The role of NO, H₂O₂, and non-NO/H₂O₂ mechanisms in acetylcholine (ACh)-induced dilation of human arterioles in the absence and presence of coronary artery disease

Abstract

This study defines the mechanisms of vasodilation to acetylcholine (ACh) in arterioles from patients with and without coronary artery disease (CAD). Human adipose arterioles (HAA) dissected from discarded surgical samples were cannulated and pressurized at 60 mmHg for measurement of diameter changes by videomicroscopy. No difference in baseline dose response to ACh was observed between non-CAD and CAD patients. L-NAME, NO synthase inhibitor, reduced dilation in both groups but to a greater extent in non-CAD. Peg-CAT, H₂O₂ scavenger, attenuated response to ACh in non-CAD but not in CAD. Inhibition of NOX4 reduced dilation in non-CAD, whereas NOX2 inhibition attenuated dilation in CAD. The SOD mimetic tempol partially normalized the NO- and H₂O₂-dependent dilation in CAD arterioles. EPR spin trapping indicated that absolute NO signal after ACh+A-23187 stimulation was higher in non-CAD than in CAD arteries. Western blot analysis revealed higher expression of monomeric eNOS but lower expression of dimeric eNOS and phosphorylated eNOS at Ser-1177 in CAD arteries. Finally, we found higher mRNA and protein expression of NOX2 in CAD arteries. These results provide new evidence that in normal human arterioles, both NO and H₂O₂ significantly contribute to ACh dilation, with substantial involvement of NOX4 in the H₂O₂-mediated response. In CAD, the contribution of both NO and H₂O₂ is diminished while an NO/H₂O₂-independent hyperpolarizing pathway becomes predominant. Mechanistically, the NOX4-to-NOX2 switch may play a key role in mediating the change of vasodilator mechanisms in human arterioles during CAD.

Introduction

The vascular endothelium plays a crucial role in regulating the tone and reactivity of underlying smooth muscles through various vasoactive factors, such as nitric oxide (NO), prostacyclin (PGI₂), and endothelium-derived hyperpolarization (EDH) factors [51]. These factors can be released in response to physical and chemical cues, including shear stress and acetylcholine (ACh), a commonly used parasympathetic vasodilator [47]. Through activating Gq-coupled subtypes of muscarinic receptors, e.g., M3 receptors widely expressed in the vascular endothelium, ACh has been shown to stimulate both Ca²⁺ release from an intracellular store and transmembrane Ca²⁺ influx in endothelial cells [47, 54]. A subsequent elevation of endothelial intracellular Ca²⁺ concentration then serves as a crucial signaling step leading to the release of endothelium-derived relaxing factors, i.e., NO, PGI_2, and EDH factors [10, 47, 51]. Impairment of ACh-induced vasodilation, especially NO-mediated dilation in large arteries, is often used as a surrogate for assessing endothelial dysfunction in a broad range of vascular diseases, such as hypertension, diabetes, and coronary artery disease (CAD) [17, 20].

Notably, the contribution of specific endothelium-derived relaxing factors to ACh-induced dilation depends on species, vessel size, and (patho)physiological conditions: in general, NO mainly contributes to ACh-induced dilation in relatively large arteries, while its role diminishes in small arteries and under disease conditions, where the involvement of EDH factors in the dilation increases [44]. Several candidates for EDH factors have been proposed, including hydrogen peroxide (H₂O₂) [36, 37] and metabolites of arachidonic acid [10]. Since previous studies have been primarily performed in vessels isolated from various animal species, the specific involvement of NO, PGI₂, and EDH factors in human arterioles remains less well understood.

Reactive oxygen species (ROS), including H₂O_2, have often been considered as damaging factors that cause oxidative stress and endothelial dysfunction. However, there is strong evidence that H₂O₂ at low physiological concentrations can regulate multiple vascular functions, from angiogenesis, endothelial apoptosis, and endothelial barrier function, to vascular inflammatory response [9]. As a diffusible vasodilator, H₂O₂ has also been shown to act as an EDH factor to regulate vascular tone in different animal and human vascular beds [36, 37]. We recently reported that NOX2 and NOX4 are the most abundantly expressed NADPH oxidase (NOX) isoforms in human coronary and adipose arterioles, with NOX4 serving as a key source of H₂O₂ in arterioles from patients without CAD [55]. However, it remains largely unexplored whether NOX2 and NOX4 contribute to H₂O₂-mediated or other mechanisms of ACh-induced dilation of human arterioles in health and disease.

In recent years, studies aimed to determine the mechanisms of dilation of the resistance vessels, especially from humans, have gained increasing attention given the proposed critical role of microvascular dysfunction in a growing number of conditions including CAD [46]. Two different approaches have been used to assay microvascular endothelial function in humans: flow-mediated dilation (FMD), triggered by shear stress, and agonist-induced dilation, initiated by receptor stimulation [5, 32, 59]. In the human microcirculation, FMD is predominantly mediated by either NO, PGI₂, or mitochondria-derived H₂O₂, depending on age and health conditions. Specifically, FMD is mediated by PGI₂ in healthy young individuals, NO in healthy adults, and H₂O₂ as a compensatory EDH factor in CAD patients [7, 35, 44, 61]. However, it remains unknown whether a similar switch of vasodilator mechanisms occur for agonist-induced dilation in human arterioles. Given the potentially different signaling mechanisms, including Ca²⁺ signaling and vasodilator factors involved in agonist- and flow-induced dilation [16, 58, 57] a better understanding of the mechanisms of agonist vs. flow-induced dilation may provide critical insight into the pathogenic processes of microvascular dysfunction in patients with CAD or other cardiovascular diseases. Importantly, the vasodilator factors released from endothelial cells can exerts diverse (patho)physiological functions beyond the regulation of microvascular tone: e.g., NO released during FMD diffuses into the surrounding parenchymal tissue, where it suppresses mitochondrial respiration, reduces ROS production, and inhibits inflammation and thrombosis, highlighting the importance of not only maintaining the magnitude of dilation but also the physiological mechanisms of dilation in the microcirculation [24, 25] . To address this gap, the present study aims to determine the vasodilatory mechanisms of ACh-induced dilation in arterioles from subjects without CAD (non-CAD) and whether these mechanisms are altered in subjects with CAD. Due to their capacity to reduce oxidative stress and improve NO bioavailability [26, 41, 56], we further test whether the antioxidants such as tempol and SOD improve endothelial function or alter the mechanisms of endothelium-dependent relaxation in human arterioles. To our knowledge, this represents the first study that addresses the critical question of how the mechanisms of receptor agonist-induced dilation change from health to the presence of CAD in the human microvasculature.

Methods

Human tissue acquisition

Human adipose tissues (pericardial, visceral, subcutaneous) were collected as discarded surgical specimens from patients without diagnosis of coronary artery disease (non-CAD) or with confirmed CAD. The non-CAD group includes patients who had no prior diagnosis of CAD and had ≤ 1 risk factor of CAD (hypertension, hyperlipidemia, obesity, diabetes, and tobacco use). The protocols were approved by the Institutional Review Board of the Medical College of Wisconsin and Froedtert Hospital and in accordance with 1964 Helsinki Declaration. The patients’ demographics were gathered from the REDCap database at the Medical College of Wisconsin and summarized in Table 1.

Table 1.

Clinical characteristics of patients included in this study

	non-CAD (n=107)	CAD (n=76)
Race
Caucasian (n)	80	67
African American (n)	13	5
Asian (n)	5	0
Hispanic (n)	4	1
Other (n)	5	3
Sex
Male (n)	12	63
Female (n)	95	13
Age (year) (mean ± SEM)	43 ± 1.2	67 ± 1.3
Body Mass Index (mean ± SEM)	29.9 ± 0.6	32.5 ± 0.7
Tissue source
Subcutaneous (n)	85 (77)	11 (8)
Visceral (n)	16 (12)	16 (15)
Pericardial (n)	0	43 (33)
No information (n)	6 (0)	6 (1)
History/Risk factors
Hypertension (n)	10	52
Hyperlipidemia (n)	7	45
Smoking (n)	6	10
Chronic heart failure (n)	0	14
Diabetes (n)	5	26
Myocardial infarction (n)	0	17
Valve disease (n)	0	4
Peripheral vascular disease (n)	0	3

Values for age and body mass index (BMI) are shown as mean ± SEM; n indicates the number of patients. Tissue source relates to the origin of the adipose tissue sample. Numbers in parentheses specify the number of tissues used for functional studies (videomicroscopy of isolated arterioles).

Videomicroscopy

Human adipose arterioles (HAA) with internal diameters of 100–250 μm were dissected from adipose tissues. Isolated vessels were carefully removed of surrounding fat and connective tissue, transferred to an organ chamber, and cannulated at both ends with glass micropipettes (inner diameter 35-45 μm). Vessels were pressurized at 60 mmHg in Krebs – physiological salt solution (PSS) composed of (in mM): 123 NaCl, 4.7 KCl, 1.2 MgSO₄, 2.5 CaCl₂, 19 NaHCO₃, 1.2 KH₂PO₄, 0.026 EDTA, and 11 glucose. The solution was consistently gassed with a mixture of 21% O₂, 5% CO₂ and 74% N₂ and kept at 37 °C to maintain pH=7.4. Changes in internal diameter were tracked by videomicroscopy as previously described [18, 40]. Vascular characteristics are shown in Table 2. To study vasodilator response, arterioles were preconstricted by adding 0.5 μL of endothelin-1 (ET-1, 0.05 nmol/L) directly to the bath and waiting 5 min. This process was repeated until 30-50% reduction of the baseline internal diameter was achieved. After ET-1-induced contraction reached steady state, relaxation responses to cumulative concentrations of ACh (10⁻⁹–10⁻⁵ M) were determined in the absence and presence of 30 minutes preincubation with one of the following modulators alone or in combination: NO synthase inhibitor NG-nitro-L-arginine methyl ester (L-NAME, 100 μM), an NO scavenger cPTIO (100 μM), a cyclooxygenase inhibitor indomethacin (Indo, 10 μM), cell-permeable H₂O₂ scavenger peg-catalase (peg-CAT; 500 U/mL) or regular catalase (CAT; 500 U/mL), NADPH oxidase NOX2 inhibitor GSK2795039 (1 μM) and NOX1/4 inhibitor GKT137831 (1 μM), high potassium (K⁺) concentration (60 mM), superoxide dismutase (SOD, 150 U/mL), SOD mimetic tempol (10 μM) or mitochondria-targeted mitoTempol (50 nM). Papaverine (100 μM) was added at the end of each experiment to assess endothelium-independent vasodilation. Only vessels showing ≥90% of their baseline diameter (measured before addition of ET-1 at the beginning of the experiment) in response to papaverine were included in the analysis. The vessel dilation response was calculated: percent maximal dilation = [((diameter with test compound) – (basal diameter))/((maximum diameter) – (basal diameter))] × 100. The percentage of maximal dilation was averaged and plotted using vessels from at least 5 independent subjects. In some experiments, the area under curve (AUC) was calculated, using SigmaPlot (version 15.0), to compare the degree of dilation.

Table 2.

Vascular characteristics

	non-CAD	CAD
Baseline ID	148.0 ± 4.2	149.5 ± 6.1
Max Diameter	149.0 ± 4.2	151.5 ± 6.3
% Tone	0.8 ± 0.2	1.2 ± 0.6

Data presented as mean ± SEM; ID indicates the internal diameter; % tone characterizes the developed tone during pressurizing at 60 mmHg preceding preconstriction with ET-1.

Determination of NO production in human adipose small arteries

Freshly dissected human adipose arteries (350-450 μm) were incubated for 60 min in 1 ml of Krebs/HEPES buffer containing 200 μM iron diethyldithiocarbamate (Fe(DETC)₂) and 1μM of A-23187 (Ca²⁺ ionophore) at 37 °C. Some vessels were also incubated with 10 μM of ACh. After incubation, vessels were gently moved into the middle of 1 mL insulin syringes. In advance, the needle-end was cut from a 0.1 mL syringe barrel, the plunger drawn out 0.1 mL from the cut end, and the syringe was filled with 0.4 mL of Krebs buffer. After putting the vessels syringes were additionally filled with 0.25 mL of Krebs buffer and then frozen in liquid nitrogen and stored at −80 °C until use. The NO-Fe(DETC)₂ complex was detected using EPR that was performed at FRIMCORE, Vanderbilt University, Nashville, TN [14]. EPR samples were placed in quartz Dewar (Corning, New York, NY) filled with liquid nitrogen. EPR spectra were recorded using an EMX EPR spectrometer (Bruker Biospin, Billerica, MA) and a super high Q microwave cavity. The EPR settings were as follows: field sweep, 160 Gauss; microwave frequency, 9.42 GHz; microwave power, 10 mW; modulation amplitude, 3 Gauss; scan time, 150 ms; time constant, 5.2 s; and receiver gain, 60 dB (n = 4 scans). Values were normalized to the lumen area of the vessels.

RNA extraction and reverse transcription-polymerase chain reaction (PCR)

Human adipose arteries (200–400 μm), freshly excised, were rapidly frozen using liquid nitrogen and maintained at −80 °C. Total RNA was extracted utilizing the RNeasy Plus Universal Mini Kit (Qiagen), and cDNA was generated using the SuperScript III reverse transcriptase system (Invitrogen), as previously described [55]. For real-time quantitative PCR (qPCR), reactions were set up in 20 μL volumes containing 2 ng of cDNA, 10 μL of SYBR Green qPCR Master Mix (Qiagen), 1 μL each of forward and reverse primers, and nuclease-free water. Amplification was conducted on a CFX96 C1000 Thermal Cycler, and quantification cycle values were assessed by using CFX Manager version 3.1. H₂O was used for negative control. Primers targeting NOX2 were custom-synthesized by Integrated DNA Technologies Inc. Primer sequences have been described in a previous study of our group [55]. For normalization, 18S rRNA primers (Qiagen,) were used as an endogenous control.

Immunoblotting

Freshly isolated human adipose arteries were homogenized in ice-cold lysis buffer consisting of 1x NP40-SDC-OG Buffer (Pierce 1% NP-40 IP lysis buffer, 0.5% sodium deoxycholate, 1.5% octyl-beta-Glucoside detergent) protease and phosphatase inhibitors and 10 mM N-Ethylmaleimide (Thermo Scientific). Proteins were extracted and analyzed for their total eNOS and its respective monomers and dimers by Western blot as described previously [30]. Further, NOX2 protein expression, were also analyzed from human adipose arteries as described previously [55].

Chemicals

All chemicals were purchased from Sigma-Aldrich company except cPTIO was purchased from Cayman Chemical Company, GSK2795039/GKT137831 from Tocris Bioscience Company, and DD1 donor from Enzo Life Sciences.

Statistical Analysis

All vasodilation data are presented as means ± SEM, except Western blot densitometry, NO measurements, and qPCR bar graphs, which are expressed as means ± SD. For vasodilation experiments, results from independent experiments were averaged, and significant differences in vasodilatory responses were analyzed by using two-way repeated measures ANOVA (factor 1, doses; factor 2, different treatments such as a vehicle or an inhibitor), followed by the Student-Newman-Keuls multiple-comparison test for pairwise comparisons or comparisons versus control. For other studies, statistical comparisons were made by Student’s t-test (either paired or unpaired t-test) or one-way ANOVA. All calculations were made using SigmaPlot (version 15.0). For western blot experiments, shown in Fig 6 and Fig 7, eNOS and NOX2 band densities were normalized to β-actin expression levels in both CAD and non-CAD samples.

Fig. 6. Measurement of NO signals and endothelial NO synthase (eNOS) protein expression in human vessels.

A EPR spectra of adipose arteries incubated for 60 min at 37 °C with 200μM colloid Fe(DETC)₂. a Representative traces. Arteries from non-CAD (blue) and CAD (red) patients stimulated with 1μM calcium ionophore A-23187. b Data are expressed as signal intensity in arbitrary units, normalized to the lumen area of the vessels. Shown as mean values ± SD from n=5 experiments in each group. P=0.0096 A-23187 non-CAD vs A-23187+ACh non-CAD, P=0.018 A-23187 CAD vs A-23187+ACh CAD, P=0.000079 A-23187 non-CAD vs A-23187 CAD, P=0.0023 A-23187+ACh non-CAD vs A-23187+ACh CAD. c Representative Western blots for eNOS dimer, monomer, phosphorylation at serine-1177 (p-eNOS) (d), and total eNOS (e) using a low-temperature gel. f Densitometry values for eNOS dimer/total eNOS ratio, phosphorylated at serine 1177 eNOS (p-eNOS)/total eNOS ratio, and eNOS monomer/total eNOS ratio. n=4 in each group. Differences between means were analyzed using either paired or unpaired t-test (b) or one-way ANOVA (f). *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001

Fig. 7. NOX2 mRNA and protein expression in human adipose arteries in non-CAD and CAD patients.

a qPCR analysis demonstrated that NOX2 mRNA expression was higher in CAD compared to non-CAD patients (n=6/group, P=0.08). b Quantitative analysis of blots showing the enhanced NOX2 protein expression in arteries from CAD patients compared to non-CAD (n=7/non-CAD and n=9/CAD group, P=0.12). c Representative blots of NOX2 in HAA from non-CAD and CAD subjects. Differences between means were analyzed with unpaired t-test (a and b)

Results

The role of NO and PGI₂ in acetylcholine-induced dilation in human adipose arterioles from non-CAD patients

We first examined the role of NO and PGI₂ in ACh-induced dilation in HAA from non-CAD subjects. As shown in Fig 1a, ACh (log 10⁻⁹ – 10⁻⁵ M) induced potent concentration-dependent dilation, with an average maximum dilation of 96.0 ± 1.0% (control, n=6). The preincubation with NO synthase (NOS) inhibitor L-NAME (100 μM) significantly reduced this dilation to the average maximum amplitude of 73.7 ± 3.2% (L-NAME, n=6, *P<0.05). Prior incubation with the cyclooxygenase (COX) inhibitor indomethacin (Indo, 10 μM) only slightly reduced ACh-induced response at log −7 M of ACh (Fig 1b, n=7, *P<0.05). Moreover, the combination of L-NAME and Indo reduced dilation to a similar extent as L-NAME alone, with the average maximum amplitude decreased from 95.5 ± 1.9% to 67.0 ± 6.1% (Fig 1c, n=8, *P<0.05). To test whether the modest effect of L-NAME to inhibit the ACh response is related to the low efficiency of L-NAME to reduce NO formation or due to low NO release by ACh, we further examined the effect of cPTIO, a NO scavenger, on the ACh response. Preincubation with cPTIO (100 μM) attenuated the dilation to the same extent as L-NAME, with the average maximum amplitude reduced from 93.3 ± 3.3% to 71.5 ± 5.5% (Fig S1a, n=6, *P<0.05). The effects of the above inhibitors on the EC50 and maximal dilation of ACh are also provided in Table S1. Thus, the NO component likely plays an important, but not predominant, role in ACh-induced dilation in non-CAD vessels.

Fig. 1. Role of NO, PGI₂, H₂O₂ and EDH in acetylcholine (ACh)-induced dilation in human adipose arterioles (HAA) in non-CAD patients.

a NOS inhibitor (L-NAME, 100 μM) significantly reduced ACh-induced dilation (n=6). b Cyclooxygenase inhibitor Indomethacin (Indo, 10 μM) attenuated ACh-induced dilation in HAA just at log −7 M of ACh (n=7). c Combination use of L-NAME and Indo reduced the ACh-induced dilation to the same extent as L-NAME by itself (n=8). d Preincubation with peg-CAT significantly reduced ACh dilation at log −7 M, −6 M and −5 M of ACh (n=8). e Peg-CAT further reduced ACh-induced dilation at log −7 M and −6 M of ACh in HAA that were previously preincubated with L-NAME and Indo (n=7). f The high potassium (K⁺, 60 mM) significantly diminished remaining dilation at log −7 M, −6 M and −5 M of ACh in preincubated with L-NAME and Indo HAA (n=5). Statistical analysis was performed using repeated measures of two-way ANOVA, followed by Student-Newman-Keuls multiple comparison test for concentration response (a-f). *P <0.05

The role of H₂O₂ and EDH in acetylcholine-induced dilation in human adipose arterioles from non-CAD patients

Next, we used peg-CAT (500 U/mL), a cell-permeable H₂O₂ scavenger, to determine the role of H₂O₂ in ACh-induced dilation in non-CAD patients. The preincubation of vessels with peg-CAT significantly reduced the ACh-induced dilation at log −7 M to −5 M of ACh, with the average maximum amplitude reduced from 95.7 ± 1.9% to 74.6 ± 8.1% (Fig 1d, n=8, *P<0.05). Moreover, in the presence of both L-NAME and Indo, peg-CAT further reduced ACh-induced dilation at log −7 M and −6 M of ACh (Fig 1e, n=7, *P<0.05). Together, these results suggest that H₂O₂ partially mediates ACh-induced dilation in HAA from non-CAD subjects, likely acting through the NO- and PGI₂-independent pathway. Interestingly, a significant portion of ACh-induced dilation remained after inhibiting NO, PGI_2, and H₂O₂ pathways. This remaining dilation was largely abolished in the presence of high potassium (K⁺) concentration (60 mM) in the Krebs solution (Fig 1f, n=5, *P<0.05), suggesting that a significant portion of the remaining dilation after inhibition of NO and PGI₂ is associated with smooth muscle hyperpolarization.

The role of NO and PGI₂ in acetylcholine-induced dilation in human adipose arterioles from CAD patients

There was no significant difference in ACh response in HAA from CAD patients as compared with non-CAD except at log −8 M and −7 M of ACh (Fig 2a, *P<0.05): the average maximum amplitude was 90.1 ± 0.9% in CAD (n=45) and 92.5 ± 1.0% in non-CAD (n=55). In HAA from CAD patients, the combination of L-NAME (100 μM) and Indo (10 μM) reduced the ACh-induced dilation at log −7 M and −6 M of ACh, with the average maximum amplitude of 81.4 ± 3.0% (Fig 2b, n=10, *P<0.05). However, this reduction of the dilation was less compared with non-CAD vessels incubated with L-NAME and Indo. cPTIO attenuated dilation to the same extent as the combination of L-NAME and Indo, with the average maximum amplitude reduced from 86.5 ± 2.1% in control to 73.5 ± 1.7% in HAA preincubated with cPTIO (Fig S1b, n=6, *P<0.05). To further quantitate the contribution of NO and PGI₂ to ACh-induced dilation, we calculated the difference between the area under the curve (AUC) in control and after the preincubation with L-NAME and Indo. As shown in Fig 2c, in HAA from CAD subjects, the NO/PGI₂ contribution to ACh dilation was lower than non-CAD arterioles (Fig 2c, n=8/non-CAD group, n=10/CAD group, *P<0.05).

Fig. 2. Role of NO, PGI_2, H₂O₂, and EDH in acetylcholine (ACh)-induced dilation in human adipose arterioles (HAA) from CAD tissues.

a There was no difference in maximum level of dilation of HAA between non-CAD (n=55) and CAD (n=45) patients. However, in CAD patients the ACh response was reduced at log −8 M and −7 M of ACh. b In HAA from CAD tissues, preincubation with both L-NAME and Indo significantly reduced ACh-induced dilation at log −7 M and −6 M of Ach (n=10). c The NO/PGI₂ contribution to ACh-induced dilation was higher in non-CAD (n=8) patients compared to CAD (n=10, P =0.049). d Peg-CAT reduced dilation only at log −7 M of ACh (n=7). e The H₂O₂ contribution to ACh-induced dilation was also larger in non-CAD (n=8) subjects compared to CAD (n=7, P=0.043). f In the presence of high potassium (K⁺, 60 mM) concentration the ACh-induced dilation was significantly abolished at log −7 M, −6 M and −5 M of ACh (n=5). Statistical analysis was performed using repeated measures of two-way ANOVA, followed by Student-Newman-Keuls multiple comparison test for concentration response (a, b, d and f). Differences between means were analyzed with unpaired t-test (c and e). *P <0.05

The role of H₂O₂ and EDH in acetylcholine-induced dilation in human adipose arterioles in CAD patients

As shown in Fig 1, we found that in HAA from non-CAD tissues, ACh-induced dilation was partially mediated by H₂O_2. The effect of peg-CAT was next examined on HAA from CAD subjects. Preincubation with peg-CAT only slightly reduced ACh-induced dilation at log −7 M of ACh (Fig 2d, n=7, *P<0.05). The analysis of AUC further confirmed that the contribution of H₂O₂ to ACh response in HAA from CAD patients is much smaller than non-CAD (Fig 2e, *P<0.05). Given that a large portion of dilation in HAA from CAD tissues remained intact after both L-NAME and Indo, or peg-CAT incubation, we further determined the role of hyperpolarization in remaining dilation by using the high K⁺ concentration. High K⁺ concentration almost abolished the dilation in HAA from CAD patients (Fig 2f, n=5, *P<0.05), indicating that the main mechanism of ACh-induced dilation involves hyperpolarization, possibly by an unknown EDH factor(s) in CAD arterioles.

The role of NOX2 and NOX4 in acetylcholine-induced dilation in human adipose arterioles in non-CAD and CAD patients

To further investigate the role of NOX2 and NOX4 in ACh-induced dilation, we preincubated HAA with the NOX2 selective inhibitor GSK2795039 (1μM) or the NOX1/4 selective inhibitor GKT137831 (1μM). GSK2795039 had no impact on the dilation in non-CAD patients (Fig 3a, n=5); however, it markedly attenuated the dilation in CAD subjects, with the average maximum amplitude reduced from 92.0 ± 2.9% to 79.5 ± 5.5% (Fig 3b, n=5, *P<0.05). In contrast to GSK2795039, preincubation of HAA with GKT137831 substantially diminished the response to ACh in non-CAD patients, with average maximum amplitude reduced from 94.4 ± 2.5% to 81.0 ± 4.3% (Fig 3c, n=5, *P<0.05), while the same pretreatment with GKT137831 did not affect the dilation in CAD subjects (Fig 3d, n=7). Thus, NOX2 contributes to the dilation in CAD subjects (Fig 3e, **P<0.01), while NOX4 partially contributes to ACh-induced dilation in HAA from non-CAD patients (Fig 3f, ***P<0.001).

Fig. 3. Role of NOX2 and NOX4 in acetylcholine (ACh)-induced dilation in human adipose arterioles (HAA) in non-CAD and CAD patients.

a In non-CAD HAA preincubation with NOX2 inhibitor GSK2795039 did not impact ACh-induced dilation (n=5). b In CAD vessels, NOX2 inhibitor GSK2795039 significantly attenuated ACh-induced dilation at log −8 M, −7 M, −6 M and −5 M of ACh (n=5). c NOX4 inhibitor GKT137831 markedly reduced ACh-induced dilation in non-CAD subjects at log −7 M, −6 M and −5 M of ACh (n=5). d NOX4 inhibitor GKT137831 did not affect ACh-induced dilation in CAD patients (n=7). e The NOX2 contribution to ACh-induced dilation was higher in CAD patients compared to non-CAD (P=0.004). f NOX4 contribution to ACh-induced dilation was higher in non-CAD subjects compared to CAD (P=0.0003). Statistical analysis was performed using repeated measures of two-way ANOVA, followed by Student-Newman-Keuls multiple comparison test for concentration response (a-d). Differences between means were analyzed with unpaired t-test (e and f). *P <0.05, **P <0.01, ***P <0.001

The effect of tempol and SOD on acetylcholine-induced dilation of human adipose arterioles in non-CAD and CAD patients

We further tested whether tempol or SOD could augment the ACh-induced dilation either in non-CAD or CAD patients. Unexpectedly, the preincubation of HAA with 10 μM tempol did not affect ACh response either in non-CAD (the average maximum amplitude, 85.6 ± 5.3% in control vs 88.8 ± 5.6% with tempol, Fig 4a, n=6) or in CAD patients except improving the dilation at log −8 M of ACh (the average maximum amplitude, 91.4 ± 3.2% in control vs 87.0 ± 4.5% with tempol, Fig 4b, n=8, *P<0.05). Consistent with tempol, treatment with the mitochondrial-targeted antioxidant mitoTempol did not significantly alter ACh-induced dilation in either non-CAD or CAD subjects (Fig S2). Treatment of vessels with SOD reduced the dilation in non-CAD patients at log −7 M of ACh (the average maximum amplitude, 93.4 ± 3.5% in control vs 90.0 ± 4.7% with SOD, Fig 4c, n=6, *P<0.05). Additionally, SOD did not affect the ACh response in CAD subjects (the average maximum amplitude, 89.0 ± 2.9% in control vs 87.6 ± 10.5% with SOD, Fig 4d, n=5).

Fig. 4. The effect of tempol and superoxide dismutase (SOD) on acetylcholine (ACh)-induced dilation in human adipose arterioles (HAA) in non-CAD and CAD patients.

a In non-CAD patients, the preincubation of HAA with tempol did not affect ACh-induced dilation (n=6). b In CAD subjects, tempol enhanced ACh dilation at log −8 M of ACh (n=8). c In HAA from non-CAD patients SOD reduced ACh dilation at log −7 M of ACh (n=6). d SOD had no effect on ACh dilation in CAD patients (n=5). Statistical analysis was performed using repeated measures of two-way ANOVA, followed by Student-Newman-Keuls multiple comparison test for concentration response (a-d). *P <0.05

The effect of tempol on NO- and H₂O₂-contribution to acetylcholine-induced dilation in human adipose arterioles from non-CAD and CAD patients

Considering that tempol may potentially increase NO bioavailability and H₂O₂ formation, we examined whether it could affect the NO- or H₂O₂-dependent dilation in response to ACh. Firstly, we tested the action of tempol on the NO-mediated component of the relaxation by examining the effect of L-NAME on the dilation in the presence of tempol. We found that L-NAME diminished the ACh response in both groups of patients (Fig 5a and b), particularly in CAD patients (the average maximum amplitude, 91.6 ± 2.4% with tempol vs 73.7 ± 2.6% with tempol+L-NAME, Fig 5b, n=6, *P<0.05). Moreover, analysis of AUC demonstrated that in the presence of tempol the NO-contribution to ACh dilation became equal to that observed in CAD patients and non-CAD (Fig 5c, n=6/group). Similarly, to determine if tempol can impact the H₂O₂-mediated component of the ACh response, we recorded the effect of CAT on the dilation in the presence of tempol. In HAA from non-CAD patients, CAT reduced the relaxation (the average maximum amplitude, 96.0 ± 2.7% with tempol vs 64.8 ± 10.7% with tempol+CAT, Fig 5d, n=8, *P<0.05). As shown in Fig 5e-f (n=8/non-CAD group, n=6/CAD group), tempol slightly restored the H₂O₂ contribution to ACh-induced dilation in CAD arterioles.

Fig. 5. The effect of tempol on NO- and H₂O₂ contribution to acetylcholine (ACh)-induced dilation in human adipose arterioles (HAA) in non-CAD and CAD patients.

a In non-CAD patients, the preincubation of HAA with L-NAME in the presence of tempol diminished ACh-induced dilation at log −7 M of ACh compared to HAA treated just with tempol (n=6). b In CAD patients the preincubation of HAA with L-NAME in the presence of tempol reduced ACh dilation at log −7 M, −6 M and −5 M of ACh compared to HAA treated just with tempol (n=6). c In the presence of tempol, the NO contribution to ACh-induced dilation of HAA was equal in non-CAD and CAD patients (n=6/group, P=0.91). d In non-CAD subjects the preincubation of HAA with CAT in the presence of tempol attenuated ACh-induced dilation at log −7 M, −6 M and −5 M of ACh compared to HAA treated just with tempol (n=8). e In CAD patients the preincubation of HAA with CAT in the presence of tempol reduced ACh dilation at log −7 M of ACh compared to HAA treated just with tempol (n=6). f In the presence of tempol the H₂O₂ contribution to ACh-induced dilation of HAA was higher but not significantly in non-CAD subjects compared to CAD (n=8/non-CAD, n=6/CAD group, P=0.34). Statistical analysis was performed using repeated measures of two-way ANOVA, followed by Student-Newman-Keuls multiple comparison test for concentration response (a, b, d, e). Differences between means were analyzed with unpaired t-test (c and f). *P <0.05

Measurement of NO concentrations in human adipose arteries

To directly test whether NO contributes differently to ACh-induced dilation in non-CAD and CAD vessels, we used EPR spin trapping to measure NO release in human arteries. A characteristic NO-Fe(DETC)₂ signal with three peaks was detected in both non-CAD and CAD resistance arteries incubated with Fe(DETC)₂ for 60 min in buffer and stimulated with A-23187 (Fig 6a, n=5/group). Quantification of the NO-Fe(DETC)₂ signal from EPR spectra showed that NO signal in A-23187-treated vessels, which were considered basal, was 1.7-fold higher in non-CAD than in CAD (Fig 6b, ****P<0.0001). Addition of ACh to A-23187 induced a 1.3-fold increase in NO signal in both non-CAD and CAD arteries (Fig 6b), indicating a comparable relative response. However, the absolute NO signal after ACh+A-23187 stimulation remained 1.7-fold higher in non-CAD than in CAD arteries (Fig 6b, **P<0.01). These results demonstrate that while the relative ACh-stimulated NO response is preserved, overall bioavailable NO is reduced in arteries from CAD patients.

Endothelial NO synthase (eNOS) protein expression in human adipose arteries from non-CAD and CAD patients

Given that one likely cause for the reduced bioavailable NO in CAD subjects is the uncoupling of eNOS, we estimated eNOS uncoupling by measuring eNOS dimers (reflecting coupled eNOS) and monomers (reflecting uncoupled eNOS), as well as eNOS phosphorylated at Ser-1177 [15]. As shown in Fig 6c-f, compared with non-CAD vessels, CAD exhibited significantly lower levels of both dimerized (****P<0.0001) and Ser-1177 phosphorylated eNOS, while the expression of eNOS monomers was higher (n=4/group, ***P<0.001). Total eNOS was similar in non-CAD and CAD vessels.

NOX2 mRNA and protein expression in human adipose arteries from non-CAD and CAD patients

Our recent studies demonstrated that NOX2 and NOX4 are two most abundantly expressed NOX isoforms in human adipose and coronary arteries [55]. Given the enhanced function of NOX2 in ACh-induced dilation during CAD, we used quantitative PCR (qPCR) to compare the NOX2 mRNA expression in adipose resistance arteries from non-CAD and CAD subjects. NOX2 mRNA expression has been found to be higher in CAD subjects (Fig 7a, n=6/group). Based on these results, we further determined the NOX2 protein expression. The NOX2 protein expression was higher in arteries from CAD (n=9) patients compared to non-CAD (n=7).

Due to the large variability of protein expression in human samples, the difference between non-CAD and CAD subjects was not significant (Fig 7a-c).

Discussion

We present the first evidence regarding the previously unknown mechanisms of the receptor agonist ACh-induced dilation in arterioles from non-CAD and CAD patients. There are several major findings in this study. First, the maximum dilation in HAA is only slightly reduced in arterioles from CAD patients compared to non-CAD patients. Second, in non-CAD subjects, NO, H₂O_2, and unidentified EDH factor(s) are equally involved in the ACh dilation. In contrast, in CAD subjects, the contribution of NO and H₂O₂ is markedly reduced, and the primary mechanism of dilation derives from an unidentified EDH factor(s). Third, NOX2 is mainly involved in ACh response in CAD patients, whereas NOX4 contributes to ACh-induced dilation in non-CAD patients. Fourth, as a SOD mimetic antioxidant, tempol does not affect ACh dilation either in non-CAD or in CAD patients. However, tempol improves the NO- and H₂O₂-mediated components of ACh-induced dilation in CAD subjects, making it closer to the non-CAD. Finally, CAD vessels exhibited enhanced expression of NOX2, which is accompanied by reduced NO release in CAD vessels, as well as eNOS uncoupling as indicated by attenuated expression of eNOS dimers and eNOS phosphorylated at Ser-1177. Together, these results demonstrate a novel mechanism of ACh-induced dilation in human arterioles that differs in subjects without and with CAD (Fig 8)

Fig. 8. Schematic diagram illustrating the proposed mechanisms by which ACh induces smooth muscle relaxation in human adipose arterioles of non-CAD and CAD patients (created in BioRender. Drachuk, K. (2025) https://BioRender.com/n9ra9df).

In non-CAD patients, ACh-induced dilation is equally mediated by NO, H₂O₂ and unknown EDH factor with the involvement of NOX4 in the dilatory response. In CAD arterioles, the contribution of both NO and H₂O₂ is diminished while an NO/H₂O₂-independent hyperpolarizing pathway becomes predominant. Mechanistically, the change in vasodilator mechanisms during CAD is associated with reduced NO release as well as an increased expression of NOX2. SOD mimetic tempol improves both NO and H₂O₂ contribution to the dilation in adipose arterioles from CAD subjects. Black text and lines represent findings obtained in the present study, while grey elements represent mechanisms derived from previously published articles. Green highlights indicate the effects of tempol observed in our experiments. Line thickness reflects the relative contribution of each pathway to ACh-induced dilation, as determined in our experiments

The role of NO in ACh-induced dilation in human arterioles

ACh is commonly used to test endothelial function in isolated vessels and in vivo in various cardiovascular diseases [17, 20, 47]. Although the NO-cGMP pathway plays a primary role in the vasodilation of large conduit vessels, a variety of other dilatory mechanisms are associated with small resistance vessels, some of which are still unidentified, especially in human arterioles. This is the first study to determine ACh-induced dilation in arterioles from patients with and without CAD and to determine the NO-dependent mechanism of such dilation. Compared with non-CAD arterioles, ACh responses were only slightly reduced in CAD arterioles at log −8 M and −7 M of ACh, but the maximal dilations were indistinguishable between the two groups. These results are consistent with those of previous studies demonstrating a similar level of FMD in coronary and adipose arterioles from non-CAD and CAD subjects [44]. However, disparate results have been reported for ACh-induced dilation. For example, human in vivo studies [51] demonstrated the impaired responses of blood flow to ACh in subjects with diabetes and hypertension; however, whether the impaired response to ACh is due to altered vasodilation at the level of resistance vessels or larger arteries was not directly examined nor was the mechanism of endothelial dysfunction in that study.

Previous ex vivo studies of FMD indicated that similar levels of dilation in non-CAD and CAD patients do not necessarily imply an identical mechanism of dilation [44]. Moreover, Najibi et al. revealed a switch from NO to EDH factor to maintain ACh response in hypercholesterolemic rabbit carotid arteries [39]. Therefore, our next goal was to determine the NO-dependent component of the ACh response in both non-CAD and CAD subjects. Our results with L-NAME indicate that NO partially contributes to ACh-induced dilation in non-CAD, while its vasodilatory role is diminished in CAD patients. These data coincide with previous observations that another NOS inhibitor L-NOARG (10 μM – 1 mM) slightly attenuated ACh response in human small subcutaneous arteries [8]. Interestingly, another study from Cohen et al. reported only a partial inhibition of ACh-induced NO release by L-NAME (30 μM) in the rabbit carotid artery [13]. In the present study, the NO scavenger cPTIO resulted in a similar partial inhibition of ACh-induced dilation as L-NAME in adipose arterioles, suggesting that the apparent small contribution of NO to ACh-induced dilation is unlikely due to an incomplete inhibition of NO release by NOS inhibitors.

To directly examine the role of NO in ACh-induced dilation, we performed the first EPR spin-trapping study to measure NO release in human vessels. Among various NO assays, EPR spin trapping is one of the most specific methods for NO detection in biological tissues [15, 29]. For example, the study from Khoo et al. demonstrated the feasibility of the EPR technique to measure NO release from isolated arteries in response to the Ca²⁺ ionophore A-23187 in mouse aorta, which was abolished by L-NAME incubation or by endothelial removal [29]. Using the same technique, Dikalova and colleagues reported that reduced ACh-induced dilation in NOX1-overexpressing mice was associated with increased ROS production and attenuated NO formation [15]. In the present study, we observed that ACh induced an increase in NO signal in both non-CAD and CAD small arteries but the absolute NO signal remained higher in non-CAD vessels. These findings are consistent with our functional data, which demonstrate that NO contributes partially to ACh-induced dilation in non-CAD arterioles, while overall bioavailable NO is reduced in CAD arterioles. Interestingly, the study from Buus et al. reported no detectable increase in NO concentration measured by NO-sensitive microelectrode during ACh dilation in healthy human arteries [8]; however, authors could not exclude the possibility that small amounts of NO, below the detection limit for the NO microelectrode, were released. In contrast, Vallance et al. demonstrated the increase of NO by ACh in human hand veins with a handmade NO sensor [53], suggesting that an appropriate method may be needed to measure NO in a specific type of vessel.

One direct cause for reduced NO release in CAD patients could be the uncoupling of eNOS, leading to the formation of O₂⁻ instead of NO. Under normal conditions, two distinct mechanisms responsible for eNOS coupling and enhanced activity to produce NO are dimerization and/or phosphorylation of Ser-1177 [11]. Accordingly, a common way to estimate eNOS coupling/uncoupling involves the evaluation of the expression of eNOS dimers and monomers, and eNOS phosphorylation at Ser-1177. Our results demonstrated the lower levels of eNOS dimers and phosphorylation at Ser-1177, along with an elevated expression of eNOS monomers in CAD, indicating that eNOS uncoupling is a likely cause for reduced bioavailable NO in CAD vessels. These data provide direct evidence of eNOS uncoupling in human arterioles under CAD conditions. Taken together, the above results demonstrate that arterioles from non-CAD and CAD patients exhibit almost similar magnitudes of ACh-induced dilation, with a partial contribution of NO to the dilatory response in non-CAD. However, in CAD arterioles, the contribution of NO to ACh response is considerably lower, likely due to eNOS uncoupling and subsequent reduced NO bioavailability that, in turn, could be associated with NOX-related mechanisms discussed below.

The role of H₂O₂ and other EDH factors in ACh-induced dilation in human arterioles

H₂O₂ has been reported as a diffusible relaxing factor contributing to the physiological regulation of the vascular tone [35-37, 61]. Matoba et al. have reported that H₂O₂ is a primary EDH factor mediating bradykinin-induced dilation in human mesenteric arteries [36]. Moreover, Thengchaisri et al. have demonstrated that H₂O₂ partially compensates for NO deficiency and maintains both VEGF-induced dilation and FMD in coronary arterioles from pigs with early type 1 diabetes [52]. Furthermore, an NO-H₂O₂ switch mediated by miR-21 has been observed during ACh-induced dilation of coronary arteries from diabetic mice [27]. We hypothesized that H₂O₂ at least partially mediates ACh-induced dilation in human arterioles, especially in CAD subjects. Unexpectedly, peg-CAT, an H₂O₂ scavenger, substantially attenuated the ACh dilation of non-CAD arterioles but had no effect in CAD except at log −7 M of ACh. It was also found that H₂O₂ acts independently from NO- and PGI₂-dependent pathways in non-CAD arterioles. Thus, the contribution of H₂O₂ to ACh-induced dilation was much higher in non-CAD arterioles compared to CAD arterioles.

Compared to vessels treated with L-NAME and Indo alone, the combination of L-NAME, Indo, and high K⁺ largely abolished ACh-induced dilation in both non-CAD and CAD vessels. These results indicate that ACh-induced dilation in human arterioles is mainly mediated by NO, H₂O₂, and another unidentified EDH factor(s). Moreover, the contribution of these vasodilator factors depends on whether the vessels are from non-CAD or CAD patients. In non-CAD patients, all three factors can be equally involved in the dilation. In contrast, the primary mediator of ACh-induced dilation is an unidentified EDH factor in CAD subjects, where H₂O₂ plays little role as opposed to FMD.

Although signaling mechanisms responsible for receptor agonist- and flow-induced dilation are likely different and remain to be explored in future studies, the above results are surprising and contrast with previous studies of FMD in human adipose arterioles [44, 61] and coronary arteries [7, 35] demonstrating an important role of H₂O₂ in maintaining FMD in patients with CAD. Indeed, previously studies have shown that FMD may require mechanosensitive ion channels such as Kir2.1/2.2, Piezo1 and TRPV4, which induce Ca²⁺ entry and activate downstream vasodilator pathways [4, 7, 59]. The primary mediator of FMD in human coronary arteries transitions with age from PGI₂ in youth to NO in adulthood and then switched to H₂O₂ in adults with CAD [5]. This switch was demonstrated in both coronary and adipose microcirculation and preserves stable tissue perfusion when NO bioavailability is reduced in CAD. Previous study further indicated NADPH-mediated redox signaling and mitochondrial complex I as the primary source of H₂O₂ during FMD [61]. By contrast, agonist-induced dilation involves receptor-specific Ca²⁺ signaling pathways that may induce the release of multiple vasodilator factors especially in the human microcirculation, for example NO in bradykinin-induced dilation of epicardial coronary arteries under healthy conditions [31], but both NO and EDH mechanisms in bradykinin-induced dilation of coronary arterioles from patients with CAD [32].

NOX2 and NOX4 contribution to ACh-induced dilation in human arterioles

NADPH oxidase family of enzymes (NOX) are major sources of ROS in the vascular system, with 4 of the seven isoforms of NOX expressed in the endothelial and smooth muscle cells [3]. Intriguingly, two of them (NOX2 and NOX4) may demonstrate opposite effects on vascular function. For example, NOX4 is a constitutively active enzyme and mediates beneficial effects by contributing to endothelium-dependent vasorelaxation under healthy conditions [49, 55]. In contrast, the pathophysiological role of NOX2 remains controversial. While several studies have implicated NOX2 in promoting endothelial dysfunction, hypertension, and inflammation [6], Zinkevich et al. reported a vasoprotective role of NOX2 in maintaining FMD and bradykinin-induced dilation in human coronary arteries from CAD patients [61]. However, during cardiovascular disease, NOX2-derived ROS may also contribute to vascular inflammation and maladaptive remodeling. This occurs, in part, through the paracrine effects of ROS-induced cytokines such as cyclophilin A, which promotes oxidative damage and inflammation in vascular smooth muscles [43]. The present study demonstrates that mRNA expression of NOX2 was higher in CAD patients. Consistent with qPCR results, NOX2 protein expression was also higher in arteries from CAD subjects. Thus, NOX2 has higher expression at both mRNA and protein levels in CAD patients, which supports the previously described role of NOX2 in FMD during CAD.

In the present study, GKT137831 (a NOX4 inhibitor) significantly reduced the dilation in non-CAD patients and did not affect vessels from CAD patients. Since NOX4-derived superoxide anion is rapidly converted to H₂O₂ within the enzyme complex [3], activation of NOX4 is likely associated with increased H₂O₂ but not superoxide levels, which may further stabilize any NO being produced in the cell [38]. Accordingly, NOX4-derived H₂O₂ likely partially mediates ACh-induced dilation in non-CAD subjects, although NOX4 may also contribute to the dilation through other mechanisms. For example, we previously demonstrated that NOX4-induced H₂O₂ formation activates TRPV4, resulting in an increased Ca²⁺ influx through this channel [55]. Moreover, Alves-Lopes et al. have recently reported that NOX4-derived H₂O₂ activates both TRPM2 and Ca²⁺ influx, leading to eNOS activation and NO release in endothelial cells [2]. Furthermore, an increase of Ca²⁺ influx may be required not only for inducing NO synthesis but also for triggering the formation of some other endothelium-derived relaxing factors, such as PGI₂ and hydrogen sulfide (as EDH factor) [12, 19]. Thus, NOX4 may contribute to all three mechanisms of ACh-induced dilation involving NO, H₂O_2, and unknown EDH factor(s).

To investigate whether NOX2 contributes to ACh-induced dilation, we evaluated the effect of NOX2 inhibition on the dilatory response to ACh. After treatment with GSK2795039, a NOX2-selective inhibitor, the dilatory response to ACh was reduced in CAD subjects and remained unchanged in non-CAD. Together with the findings of higher mRNA and protein expression of NOX2 in vessels from CAD patients, these results support a dual role for NOX2 in CAD: while it may maintain dilatory response, its upregulation could also escalate vascular inflammation and remodeling. In contrast to our findings, Munoz et al. suggested the involvement of NOX2-derived H₂O₂ in ACh-induced dilation of both rat and human renal arteries under healthy conditions [38]. Consistent with our results, Larsen et al. reported that NOX2-derived H₂O₂ mediates bradykinin-induced dilation in human coronary arterioles isolated mostly (73%) from CAD patients [32]. However, considering that peg-CAT had no effect on ACh-induced dilation in CAD patients except at log −7 M of ACh, NOX2 contribution to ACh response in CAD subjects may be mainly associated with the H₂O₂-independent mechanisms.

Numerous studies have revealed the distinct contribution of mitochondria-derived O₂⁻ to flow-dependent H₂O₂ formation [35, 61]. The potential mechanism of flow-induced release of O₂⁻ from mitochondria in endothelial cells can include cytoskeletal microfilaments that connect sarcolemmal and mitochondrial membranes [57]. However, whether mitochondria-derived ROS contribute to ACh-induced dilation remains to be determined in future studies.

The effect of SOD and tempol on Ach-induced dilation in human arterioles

There is abundant evidence that cardiovascular diseases are associated with excessive ROS production, leading to lower NO bioavailability. Under normal physiological conditions, the rate of ROS generation is under tight control by enzymatic and non-enzymatic antioxidant systems [22]. SOD is one of the major antioxidant enzymes that regulate the levels of superoxide anion and other ROS [56]. Yan et al. demonstrated the increased superoxide level and decreased eNOS activity leading to reduced FMD in resistance arteries from Mn-SOD-deficient mice [56]. Moreover, Itoh et al. indicated that SOD enhanced the ACh-induced NO-dependent relaxation in rabbit mesenteric resistance arteries [26]. Surprisingly, we found that SOD slightly reduced the dilation in non-CAD subjects and had no effect on CAD. A possible explanation of this SOD action can be found in the study from Adachi et al. reporting that SOD overexpression attenuated ACh-induced dilation [1]. Under normal conditions, NO mediates smooth muscle relaxation by one or more pathways, including cGMP-dependent relaxation via cGC formation and cGMP-independent relaxation through peroxynitrite-mediated glutathionylation of SERCA – a reversible modification that activates SERCA and thus increases Ca²⁺ uptake to induce smooth muscle relaxation [1]. Consequently, in non-CAD patients, SOD may disrupt the peroxynitrite-dependent component of SERCA activation and dilation. An opposite effect can occur in subjects with cardiovascular diseases because excessive superoxide production leads to a high level of peroxynitrite, which in turn causes the irreversible SERCA modification to reduce its activity [45]. Future studies will be required to determine the role of peroxynitrite in ACh-induced dilation in human small vessels.

To further corroborate the results obtained with SOD, we tested the effect of an SOD mimetic – tempol which can more easily penetrate cell membranes than SOD and hence react with both intracellular and extracellular superoxide anions [42]. As a result of this feature, tempol is considered a particularly appealing treatment for cardiovascular disease related to oxidative stress [50]. Although tempol had not affected the magnitude of the dilation, we demonstrated that tempol increased NO and H₂O₂ contribution to ACh response in CAD individuals, making it closer to non-CAD. Our results are consistent with the findings of Kang et al. that tempol did not change FMD in coronary arterioles from old rats [28]. However, in the same study tempol surprisingly impaired the flow response in vessels from young rats [28]. The ability of tempol to enhance NO component of the dilation in CAD subjects can be explained by the capacity of tempol to mimic SOD. Superoxide anion dismutation to H₂O₂ prevents NO interaction with superoxide, which potentially increases NO bioavailability and enhances the contribution of both NO and H₂O₂ to the vasodilation [23]. Our results are partially in line with the study from Fleenor et al., indicating the ability of tempol to restore eNOS expression with a subsequent increase of NO bioavailability and an improvement of ACh-induced dilation in old mice [21]. Collectively, our results indicate that tempol may not improve ACh-induced dilation but can increase NO- and H₂O₂ contribution to the dilation in CAD subjects. Thus, tempol or related compounds have the potential to enhance NO contribution to human arteriolar dilation under cardiovascular diseases.

Beyond its vasoactive role, NO has been recognized as a critical regulator of endothelial cell metabolism. Specifically, physiological levels of NO prevent the hyperactivation of glycolysis by inhibiting pyruvate kinase M2, thereby promoting glucose flux through oxidative branch of the pentose phosphate pathway to produce NADPH for redox protection [43]. This pathway is vital in conditions of oxidative stress, such as CAD, where enhanced NADPH production contributes to antioxidative defense. Moreover, NO plays a protective role in preventing endothelial-to-mesenchymal transition, a process that contributes to vascular inflammation and remodeling [24, 43]. These concepts coordinate with another recent review addressing how redox imbalance, impaired endothelial and smooth muscle cell function and vascular inflammation contribute to vascular aging and CAD progression[25, 34]. Therefore, normalizing NO signaling in CAD patients may improve not only vasodilation but also metabolic reprogramming and redox homeostasis.

Conclusions

Despite the growing evidence that microvascular dysfunction often precedes cardiovascular diseases, our current understanding of the dilatory mechanisms of human microvessels remains limited. In this study, we reveal a novel mechanism of agonist-induced dilation in arterioles of non-CAD and CAD patients. In non-CAD arterioles, NO and H₂O₂ significantly contribute to ACh dilation, with substantial involvement of NOX4 in H₂O₂-mediated response. In CAD arterioles, the contribution of both NO and H₂O₂ is diminished while an NO/H₂O₂-independent hyperpolarizing pathway becomes predominant. Moreover, the change in vasodilator mechanisms during CAD is associated with reduced NO release and eNOS uncoupling, as well as an increased expression of NOX2. Therefore, a better strategy for improving endothelial function should not only maintain the magnitude of the dilation but also protect or restore the physiological mechanisms of the dilation. In this regard, SOD mimetic tempol or related compounds show the potential to partially improve NO and H₂O₂ contribution to the dilation of adipose arterioles in CAD patients.

Study limitations

The human adipose tissues used for that study were collected from patients with different conditions and medication histories that can impact vasodilator response. All non-CAD patients underwent surgery and therefore cannot be considered as healthy control. To reduce the effect of cardiovascular diseases on the vascular response, our non-CAD patients include subjects with ≤ 1 cardiovascular risk factor although there may be other unidentified diseases that could affect the vascular function. Detailed information about prescribed medication was not available for all patients. Therefore, we could not exclude its potential effect on vascular dilation. The fact is that most patients in the CAD group were taking at least 1 medication to treat cardiovascular disease. Thus, the study could not compare ACh-dilation between untreated non-CAD and CAD subjects. The number of patients in each study group is small (n≥5), however, previous studies performed on HAA successfully used such sample-size groups to compare FMD [44, 61], and A-23187-induced dilation [60] between non-CAD and CAD subjects.

The human adipose tissues include different types of adipose tissue: subcutaneous vs visceral vs pericardial. However, we did not find the effect of the type of tissue on the ACh-induced dilation. Notably, the ACh dilatory response assessed in isolated vessels does not fully represent the complexity of vascular function regulation in vivo. Specifically, it does not account the influence of in vivo factors such as neurohormonal mechanisms, as described in the review by Liberale et al [33], nor does it consider the modulatory role of circulating pro-inflammatory, anti-inflammatory and pro-thrombotic factors released by macrophages, as discussed in the review by Sansonetti et al [48].

Another limitation may be related to the difference in such demographic factors as age (non-CAD, 43 ± 1.2 and CAD, 67 ± 1.3), BMI (non-CAD, 29.9 ± 0.6 and CAD, 32.5 ± 0.7), and gender distribution (non-CAD, male n=12, female n=95; CAD, male n=63, female n=13). However, a secondary analysis of demographic factors did not demonstrate an impact of those factors on vascular dilation. We have also observed no effect of cardiovascular risk factors on ACh-induced dilation in non-CAD patients (Fig S3). However, we found that in CAD group patients with hypertension exhibited higher dilation response to ACh compared to subjects without arterial hypertension (Fig S4). How the status of hypertension affects ACh dilation in CAD group remains unclear but could be related to the potential vessel protective effects of anti-hypertension medications routinely taken by those patients.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request. Uncut Western blot images are provided in Fig S5.