Calcium-activated potassium channel family in coronary artery bypass grafts

Journal Pre-proof Calcium-activated Potassium Channel Family in Coronary Artery Bypass Grafts Wen-Tao Sun, PhD, Hai-Tao Hou, MPhil, Huan-Xin Chen, MPhil, Hong-Mei Xue, PhD, Jun Wang, MPhil, Guo-Wei He, MD, PhD, DSc, Qin Yang, MD, PhD PII:

S0022-5223(19)33456-7

DOI:

https://doi.org/10.1016/j.jtcvs.2019.11.016

Reference:

YMTC 15380

To appear in:

The Journal of Thoracic and Cardiovascular Surgery

Received Date: 24 July 2019 Revised Date:

6 November 2019

Accepted Date: 8 November 2019

Please cite this article as: Sun W-T, Hou H-T, Chen H-X, Xue H-M, Wang J, He G-W, Yang Q, Calciumactivated Potassium Channel Family in Coronary Artery Bypass Grafts, The Journal of Thoracic and Cardiovascular Surgery (2019), doi: https://doi.org/10.1016/j.jtcvs.2019.11.016. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Copyright © 2019 Published by Elsevier Inc. on behalf of The American Association for Thoracic Surgery

Sun et al, 2019 1

7 November 2019

2

JTCVS-19-1932R2

3

For Resubmission to the Journal of Thoracic and Cardiovascular Surgery

4

Calcium-activated Potassium Channel Family in Coronary Artery Bypass Grafts

5 6

Wen-Tao Sun, PhD1; Hai-Tao Hou, MPhil1; Huan-Xin Chen, MPhil1; Hong-Mei Xue, PhD1;

7

Jun Wang, MPhil1; Guo-Wei He, MD, PhD, DSc1,2,3; Qin Yang, MD, PhD1*

8 9

1

Center for Basic Medical Research, TEDA International Cardiovascular Hospital, Chinese

10

Academy of Medical Sciences & Peking Union Medical College, Tianjin, China

11

2

School of Pharmacy, Wannan Medical College, Wuhu, Anhui, China

12

3

Department of Surgery, Oregon Health and Science University, Portland, Oregon, USA.

13 14 15

Conflict of interest statement: none

16

Source of funding: supported by National Natural Science Foundation of China 81870227 &

17

81870288, Tianjin Municipal Science and Technology Commission 18PTZWHZ00060, and

18

Key Medical Program of Tianjin Binhai New Area Health Bureau 2018BWKZ005.

19 20

*

Correspondence to

21

Professor Qin Yang, MD, PhD

22

TEDA International Cardiovascular Hospital,

23

Chinese Academy of Medical Sciences & Peking Union Medical College

24

Tel: 86-22-65209089

25

E-mail: [email protected]

26

Article word count: 3340 1

Sun et al, 2019 27

Glossary of Abbreviations

28

Calcium-activated potassium channels: KCa channels

29

Coronary artery bypass graft: CABG

30

Endothelium-dependent relaxing factor: EDRF

31

Endothelium-derived hyperpolarizing factor: EDHF

32

Intermediate-conductance KCa channel: IKCa channel

33

Internal mammary artery: IMA

34

Large-conductance KCa channel: BKCa channel

35

Nitric oxide: NO

36

Saphenous vein: SV

37

Small-conductance KCa channel: SKCa channel

2

Sun et al, 2019 38

Central Message

39

All KCa subtypes are distributed in the endothelial and smooth muscle layers of IMA and

40

SV. IMA and SV differ in KCa subtype abundance in the two layers. BKCa plays a critical role

41

in IMA dilatation.

42 43

Perspective Statement:

44

By revealing the distribution profile and physiological role of the KCa family in IMA and

45

SV, this study provided new mechanistic insight into the intrinsic biological differences

46

between the arterial and venous CABG that may impact on the postoperative graft function

47

and suggested the potential of development of BKCa modulators for the prevention and

48

treatment of IMA spasm during/after CABG surgery.

49

3

Sun et al, 2019 50

Abstract

51

Objectives: We examined the expression, distribution, and contribution to vasodilatation of

52

the calcium-activated potassium channel (KCa) family in the commonly used coronary artery

53

bypass graft (CABG) internal mammary artery (IMA) and saphenous vein (SV) to understand

54

the role of large-, intermediate-, and small- conductance KCa (BKCa, IKCa, SKCa) subtypes in

55

graft dilatory properties determined by endothelium-smooth muscle interaction that is

56

essential to the postoperative performance of the graft.

57

Methods: Real-time polymerase chain reaction and western blot were employed to detect the

58

mRNA and protein level of KCa subtypes. Distribution of KCa subtypes was examined by

59

immunohistochemistry. KCa subtype-mediated vasorelaxation was studied using wire

60

myography.

61

Results: Both IMA and SV express all KCa subtypes with each subtype distributed in both

62

endothelium and smooth muscle. IMA and SV do not differ in the overall expression level of

63

each KCa subtype, corresponding to comparable relaxant responses to respective subtype

64

activators. In IMA, BKCa is more abundantly expressed in smooth muscle than in endothelium

65

while SKCa exhibits more abundance in the endothelium. In comparison, SV shows even

66

distribution of KCa subtypes in the two layers. The BKCa subtype in the KCa family plays a

67

significant role in vasodilatation of IMA whereas its contribution in SV is quite limited.

68

Conclusions: KCa family is abundantly expressed in IMA and SV. There are differences

69

between these two grafts in the abundance of KCa subtypes in the endothelium and the smooth

70

muscle. The significance of the BKCa subtype in vasodilatation of IMA may suggest the

71

potential of development of BKCa modulators for the prevention and treatment of IMA spasm

72

during/after CABG surgery.

73

Keywords: Calcium-activated potassium channels, Coronary artery bypass graft, Internal

74

mammary artery, Saphenous vein 4

Sun et al, 2019 75

Introduction

76

Internal mammary artery (IMA) and saphenous vein (SV) are the most commonly used

77

coronary artery bypass grafts (CABG) for the treatment of ischemic heart disease. IMA is

78

superior to SV in the long-term patency while has tendency to develop spasm during surgical

79

dissection and the perioperative period. In addition to the intrinsic structure difference, IMA

80

and SV are known to be different in native matrix metalloproteinase characteristics1 and

81

endothelial function. Previous studies from our group have demonstrated that compared with

82

SV, IMA produces more nitric oxide (NO) and shows greater biofunctionality of

83

endothelium-derived hyperpolarizing factor (EDHF)2. These divergent properties are thought

84

to underlie the differences between IMA and SV in the postoperative graft function and

85

long-term patency rates.

86

Calcium-activated potassium (KCa) channels, including large-, intermediate-, and

87

small-conductance KCa (BKCa, IKCa, and SKCa), are involved in the regulation of vascular tone.

88

Being composed of a symmetrical tetramer containing four pore-forming α subunits and four

89

regulatory β subunits3, BKCa channels in the vasculature were reported to be mainly

90

distributed in the plasma membrane of smooth muscle cells4, while the existence and function

91

of BKCa in the vascular endothelium remains controversial5, 6. In comparison, SKCa and IKCa

92

channels were found to be predominantly expressed in endothelial cells in various

93

vasculatures7-9, with sporadic reports of function and expression in the smooth muscle10, 11.

94

Opening of BKCa channels in smooth muscle cells causes hyperpolarization, which limits Ca2+

95

entry through voltage-activated Ca2+ channels, providing a negative-feedback mechanism

96

opposing vasoconstriction12. BKCa channels act as a smooth muscle effector of

97

endothelium-derived relaxing factors (EDRF)13, 14, in particular NO. Numerous studies have

98

demonstrated that activation of endothelial IKCa (KCa3.1) and SKCa (KCa2.3) underlies the

99

classical EDHF pathway15, and these channels are also involved in the regulation of NO 5

Sun et al, 2019 100

bioavailability16.

101

Despite extensive research in animal vasculatures, studies regarding the role of KCa

102

channels in human vessels are limited. It was reported that BKCa channels are expressed in

103

smooth muscle cells of human coronary arteries and mediate coronary artery dilation17.

104

Modulation of endothelial SKCa and IKCa channels were found to take part in coronary

105

arteriolar dysfunction in pathological/diseased conditions such as laminar shear stress,

106

ischemia-reperfusion, and diabetes18-20. Expressions or function of KCa channels were also

107

demonstrated in human mesenteric7, gastroepiploic21 and pulmonary arteries22. Nevertheless,

108

these studies either focused on specific KCa subtype or investigated the overall function of KCa

109

channels, none of them looked into the distribution profile and the respective contribution to

110

vasorelaxation made by KCa subtypes.

111

There has been a certain amount of studies concerning the effects of chemically

112

synthesized compounds and natural substances on CABG. By using pharmacological tools,

113

some studies including those from our group have identified the functional role of KCa

114

channels in human IMA. For example, blockade of KCa channels with tetraethylammonium

115

enhanced the contractile effect of endogenous vasopressin23. Relaxation induced by propofol

116

and flavanol compound (-)-epicatechin was significantly inhibited by the selective BKCa

117

blocker iberiotoxin in IMA24,

118

reduced by the IKCa inhibitor TRAM-34 whereas almost abolished by iberiotoxin26. Few

119

studies investigated the role of KCa channels in the venous graft SV. Zhang and colleagues

120

reported that the total patch current was significantly suppressed by tetraethylammonium and

121

iberiotoxin in smooth muscle cells isolated from human SV27. So far, no efforts have been

122

made to compare the localization and function of different subtypes of KCa channels between

123

IMA and SV.

124

25

. Procyanidin B2-induced relaxation of IMA was slightly

The present study aimed to examine the expression, localization, and the respective 6

Sun et al, 2019 125

contribution to relaxation of KCa subtypes in both IMA and SV. By using multiple methods

126

including western blot, real-time polymerase chain reaction (Q-PCR), immunohistochemistry,

127

and myograph recording, we were able to clarify the differences between the arterial and

128

venous grafts regarding the distribution profile and physiological role of the KCa family.

129

Materials and Methods

130

IMA (n=50) and SV (n=50) grafts were harvested from patients undergoing CABG

131

surgery. The residual segments of IMA and SV that would otherwise be discarded were

132

collected with the consent of the patients. The study protocol was approved by the

133

Institutional Ethics Review Board of TEDA International Cardiovascular Hospital

134

([2018]-0626-2).

135

Q-PCR

136

Total RNA was extracted from IMA and SV segments using TRIzol reagent and reverse

137

transcription and quantitative PCR amplification were performed in LightCycler 96

138

employing TransScript Green Two-Step qRT-PCR SuperMix system. The details of PCR

139

cycle conditions, primers used for amplification of BKCaα, BKCaβ1, KCa2.3, and KCa3.1, and

140

quantification methods were described in the Supplementary Material.

141

Western blot

142

Whole cell protein of IMA and SV segments were extracted and the protein of interests

143

including BKCaα, BKCaβ1, KCa2.3, and KCa3.1 were detected by using specific primary

144

antibodies followed by horseradish peroxidase (HRP)-conjugated secondary antibodies, as

145

published elsewhere28, 29. The details of the procedure were described in the Supplementary

146

Material.

147

Histology and immunohistochemistry staining

148

IMA and SV segments were fixed in 4% paraformaldehyde, decalcified, dehydrated, and

149

embedded in paraffin. The embedded tissue samples were sliced into 4-5 µm sections and 7

Sun et al, 2019 150

stained with haematoxylin and eosin after deparaffination. Immunohistochemistry staining

151

was performed with the primary antibody against BKCaα, BKCaβ1, KCa2.3, or KCa3.1, as

152

described in details in the Supplementary Material.

153

Immunostaining was evaluated by two independent pathologists using a blind protocol

154

design. For each specimen, the histoscore (H-score) of each KCa subtype was calculated as the

155

sum of staining intensity (0-3, negative staining: 0; weak staining: 1; moderate staining: 2;

156

strong staining: 3) multiplied by the percentage of stained cells (0-100%).

157

Isometric force study

158

IMA and SV segments were cut into 3-mm rings and mounted in a four-channel Mulvany

159

myograph (Model 620M, J.P.Trading, Aarhus, Denmark). The details of myograph experiment

160

have been extensively published in our previous studies28-30. Cumulative dose-response

161

curves to (a) BKCa activator NS1619 (-9~-4.5 LogM), (b) IKCa/SKCa activator NS309 (-9~-4.5

162

LogM), and (c) acetylcholine (-10~-4.5 LogM) were established in IMA and SV rings

163

precontracted by U46619. In the study of acetylcholine-induced relaxation, vessels were

164

incubated with the specific blocker for BKCa (iberiotoxin, 100 nmol/L), IKCa (TRAM34, 1

165

µmol/L), or SKCa (apamin, 100 nmol/L) prior to U46619-precontraction. Relaxant responses

166

of IMA and SV to acetylcholine were also studied after combined application of TRAM34

167

and apamin, or TRAM34 and apamin plus iberiotoxin.

168

Statistical analysis

169

Expression of targets of interest was normalized to the expression of β-actin for protein

170

analysis, and to the level of GAPDH for mRNA analysis. Relaxation was expressed as the

171

percentage decrease in isometric force induced by U46619. Data were expressed as

172

mean±SEM. One-way ANOVA was used to assess differences among groups followed by

173

Scheffe post-hoc test (SPSS, version 20). When two groups were compared, differences were

174

assessed by Student’s t-test. p<0.05 was considered statistically significant. 8

Sun et al, 2019 175

Results

176

KCa channel expression in IMA and SV

177

mRNA expression of KCa subtypes: IMA vs. SV

178

Q-PCR experiments enrolling five pairs of IMA and SV (each pair from the same

179

individual) showed that all three subtypes of KCa channels are expressed in both the arterial

180

and venous grafts. Further analysis indicated that the mRNA level of the BKCa subtype is

181

significantly higher than that of the IKCa and SKCa in IMA (Fig.1A), whereas in SV, the

182

mRNA expressions of the three KCa subtypes are of no statistical difference (Fig.1B). In

183

comparison with IMA, SV expresses less BKCa subtype, shown by lower mRNA level of α

184

and β1 subunits of BKCa (Fig.1C).

185

Protein expression of KCa subtypes: IMA vs. SV

186

Protein detection using eight pairs of IMA and SV further demonstrated the endogenous

187

expression of the KCa family in these grafts (Fig.2A&B). The protein level of BKCa, IKCa, and

188

SKCa was not significantly different from each other in IMA, which was inconsistent with the

189

results from Q-PCR showing significant higher mRNA expression of BKCa, suggesting a

190

significant translational control of KCa expression in this arterial graft. In comparison, the

191

protein expression level of the three KCa subtypes in SV was in line with the mRNA

192

expression pattern, showing no differences in the protein content of BKCa, IKCa, and SKCa.

193

Further comparison between IMA and SV showed a relatively lower protein level of BKCa in

194

SV whereas the difference was statistically insignificant.

195

Distribution profile of KCa subtypes in IMA and SV

196

Immunohistochemistry staining revealed the localization of KCa family in IMA and SV,

197

showing the distribution of all three subtypes of KCa in both endothelium and smooth muscle

198

layers (Fig.3A). H-score calculated from five independent pairs of IMA and SV samples

199

indicated no significant differences in the overall staining of each KCa subtype (Fig.3B), 9

Sun et al, 2019 200

which was in agreement with the results of western blot experiment showing comparable

201

protein level of each KCa subtype in the two grafts (Fig.2B). Further analysis of the

202

distribution of each KCa subtype in the endothelial and the smooth muscle layer showed that

203

in IMA, BKCa is more abundantly expressed in smooth muscle cells than in endothelial cells

204

(p=0.018) while SKCa exhibits more abundance in the endothelial layer (p=0.020) (Fig.3C).

205

Different from IMA, SV exhibited no significant differences in the KCa distribution in the

206

endothelial and the smooth muscle layer, though the abundance of IKCa tends to be slightly

207

more in the endothelium than in the smooth muscle (p=0.058) (Fig.3D). Comparison between

208

IMA and SV showed a significant lower H-score of BKCa β1 in the smooth muscle layer of

209

the SV (p=0.021) (Fig.3E).

210

Role of KCa subtypes in the vasoactivity of IMA and SV

211

Relaxant response of IMA and SV to KCa channel openers

212

The BKCa channel agonist NS1619 and the IKCa/SKCa channel opener NS309 elicited

213

dose-dependent relaxations in both IMA and SV. Relaxant responses of IMA and SV to

214

NS1619 were similar in terms of the maximal response (51.0±3.1% and 56.4±8.6%

215

respectively) (Fig.4A). The EC50 value for NS1619 in IMA was -5.45±0.11 LogM and

216

-5.84±0.08 LogM in SV (p=0.033). The maximal response and sensitivity to NS309 did not

217

differ between IMA and SV (Rmax: 79.0±6.7% vs. 75.0±7.3%, EC50: -5.38±0.04 vs.

218

-5.35±0.08 LogM) (Fig.4B).

219

KCa-mediated relaxation in response to acetylcholine in IMA and SV

220

The contribution of KCa subtypes in the relaxant response of IMA and SV to acetylcholine

221

was further studied with the use of specific KCa blockers. Acetylcholine induced

222

dose-dependent relaxation in both grafts, with maximal response of 76.2±3.0% in IMA and

223

23.2±1.5% in SV (Fig.4C&D). Application of the BKCa blocker iberiotoxin significantly

224

decreased acetylcholine-induced relaxation in IMA from 76.2±3.0% to 41.9±5.4% (p<0.01). 10

Sun et al, 2019 225

In comparison, inhibition of IKCa with TRAM-34 or inhibition of SKCa with apamin did not

226

significantly suppress acetylcholine-induced relaxation in IMA though minor decreases were

227

observed. Except a slight decrease in the maximal response, saying 71.2±3.0% in

228

TRAM-treated group and 66.0±4.6% in apamin-treated group, inhibition of IKCa and SKCa

229

tended to blunt the response of IMA to acetylcholine at the low concentration range (Fig.4C).

230

Application of TRAM-34 in conjunction with apamin showed no further inhibition on the

231

relaxation as compared to the use of TRAM-34 or apamin alone. Further combination of

232

TRAM-34,

233

acetylcholine-induced relaxation (36.9±6.1%), which however was only slightly less than the

234

relaxation in the IMA treated with iberiotoxin alone (Fig.4C), suggesting the predominant

235

role of BKCa subtype in the KCa family in mediating the relaxant response of IMA to

236

endogenous vasodilators.

apamin,

and

iberiotoxin

resulted

in

a

dramatic

attenuation

in

237

Compared with IMA, SV showed significantly less relaxant response to acetylcholine

238

(23.2±1.5% vs. 76.2±3.0%, p<0.001). Blockade of different KCa subtype respectively caused

239

no obvious inhibition on acetylcholine-induced relaxation in SV. Even triple application of

240

BKCa, IKCa, and SKCa blockers showed quite limited inhibitory effect on the relaxant response,

241

with significant suppression only observed at the highest concentration of acetylcholine

242

(15.2±2.8% vs. 23.2±1.5% in control, p=0.02).

243

Treatment with respective KCa blockers individually did not significantly affect the

244

resting force and U46619-precontraction of IMA and SV except that in the IMA, combined

245

use of BKCa, IKCa, and SKCa blockers enhanced U46619-induced precontraction (Table 1).

246

The sensitivity of IMA and SV to acetylcholine was barely affected by the KCa blockers, as

247

indicated by the unaltered EC50 values (Table 2).

248

Discussion

249

In this study, by using Q-PCR, western blot, and immunohistochemistry for localization 11

Sun et al, 2019 250

and quantification of KCa subtypes, and isometric force measurement for functional

251

investigation, we for the first time compared the distribution profile and contribution to

252

vasodilatation of the members of KCa family between IMA and SV. Results derived from this

253

study demonstrated that 1) both IMA and SV express all three subtypes of KCa channels,

254

which are distributed in both endothelial and smooth muscle layers; 2) IMA does not differ

255

from SV in the overall expression level of each KCa subtype; 3) IMA and SV show differences

256

in the abundance of KCa subtypes in endothelial and smooth muscle layers; 4) the BKCa

257

subtype in the KCa family plays a significant role in endothelium-dependent vasodilatation of

258

IMA whereas its contribution in SV is quite limited. Figure 5 provided a schematic summary

259

of the present study.

260

As the commonly used CABG, IMA and SV are obviously different in anatomic structure

261

and hemodynamic characteristics. In addition, they are different in the response to constrictor

262

and dilator agents. For example, the inotropic/vasodilator compound levosimendan relaxed

263

IMA but failed to relax SV completely31. PGE(2) induced vasoconstriction in IMA whereas

264

vasodilatation in SV, which was attributed to the mediation of different EP receptors32.

265

Compared with SV, IMA exhibited greater endothelium-dependent relaxation33, which was

266

again demonstrated in the present study. The EDRF stimulus acetylcholine evoked

267

approximately 80% relaxation in IMA but less than 25% relaxant response in SV. Our

268

previous efforts to understand the differences between IMA and SV in the degree of

269

endothelium-dependent relaxation revealed that IMA releases more NO and exhibits more

270

hyperpolarization than SV to EDRF stimuli2. In this study, with a further attempt to dissect the

271

molecular basis of such intrinsic differences in endothelial function between the two grafts,

272

we studied and compared the expression and function of the KCa channel family, which has

273

been suggested to largely take part in the generation/action of EDRF including EDHF and

274

NO8, 13, 16, 34. 12

Sun et al, 2019 275

In order to gain a precise knowledge about whether KCa channels differ between IMA and

276

SV in the expression and distribution, Q-PCR, western blot, and immunohistochemistry

277

experiments were performed using “paired grafts”, namely, each pair of IMA and SV samples

278

was taken from the same patient. This ruled out the influence of clinical characteristics of

279

different patients on the KCa channels and thus, enabled the revealing of the intrinsic

280

differences, if any, in this channel family between IMA and SV. Our results suggested that

281

IMA and SV do not differ from each other with respect to the overall expression level of the

282

members of the KCa family, evidenced by the comparable protein level and H-score of each

283

KCa subtype in these two grafts. Interestingly, at mRNA level, IMA was found to express more

284

BKCa subtype than SV. The inconsistent mRNA and protein data may suggest that IMA differs

285

from SV in the translational control of BKCa expression. Whether this is due to the differences

286

in oxygen tension and shear stress in the arterial and venous system is a topic for future

287

investigation. The higher BKCa mRNA content in IMA may serve as a reservoir for

288

maintaining or repairing BKCa channel function in the arterial graft in pathological conditions.

289

The comparable expression of KCa subtypes in IMA and SV was also supported by the similar

290

degree of responses of the two grafts to the BKCa opener NS1619 and the IKCa/SKCa opener

291

NS309.

292

Immunohistochemistry experiments showed that in IMA and SV, IKCa, SKCa, and BKCa

293

subtypes are localized in both endothelial and smooth muscle cells and they present in similar

294

abundance in these two grafts, as suggested by western blot and H-score analysis. Though

295

IMA and SV share similar expression and localization profile for KCa channels, they are

296

different in the distribution abundance of KCa subtypes. In IMA, BKCa is more abundantly

297

expressed in smooth muscle cells than in endothelial cells while SKCa exhibits more

298

abundance in the endothelial layer, which is in accordance with the finding of BKCa and SKCa

299

distribution in various vasculatures. In comparison, SV exhibits even distribution of KCa 13

Sun et al, 2019 300

subtypes in the endothelial and the smooth muscle layer, though IKCa tends to be slightly

301

more in abundance in the endothelium. Another important finding of the present study is that

302

as compared to IMA, SV expresses less BKCa β1 in the smooth muscle, shown by the lower

303

H-score of BKCa β1 staining. Considering the essential role of the β1 subunit in the function

304

of BKCa channels, this may indicate a difference in the physiological performance of BKCa

305

channels between IMA and SV.

306

Functional studies of the KCa channels using acetylcholine suggested that in IMA, BKCa

307

subtype plays a significant role in endothelium-dependent vasodilatation whereas the

308

participation of IKCa and SKCa is limited. Single blockade of IKCa or SKCa and combined

309

blockade of these two subtypes only inhibited the relaxation to a small extent. The

310

insignificant contribution of IKCa and SKCa was not in line with our expectation. Taking into

311

account of the interaction between NO and EDHF35, we assumed that under physiological

312

condition, the predominant eNOS-NO signaling masks the role of IKCa and SKCa channels, the

313

molecular basis for EDHF, in the vasodilatation of IMA. In fact, previous studies including

314

our study in the IMA showing the role of EDHF2 were conducted in the presence of eNOS

315

and PGI2 inhibitors. The abundant expression of BKCa channels in the smooth muscle of IMA

316

may also imply the critical role of NO in acetylcholine-induced relaxation in this arterial graft

317

as BKCa is a well-recognized smooth muscle effector of NO.

318

We previously reported that SV produces less NO and EDHF than IMA2, which explains

319

the dramatically weaker relaxation of SV compared to IMA in response to acetylcholine.

320

Similar to IMA, blockade of IKCa and SKCa channels in SV barely affected the

321

endothelium-dependent relaxation induced by acetylcholine. The BKCa channel blocker

322

iberiotoxin only caused minor inhibition on acetylcholine-induced relaxation in SV, which

323

was in contrast to the significant inhibition by iberiotoxin in IMA. Less expression of BKCa

324

β1 in the smooth muscle of SV may provide an explanation for the weaker inhibitory effect of 14

Sun et al, 2019 325

iberiotoxin in this graft.

326

Future perspectives and implications

327

Inconsistent with the common notion that IKCa and SKCa channels are generally

328

specifically expressed in vascular endothelium while BKCa in the smooth muscle, the present

329

study provided solid evidence that all three subtypes of KCa channels are abundantly

330

expressed in both endothelial and smooth muscle cells of IMA and SV. This raises two

331

questions for future research: 1) the respective role and contribution of endothelial KCa and

332

smooth muscle KCa channels to the vasoactivity of IMA and SV; 2) does certain KCa subtype

333

in the endothelium behave the same as the subtype in the smooth muscle? Further studies

334

using endothelium-denuded preparation and electrophysiological approaches are warranted to

335

provide more comprehensive understanding of the role of KCa family in the function of CABG.

336

In addition, although in this study IKCa and SKCa channels were proved to be unessential to

337

relaxation of IMA and SV, considering their abundance in the grafts, whether these channels

338

are involved in the regulation of inflammatory process to impact on the postoperative

339

performance and long-term patency of the graft is worthy of further studies. Moreover, the

340

finding of the differences in BKCa subtype (significant vs. minor contribution to

341

vasorelaxation in IMA and SV, correlated to higher smooth muscle BKCa β1 level in IMA

342

than SV) suggests the potential of BKCa modulator in the prevention and treatment of IMA

343

spasm, which whereas cannot be achieved by IKCa and SKCa channel modulators.

344

Taken together, by revealing for the first time the distribution profile and the respective

345

contribution of KCa subtypes in vasorelaxation in IMA and SV, this study demonstrated that

346

KCa channel family is abundantly expressed in IMA and SV. There are differences between

347

these two grafts in the abundance of KCa subtypes in the endothelium and the smooth muscle.

348

The significance of the BKCa subtype in vasodilatation of IMA may suggest the potential of

349

development of BKCa modulators for the prevention and treatment of IMA spasm during/after 15

Sun et al, 2019 350

CABG surgery.

351 352

Acknowledgements: We thank Ms. Jing Zhang from Yuebin Medical Research Laboratory

353

for technical support in immunohistochemistry experiments.

16

Sun et al, 2019 354

References

355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391

1.

2.

3.

4. 5. 6. 7.

8.

9.

10.

11.

12.

13.

Anstadt MP, Franga DL, Portik-Dobos V, Pennathur A, Bannan M, Mawulawde K, et al. Native matrix metalloproteinase characteristics may influence early stenosis of venous versus arterial coronary artery bypass grafting conduits. Chest. 2004;125:1853-1858. Liu ZG, Ge ZD, He GW. Difference in endothelium-derived hyperpolarizing factor-mediated hyperpolarization and nitric oxide release between human internal mammary artery and saphenous vein. Circulation. 2000;102:III296-301. Knaus HG, Eberhart A, Glossmann H, Munujos P, Kaczorowski GJ, Garcia ML. Pharmacology and structure of high conductance calcium-activated potassium channels. Cell Signal. 1994;6:861-870. Ledoux J, Werner ME, Brayden JE, Nelson MT. Calcium-activated potassium channels and the regulation of vascular tone. Physiology (Bethesda). 2006;21:69-78. Sandow SL, Grayson TH. Limits of isolation and culture: intact vascular endothelium and BKCa. Am J Physiol Heart Circ Physiol. 2009;297:H1-7. Yang Q, Underwood MJ, He GW. Calcium-activated potassium channels in vasculature in response to ischemia-reperfusion. J Cardiovasc Pharmacol. 2012;59:109-115. Kohler R, Degenhardt C, Kuhn M, Runkel N, Paul M, Hoyer J. Expression and function of endothelial Ca(2+)-activated K(+) channels in human mesenteric artery: A single-cell reverse transcriptase-polymerase chain reaction and electrophysiological study in situ. Circ Res. 2000;87:496-503. Yang Q, Huang JH, Man YB, Yao XQ, He GW. Use of intermediate/small conductance calcium-activated potassium-channel activator for endothelial protection. J Thorac Cardiovasc Surg. 2011;141:501-510, 510 e501. McNeish AJ, Sandow SL, Neylon CB, Chen MX, Dora KA, Garland CJ. Evidence for involvement of both IKCa and SKCa channels in hyperpolarizing responses of the rat middle cerebral artery. Stroke. 2006;37:1277-1282. Kohler R, Wulff H, Eichler I, Kneifel M, Neumann D, Knorr A, et al. Blockade of the intermediate-conductance calcium-activated potassium channel as a new therapeutic strategy for restenosis. Circulation. 2003;108:1119-1125. Wong KL, Chan P, Huang WC, Yang TL, Liu IM, Lai TY, et al. Effect of tetramethylpyrazine on potassium channels to lower calcium concentration in cultured aortic smooth muscle cells. Clin Exp Pharmacol Physiol. 2003;30:793-798. Eichhorn B, Dobrev D. Vascular large conductance calcium-activated potassium channels: functional role and therapeutic potential. Naunyn Schmiedebergs Arch Pharmacol. 2007;376:145-155. Bolotina VM, Najibi S, Palacino JJ, Pagano PJ, Cohen RA. Nitric oxide directly activates calcium-dependent potassium channels in vascular smooth muscle. Nature. 17

Sun et al, 2019 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429

14.

15. 16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

1994;368:850-853. Weston AH, Feletou M, Vanhoutte PM, Falck JR, Campbell WB, Edwards G. Bradykinin-induced, endothelium-dependent responses in porcine coronary arteries: involvement of potassium channel activation and epoxyeicosatrienoic acids. Br J Pharmacol. 2005;145:775-784. Edwards G, Feletou M, Weston AH. Endothelium-derived hyperpolarising factors and associated pathways: a synopsis. Pflugers Arch. 2010;459:863-879. Sheng JZ, Braun AP. Small- and intermediate-conductance Ca2+-activated K+ channels directly control agonist-evoked nitric oxide synthesis in human vascular endothelial cells. Am J Physiol Cell Physiol. 2007;293:C458-467. Tanaka Y, Meera P, Song M, Knaus HG, Toro L. Molecular constituents of maxi KCa channels in human coronary smooth muscle: predominant alpha + beta subunit complexes. J Physiol. 1997;502 ( Pt 3):545-557. Feng J, Liu Y, Clements RT, Sodha NR, Khabbaz KR, Senthilnathan V, et al. Calcium-activated potassium channels contribute to human coronary microvascular dysfunction after cardioplegic arrest. Circulation. 2008;118:S46-51. Liu Y, Xie A, Singh AK, Ehsan A, Choudhary G, Dudley S, et al. Inactivation of Endothelial Small/Intermediate Conductance of Calcium-Activated Potassium Channels Contributes to Coronary Arteriolar Dysfunction in Diabetic Patients. J Am Heart Assoc. 2015;4:e002062. Takai J, Santu A, Zheng H, Koh SD, Ohta M, Filimban LM, et al. Laminar shear stress upregulates endothelial Ca(2)(+)-activated K(+) channels KCa2.3 and KCa3.1 via a Ca(2)(+)/calmodulin-dependent protein kinase kinase/Akt/p300 cascade. Am J Physiol Heart Circ Physiol. 2013;305:H484-493. Urakami-Harasawa L, Shimokawa H, Nakashima M, Egashira K, Takeshita A. Importance of endothelium-derived hyperpolarizing factor in human arteries. J Clin Invest. 1997;100:2793-2799. Bonnet S, Archer SL. Potassium channel diversity in the pulmonary arteries and pulmonary veins: implications for regulation of the pulmonary vasculature in health and during pulmonary hypertension. Pharmacol Ther. 2007;115:56-69. Novella S, Martinez AC, Pagan RM, Hernandez M, Garcia-Sacristan A, Gonzalez-Pinto A, et al. Plasma levels and vascular effects of vasopressin in patients undergoing coronary artery bypass grafting. Eur J Cardiothorac Surg. 2007;32:69-76. Dogan MF, Arslan SO, Yildiz O, Kurtoglu M, Parlar A. Propofol-Induced Vasodilation in Human Internal Mammary Artery: Role of Potassium Channels. J Cardiothorac Vasc Anesth. 2019;33:2183-2191. Novakovic A, Marinko M, Vranic A, Jankovic G, Milojevic P, Stojanovic I, et al. Mechanisms underlying the vasorelaxation of human internal mammary artery 18

Sun et al, 2019 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

induced by (-)-epicatechin. Eur J Pharmacol. 2015;762:306-312. Novakovic A, Marinko M, Jankovic G, Stojanovic I, Milojevic P, Nenezic D, et al. Endothelium-dependent vasorelaxant effect of procyanidin B2 on human internal mammary artery. Eur J Pharmacol. 2017;807:75-81. Zhang H, Li P, Almassi GH, Nicolosi A, Olinger GN, Rusch NJ. Single-channel and functional characteristics of a KCa channel in vascular muscle membranes of human saphenous veins. J Cardiovasc Pharmacol. 1996;28:611-617. Sun WT, Wang XC, Mak SK, He GW, Liu XC, Underwood MJ, et al. Activation of PERK branch of ER stress mediates homocysteine-induced BKCa channel dysfunction in coronary artery via FoxO3a-dependent regulation of atrogin-1. Oncotarget. 2017;8:51462-51477. Sun WT, Wang XC, Novakovic A, Wang J, He GW, Yang Q. Protection of dilator function of coronary arteries from homocysteine by tetramethylpyrazine: Role of ER stress in modulation of BKCa channels. Vascul Pharmacol. 2019;113:27-37. Hou HT, Wang J, Zhang X, Wang ZQ, Chen TN, Zhang JL, et al. Endothelial nitric oxide synthase enhancer AVE3085 reverses endothelial dysfunction induced by homocysteine in human internal mammary arteries. Nitric Oxide. 2018;81:21-27. Mirkhani H, Shafa M, Khazraei H. Comparison of the effects of levosimendan and papaverine on human internal mammary artery and saphenous vein. Cardiovasc Drugs Ther. 2009;23:355-359. Foudi N, Kotelevets L, Gomez I, Louedec L, Longrois D, Chastre E, et al. Differential reactivity of human mammary artery and saphenous vein to prostaglandin E(2) : implication for cardiovascular grafts. Br J Pharmacol. 2011;163:826-834. Luscher TF, Diederich D, Siebenmann R, Lehmann K, Stulz P, von Segesser L, et al. Difference between endothelium-dependent relaxation in arterial and in venous coronary bypass grafts. N Engl J Med. 1988;319:462-467. Archer SL, Gragasin FS, Wu X, Wang S, McMurtry S, Kim DH, et al. Endothelium-derived hyperpolarizing factor in human internal mammary artery is 11,12-epoxyeicosatrienoic acid and causes relaxation by activating smooth muscle BK(Ca) channels. Circulation. 2003;107:769-776. Nishikawa Y, Stepp DW, Chilian WM. Nitric oxide exerts feedback inhibition on EDHF-induced coronary arteriolar dilation in vivo. Am J Physiol Heart Circ Physiol. 2000;279:H459-465.

19

Sun et al, 2019 464 465 466

Table 1. Resting force and U46619-induced contraction of internal mammary artery (IMA) and saphenous vein (SV) Group

467 468 469

Resting Force (mN)

U Contraction (mN)

IMA

SV

IMA

SV

Control

16.3±1.3

5.0±0.6

53.3±4.7

29.4±4.4

Iberiotoxin

11.7±2.2

6.2±1.0

66.9±10.7

21.4±6.3

TRAM34

14.5±3.7

4.9±0.5

38.4±7.0

21.3±4.8

Apamin

17.5±3.1

4.0±0.9

49.3±6.7

20.4±2.9

TRAM34+Apamin

12.0±1.8

4.8±0.9

47.5±9.5

18.8±2.9

Iberiotoxin+TRAM34+Apamin 18.1±2.5

6.0±2.1

76.3±10.2*

27.5±6.5

*p<0.05 vs. Control. Iberiotoxin: BKCa blocker; TRAM-34: IKCa blocker; apamin: SKCa blocker.

20

Sun et al, 2019 470 471 472

Table 2. EC50 values for acetylcholine in internal mammary artery (IMA) and saphenous vein (SV) with or without pretreatment with different KCa blockers Group

EC50 (-LogM) IMA

SV

Control

6.97±0.19

6.73±0.29

Iberiotoxin

7.02±0.25

7.21±0.38

TRAM34

6.73±0.08

7.13±0.25

Apamin

6.67±0.08

7.35±0.31

TRAM34+Apamin

6.99±0.21

7.13±0.34

Iberiotoxin+TRAM34+Apamin

6.45±0.21

7.21±0.31

473 474

Iberiotoxin: BKCa blocker; TRAM-34: IKCa blocker; apamin: SKCa blocker.

21

Sun et al, 2019 475 476 477

Figure Legends

478

channels, including large-conductance (BKCa, composed of α and β1 subunits),

479

intermediate-conductance (IKCa), and small-conductance KCa (SKCa) channels in internal

480

mammary artery (IMA) (A) and saphenous vein (SV) (B). Comparison between IMA and SV

481

with respect to the mRNA expression level of each KCa subtype (C). Both IMA and SV

482

express all three subtypes of KCa channels. IMA shows more mRNA abundance of the BKCa

483

subtype than the IKCa and SKCa subtypes while in SV the mRNA expression level of the three

484

subtypes are of no statistical difference. At mRNA level, IMA expresses more BKCa subtype

485

than SV. Data are representative of 5 independent experiments. In each experiment, the IMA

486

and the SV samples were taken from the same patient. Values are expressed as fold-difference

487

relative to BKCaα for comparison within the same graft (A & B) and to corresponding KCa

488

subtype in the IMA for comparison between IMA and SV (C). **p<0.01.

489

Figure 2. The protein expression of calcium-activated potassium (KCa) channel subtypes

490

in internal mammary artery (IMA) and saphenous vein (SV). (A) Representative blots of

491

KCa subtypes in IMA and SV, including large-conductance (BKCa, composed of α and β1

492

subunits), intermediate-conductance (IKCa), and small-conductance KCa (SKCa). (B) Protein

493

expression levels of KCa subtypes obtained from 8 independent experiments, each employing

494

IMA and SV samples from the same patient. No significant differences were observed in the

495

overall protein expression level of each KCa subtype between IMA and SV.

496

Figure 3. Distribution of calcium-activated potassium (KCa) channel subtypes in internal

497

mammary artery (IMA) and saphenous vein (SV). (A) Immunohistochemistry images

498

showing the distribution of KCa subtypes including large-conductance (BKCa, composed of α

499

and β1 subunits), intermediate-conductance (IKCa), and small-conductance KCa (SKCa) in both

500

the endothelium and smooth muscle layers in IMA and SV. Positive staining of specific KCa

Figure 1. Comparison of the mRNA expression level of calcium-activated potassium (KCa)

22

Sun et al, 2019 501

subtype was pointed with black arrows. (B) Comparison between IMA and SV with respect to

502

the overall histoscore (H-score) of each KCa subtype in both endothelial and smooth muscle

503

cells (EC+SMC), which shows no significant differences. (C) & (D) Comparison of the

504

H-score of each KCa subtype between the endothelial layer and the smooth muscle layer in

505

IMA (C) and SV (D). BKCa is more abundantly expressed in SMC than in EC in the IMA

506

while SKCa exhibits more abundance in the endothelial layer (C). SV exhibited no significant

507

differences in the KCa distribution in the endothelial and the smooth muscle layer (D). (E)

508

Comparison between IMA and SV of the H-score of each KCa subtype in endothelial cells (EC)

509

and smooth muscle cells (SMC) respectively shows a significant lower expression of BKCa β1

510

in the smooth muscle layer of the SV. Data were obtained from 5 independent experiments,

511

each employing IMA and SV samples from the same patient. *p<0.05; **p<0.01.

512

Figure 4. Vasorelaxant response mediated by calcium-activated potassium (KCa)

513

channels in internal mammary artery (IMA) and saphenous vein (SV). (A) & (B)

514

Cumulative concentration-response curves of IMA and SV to the large-conductance (BKCa)

515

channel opener NS1619 (A) and the intermediate/small-conductance (IKCa/SKCa) channel

516

opener NS309 (B). (C) & (D) Role of KCa subtypes in IMA and SV in acetylcholine-induced

517

relaxation. The relaxant response to acetylcholine was studied in IMA (C) and SV (D) with or

518

without a pre-treatment with selective channel blockers either individually or in combined use.

519

BKCa subtype plays a predominant role in mediating the relaxant response of IMA to

520

acetylcholine whereas its role in SV is minor. IKCa and SKCa subtypes are unessential to the

521

relaxation induced by acetylcholine in both IMA and SV. Iberiotoxin: BKCa blocker;

522

TRAM-34: IKCa blocker; apamin: SKCa blocker. *p<0.05, **p<0.01, vs. Control. For each

523

group, data represents 8 independent experiments.

524

Figure 5. Schematic summary of the expression and distribution profile of

525

calcium-activated potassium (KCa) channel family in coronary artery bypass grafts 23

Sun et al, 2019 526

(CABG) and their role in the dilatory function of the grafts.

527

intermediate-, and small-conductance KCa; EC: endothelial cell, SMC: smooth muscle cell.

BKCa, IKCa, and SKCa: large-,

528 529

Video: a videoclip depicting the aim, methods, results, and clinical implication of the

530

study on calcium-activated potassium (KCa) channel family in coronary artery bypass

531

grafts.

24

JTCVS-19-1932R2

Supplementary Material Materials and Methods Q-PCR Extraction of total RNA from IMA and SV segments using TRIzol reagent (ThermoFisher Scientific, USA) was performed according to manufacturer's instructions. Reverse transcription and quantitative PCR amplification were performed in LightCycler 96 (Roche, Germany) employing TransScript Green Two-Step qRT-PCR SuperMix system (Transgen, China) under optimal PCR cycle conditions: 94°C for 30s, 45 cycles of 5s at 94°C, 50°C for 15s, 72°C for 10s, and melting at 95°C for 10s, 65°C for 10s and 97°C for 1s. The following primers were used for amplification: BKCaα, 5′-ACAGCATCGGAGTCTTG-3′ (forward)

and

5′-GTCATCATCATCGTCTTGG-3′

(reverse);

BKCaβ1,

5′-GACCAGAACCAGCAGTG-3′ (forward) and 5′-GGAGAAGCAGTAGAAGACC-3′ (reverse);

SKCa

(SK2.3)

5′-TGGAGAAGCAGATTGG-3′

(forward)

and

5′-GATGATGGCAGACAGG-3′ (reverse); IKCa, 5′-AGCCTGGATGTTCTAC-3′ (forward) and 5′-ATGGAGTTCACTTGTTC-3′ (reverse). GAPDH was amplified in parallel as an internal loading control with the primers 5′-CATCCCTGCCTCTACTG-3′ (forward) and 5′-GCTTCACCACCTTCTTG-3′ (reverse). Q-PCR was performed in triplicate for each gene. The Ct (threshold cycle) values were calculated and statistically evaluated by SPSS (Version 20, IBM, USA). Expression of the target mRNAs was normalized to GAPDH levels and relative differences were determined using the comparative Ct (∆∆Ct) method, and fold expression was calculated as 2–∆∆Ct, where ∆∆Ct represents ∆Ct values normalized with the mean ∆Ct of control samples. Western blot Whole cell protein of IMA and SV segments were extracted with RIPA buffer containing

protease inhibitor (Solarbio, China). Protein extracts were determined for concentration by bicinchoninic acid assay, aliquoted, and fractionated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (40µg/lane), followed by electro-transferred to polyvinylidene difluoride (PVDF) membrane (Thermo Fisher Scientific, USA) for detection of the protein of interest. Details of the procedures were published elsewhere28, 29. The PVDF membrane was probed overnight at 4°C with primary antibodies (Abcam, USA) against the protein of interest, including BKCaα (1:1000), BKCaβ1 (1:200), KCa2.3 (1:500), and KCa3.1 (1:500), followed by horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG or horse anti-mouse IgG secondary antibodies (1:3000, Cell signaling, USA) for 1h at room temperature. β-actin (1:2000, Absin) was used as internal loading control. Color development was performed with electrochemiluminescence kit (Beyotime, China), followed by imaging using G:BOX gel doc system (Syngene, UK), and analyzing by Quantity One imaging system (Version 4.6.6, Bio-Rad). Histology and immunohistochemistry staining IMA and SV segments were fixed in 4% paraformaldehyde, decalcified, dehydrated, and embedded in paraffin. The embedded tissue samples were sliced into 4-5 µm sections and stained with haematoxylin and eosin after deparaffination. For immunohistochemistry staining, slides were heated at 100°C in 10 mmol/L tannic acid/1 mmol/L EDTA solution (ZSGB-BIO, China) to retrieve antigens. Endogenous peroxidase activity was quenched by a 10 min-incubation with hydrogen peroxide (ZSGB-BIO, China) (3% in PBS). Sections were incubated for 60 min at 37°C with primary antibodies against BKCaα (1:4000), BKCaβ1 (1:500), KCa2.3 (1:800), and KCa3.1 (1:800) and followed by 30-min incubation with goat anti-rabbit IgG HRP-conjugated secondary antibody at room temperature for signal detection of KCa subtypes. Afterward, the specimens were washed in PBS and stained with a liquid 3,3'-diaminobenzidine (DAB) peroxidase substrate

kit (ZSGB-BIO, China). Counterstaining was performed with Mayer’s haematoxylin (ZSGB-BIO, China). Negative controls were immunostained without the primary antibody. The immunostained images were captured using a microscope (Olympus BX43, Japan). Immunostaining was evaluated by two independent pathologists using a blind protocol design. For each specimen, the histoscore (H-score) of each KCa subtype was calculated as the sum of staining intensity (0-3, negative staining: 0; weak staining: 1; moderate staining: 2; and strong staining: 3) multiplied by the percentage of stained cells (0-100%). Isometric force study IMA and SV segments were cut into 3-mm rings and mounted in a four-channel Mulvany myograph (Model 620M, J.P.Trading, Aarhus, Denmark). The details of myograph experiment have been extensively published in our previous studies28-30. Cumulative dose-response curves to (a) BKCa activator NS1619 (-9~-4.5 Log M), (b) IKCa/SKCa activator NS309 (-9~-4.5 Log M), and (c) acetylcholine (-10~-4.5 Log M) were established in IMA and SV rings precontracted by U46619. In the study of acetylcholine-induced relaxation, vessels were incubated with the specific blocker for BKCa (iberiotoxin, 100 nmol/L), IKCa (TRAM34, 1 µmol/L), or SKCa (apamin, 100 nmol/L) prior to U46619-precontraction. Relaxant responses of IMA and SV to acetylcholine were also studied after combined application of TRAM34 and apamin, or TRAM34 and apamin plus iberiotoxin.