Introduction

Cigarette smoking is a risk factor for vascular diseases, including hypertension, atherosclerosis, and aneurysm [1, 2]. A crucial feature of smoking-induced vascular injury is endothelial dysfunction, defined as decreased nitric oxide (NO) bioavailability in the vascular wall [3]. Cigarette smoke contains many harmful chemicals and consists of a particle phase, which contains nicotine and tar, and a gas phase. Although nicotine is recognized as a component responsible for adverse vascular effects caused by cigarette smoking, several lines of evidence show that nicotine-free cigarette smoke extract (N(−)-CSE) substantially impairs endothelial NO-dependent relaxation [4,5,6]. However, there is also a report showing that nicotine-containing cigarette smoke extract (N(+)-CSE) is more harmful to endothelial integrity than N(−)-CSE [7]. Various components in cigarette smoke are suggested to contribute to the decreased NO bioavailability in the vascular system.

NO regulates vascular tone by activating soluble guanylate cyclase (sGC) [8]. However, there are several forms of sGC; the reduced form (which contains a heme moiety with ferrous iron) is NO-sensitive, while the oxidized form (which contains a heme moiety with ferric iron) and the heme-free form (which does not contain a heme group) are NO-insensitive [9]. Briefly, NO activates sGC by binding to its ferrous heme iron. The balance of these forms has been shown to be shifted from the NO-sensitive form to the NO-insensitive form by reactive oxygen species (ROS) and to be disrupted under some pathological conditions [10, 11].

NO donors serve as NO sources in the body, activating the reduced form of sGC [12]. In contrast, sGC activators can increase the enzyme activity of the oxidized/heme-free form in an NO-independent manner [9]. Since these drugs induce potent vasorelaxation, both are practically used or tested in the clinical setting. Of importance, the vasorelaxant effects increase or decrease under pathological conditions [10, 11]. However, there is still not enough evidence that cigarette smoking affects vascular responsiveness to sGC-targeting drugs, particularly sGC activators. Therefore, this study examined this issue in animal and human blood vessels.

Materials and methods

Animal samples

A total of 20 male Sprague-Dawley rats (6 weeks old, Charles River Laboratory Japan Inc., Yokohama, Japan) were used for the chronic administration experiments. The animals were housed in a light-controlled room with a 12-h light/dark cycle and were allowed access to food and water ad libitum. The Animal Research Committee of Kanazawa Medical University approved the use of rats along with the experimental protocols in this study (Permit no.: 2017-100). The rats were subcutaneously administered 0.5 mL of phosphate-buffered saline (PBS, pH 7.4), N(−)-CSE or N(+)-CSE once a day. After 4 weeks, each rat was anesthetized with sodium pentobarbital (50 mg/kg i.p.), and blood samples were collected from the inferior vena cava. The rats were then injected with heparin (500 U/kg i.v.) and were sacrificed by bleeding from the abdominal aorta. The thoracic aorta and the extralobar pulmonary artery were dissected and excised.

A total of 6 male and 6 female Wistar rats (2–4 months old, Charles River Laboratory Japan Inc.) were used for experiments to study sex differences. The thoracic aorta was isolated as described above.

Preparation of N(−)-CSE and N(+)-CSE

N(−)-CSE was prepared by passing mainstream smoke from 20 cigarettes (Winston Caster White One 100’s Box; Japan Tobacco Inc., Tokyo, Japan) through a Cambridge filter and bubbling into 7 mL of PBS. The same protocol was used for the preparation of N(+)-CSE, except that a Cambridge filter was not used. The prepared solution was then sterilized through a 0.22-µm filter. N(−)-CSE and N(+)-CSE were dispensed and kept at −80 °C until used. The nicotine concentrations in N(−)-CSE and N(+)-CSE were 0.65 and 590 µg/mL, respectively.

Human samples

A total of 20 patients undergoing gastrectomy or coronary artery bypass grafting (CABG) at the Shiga University of Medical Science Hospital were enrolled. The Human Ethics Committee of Shiga University of Medical Science approved the use of human blood vessels along with the experimental protocols in this study (Permit no.: 25-169). Informed consent was obtained from all patients. The right gastroepiploic artery was isolated from the stomach removed by gastrectomy, or the redundant portion of the artery used as a coronary bypass conduit was isolated.

Assessment of ROS formation

Plasma lipid peroxide levels were estimated using a commercial thiobarbituric acid reactive substances (TBARS) assay kit (Cayman Chemical Co., Ann Arbor, MI, USA), as previously described [13].

Vascular reactivity

The isolated rat aorta and pulmonary artery and human gastroepiploic artery were used in organ chamber experiments as previously described [14, 15]. Briefly, the arteries in which the resting tension was adjusted to 1.5 g for the aorta and the gastroepiploic artery and 0.7 g for the pulmonary artery were contracted using phenylephrine (0.1–1 µM for the aorta and the pulmonary artery) or KCl (30 mM for the gastroepiploic artery). KCl was used instead of phenylephrine for the gastroepiploic artery because phenylephrine often induces rhythmic but not tonic contraction in this artery [16]. After the contraction reached a plateau phase, the concentration–response curves for sodium nitroprusside (SNP, NO donor), BAY 60-2770 (sGC activator), and acetylcholine (ACh) were obtained by adding the drug directly to the bathing media in cumulative concentrations. At the end of each experiment, papaverine (0.1 mM) was added to induce maximal relaxation, which was considered as 100% relaxation induced by the agonists. In the case with the gastroepiploic artery, two preparations per patient were analyzed per experimental run.

In the sex difference experiments, the aorta was exposed to the heme oxidant ODQ (10 µM; referred to as “ODQ(+)”) or its solvent (referred to as “ODQ(−)”) for 30 min. Afterward, vascular reactivity to SNP and BAY 60-2770 was observed in phenylephrine-contracted preparations as described above.

Drugs

The following drugs were used: sodium pentobarbital (Kyoritsu Seiyaku Co., Tokyo, Japan); heparin (Mitsubishi Tanabe Pharma Co., Osaka, Japan); phenylephrine (Sigma Chemical Co., St. Louis, MO, USA); KCl and SNP (Nacalai Tesque, Kyoto, Japan); BAY 60-2770 (kindly provided by Dr Johannes-Peter Stasch of the Institute of Cardiovascular Research, Pharma Research Centre, Bayer AG, Wuppertal, Germany); ACh (Daiichi-Sankyo Co., Tokyo, Japan); papaverine hydrochloride (Dainippon-Sumitomo Pharma Co., Osaka, Japan); and ODQ (Tocris Bioscience, Bristol, UK). Dimethyl sulfoxide and ethanol were used as solvents for BAY 60-2770 and ODQ, respectively. Distilled water was used to dissolve all other drugs and to prepare serial dilutions, as required, from stocks available on the day of the experiment.

Statistics

All values are expressed as the mean ± SEM. The concentration–response curves were fitted by nonlinear regression analysis using Graph Pad Prism 7.0 software (Graph Pad Software Inc., San Diego, CA, USA). The negative logarithm of the dilator concentration that produced the half-maximal response (pEC50) was determined. The concentration–response curves were analyzed with two-way repeated measures analysis of variance (ANOVA) and Bonferroni post hoc test. The pEC50 and TBARS levels were compared using Bonferroni multiple comparisons test after one-way ANOVA. Univariate comparisons were performed using an unpaired two-tailed Student’s t test for continuous variables and Fisher’s exact test for categorical variables. Factors associated with vascular reactivity were assessed by calculating the 95% confidence interval (CI). Differences were considered to be significant at p < 0.05.

Results

Effects of N(−)-CSE or N(+)-CSE administration on TBARS levels

Plasma TBARS concentrations significantly increased in the N(−)-CSE (3.26 ± 0.22 µM, n = 6, p < 0.01) and N(+)-CSE groups (4.33 ± 0.25 µM, n = 6, p < 0.01) compared with that in the PBS group (2.08 ± 0.22 µM, n = 6). In addition, the TBARS levels in the N(+)-CSE group were significantly higher than those in the N(−)-CSE group (p < 0.05).

Effects of N(−)-CSE or N(+)-CSE administration on the responsiveness to relaxant agonists

The concentration–response curve (Fig. 1a) and the pEC50 values (Table 1) showed that the relaxant response of the aorta to SNP was identical among the PBS, N(−)-CSE, and N(+)-CSE groups. The results were similar to those obtained in the pulmonary artery, where SNP-induced relaxation did not differ significantly among the groups (Fig. 2a and Table 1).

Fig. 1
figure 1

Relaxant responses of the rat aorta to SNP (a), BAY 60-2770 (b), and ACh (c). Each point and bar represent the mean ± SEM of six experiments. **p < 0.01, compared with the PBS group. ACh acetylcholine, N(+)-CSE nicotine-containing cigarette smoke extract, N(−)-CSE nicotine-free cigarette smoke extract, PBS phosphate-buffered saline, SNP sodium nitroprusside

Full size image
Table 1 pEC50 to relaxant agonists
Full size table
Fig. 2
figure 2

Relaxant responses of the rat pulmonary artery to SNP (a), BAY 60-2770 (b), and ACh (c). Each point and bar represent the mean ± SEM of six experiments. **p < 0.01, compared with the PBS group. ACh acetylcholine, N(+)-CSE nicotine-containing cigarette smoke extract, N(−)-CSE nicotine-free cigarette smoke extract, PBS phosphate-buffered saline, SNP sodium nitroprusside

Full size image

Irrespective of the type of vessel, BAY 60-2770-induced relaxation curves in the N(−)-CSE and N(+)-CSE groups were not different from the curve in the PBS group (Fig. 1b, the aorta and Fig. 2b, the pulmonary artery). There was also no difference in the relaxant potency of BAY 60-2770 among the groups, as evidenced by similar pEC50 values (Table 1).

The addition of ACh produced a concentration-dependent relaxation both in the aorta (Fig. 1c) and the pulmonary artery (Fig. 2c). The relaxant responses at 10–7 M in the aorta and at 10–6 M in the pulmonary artery were significantly attenuated in the N(−)-CSE and N(+)-CSE groups. In addition, the pEC50 values for ACh in the N(+)-CSE group were significantly decreased compared with the values in the PBS group (Table 1).

Patient characteristics

The characteristics of the study groups are presented in Table 2. Twenty patients were classified as nonsmokers or smokers. Among smokers, with the exception of 1 or 2 participants whose details were unknown, the average amount of smoking was 25 (range: 20–40) cigarettes per day, and the average duration of smoking was 38 (range: 19–53) years. The proportion of men in the smokers group was notably higher than that in the nonsmokers group, but there were no significant differences in the other indexes.

Table 2 Clinical and demographic characteristics of patients
Full size table

Influence of smoking on the responsiveness to relaxant agonists

SNP evoked a concentration-dependent relaxation of the gastroepiploic artery, which was not different between nonsmokers and smokers (Fig. 3a). The pEC50 values for SNP were also identical [6.75 ± 0.18 (nonsmokers) vs. 6.86 ± 0.09 (smokers), p = 0.555].

Fig. 3
figure 3

Relaxant responses of the human gastroepiploic artery to SNP (a) and BAY 60-2770 (b). Each point and bar represent the mean ± SEM of sixteen (nonsmokers) or twenty-four (smokers) experiments. SNP sodium nitroprusside

Full size image

As was the case with SNP responsiveness, there were no significant differences in the concentration–response curve for BAY 60-2770 (Fig. 3b) and relaxant potency between nonsmokers and smokers [pEC50 values, 8.58 ± 0.15 (nonsmokers) vs. 8.82 ± 0.16 (smokers), p = 0.298].

Factors influencing the responsiveness to relaxant agonists

Due to the limited group size, multiple regression analysis was not applied, but the respective group means and 95% CIs were calculated. The results show that smoking and sex had no influence on the pEC50 values for SNP-induced relaxation of the gastroepiploic artery, and only dyslipidemia had a significant influence (Fig. 4a). On the other hand, none of the factors, including smoking and sex, had any apparent effect on the values for BAY 60-2770 (Fig. 4b).

Fig. 4
figure 4

pEC50 to SNP (a) and BAY 60-2770 (b) in the different subgroups. Each point and bar represent the mean ± 95% confidence interval. The vertical dotted line is an indicator of the mean value of all samples. CKD chronic kidney disease, DL dyslipidemia, HT hypertension, OB obesity, pEC50 the negative logarithm of the dilator concentration that produced the half-maximal response, SNP sodium nitroprusside, T2DM type 2 diabetes mellitus

Full size image

Influence of sex on the responsiveness to relaxant agonists

SNP induced a relaxation within the aorta, and this response was not different between age-matched male and female rats (Supplementary Fig. 1a). Although the relaxation produced in ODQ(+) was significantly attenuated compared with that produced in ODQ(−), there was also no sex difference in the degree of attenuation.

BAY 60-2770 administration produced a concentration-dependent relaxation response in the aorta of male rats; this response was identical in the aorta of female rats (Supplementary Fig. 1b). ODQ(+) significantly augmented the relaxation response to the same extent in both male and female rats.

Discussion

sGC is a key drug target in hypertension and ischemic heart disease, for which smoking is a known risk factor [9]. Therefore, it is of great importance to understand the impact of smoking on sGC-mediated vascular tone regulation. To this end, we first examined the effects of N(−)-CSE administration in rats. Although increased oxidative stress and arterial endothelial dysfunction were observed in rats administered N(−)-CSE as previously reported [4, 17], the relaxant responses to SNP and BAY 60-2770 were normal both in the aorta and the pulmonary artery. The result that chronic administration of N(−)-CSE did not affect SNP-induced vasorelaxation is in line with our previous report [4]. Altogether, NO-sensitive and NO-insensitive sGC-mediated vascular tone regulation is considered to be preserved in rats administered N(−)-CSE.

Many studies have shown that nicotine is a detrimental component responsible for cigarette smoking-induced oxidative stress and vascular injury [18,19,20]. For this reason, we next examined the effects of N(+)-CSE administration in rats. The results confirmed that oxidative stress was higher in the N(+)-CSE group than that in the N(−)-CSE group and that endothelial dysfunction occurred in rats administered N(+)-CSE. However, similar to the rats administered N(−)-CSE, there were no differences in the responsiveness of the aorta and the pulmonary artery to SNP and BAY 60-2770 even when the rats were chronically administered N(+)-CSE. A similar result has been reported by Carlo et al., who demonstrated that NO-induced vascular smooth muscle relaxation is not impaired by chronic exposure to N(+)-CSE, whereas endothelium-dependent relaxation is impaired [5]. In conclusion, this suggests that NO-sensitive and NO-insensitive sGC-mediated vascular tone regulation is not disturbed even in rats administered N(+)-CSE.

The results obtained from studies using CSE may not always be extrapolated to cigarette smoking in humans. Therefore, we finally examined the differences in vascular reactivity to sGC-targeting drugs between nonsmokers and smokers. Importantly, SNP- and BAY 60-2770-induced relaxations of the gastroepiploic artery were not affected by smoking status. This is in accordance with our previous study showing that smoking had no impact on the relaxant responses to nitroglycerin and BAY 60-2770 in the internal thoracic artery obtained from patients undergoing CABG [21]. Furthermore, NO-induced relaxation in smokers has been confirmed to be normal in many different arteries, including the coronary [22], pulmonary [23], brachial [24], radial [25], and uterine artery [26]. These findings suggest that NO-sensitive and NO-insensitive sGC functions normally to regulate vascular tone even in smokers.

In this study, the ratio of men and women was different between nonsmokers and smokers. The analysis of confounding factors showed that sex, as well as smoking, had no significant effect on the responsiveness of the gastroepiploic artery to SNP and BAY 60-2770. Interestingly, Witte et al. have demonstrated that there are no sex differences in the relative increases in cGMP formation after stimulation by an NO donor and sGC activator in the human internal thoracic artery [27]. In addition, we observed that regardless of the presence or absence of ODQ, the relaxant responses of the aorta to SNP and BAY 60-2770 were identical between male and female rats. This finding suggests that there are no sex-dependent differences in the balance between NO-sensitive and NO-insensitive forms of sGC. Considering these facts, it is unlikely that in this study, sex led to a bias in data interpretation.

The imbalance between NO-sensitive and NO-insensitive sGC has been confirmed in diseased blood vessels. For example, Stasch et al. reported that the NO-sensitive form changes into the NO-insensitive form in the aorta of aged spontaneously hypertensive rats, in the aorta of high-fat diet-fed ApoE−/− mice, and in the saphenous artery of Watanabe hereditable hyperlipidemic rabbits [10]. In addition, we recently reported that the balance between NO-sensitive and NO-insensitive sGC is disrupted in the balloon-injured rat carotid artery [28]. In contrast, such a phenomenon is not observed in the aorta of older mice [13] or in the aorta of streptozotocin-induced diabetic rats [29]. Therefore, only severely diseased blood vessels might carry a risk for disrupting the balance in sGC. That is, vascular conditions in rats administered N(−)-CSE or N(+)-CSE and in smokers recruited in this study are thought to not be as serious. If that is the case, it would be interesting to examine how this balance changes in heavier smoking conditions.

One important clinical implication of this study is that sGC-targeting drugs can exert normal vasodilatory effects even in smokers. An NO donor is clinically used for dilating blood vessels in hypertension and ischemic heart disease, and an sGC activator is currently being tested in pulmonary hypertension (clinicalTrials.gov identifier: NCT03754660). Since some patients with such diseases are most likely smokers, the findings obtained from this study will be of value in predicting drug efficacy.

One limitation of this study is that we did not examine whether oxidative stress is greater in smokers than in nonsmokers. This finding would better help us understand the relationship between cigarette smoking, oxidative stress, and vascular dysfunction. Another limitation is that in experiments using the human gastroepiploic artery, enrollment into each group was not very large. Future studies with a larger sample size would make our conclusions more robust. In addition, although most of the subjects recruited in this study had vascular risk factor(s), the comparison of healthy nonsmokers vs. healthy smokers would be more preferable if possible. The findings obtained from such studies would allow a clear understanding of the impact of smoking.

In conclusion, this study demonstrated that NO-sensitive and NO-insensitive sGC-mediated vascular tone regulation is not affected by cigarette smoking. As summarized in Fig. 5, it is considered that the balance between the two sGCs is maintained even in the blood vessels damaged by cigarette smoking.

Fig. 5
figure 5

Summary scheme showing cigarette smoking-induced vascular dysfunction within the NO/sGC/cGMP pathway. Cigarette smoking induces endothelial dysfunction but does not affect the balance between NO-sensitive and NO-insensitive forms of sGC. ACh acetylcholine, eNOS endothelial nitric oxide synthase, NO nitric oxide, sGC soluble guanylate cyclase, SNP sodium nitroprusside

Full size image