Homocysteine impairs the anticontractile/vasorelaxing activity of perivascular adipose tissue surrounding human internal mammary artery

Abstract

Open in new tabDownload slide

OBJECTIVES

Perivascular adipose tissue (PVAT) surrounding human internal mammary artery (IMA) possesses anticontractile property. Its function under pathological conditions is barely studied. We previously reported that homocysteine impairs the vasodilator function of IMA through endothelium and smooth muscle-dependent mechanisms. This study investigated the effect of homocysteine on the function of PVAT and the associated mechanisms.

METHODS

Residual IMA tissues were collected from patients undergoing coronary artery bypass grafting. Vasoreactivity was studied using myograph. Adiponectin was measured by ELISA. Expressions of adiponectin receptors (AdipoRs), eNOS and p-eNOS were determined by RT-qPCR and Western blot.

RESULTS

Exposure to homocysteine augmented the contractile responses of PVAT-intact IMA to U46619 and potassium chloride, regardless with or without endothelium. Such augmentation was also observed in skeletonized IMA with transferred, homocysteine-exposed PVAT. Homocysteine attenuated the relaxant response of PVAT-intact while endothelium-denuded vessels to acetylcholine. Homocysteine lowered adiponectin content in the PVAT, downregulated the expression of AdipoR1 and AdipoR2 as well as eNOS and p-eNOS in skeletonized IMA. The relaxant response of skeletonized IMA to AdipoR agonist AdipoRon was blunted by homocysteine or eNOS inhibitor, and homocysteine significantly attenuated the inhibitory effect of eNOS inhibitor on AdipoRon-induced relaxation.

CONCLUSIONS

Homocysteine impairs the anticontractile/vasorelaxing activity of PVAT surrounding the IMA through inhibiting adiponectin/AdipoR/eNOS/nitric oxide signalling pathway.

INTRODUCTION

Perivascular adipose tissue (PVAT) was once considered a non-physiological supporting scaffold for blood vessels. Later studies revealed that PVAT may secret various adipokines, cytokines and gaseous molecules such as adiponectin, leptin, resistin, angiotensin 1–7, nitric oxide (NO) and hydrogen sulphide to actively regulate vascular tone and exert antiatherogenic effects [1, 2]. PVAT dysfunction is implicated in the development of atherosclerosis, hypertension and ischaemic heart disease, which makes it a novel predictor for prognosis and potential therapeutic target for cardiovascular disease [3, 4].

Abnormally raised plasma homocysteine level (>15 μmol/l), called hyperhomocysteinemia, is significantly associated with an increased risk of coronary atherosclerotic heart disease, hypertension, myocardial infarction, etc. [5, 6]. Studies concerning the unfavourable effect of homocysteine on vascular function demonstrated the role of overproduction of reactive oxygen species, inhibition of endothelial nitric oxide synthase (eNOS) /NO pathway and promotion of smooth muscle cell proliferation [7, 8]. We previously reported that endoplasmic reticulum stress-mediated inhibition of calcium-activated potassium (K_Ca) channels, i.e. intermediate- and small-conductance K_Ca (IK_Ca and SK_Ca) channels in endothelial cells and large-conductance K_Ca (BK_Ca) channels in smooth muscle cells, is involved in homocysteine-induced dilatory dysfunction of porcine coronary arteries [9, 10]. Compared with the growing knowledge of the effect of homocysteine on vascular endothelium and smooth muscle, whether homocysteine affects PVAT function remains unknown.

Internal mammary artery (IMA) is one of the most commonly used grafts for coronary artery bypass grafting (CABG) in patients with ischaemic heart disease. As an arterial graft, IMA is tended to develop spasm during surgical dissection and the perioperative period. Previously, in PVAT-removed IMA segments, we demonstrated that homocysteine impairs the vasodilator function of IMA by reducing NO bioavailability [11] and lowering the expression of BK_Ca channel β1-subunit thus compromising BK_Ca channel-mediated smooth muscle relaxation [12]. To date, studies on the regulatory effect of PVAT on IMA function are limited, and there is a lack of studies concerning its functionality under pathological conditions. Malinowski et al. [13] reported that PVAT of IMA may release a NO and prostacyclin-independent anticontractile factor that acts with K_Ca channels to modulate the tone of IMA. In the PVAT of IMA isolated from obese patients, Cybularz et al. [14] observed increased gene expression of adiponectin, which they thought may explain the maintenance of endothelial function in obese patients with a body mass index (BMI) of above 30 kg/m². There are no studies yet addressing the role of PVAT in homocysteine-induced vasoreactivity change of the IMA.

Adiponectin receptor (AdipoR) 1 and 2 are homoreceptors for adiponectin [15]. Adiponectin deficiency occurs in cardiovascular disorders such as arterial calcification, atherosclerosis and obesity, along with down-expression of AdipoR1 and/or AdipoR2 [15, 16]. Plasma adiponectin level was found to be inversely correlated with plasma homocysteine level in hypertensive patients [17]. Whether homocysteine influences adiponectin production in PVAT and whether the receptor of adiponectin in the vascular wall is affected by homocysteine remain unknown.

Therefore, in this study, we aimed to investigate the regulatory effect of PVAT on IMA vasoreactivity, with focus on the influence of homocysteine and the underlying mechanisms.

MATERIALS AND METHODS

Ethical statement

The study protocol was approved by the Institutional Ethics Review Board of TEDA International Cardiovascular Hospital ([2021]-0617-1).

Experimental protocols

The remnant IMA tissues were collected with consent from patients undergoing CABG (n = 87). Fresh IMA tissues were then prepared as follows for studying: (i) regulatory effect of PVAT on IMA vasoreactivity and the influence of homocysteine, by myograph experiments (groups Ia, Ib and Ic), and (ii) the role of adiponectin/AdipoR/eNOS/NO in homocysteine-induced functional change of PVAT, by ELISA measurement of adiponectin in PVAT (group IIa), RT-qPCR and Western blot analysis of AdipoR, eNOS and p-eNOS expressions in IMA (groups IIb and IIc), and myograph study of the vasorelaxant response of IMA to the AdipoR agonist in the absence or presence of eNOS inhibitor (group IIb).

Group Ia: PVAT+E+ IMA: IMA rings with intact endothelium and PVAT.

Group Ib: PVAT+E− IMA: IMA rings with intact PVAT but without endothelium.

Group Ic: IMA with transferred PVAT: skeletonized IMA rings with isolated (transferred) PVAT.

Group IIa: PVAT: PVAT removed from IMA.

Group IIb: PVAT−E+ IMA: IMA rings with intact endothelium but without PVAT.

Group IIc: PVAT−E− IMA: IMA rings with neither endothelium nor PVAT.

The IMA segments (groups Ia, Ib, IIb and IIc) and the PVAT (group IIa) from each individual patient were divided into 2 parts and randomly allocated to the control group and the homocysteine-exposed group, respectively. For group Ic, PVAT isolated from each IMA was evenly divided into 2 parts (∼200 mg each) and treated with or without homocysteine. The PVAT was then transferred to the myograph chamber with untreated IMA (without PVAT) to create the condition: control IMA + control PVAT and control IMA + homocysteine-exposed PVAT. For homocysteine exposure, the IMA ring or PVAT was incubated with 100 μmol/l homocysteine (Sigma-Aldrich, Co., St Louis, MO, USA) in Dulbecco's Modified Eagle’s Medium (Gibco, USA) containing 1% penicillin/streptomycin (Solarbio, Beijing, China) at 37°C in a 5% CO₂ incubator for 24 h. The control ring or PVAT was incubated in Dulbecco's Modified Eagle’s Medium containing equal volume of vehicle and antibiotics.

Myograph study

The IMA rings were mounted, equilibrated and normalized in a four-channel Mulvany myograph (Model 610M, J.P. Trading, Aarhus, Denmark) as previously described [9–12].

Vasoconstriction

Cumulative dose–response curves for U46619 (−11 to −4.5 LogM) (Cayman Chemical, Michigan, USA) and potassium chloride (KCl) (5–120 mmol/l) (Sigma-Aldrich) were constructed in PVAT+E+ (group Ia) and PVAT+E− (group Ib) IMA rings with or without homocysteine exposure. The U46619- and KCl-induced vasoconstriction was further studied in ‘control IMA + control PVAT’ and ‘control IMA + homocysteine-exposed PVAT’ preparations (group Ic). To ensure the comparability of the relaxant response between groups, similar extent of precontraction was achieved by using varied concentrations of U46619 ranging from −7.2 to −6 LogM.

Vasorelaxation

Cumulative dose–response curves for acetylcholine (ACh) (−10 to −4.5 LogM) (group Ib) and AdipoRon (−8 to −4 LogM) (MCE, New Jersey, USA) (group IIb) were established in control and homocysteine-exposed IMA rings preconstricted with U46619. The relaxant response to AdipoRon was further studied in the presence of eNOS inhibitor L-NNA (300 μmol/l).

Real-time Quantitative Polymerase Chain Reaction (RT-qPCR)

Total RNA extraction and cDNA synthesis were carried out as described previously [11]. Quantitative PCR amplification was performed with TransStart Tip Green qPCR SuperMix (TransGen, Beijing, China) in LightCycler 96 system (Roche, Basel, Switzerland) under optimal PCR cycle conditions: 95°C for 2 min, 40 cycles of 95°C for 15 s and 60°C for 1 min. Primers used were: AdipoR1: 5′-TTCTTCCTCATGGCTGTGATGT-3′ (forward), 5′-AAGAAGCGCTCAGGAATTCG-3′ (reverse); AdipoR2: 5′-ATAGGGCAGATAGGCTGGTTGA-3′ (forward), 5′-GGATCCGGGCAGCATACA-3′ (reverse); and GAPDH: 5′-CATCCCTGCCTCTACTG-3′ (forward), 5′-GCTTCACCACCTTCTTG-3′ (reverse). Fold changes of AdipoR1 and AdipoR2 were determined using the 2^−ΔΔCT method and normalized with GAPDH.

Western blot

Details of the Western blot procedures were published elsewhere [9–12]. Protein samples (50 μg/lane) were fractionated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) for 90 min at 110 V, electro-transferred to polyvinylidene difluoride membrane for 70 min at 120 V and then blocked with 5% non-fat milk for 2 h at room temperature, followed by incubation with AdipoR1 (1:1000) or AdipoR2 (1:200) primary antibody (Santa Cruz Biotechnology, Dallas, TX, USA) overnight at 4°C. After washed with 1× TBST for 3 times, the membrane was incubated with anti-mouse or anti-rabbit IgG, HRP-linked antibody (1:2000, CST, Danvers, MA, USA) for 2 h at room temperature. GAPDH (1:1000; CST, Danvers, MA, USA) was used as the internal loading control. Colour development was performed with BeyoECL Plus kit (Beyotime, Shanghai, China), imaged using G: BOX gel doc system (Syngene, Cambridge, UK) and analysed by ImageJ software.

Adiponectin measurement

PVAT isolated from IMA (100 mg) was homogenized in 1 ml of phosphate-buffered solution (PBS) and centrifuged for 5 min at 5000 × g at 4°C. The upper layer grease and the bottom sediment were removed, and the clear liquid in the intermediate layer was retained for assay immediately using ELISA kit (Cusabio, Wuhan, China). Adiponectin level was calculated from the corresponding standard curve and expressed as ng/µg protein.

Statistical analysis

Relaxation was expressed as the percentage decrease in isometric force induced by U46619. Shapiro–Wilk test was performed for normality assessment. Data were expressed as mean (SD) and analysed with paired Student’s t-test and two-way repeated measures analysis of variance (two-way repeated measures ANOVA) when appropriate. P < 0.05 was considered statistically significant. All statistical analyses were performed with Prism 8.0.2 software.

RESULTS

Homocysteine impairs the regulatory effect of perivascular adipose tissue on IMA vasoreactivity

Homocysteine attenuates the anticontractile activity of perivascular adipose tissue

In PVAT-intact IMA with or without endothelium, no changes in resting force were observed after exposure to homocysteine (Table 1). Homocysteine exposure significantly strengthened the vasoconstriction response to U46619 in IMAs that have intact PVAT and endothelium (PVAT+E+) [P < 0.001, two-way repeated measures ANOVA; Rmax: 98.73 (35.62) vs 62.96 (19.90) mN in control, P = 0.044] (Fig. 1A and Table 1). The contractile response induced by KCl in PVAT+E+ rings was also enhanced by homocysteine exposure [P = 0.001, two-way repeated measures ANOVA; Rmax: 46.37 (18.81) vs 27.25 (16.56) mN in control, P = 0.011] (Fig. 2A and Table 1).

Further experiments with PVAT-intact but endothelium-denuded IMAs (PVAT+E−) showed that homocysteine also significantly augmented the contractile response to both U46619 [P < 0.0001, two-way repeated measures ANOVA; Rmax: 123.30 (18.42) vs 102.60 (18.45) mN in control, P = 0.016] (Fig. 1B and Table 1) and KCl [P < 0.0001, two-way repeated measures ANOVA; Rmax: 64.28 (17.03) vs 41.81 (18.71) mN in control, P = 0.001] (Fig. 2B and Table 1). As compared to PVAT+E+ IMA segments, PVAT+E− segments showed greater contractile force in response to U46619 and KCl both in control and homocysteine-exposed condition.

The ‘PVAT transfer’ experiment showed that U46619 and KCl elicit greater contractile response in the ‘IMA + homocysteine-exposed PVAT’ group than in the ‘IMA + control PVAT’ group, which further suggests that homocysteine compromises the anticontractile function of PVAT (Fig. 3 and Table 1).

Homocysteine compromises perivascular adipose tissue-mediated relaxation in IMA

ACh, a well-known endothelium-dependent vasodilator, can still elicit a certain degree of relaxant response in the IMA without endothelium but with PVAT. The maximal relaxation induced by ACh was 18.59 (8.09)% in the PVAT+E− IMA. Homocysteine significantly suppressed PVAT-mediated relaxation (P < 0.0001, two-way repeated measures ANOVA). Both the response magnitude [Rmax: 8.28 (4.04)%, P = 0.002 versus control] and the sensitivity to ACh were decreased [EC50: −6.10 (0.61) vs −6.79 (0.54) LogM in control, P = 0.025] in PVAT+E− IMAs underwent homocysteine exposure (Fig. 4 and Table 2).

Homocysteine impairs perivascular adipose tissue function through adiponectin/AdipoR/eNOS/NO pathway

Homocysteine attenuates the relaxant response of IMA to AdipoRon

AdipoRon, an active agonist for AdipoR1 and AdipoR2, evoked dose-dependent relaxation in skeletonized (PVAT−E+) IMA rings, which was blunted by homocysteine exposure (P < 0.001, two-way repeated measures ANOVA). The maximal relaxation decreased from 65.14 (12.97)% to 45.69 (15.85)% (P = 0.011) (Fig. 5A and Table 2), though the sensitivity to AdipoRon remained unaltered [EC50: −5.30 (0.43) vs −5.35 (0.56) LogM in control, P = 0.79] (Table 2). Pretreatment with eNOS inhibitor L-NNA attenuated AdipoRon-induced relaxation in skeletonized IMAs and the inhibitory effect of L-NNA was blunted in vessels subjected to homocysteine exposure (Fig. 5B and C and Table 2). These results suggest that NO acts downstream of AdipoR to mediate IMA relaxation and homocysteine-inhibited AdipoR/eNOS/NO signalling.

Homocysteine inhibits adiponectin production in the perivascular adipose tissue of IMA

Adiponectin content in homocysteine-exposed PVAT was significantly lower than that in the PVAT without homocysteine exposure, indicating the inhibitory effect of homocysteine on adiponectin production in PVAT (Fig. 6A).

Homocysteine inhibits the expression of AdipoR, eNOS and p-eNOS in IMA

Both AdipoR1 and AdipoR2 in IMA (PVAT−E+ and PVAT−E−) were significantly decreased after incubation with homocysteine, indicating that homocysteine inhibits the expression of AdipoR in the vascular wall of IMA (Fig. 6B–D). Meanwhile, homocysteine lowered eNOS and p-eNOS protein levels in the vascular wall of IMA (PVAT−E+) (Fig. 6D), which explains, at least partially, the loss of the inhibitory effect of L-NNA on the relaxation in response to AdipoR activation (Fig. 5C).

DISCUSSION

To our knowledge, this is the first study investigating the effect of homocysteine on PVAT function. We demonstrated that homocysteine compromises the anticontractile/vasorelaxing activity of PVAT surrounding the IMA. A decrease in adiponectin production in PVAT and inhibition of the expression and activity of AdipoR in the vascular wall of IMA contribute to homocysteine-induced impairment of PVAT function.

IMAs used in this study were from 87 CABG patients (Supplementary Material, Table S1) who have no history of diabetes mellitus, 67.82% of them are hypertensive and 19.54% are dyslipidaemic. Half of them (50.57%) are current smokers. Nearly all of the patients received lipid-lowering, antithrombotic and coronary dilating drugs. Taking into account the possible impact of different clinical characteristics on PVAT function, we used ‘paired design’ to rule out the influence posed by risk factors and pharmacotherapies on data interpretation. In each experiment, the sample (IMA segments with surrounding PVAT, or PVAT isolated from IMA) taken from each individual patient was divided into 2 parts and incubated with or without homocysteine. In this way, we were able to assess the effect of homocysteine per se on the functionality of PVAT, saying, release of vasoactive substances and control of the vascular tone of IMA. The basal characteristics of the patients enrolled in each experiment are shown in Supplementary Material, Table S2.

We observed that in PVAT+E+ IMA, homocysteine exposure results in contractile augmentation. Such augmentation was also observed in PVAT+E− IMA exposed to homocysteine, and in skeletonized IMA ‘transplanted’ with homocysteine-exposed PVAT, which collectively indicate impairment of the anticontractile activity of PVAT by homocysteine. The greater contractility of PVAT+E− IMA compared to PVAT+E+ IMA is in support of the significance of endothelium in the tone regulation of IMA. Meanwhile, homocysteine significantly suppressed the relaxant response of PVAT+E− IMA to ACh. Taken together, the increased contractility and decreased relaxation of PVAT+E− IMA may suggest a loss of PVAT-derived relaxing factors caused by homocysteine. This was confirmed by the finding of decreased adiponectin content in the PVAT following homocysteine incubation. Previous studies in various vascular beds including human vessels have attributed the vasorelaxing action of adiponectin to eNOS phosphorylation promotion [18] and hyperpolarization resulting from BK_Ca channel opening [19]. We previously demonstrated that homocysteine impairs the vasodilator function of skeletonized IMA by reducing eNOS and NO bioavailability [11] and lowering the expression of BK_Ca channel β1-subunit thus compromising BK_Ca channel-mediated smooth muscle relaxation [12]. The effector role of eNOS and BK_Ca channels in the action of adiponectin and the inhibition of eNOS and/or BK_Ca channels by homocysteine therefore may provide a mechanistic explanation for the compromised anticontractile/relaxing activity of PVAT due to less adiponectin release following homocysteine exposure. It is worth to mention that although adiponectin is released primarily by adipose tissue, non-adipose sources especially endothelium can also produce adiponectin [20]. Being aware of this, in this study, we isolated PVAT from the IMA and only the PAVT was used for adiponectin measurement; thus, the influence of homocysteine on PVAT adiponectin release could be ascertained. Additional experiments using cysteine, a metabolic product of homocysteine, showed no effect on adiponectin level and vasocontractility of PVAT-intact IMA, further suggesting the significance of homocysteine in the impairment of PVAT function (Supplementary Material, Fig. S1).

In addition to the reduction of adiponectin release from PVAT, we observed a significant decrease in AdipoR in the vascular wall of IMA. Expressions of both AdipoR1 and AdipoR2 were downregulated in PVAT−E+ and PVAT−E− IMAs that were exposed to homocysteine, along with a decline in receptor activity, as evidenced by the attenuated relaxant response of the vessels to the AdipoR agonist AdipoRon. Binding of adiponectin to AdipoR1 and AdipoR2 is known to trigger Ca²⁺ influx in endothelial cells and activate multiple downstream targets including AMPK and PI3K/Akt to promote NO production [21]. In arterioles collected from patients with coronary artery disease, the restoration of NO-mediated flow-induced dilation achieved by adiponectin administration was found to be abolished by AdipoR1 knockdown [22]. Together these data highlight the role of AdipoR in adiponectin-induced NO signalling. In this study, we observed that inhibition of eNOS significantly attenuates the relaxant response of IMA to AdipoRon and such inhibitory effect is blunted in homocysteine-exposed vessels. These data in conjunction with the findings on the downregulation of eNOS and p-eNOS in the vascular wall of IMA suggest that eNOS/NO acts downstream of AdipoR and homocysteine impairs PVAT function through inhibiting adiponectin/AdipoR/eNOS/NO signalling pathway.

It has been shown that preservation of PVAT in the venous graft for CABG can significantly increase long-term graft patency. Saphenous vein graft (SVG) harvested using ‘no-touch’ technique, in which PVAT is retained around the vein, showed superior graft patency compared with conventional preparation after 16 years of follow-up [23]. A recent study revealed that PVAT surrounding SVG expresses eNOS and a high level of arginosuccinate synthase 1, a rate-limiting enzyme in the citrulline–arginosuccinate–arginine cycle, which thereby continuously provides the substrate for NO synthesis. The higher NO production in the PVAT-preserved SVG than the PVAT-removed SVG is thought to greatly contribute to the superior long-term patency of the SVG harvested via ‘no-touch’ technique [24]. For CABG using arterial grafts, mainly IMA and radial artery, skeletonization technique has been popularized to improve the results of arterial revascularization. However, since PVAT surrounding IMA and radial artery exhibits anticontractile/vasorelaxing properties [13, 25, 26], and IMA PVAT shows a protective phenotype from metaflammation that likely contributes to the resistance of IMA to atherosclerosis [27], it may provoke a discussion that if preserving the PVAT could help prevent postoperative vasospasm of the arterial grafts and potentially improve their physiological properties. In the present study, we demonstrated that homocysteine impairs the anticontractile/vasorelaxing activity of PVAT surrounding the IMA, together with our previous findings of homocysteine-induced dilator dysfunction of skeletonized IMA, we suggest that attention should be paid to CABG patients with hyperhomocysteinemia, who may be more prone to develop postoperative IMA spasm. Previous studies have associated hyperhomocysteinemia with the development and severity of coronary artery disease [6, 28] and proved it to be a significant predictor for immediate postoperative adverse events after CABG [29]. These reports, in conjunction with our findings on the detrimental effect of homocysteine on IMA function, suggest the necessity of taking precautions and optimizing management strategies for coronary artery disease patients with hyperhomocysteinemia who are high-risk subsets.

Limitations

We are aware that this study has limitations. First, we demonstrated that homocysteine inhibits the expression of AdipoR1 and AdipoR2 in the vascular wall of IMA but did not further clarify the localization and distribution of these AdipoR and the influence of homocysteine. The respective role of AdipoR1 and AdipoR2 in homocysteine-induced PVAT dysfunction needs to be elucidated in future studies. Second, the unfavourable effect of homocysteine on IMA PVAT function was demonstrated using an in vitro incubation model. Given that patients with coronary artery disease have a certain incidence of hyperhomocysteinemia, it is likely that some of the IMAs had already been subjected to high concentration of homocysteine in vivo before the in vitro exposure. Since CABG patients are not routinely checked for homocysteine level in our hospital, we were unable to know the plasma homocysteine status of these IMA donors. Although with the paired design strategy we minimized the impact of plasma homocysteine status on results interpretation, further study comparing the PVAT function of IMA between hyperhomocysteinemic and non-hyperhomocysteinemic CABG patients shall strengthen the significance of the findings derived from the present work.

CONCLUSION

In conclusion, the present study demonstrated that homocysteine impairs the anticontractile/vasorelaxing activity of PVAT surrounding the IMA through inhibiting adiponectin/AdipoR/eNOS/NO pathway.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

This study was supported by the National Natural Science Foundation of China (81870227), Tianjin Science and Technology Committee (20JCZDJC00510), Key Medical Program of Tianjin Binhai New Area Health Bureau (2018BWKZ005) and Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-020A).

Conflict of interest: none declared.

DATA AVAILABILITY

Data underlying this study will be shared on reasonable request to the corresponding author.

Author contributions

Jia-Hui Wei: Data curation; Formal analysis; Investigation; Writing—original draft. Hang Qi: Investigation. Yang Zhou: Investigation. Hai-Tao Hou: Investigation. Guo-Wei He: Resources; Supervision. Qin Yang: Conceptualization; Data curation; Funding acquisition; Supervision; Visualization; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Marco Damiano Pipitone, Zsuzsanna Arnold, Tadashi Kaname and the other anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• ACh

  Acetylcholine

 
• AdipoR

  Adiponectin receptor

 
• CABG

  Coronary artery bypass grafting

 
• IMA

  Internal mammary artery

 
• K_Ca

  Calcium-activated potassium

 
• KCl

  Potassium chloride

 
• NO

  Nitric oxide

 
• PVAT

  Perivascular adipose tissue

 
• SVG

  Saphenous vein graft