioles from non-CAD and CAD subjects.

Isoproterenol-induced dilation was significantly reduced by selective K_V1.5 channel blocker DPO-1 (10⁻⁶ M) in non-CAD arterioles (A). The dilation was also reduced by DPO-1 in CAD arterioles (B). *P<0.05 versus control; n=3–6 vessels/group. C, Dot density and associated box plots depicting the isoproterenol-induced maximal dilation of individual CAD arterioles used in two vessel groups for determining 4-AP- (Figure 4C) or DPO-1 (Figure 6B)-sensitive component of vasodilation, respectively. Mean dilation is indicated by red lines.

2.3 RT-PCR and qPCR

Human coronary arteries and arterioles (100–500 μm) were freshly dissected, cleaned of adherent tissue, and snap-frozen in liquid nitrogen and stored at −80°C until use. Some vessels were denuded of the endothelium to determine the mRNA expression of smooth muscle cells. Total RNA was extracted with RNeasy Fibrous Tissue Mini Kit (Qiagen), and cDNA was synthesized using SuperScript III reverse transcriptase (Invitrogen). cDNA (2 ng) was amplified using Platinum PCR Supermix (Invitrogen) and a 45-cycle touch-down protocol with gene-specific primers as previously described.²⁹ Detailed sequence information is also shown in Table 2. Two negative controls, without reverse transcription (RT-) and without template (H₂O), were amplified in parallel by using K_Vα9.3 primers. Human brain total RNA extracted from a normal donor (Agilent Technologies, Santa Clara, CA) was included as a positive control. Human brain tissue is known to express all 11 subunits of K_V channels examined in this study.³⁰

Table 2.

Nucleotide sequence of primers used for PCR

For end-point PCR
Transcript	Forward and Reverse primers	Amplicon size, bp
KV α1.1	5′-AGT CGC ACT TCT CCA GTA TCC-3′ 5′-TCG GTC AGT AGC TTG CTC TTA T-3′	424
KV α1.2	5′-TCA AGT TGT CCA GAC ACT CCA A-3′ 5′-TGT GTT AGC CAA GGT ACA GTT G-3′	548
KV α1.3	5′-CTT CAG GTT TCA GCA GCA TCC-3′ 5′-GCA GGT GGC AGT GGA ATT G-3′	395
KV α1.4	5′-GAG GCG GAT GAA CCT ACT ACC-3′ 5′-CCC TTT CCC TGA CAC TTC TCC-3′	401
KV α1.5	5′-CCG TCT ACT TCG CAG AGG CT-3′ 5′-CCA GGC AGA GGG CAT AAA GG-3′	465
KV α1.6	5′-AGA GGC TGA CGA TGA CGA TTC-3′ 5′-CCG ATG TGG TGT AGG AAG GTA G-3′	364
KV α2.1	5′-GGC TTG CTC ATC CTC TTC CTT-3′ 5′-TCT CCT GTC TCT TCT GCT CCT T-3′	289
KV α9.3	5′-GGC TTC TGC TTC TCT TCC TCT-3′ 5′-ACT GGT CCA CAT CAA TGT CCT T-3′	297
KV β1.1	5′-CAG CAG CCT TAG TCC CTC AG-3′ 5′-GCC GTT CAG CAA CCT CAT CT-3′	239
KV β1.2	5′-GTG GAG CAG CCG AAC AGA A-3 5′-GCC GTT CAG CAA CCT CAT CT-3′	192
KV β1.3	5′-ACC TCT CGT GAC TGC CTG T-3′ 5′-GCC GTT CAG CAA CCT CAT CT-3′	378
BKCaα	5′-ATG CGG AAC TCA CCC AAC A-3′ 5′-TCG CCA AAG ATG CAG ACC AC-3′	224
β-actin	5′-GCT CGT CGT CGA CAA CGG CTC-3′ 5′-CAA ACA TGA TCT GGG TCA TCT TCT C-3′	353
For real-time PCR (qPCR)
Transcript	Forward and Reverse	Amplicon size, bp
KV α1.5	5′-CGTCATCGTCTCCAACTTCAAC-3′ 5′-TTCCGCTGGACTCCTCTGT-3′	126
ATP synthase O subunit	5′-TCTCCTTTAGAAGAAGCCACACT-3′ 5′-GCACAATCATTCCACCCAAGAT-3′	124
β-actin	5′-GCACTCTTCCAGCCTTCCTT-3′ 5′-TGTGTTGGCGTACAGGTCTT-3′	114

Quantitative PCR was performed using CFX96 C1000 Thermal Cycler (BioRad). The reactions were conducted using 2x SsoAdvanced Universal SYBR Green Supermix (BioRad), K_Vα1.5-specific primers (Table 2), nuclease-free water and pooled sample cDNA for each non-CAD and CAD group (n=3/group, 2 ng each). Cycle threshold (Ct) values of products were determined using Bio-Rad CFX Manager 3.1 software (BioRad). Relative K_Vα1.5 gene expression was normalized to average of β-actin and ATP5o genes.³¹ Comparative Ct method was used to calculate relative expression of K_Vα1.5 subunit, where CAD was normalized to non-CAD for both intact and denuded pooled samples.

2.4 Immunoblotting

Human coronary arteries and arterioles (200–1000 μm) were freshly dissected, and membrane fraction proteins were isolated by using differential centrifugation method as described previously. ³² Membrane proteins of human brain tissue and total lysates of HEK-293 with or without human K_Vα1.5 overexpression were included as positive controls as previously described.²⁴ Proteins (20 μg) were separated by 10% SDS-PAGE and membrane were blotted with a primary antibody against a specific K_V1 α-subunit (1:2,000 dilution), followed by a horseradish-peroxidase conjugated secondary antibody (1:20,000 dilution). Membranes were developed using the ECL Prime reagent (Amersham). Detailed information on the K_V1 antibodies and immunoblotting protocol has been presented in our recent study.²⁴

2.5 Immunohistochemistry

Frozen sections (10 μm) of coronary arteries and arterioles were blocked with 5% normal goat serum and probed with primary antibodies that are specific to each K_V1 α-subunit (1:200 dilution) and secondary antibodies (Alexa Fluor 568-conjugated goat anti-rabbit or anti-mouse IgG, 1:400 dilution) as previously described.²⁴ Sections were counterstained with DAPI and images were captured using a confocal fluorescence microscope (model A1-R, Nikon).

2.6 Chemicals

Forskolin and DPO-1 were obtained from Tocris. All other chemicals were purchased from Sigma. Stock solutions were prepared in distilled water, except for the following: forskolin (DMSO); DPO-1 (ethanol); 4-AP (HCl, pH readjusted to 7.4); and indomethacin (10⁻¹ M Na₂CO₃).

2.7 Statistical Analysis

All data are presented as mean±SEM. Comparisons of concentration-response curves of isolated vessels were performed using 2-way repeated measures analysis of variance (ANOVA), followed by the Student-Newman-Keuls multiple comparison test. P values <0.05 were considered statistically significant.

3. RESULTS

3.1 K_V1 subunit mRNA and protein expression in non-CAD human coronary arteries

K_V1 channels (shaker-related family) have been shown to mediate an important part of K_V channel-related function in various vascular beds.²⁵ Using non-CAD human adipose arteries, we recently reported that several K_V1 α-subunits, including 1.1, 1.2, 1.4, and 1.5, were consistently found in different arterial samples, with K_V1.5 being the most abundantly expressed K_V1 α-subunit.²⁴ In this study, we first examined mRNA transcripts of K_V1 α and β subunits in HCAs from non-CAD subjects. Selected arteries were denuded of the endothelium with air bubble or 50μm-diameter stainless steel wire to examine the relative mRNA expression of these subunits in smooth muscle cells (n=5/group). As shown in Figure 1A, several K_V1 subunits were found to be consistently expressed in both endothelium-intact and -denuded HCAs from non-CAD and CAD subjects, including K_Vα 1.2, 1.3, 1.5 (most abundant), and K_Vβ 1.1–1.3. K_Vα1.1, 1.4 and 1.6 were variably detected in some but not all samples. This mRNA expression profile is generally consistent with that of adipose arterioles, except that variable expression was noticed for K_Vα1.1 in HCAs but for K_Vα1.3 in adipose arterioles. As a positive control, the human brain tissue from a normal subject expressed all K_V channel subunits examined in this study (Figure 1A). Analysis of relative mRNA expression of K_V1.5 normalized to the mean of two housekeeping genes (β-actin and ATP5o) showed a slight decrease (approximately 22%) both in intact and denuded CAD vessels (Figure 1B, pooled sample, n=3/group).

Figure 1. The mRNA expression of K_V1 channel subunits in human coronary arteries.

Representative gel images of RT-PCR amplification products from one endothelium-intact and one endothelium-denuded from non-CAD and CAD HCAs are shown in figure 1A. K_Vα1.2, 1.3, 1.5, 2.1, 9.3 and K_Vβ1.1–1.3 were consistently found in different arterial samples, whereas K_Vα1.1 1.4 and 1.6 subunits were variably detected. As a positive control, human brain tissue from a normal subject was found to express all K_V channel subunits studied (top). RT-, without reverse transcription; H₂O, without template; M, marker. Quantitative PCR revealed a slight reduction of K_Vα1.5 mRNA in CAD arteries (B).

Next, we assessed the protein expression of K_V1 channel forming α-subunits using the membrane fraction prepared from non-CAD HCAs. Consistent with mRNA expression from intact vessels, K_V1.5 protein, as well as K_V1.3 and K_V1.4, was detected in all HCAs samples (Figure 2). K_V1.6 protein was variably found in non-CAD HCAs, whereas K_V1.1 and 1.2 proteins were not detected. As positive controls, the membrane fraction of human brain tissue and total cell lysates of HEK-293 cells overexpressing human K_V1.5 were found to express K_V1.1–1.2, and K_V1.5, respectively.

Figure 2. The protein expression of K_V1 channel-forming α-subunits in human coronary arteries.

Western blot analysis of K_V1 α-subunits was performed using the membrane fraction of non-CAD HCAs. Consistent with high mRNA expression, K_Vα1.5 protein was detected in HCAs. K_Vα1.3, 1.4 and 1.6 were also detected in these samples. Arrow indicates mature forms of K_V1 proteins, with the approximate molecular mass values as follows: 75 kDa doublet (1.1), 75 kDa (1.2), 75 kDa doublet (1.3), 110 kDa doublet (1.4), 75 kDa (1.5), and 60 kDa (1.6). Proteins prepared from human brain tissue and HEK-293 with (+) or without (-) human K_V1.5 overexpression were included as positive controls.

3.2 Immunofluorescence localization of K_V1 α-subunits in human coronary arteries

To determine the cell-type specific localization of K_V1 α-subunits, we performed immunofluorescence on frozen sections (10 μm) of human coronary arteries and arterioles. K_V1 proteins were labeled with Alexa-568 conjugated secondary antibodies which emit red fluorescence upon excitation and images were captured with a confocal fluorescence microscope. Autofluorescence, intrinsic to the internal elastic lamina could be detected in the green fluorescein isothiocyanate channel and was used to visually separate the endothelial cell (EC) from the SMC layer. As shown in Figure 3, K_V1.5 was readily detected in both the SMC and EC layers. K_V1.4, however, was modestly expressed in SMCs, with little or no EC expression. K_V1.3 and K_V1.6 proteins were faintly detected in SMCs, with little to no EC expression. Enlarged regions are indicated by the outline (bottom panels). No fluorescent signal was observed in the negative control specimen (2° antibody only), demonstrating specificity of 1° antibodies for their respective K_V1 proteins.²⁵

Figure 3. Immunofluorescence localization of K_V1 α-subunits in human coronary arteries.

Representative confocal images show K_V1 α-subunit proteins (red) in cross tissue sections (10 μm) of an intact non-CAD HCA. Amplified regions indicated by dotted outline are shown in bottom panels. K_V1.5 was highly expressed in both smooth muscle cells (SMCs) and endothelial cells (ECs), with a modest expression of K_V1.4 in SMCs. K_V1.3 and K_V1.6 were faintly detected in SMCs and/or ECs. Cell nuclei were stained with DAPI (blue). IEL, internal elastic lamina (IEL, green autofluorescence).

3.3 Role of K_V channels in isoproterenol-induced dilation of human coronary arterioles

We next examined the functional role of K_V1 channels in mediating physiological vascular responses in HCAs. Previous studies have found that 4-AP-sensitive K_V1 channels mediate a major portion of the physiological dilation to the mixed β₂-adrenergic receptor agonist isoproterenol in rat small coronary arteries. Furthermore, exposure of these coronary arteries to high glucose (a major risk factor for CAD) impairs the activity and vasodilator function of K_V channels in coronary SMCs ^19,33–35. It remains unknown whether K_V1 channels mediate isoproterenol-induced dilation in HCAs and whether this vasomotor function is altered in diseases such as CAD. As shown in Figure 4A, isoproterenol-induced dilation was markedly reduced in CAD compared with non-CAD arterioles (% dilation at 10⁻⁷, 10⁻⁶, and 10⁻⁵ M isoproterenol, 33±12, 40±13, and 37±10 versus 78±7, 86±5, and 83±7 in non-CAD, respectively, n=5–9; P<0.05). In non-CAD arterioles, the dilation was markedly reduced by 4-AP (5 × 10⁻³ M), a general K_V1 channel blocker (Figure 4B, % dilation at 10⁻⁷, 10⁻⁶, and 10⁻⁵ M isoproterenol, 25±8, 50±11, and 53±11, n=9; P<0.05 versus non-CAD control). In CAD arterioles, however, the dilation was not affected by 4-AP (Figure 4C, n=5). In contrast, dilation to the NO donor sodium nitroprusside (10⁻⁹ to 10⁻⁴ M) was not significantly altered in CAD arterioles (Figure 4D; n=5/group), suggesting that potential non-specific impairment of vasomotor function in CAD arterioles is unlikely. Together, these results indicate that K_V1 channels mediate a significant component of beta-adrenergic dilation in non-CAD HCAs and this K_V1-mediated dilation is impaired during CAD.

3.4 Role of K_V channels in forskolin-induced dilation of human coronary arterioles

To further confirm the functional role of 4-AP-sensitive K_V1 channels, we examined the vasodilator response to forskolin, a pharmacological adenylate cyclase activator that induces non-receptor-mediated vasodilation through a similar cAMP-mediated pathway. Forskolin elicited a potent concentration-dependent dilation of HCAs from non-CAD subjects, whereas the dilation was significantly reduced in CAD subjects (Figure 5A, % dilation at 10⁻⁷ and 10⁻⁶ M forskolin, 23±13 and 71±16 versus 61±7 95±2 in non-CAD, respectively, n=4–8; P<0.05). 4-AP (5 × 10⁻³ M) markedly reduced forskolin-induced dilation in non-CAD arterioles (Figure 5B, % dilation at 10⁻⁷ and 10⁻⁶ M forskolin, 3±2 and 26±5, n=8; P<0.05 versus non-CAD control) but by much less extent in CAD arterioles (Figure 5C, % dilation at 10⁻⁷ and 10⁻⁶ M forskolin, 3±1 and 34±10, n=4; P<0.05 versus CAD control). These results suggest that the reduced dilation to forskolin in CAD HCAs is partially due to an impairment of 4-AP sensitive K_V1 channels.

3.5 Role of K_V1.5 channel in isoproterenol-induced dilation of human coronary arterioles

We further evaluated the role of Kv1.5, as a major K_V1 α-subunit expressed in HCAs (Figures 1–3), in mediating the vasodilator response to isoproterenol in HCAs. Although our recent study shows that the Kv1.5 channel represents a major isoform of K_V1 channels that is functionally impaired in human adipose arterioles from CAD subjects,²⁴ it is important to know whether this impairment extends to the human coronary circulation. Human coronary arterioles were preincubated with the selective K_V1.5 blocker DPO-1 (10⁻⁶ M). This pretreatment markedly reduced isoproterenol-induced dilation in non-CAD vessels (Figure 6A, % dilation at 10⁻⁸, 10⁻⁷, 10⁻⁶, and 10⁻⁵ M isoproterenol, 18±2, 37±7, 43±8, and 43±8 versus 49±7, 67±5, 78±4, and 80±3 in non-CAD control, respectively, n=3; P<0.05). This suggests that K_V1.5 serves as an important mediator of β adrenergic receptor-induced dilation in the human coronary microcirculation. Unexpectedly, DPO-1 also significantly reduced the dilation to isoproterenol in CAD vessels (Figure 6B, % dilation at 10⁻⁸, 10⁻⁷, 10⁻⁶, and 10⁻⁵ M isoproterenol, 8±3, 13±3, 22±2, and 33±3 versus 21±4, 38±4, 48±6, and 66±5 in CAD control, respectively, n=3; P<0.05). The DPO-1 data seem inconsistent with the apparent loss of 4-AP-sensitive dilation to isoproterenol during CAD (Figure 4C). We examined the contribution of the potential heterogeneous patient groups that have been included in each experimental protocol. As indicated in Figure 6C, baseline vasodilator responses are significantly higher in arterioles included for DPO-1 versus 4-AP protocol.

4 DISCUSSION

Bolstered by our recent findings that K_V channels contribute less and BK_Ca channels contribute more to vasodilation of adipose arterioles from patients with CAD,²⁴ the present study extends these observations by defining the expression and vasomotor contribution of K_V1 channel subtypes in HCAs. The main new findings are as follows: (1) K_V1.5 is a major isoform of K_V1 channels expressed in VSMCs of coronary arteries from non-CAD subjects; (2) Isoproterenol and forskolin, two distinct cAMP-mediated vasodilators, induce potent dilation of non-CAD coronary arterioles via activation of 4-AP-sensitive K_V (especially K_V1.5) channels; and (3) Responses to both cAMP-mediated vasodilators are impaired in CAD coronary arterioles, which is associated with a loss of or reduced contribution of 4-AP-sensitive K_V channels. Together, these results provide further support for the important role of VSMC K_V1.5 channels in the regulation of vascular tone in the human microcirculation. The impairment of 4-AP-sensitive K_V channel function and associated vasodilation is a consistent finding in both the coronary and adipose microvasculature, but a significant 4-AP-sensitive K_V (and notably K_V1.5) channel component remains in coronary arterioles during CAD.

K_V1 channel expression and function in HCAs

The K_V channel family consists of 12 subfamilies with over 40 known protein-encoding genes in humans and animals.³⁶ Previous studies have shown the expression and function of specific K_V channels are rather heterogeneous among species and vascular beds.²⁵ However, there is a consensus that 4-AP-sensitive shaker-family K_V1 are functionally important K_V channels in many animal vascular beds and in a few human vasculature studied.²⁵ Using freshly isolated human adipose arteries/arterioles, we recently reported that K_V1.5 is the major K_V1 channels expressed in VSMCs.²⁴ This is important since K_V1.5 mediates metabolic vasodilation in the heart.⁸ In the present study, immunofluorescence revealed that K_V1.5 is also the most abundant K_V1 channel-forming α subunit expressed in VSMCs of HCAs. Additional RT-PCR analysis showed that mRNA transcripts of K_Vα1.2, 1.3, 1.4, 1.5 (most abundant), as well as K_Vβ1.1–1.3, are present in HCAs from non-CAD subjects. Using immunoblotting analysis, K_Vα1.3–1.6 but not K_V1.1–1.2 were also detected at the protein level. Overall, these observations are consistent with those of our previous study in human adipose arterioles, thus suggesting that the transcriptional and translational regulation of K_V1 subunits may be relatively conserved in different vascular beds in humans. Several questions remain to be solved in future studies. As reported in several previous studies,^25,37 we found that the number of K_V1 α subunits expressed at the mRNA level is larger than that of the functional K_V1 channels observed in human arterioles. The role of endothelial K_V1 channels such as K_V1.5, which was detected in ECs of HCAs, is largely unknown. In addition, the vascular and cell-specific expression of K_V1.3, 1.4, and 1.6 channel subunits may require verification with different antibodies.

In this study, we used two different coronary vasodilators to elicit K_V-mediated vasodilation. Currently, there are no specific direct K_V1 channel openers. We instead used the mixed β₂-adrenoceptor agonist isoproterenol, which activates 4-AP-sensitive K_V channel currents in rabbit portal vein and rat coronary artery.^18,19 Isoproterenol activates the adenylate cyclase-cAMP-protein kinase A pathway that in turn phosphorylates and opens K_V channels.¹⁸ The adenylate cyclase activator forskolin also induces vasodilation by opening K⁺ channels, such as K_V channels, ^38,39 but its action requires no receptor activation. Forskolin increases whole-cell K⁺ currents in rabbit coronary arteries in a 4-AP-sensitive fashion ^38,39 Both isoproterenol and forskolin induced potent dilation in non-CAD coronary arterioles, with a large portion of the dilation inhibited to a similar extent by both the general K_V1 blocker 4-AP and the selective K_V1.5 blocker DPO-1. Although isoproterenol- and forskolin-induced dilations may involve multiple K⁺ channel types or other mechanisms, the data with the selective K_V1.5 blocker DPO-1 indicate that K_V1.5 plays a major functional role in cAMP-mediated coronary dilation. Whether K_V1.5 forms a homotetrameric channel complex or associates with other K_V subunits remains unclear.

K_V1 channel function in coronary arterioles from subjects with CAD

Alterations in K_V channel expression and function are key features of several pathological conditions including hypertension, metabolic syndrome, and diabetes.^4,6,20–22 In coronary arterioles from patients with CAD, we observed a significant reduction in cAMP-mediated vasodilation that was largely 4-AP-sensitive. The findings that both receptor (isoproterenol) and non-receptor (forskolin) agonist-induced vasodilator responses were impaired in CAD are consistent with the involvement of signaling components downstream of cAMP generation such as K_V channels, an important end-effector of vasomotor control. These results, together with our previous findings,²⁴ indicate that 4-AP-sensitive K_V channel function is impaired broadly (in both coronary and adipose arterioles) from subjects with CAD. Importantly, this impairment of K_V channel function contributes to reduced vasodilation in response to physiological stimuli. In contrast to altered K_V-dependent dilation, the NO donor sodium nitroprusside produced dilation that was similar in CAD and non-CAD coronary arteries, indicating specificity for cAMP-mediated dilation in CAD. Interestingly, activation of BK_Ca channels contribute to NO-induced dilation in several vascular beds,⁴⁰ and BK_Ca-mediated dilation of human adipose arterioles was not affected in CAD.²⁴

It is interest that in subjects with CAD, a greater 4-AP-sensitive dilation is observed in coronary compared to adipose arterioles.²⁴ In adipose arterioles, neither 4-AP nor DPO-1 further reduced H₂O₂-induced vasodilation in CAD. However, in coronary arterioles from CAD subjects, 4-AP-sensitive dilation was still observed. Furthermore, the selective K_V1.5 blocker DPO-1 significant inhibited isoproterenol-induced dilation in CAD coronary arterioles. These results suggest there may be vascular bed-specific differences in the extent of K_V channel dysfunction in patients with CAD. Previous studies in animal models also suggest that certain vascular beds (e.g., coronary circulation) may adapt in the presence of disease to limit the deleterious effects on vascular function.^26,27 The reasons for the persistent 4-AP/DPO-1-sensitive dilator component in coronary but not adipose vessels from subjects with CAD are unclear but could relate to potential heterogeneous patient groups included in different experimental protocols (Figure 6C), biased receptor signaling to isoproterenol, heterogeneous oxidative damage to K_V channels in different vascular beds, or compensation by upregulation of other K_V α or β subunits in the heart.

4.1 Study limitations

The present study used blood vessels isolated from surgically removed atrial appendages and unused donor hearts of subjects with a variety of conditions, and thus the non-CAD tissue samples are not truly normal controls. Ethical considerations preclude obtaining tissue from those subjects. To minimize potential confounding effects of the underlying disease, all of the non-CAD tissue samples included in this study are from subjects with no evidence of CAD and no more than 1 risk factor for CAD. Our determination of CAD is based on the patient’s medical record or direct observation of the conduit coronary arteries. We do not pre-select subjects; the majority of CAD patients are undergoing coronary bypass surgery and thus have significant coronary atherosclerosis (e.g., >50% occlusion of the left main coronary or of multiple vessels). We recognize that as a limitation the inherent heterogeneous nature of the CAD population, which is also evident in our analysis of isoproterenol-elicited dilation (Figure 6C). Additional studies to augment sample size would allow the use of statistical approaches to assess the effect of specific risk factors such as sex and age as well as other underlying cardiac conditions on K_V channel expression and function.

In the present study, we focused on the role of K_V1 channels in cAMP-mediated dilation and did not address the potential contribution of other K⁺ channels or non-ion channel mechanisms in this dilation. For example, there is evidence that K_V7 channels, another distinct family of VSMC K_V channels, can be activated by cAMP/protein kinase A and thus contribute to isoproterenol-induced dilation in some vascular beds.^41,42 We also did not further explore the mechanisms underlying altered K_V1 channel function in CAD coronary arterioles. Consistent with previous findings in human adipose arterioles,²⁴ quantitative PCR suggested a slight reduction mRNA expression of K_V1.5 (Figure 1B).

4.2 Clinical Implications

Dysregulation of microvascular function has been implicated in a variety of pathological conditions such as inflammation in visceral fat, obesity-associated insulin resistance, and ischemic heart disease.^43–45 This microvascular dysfunction is of particular importance in the coronary circulation, where it serves as an important contributor to myocardial ischemia as well as an independent risk factor for adverse cardiovascular outcomes in patients with CAD.^44,45 The heart has near-maximal extraction of oxygen at rest (70–80%) at rest and requires tight, beat-to-beat regulation of blood flow to nearly instantaneously match oxygen supply with its metabolic demand.^8,46 Thus, the coronary vascular tone is tightly regulated by endothelial factors (e.g., NO, prostacyclin, and endothelium-derived hyperpolarization factors), neurohumoral factors, metabolic factors, as well as myogenic mechanisms,^17,47–54 and this integral regulation allows for the precise control of blood flow in the heart. As key end-effectors of vasomotor control in the coronary circulation, dysfunctional K_V channels will negatively impact local blood flow regulation in response to not only endothelial factors such as H₂O₂ but also to tissue factors such as β-adrenergic transmitters and other metabolic factors.^19,20,46 This may induce deficit in regional blood supply and lead to a mismatch in oxygen supply and demand, which over time can impair cardiac function, induce local inflammation, fibrosis, and even heart failure.²³

4.3 Perspectives

This study provides evidence that K_V1.5, as a major 4-AP-sensitive K_V1 channel expressed in human coronary VSMCs, plays an important role in mediating cAMP-mediated dilation of coronary arterioles in the absence of CAD. The cAMP-mediated dilation is reduced in coronary arterioles during CAD, and this impairment is associated with impaired 4-AP-sensitive K_V channel function. Future studies are required to elucidate the mechanisms responsible for the regulation of K_V1 channels in health and disease and for the potential difference in K_V1 channel function and regulation between coronary and other vascular beds of humans.

Abbreviations

4-AP

4-aminopyridine

BK_Ca channel

large-conductance Ca²⁺-activated K⁺ channel

CAD

coronary artery disease

DPO-1

diphenyl phosphine oxide-1

EC

endothelial cell

HEK-293

human embryonic kidney 293

HCA

human coronary artery/arteriole

H₂O₂

hydrogen peroxide

K_Ca channel

Ca²⁺-activated K⁺ channel

K_V channel

voltage-gated K⁺ channel

K_V1 channel

shaker-related voltage-gated K⁺ channel

RT-PCR

reverse transcription-polymerase chain reaction

VSMC

vascular smooth muscle cell