Vasorelaxant Effect of Moroccan Cannabis sativa Threshing Residues on Rat Mesenteric Arterial Bed is Endothelium and Muscarinic Receptors Dependent

Introduction Ethanolic fraction of Moroccan Cannabis sativa threshing residues (EFCS) was evaluated for its vasorelaxant activity. The current work aims to identify the active metabolites in the ethanolic fraction of the EFCS and illustrate their mechanism of action. Methods Free radical scavenging capacity of EFCS was assessed using DPPH method. The EFCS vasodilation activities in phenylephrine-precontracted isolated rat mesenteric arterial beds were investigated in presence of L-NAME (nitric oxide synthase inhibitor), indomethacin (cyclooxygenase inhibitor), potassium channel blockers (namely tetraetylamonium, barium chloride, and glibenclamide), and atropine. Nitric oxide vascular release was measured by electron paramagnetic resonance (EPR) using a spin trap in rat aortic rings. Results EFCS induced dose-dependent vasorelaxation on mesenteric vascular bed. Incubation of the preparations with L-NAME, ODQ (a soluble guanylyl cyclase inhibitor), or potassium channel blockers reduced the fall of perfusion pressure caused by EFCS. Endothelial denudation or atropine abolished the EFCS's vasorelaxant effect, suggesting involvement of muscarinic receptors and endothelium-relaxing factors. The extract induced nitric oxide release in aortic rings in a similar manner as acetylcholine suggesting an effect of EFCS on the muscarinic receptor and the conductance arteries. Chemical investigation of EFCS identified potential active components namely apigenin and derivatives of luteolin skeleton and also additional components such as neophytadiene, squalene, and β-sitosterol. In conclusion, the vasorelaxant effect of EFCS on rat mesenteric arterial bed, which is dependent of muscarinic receptor activation, nitric oxide, and EDHF, can account for potential therapeutic use against high blood pressure related cardiovascular diseases.


Introduction
Hypertension is a serious cardiovascular complication and is a leading cause of death in adults worldwide [1]. Hence, modulation of arteries' tone is very important to regulate blood fow and pressure in the whole body.
Endothelial cells play a main role in vascular tone control through muscarinic receptor activation [2] and release of several factors such as nitric oxide (NO), prostacyclin (PGI 2 ), and endothelium-derived hyperpolarizing factor (EDHF) [3]. Tese factors illicit relaxation of adjacent smooth muscle cells in the arteries. NO and PGI 2 may induce the relaxation of arteries by stimulating guanosine 3′, 5′ cyclic monophosphate (cyclic GMP) and adenosine 3′, 5′ cyclic monophosphate (cyclic AMP) productions [3,4] reducing cytoplasmic Ca 2+ in vascular smooth muscle cells and decreasing Ca 2+ sensitivity of the contractile apparatus. Tese vasodilators may also open K + channels either directly or through their respective second messengers [5]. Te endothelium relaxation independent of NO and PGI 2 is imputed to EDHF which may also act through activation of K + channels opening with subsequent hyperpolarization of smooth muscle cells [3].
Many studies have reported that the vasodilator efect of medicinal plants is attributed to endothelial function or endothelium integrity improvement through muscarinic receptors activation [6][7][8].
Cannabis is one of the oldest medicinal plants reported to be used in traditional medicine [9]. Cannabis has been shown to have therapeutic potential in diseases associated with infammation, gastrointestinal disturbance, oxidative stress, neurodegenerative diseases, and metabolic diseases such as diabetes [10,11].
Cannabis plant produces several hundreds of chemical compounds including cannabinoids, a family of specifc substances that exert their biological efects by interacting with endogenous cannabinoid receptors [9,12,13].
Tis plant could be classifed into drug and nondrug varieties which have as main diference in the content of the psychoactive cannabinoid namely tetrahydrocannabinol (THC). Te drug type, best known in Morocco as kif (khardala), contains THC in concentration between 1 and 20% [14]. Te nondrug type, commonly referred as industrial hemp, has no psychoactive activities because it contains less than 0.2% of THC [15].
In the latest decade, nonpsychoactive cannabinoids has been a target for pharmaceutical evaluation owing to their new therapeutic applications.
Several studies demonstrated that CBD can relieve ailments caused by a wide spectrum of diseases such as epilepsy neurodegenerative diseases, neuropsychiatric disorders, and cancer diseases [17,18]. Recently, many benefcial efects of CBD have been evidenced in heart diseases (myocardial infarction, cardiomyopathy, and myocarditis), stroke, and against ischemia/reperfusion injuries. In these pathological conditions, CBD decreased oxidative stress, apoptosis, and damage to organs and also improved the endothelium function [19][20][21].
On the other hand, polyphenols (favonoids) and terpenes have been intensively investigated because of the large spectrum in their biological activities and specifcally for their cardiovascular benefcial efects [22,23].
However, unlike other cannabinoids, the direct vascular efects of nonpsychoactive secondary metabolites from cannabis have not been fully investigated. Tis study was conducted to evaluate the antioxidant and vasorelaxant efects of nonpsychoactive compounds from Moroccan C. sativa. We expected that such approach may be a promising way for converting cannabis or hemp processing by-products into novel bioactive ingredients with cardiovascular benefts. To the best of our knowledge, this is the frst study conducted on the Moroccan C. sativa leaves and inforescences threshing residues to test their vasorelaxant efect on rat mesenteric arterial bed (MAB) and NO releasing in rat aorta.

Plant
Material. Te C. sativa plant was harvested in the region of Tafrante in Morocco (37°38′38.67396″N, 5°5′20.92272″W) and transported to the national agency of aromatic and medicinal plants (Taounate, Morocco). Since we are interested to study the nonpsychoactive products of C. sativa, only the leaves and inforescences of the female plant were used, without trichomes and seeds. Te trichomes were previously eliminated with a traditional threshing method [24] to extract the cannabis resin fully rich in THC, the psychoactive compound.

Extraction.
Leaves and inforescences of dried threshing residues of C. sativa were sequentially extracted using a Soxhlet extractor with 300 mL of hexane, dichloromethane, ethyl acetate, ethanol, and water. Extracts were then fltered through the flter paper with pore size 4-12 μm under vacuum; solvent was evaporated using a rotary vacuum evaporator and stored at 4°C until subsequent testing and analyses.

Quantitative Analysis of Total Phenols (TPC) and Total
Flavonoids (TFC). TPC was determined by the Folin-Ciocalteu colorimetric assay [25], then was expressed as milligrams of gallic acid equivalent per gram dry weight. TFC was determined according to the Harborne method [26] and expressed as milligrams of quercetin equivalent per gram dry weight (mg QE/g dw).
Te inhibition percentage was expressed by the following equation: At least three independent experiments were carried out and expressed as mean ± SEM values.

Isolated MAB Experiments.
MABs were isolated as described by Cheikh et al. [28]. Rats were anesthetized with sodium pentobarbital (50 mg/kg, i. p.). After opening the abdominal cavity through an abdominal midline incision, the superior mesenteric artery was rapidly cannulated with a heparinized hypodermic needle at its origin from the abdominal aorta and immediately fushed with warm Krebs-Henseleit solution. Te whole cannulated arterial bed was carefully separated from the intestines by cutting close to the intestinal wall and placed into a Petri dish. Once the MAB was isolated, the Krebs-Henseleit solution was pumped into the MAB using a peristaltic pump (Pharmacia Biotech, USA) at a continuous constant fow of 2 mL/min. Te Krebs-Henseleit solution consisted of 118 mmol/L NaCl, 15 mmol/L NaHCO 3 , 4.7 mmol/L KCl, 1.2 mmol/L MgCl 2 , 2.5 mmol/L CaCl 2 , 1.2 mmol/L KH 2 PO 4 , and 11.0 mmol/L glucose and was maintained at 37°C and aerated with carbogen (5% CO 2 in O 2 ; fnal pH 7.4). Te MAB perfusion pressure was continuously measured using a pressure transducer (Capto SP844) and recorded on a universal oscillograph (50-8622, Harvard Apparatus Limited, UK). Te basal perfusion pressure was between 6 and 16 mmHg at the beginning of the study and remained stable throughout the experiment.

Efect of EFCS on the Precontracted MAB.
After 30 min equilibration, the MAB was constricted by continuous infusion of phenylephrine (PHE) (10-20 μM) using a syringe pump (Orion M365) to produce a consistent PP of 80-100 mmHg [7]. When the plateau of contraction was reached, the endothelium MAB's integrity was confrmed by 5.5 nmol of acetylcholine (ACh) [7,28]. Cumulative-doseresponse curves were started by bolus (10-100 μl) injections of EFCS (10 to 500 μg), before and after incubation with various pharmacological inhibitors. Assuming that a 100% relaxation represents a return to the baseline, relaxations were reported as a percentage of the drop in perfusion pressure.

Role of Endothelium in the Vascular Relaxant Efect of EFCS.
Te vascular relaxant efect of EFCS was examined on endothelium-denuded preparations (perfusing MAB with distilled water for 5-6 minutes to remove the endothelium) as described by Adaegbo et al. [29]. Te absence of functional endothelium was confrmed by the inability of ACh (5.5 nmol) to induce more than 15% relaxation in precontracted vessels with PHE. To verify the integrity of smooth muscle cells, the MAB was challenged with an exogenous donor of NO, sodium nitroprusside (SNP, 10 nmol) [7,28].

Role of NO/c-GMP in Vascular Relaxant Efect of EFCS.
To investigate the involvement of NO/c-GMP pathway in the observed vasorelaxant efect of EFCS, the endotheliumintact MAB was incubated with a NO-synthase inhibitor, Nnitro-L-arginine methyl ester (L-NAME, 100 μM), or a guanylate cyclase inhibitor (ODQ 1 μM), for 20 min, prior precontracting with a lower concentration of PHE (2.5-5 μM) to get a perfusion pressure plateau identical to that obtained in controls [7,28].

Role of PGI 2 in Vascular Relaxant Efects of EFCS.
To determine the contribution of PGI 2 in the EFCS's efect, cyclooxygenase inhibitor (indomethacin, 10 μM) was applied to the MAB with intact endothelium for 20 min prior precontraction with PHE.

Role of K + Channels in Vascular Relaxant Efect of EFCS.
Te MAB was perfused with Krebs-Henseleit solution containing 100 mM of KCl in order to determine whether the vascular relaxant efect of EFCS is caused by membrane hyperpolarization. Under such condition, the preparation is rather depolarized, preventing any membrane hyperpolarisation and vascular relaxation due to K + channels opening. Terefore, the possibility that EFCS may act as an opener of K + channels was explored by changing the extracellular concentration of K + from 4.7 mM to 100 mM KCl by equimolar replacement of NaCl with KCl in the Krebs-Henseleit solution.
Tree potassium channel inhibitors namely tetraethylamonium (TEA) (Ca 2+ -activated K + channel inhibitor), glibenclamide (ATP-sensitive K + channel inhibitor), and barium chloride (BaCl 2 ) (inward rectifying potassium channels blocker) were used in a new series of experiments in endothelium-intact MAB to confrm the potential role of K + channels in the EFCS's activity.

Role of Muscarinic Receptors in the Vascular Relaxant
Efect of EFCS. To assess if EFCS produced vascular relaxation through the activation of muscarinic receptors, endothelium-intact MAB was incubated with 1 μM atropine (a muscarinic receptor antagonist) for 20 min before the exposure to the EFCS.

Implication of NO in Vascular Relaxant Efect of EFCS.
To assess the ability of EFCS to induce NO release in rat vessels, we measured the NO production by electron paramagnetic resonance (EPR) in conductance arteries, the rat aortic rings, exposed or not to ACh (1 μM) and/or EFCS (100 μg/ml), or to the NOS inhibitor L-NAME (0.1 M) as negative control. Aortas were briefy incubated for 45 min at 37°C in a Krebs-Hepes colloid solution with Na-DETC (Sigma-Aldrich) mixed with FeSO4 as a spin trap for NO detection. Each sample was then instantly frozen in liquid nitrogen and analyzed using an EPR Miniscope MS5000 in a Dewar fask at 77°K (Frieberg Instruments, Germany). Te instrument settings were microwave power of 10 MW, 1 mT of amplitude modulation, 100 kHz modulation frequency, sweep time of 150 s, and 3 scans. Data were obtained by measuring the overall peak amplitude of the generated spectra and normalizing it to the sample's dry weight in arbitrary units (A.U.) [30].

GC-MS Analysis.
Analysis of the C. sativa ethanolic extract volatile compounds was completed on an Agilent 8890 GC apparatus coupled to an Agilent GC/MSD 5977B mass spectrometry detector. Sample was injected using a 7693A autosampler. A split mode injection was used with a split ratio of 50 : 1. Helium was used as carrier gas, and the injector temperature was set at 250°C. Compounds were separated on a HP-5MS 30 m, 0.25 mm, and 25 μm capillary column. Te column oven temperature program started at 60°C for 2 min then ramped from 60 to 295°C during 15 min at a rate of 15°C/min. Results and data were recorded and monitored using Mass Hunter workstation acquisition software.

High-Performance Liquid Chromatography Coupled with Diode-Array Detection and Electrospray Ionization Tandem Mass Spectrometry (HPLC-DAD-MS) Analysis.
In order to investigate the nonvolatile phytoconstituents extracted with ethanol from C. sativa, an aliquot volume was injected into a reverse phase HPLC apparatus coupled to both UV-visible and mass spectrometry detectors. Te LC-DAD-ESI-MS system consisted of a TermoFinnigan liquid chromatography apparatus, a UV-visible Termo Surveyor photodiode array detector, and an ion trap LCQ Deca Termo Finnigan mass spectrometry detector equipped with an electrospray ionization source. Nitrogen was used as nebulizer gas and helium was used as collision gas. Separation was performed on a Merck Hibar ® HR Purospher ® STAR RP-18 end capped UHPLC column (2.1 × 150 mm). Data were acquired using Xcalibur software. Compounds were detected and tentatively identifed from their UVvisible spectra and their MS, MS/MS spectra recorded in the negative ion mode.

Chemicals and Drugs.
Distilled water was used to prepare daily ACh, L-NAME, PHE, TEA, glibenclamide, BaCl 2 , and atropine which were kept on ice until their use. Indomethacin was dissolved in NaHCO 3 (150 mM). ODQ was dissolved in dimethyl sulfoxide (DMSO). Drugs were purchased from Sigma-Aldrich.

Statistical Analysis.
All values were expressed as mean ± SEM. For dose-response curves and NO EPR measurements, a one or two way ANOVA analysis for repeated measurements, followed by the Bonferroni's post hoc test were applied. A diference was considered statistically signifcant at p < 0.05. Statistical analyzes were assessed using GraphPad Prism 5.0 software.

Antioxidant Efect of EFCS.
Te results obtained from the evaluation of the antioxidant efect by the DPPH method as well as total phenolic content (TPC), and total favonoid content (TFC) of ethanolic extract is reported in Table 1 below.
Te percentage inhibition of antioxidant activity by trapping free radicals was about 43.08 ± 0.65, while the total content of phenols for TPC and TFC were about 8.43 ± 0.28 mg GAE/g dw and 3.13 ± 0.43 mg GAE/g dw, respectively (Table 1), with DPPH activity between 41.44% and 24.22%. In all investigated parameters, the highest values were found in the ethanolic extract and decreased in the following order: ethanol, hexane, dichloromethane, ethyl acetate, and water.

Vasorelaxant Efect and Mechanism of Action of EFCS.
Te bolus injections of EFCS (10-500 μg) caused dosedependent relaxation of endothelium-intact MAB, which was detectable as a decrease of the perfusion pressure ( Figure 1). Te maximal relaxation was about 74.17 ± 1.44% of perfusion pressure (PP) decrease and occurred with the perfusion of a solution containing 500 μg of EFCS while the EC 50 (the dose of EFCS that provides 50% of fall in pressure) was about 47,72 ± 5,664 μg (n � 10).
To investigate the involvement of endothelium in the EFCS's vasorelaxation, new experiments were performed on endothelium-denuded preparations. Te removal of endothelium abolished radically the vasorelaxant responses of EFCS (Figure 2), indicating that EFCS-induced vasorelaxation was totally dependent on the endothelial function (n � 6).
Under the same experimental conditions, SNP (exogenous NO-donor) confrms the integrity of the smooth muscle cells of the MAB. Tis verifcation excludes a possible alteration of smooth muscle cells and highlights that the response to EFCS extract was attenuated because of the destruction of endothelium.
Interestingly, under NO-sGC blockade, the EFCSinduced vasorelaxation was signifcantly attenuated (Figure 3) suggesting that response to EFCS was mediated mainly by the NO-sGC-dependent signalling.
To investigate if PGI 2 mediated the EFCS-induced vasorelaxation we used indomethacin (10 μM) as an inhibitor of cyclooxygenase, an enzyme responsible for synthesis of PGI 2 in endothelium. As shown in Figure 3, there was no signifcant reduction in the EFCS-induced vasorelaxation by indomethacin, suggesting that this efect was independent of PGI 2 . However, the simultaneous inhibition of NO and PGI 2 with both L-NAME and indomethacin seems to reduce the EFCS-induced vasorelaxation further and in a signifcant manner (Figure 3).
To fnd out the possible implication of K + channels opening in the EFCS's vasorelaxation, experiments were performed on preparations precontracted with Krebs high K + (100 mM KCl). As shown in Figure 4, vasorelaxation caused by EFCS on MAB vessels was signifcantly diminished on preparations depolarized and precontracted with high K + comparing to the preparations precontracted with PHE, suggesting the involvement of activated K + channels associated to EDHF relaxation.
Furthermore, the involvement of these channels was assessed in the presence of three K + channel inhibitors namely TEA (an inhibitor of Ca 2+ -activated K + channels), glibenclamide (an inhibitor of ATP-sensitive K + channels), and BaCl 2 (a blocker of inwardly rectifying potassium channels). Te obtained data revealed that preincubation of MAB with 10 mM of either TEA, glibenclamide, or BaCl 2 caused an inhibitory efect on EFCS-induced vasorelaxation suggesting that the observed EFCS vasorelaxant efect requires at least the involvement of the three types of K + channels ( Figure 5(a)). Te inhibition of these channels simultaneously in the presence of L-NAME and indomethacin completely abolished the EFCS-induced relaxation ( Figure 5(b)).
To verify if EFCS act through muscarinic receptors, the MAB was incubated with atropine. Figure 6 shows that 1 μM atropine almost completely abolished the relaxant efect of EFCS, suggesting that muscarinic receptors are activated to induce relaxation of MAB when incubated with EFCS. Te functionality of the preparation was confrmed by evaluating the vasorelaxation induced by 10 nmol of the NO-donor SNP (data not shown).
To verify if the EFCS is able to induce potentially vascular relaxation in conductance arteries such as aorta, we measured the release of NO in aortic rings by EPR using Fe (DETC) 2 as spin trap. Indeed, NO in conductance arteries is the main vasorelaxant factor involved. As expected, NO release is signifcantly increased when the aortic ring is incubated with ACh (1 μM). Interestingly, the vessel is able to release a similar amount of NO if incubated with a solution of spin trap containing 100 μg of EFCS, suggesting eNOS involvement in this efect (Figure 7). If added to ACh, EFCS is not able to improve ulterior NO release induced by ACh (1 μM) suggesting a maximal efect of the EFCS at the dose used. Finally, the use of L-NAME also abolished the basal level of NO released by control vessels suggesting that the signal evaluated is NO of NOS origin in all the groups of aortic rings (Figure 7).    or L-NAME + indomethacin. Values are expressed as mean ± SEM (n � 10 for control curve without inhibitors and n � 6 for all others groups). ns: not signifcant (control vs indomethacin); * * * p < 0.001 (control vs L-NAME, ODQ, or L-NAME plus indomethacin).

Phytochemical Analysis of EFCS.
Te phytochemical composition of the studied C. sativa ethanolic extract was explored through hyphenated coupled chromatographic techniques. While the nonvolatile fraction of the extract was explored through high performance liquid chromatography coupled to both mass spectrometry and diode array detectors, the volatile compounds were identifed by gas chromatography coupled to mass spectrometry (GC-MS). Te generated data from both analyses are discussed below.
Te generated data for the volatile metabolites representing the major detected compounds by GC-MS are gathered in the Table 2 Figure 5: Efects of EFCS extract on PP of rat MAB precontracted with PHE in the absence and presence of of K + channel blockers (10 mM of TEA, 1 μM of glibenclamide, and 100 μM of BaCl 2 ) (a) and L-NAME + indomethacin + K + channels blockers (b) Values are expressed as mean ± SEM (n � 10 for control curve without inhibitors and n � 6 for all others groups). * p < 0.05 (control vs. BaCl 2 ); * * p < 0.01 (control vs. glibenclamide or TEA); * * * p < 0.001 (control vs. L-NAME + Indomethacin + glibenclamide + TEA + BaCl 2 ). 6 Evidence-Based Complementary and Alternative Medicine Five compounds were tentatively identifed as apigenin and derivatives of luteolin skeleton. Tus, the presence of neophytadiene, squalene, β-sitosterol, apigenin, and luteolin derivatives in high amounts in EFCS could be responsible for the antioxidant and the vasorelaxant efects of EFCS.

Discussion
Te goal of this study focused on understanding the actions of nonpsychoactive Cannabis compounds on vascular tone modulation. To our knowledge, this is the frst study that demonstrates how EFCS of C. sativa threshing residues causes vasorelaxation in rat MAB, which is dependent on muscarinic receptor activation, endothelial activity, NO, and potassium channel activity modulation.
Te C. sativa threshing ethanolic extract contained a high level of total polyphenols (TP) and total favonoids (TF), 8.43 ± 0.28 mg GAE/g dw and 3.13 ± 0.43 mg GAE/g dw, respectively. In agreement with a previous study reporting similar levels of TP and TF content of aerial parts of mature hemp that ranged from 5.85 to 9.25 mg GAE/g dw and from 1.83 to 5.21 mg CE/g dw, respectively [31].
Our data agreed with a study by Kitrytė et al. [32] demonstrating the highest antioxidant capacity in the ethanolic fraction harvested from C. sativa threshing residues. Numerous epidemiological and in vitro investigations suggested that polyphenol-rich herbal medicine extracts play a signifcant role in maintaining vascular health and preventing cardiovascular diseases [22,23]. Indeed previous studies have shown that nonpsychoactive compounds from Cannabis have antioxidant efects [16,33]. When confronted by cardiovascular diseases, antioxidant therapy has been recognized as a reasonable strategy to improve vascular function [34,35]. Te scavenging efect displayed by Cannabis threshing compounds could be due to the high content of total polyphenols and favonoids in EFCS.
Furthermore, our study demonstrates that EFCS induced a dose-dependent vasorelaxation on preconstricted MAB with a pEC50 in the midmicromolar range. Similar fndings have been reported in the rat mesenteric artery with other medicinal plant extracts [7,36] and also with Cannabisderived products [20,21].
Te endothelium is well known to play a main role in the vasorelaxation process and implicates activation of the muscarinic receptor [2,3]. EFCS-induced vasorelaxation seems to be totally endothelium dependent, since the removal of endothelial cells by water perfusion radically abolished the fall in perfusion pressure. Endotheliumdependent vasorelaxation in mesenteric arteries implicates release of endothelium-derived relaxing factors, such as NO, PGI 2 , and EDHF [3]. Te NO is synthesized from L-arginine by endothelial-NOS. L-NAME, a L-arginine analog is able to inhibit the 3 isoforms of NOS in a nonspecifc way. PGI 2 is synthesized by endothelial cyclooxygenase 1 (COX1) using arachidonic acid as substrate [3], and we can inhibit the PGI 2 production using indomethacin, a nonselective COX inhibitor able to inhibit also the inducible isoform (COX2).
We next investigated which endothelium-derived relaxing factors, namely NO, PGI 2 , and EDHF was/were implicated in EFCS-induced vasorelaxation. Here again, the incubation of the MAB with L-NAME attenuated signifcantly but not totally the EFCS-induced vasorelaxation. Tis fnding suggests that the latter is at least NO dependent and that the remaining part is caused by additional endothelium relaxing substances like PGI 2 and/or EDHF. Also, we observed that NO inhibition release by L-NAME potentiated the PHE-induced contraction, explaining the reason why we had lowered PHE concentrations from 10-20 μM to 2.5-5 μM to achieve a PP close to 80-100 mmHg.
Tis is consistent with numerous other studies demonstrating that NO inhibition enhances the MAB's      [29,37]. Moreover, the blockage with ODQ of the NO second messenger, the cGMP, afected signifcantly EFCS's efect, confrming that this vasorelaxation requires at least the NO-cGMP pathway. We tried to assess if PGI 2 mediated the EFCS-induced vasorelaxation by incubating the intact MAB with 10 μM indomethacin to inhibit COX. In this condition, the induced vasorelaxation was not signifcantly modifed, suggesting a marginal contribution of PGI 2 in the relaxation. In contrast, the association of indomethacin and L-NAME showed a signifcant attenuation of the vasorelaxation, suggesting that COX inhibition potentiates the NO inhibition release. Similar fndings were reported in our previous studies [7,28].
In order to investigate the involvement of EDHF in EFCS's efect, vasorelaxation mediated by the opening of K + channels was prevented by precontracting MABs with a solution containing high concentration of KCl (100 mM). In this condition, the preparation is rather depolarized preventing the K + -dependent relaxation [38]. Te EFCSinduced vasorelaxation on MABs precontracted with high potassium preparations was signifcantly diminished compared to the preparations precontracted with PHE. Tis suggests that activation of K + channels is also involved in the C. sativa extract efect.
To confrm the involvement of K + channels in the EFCS's vasorelaxant efect, additional assays were performed in the presence of 3 potassium channel inhibitors (e.g., TEA, glibenclamide, and BaCl 2 ). Our results revealed that simultaneous preincubation of the MAB with these blockers, drastically reduced the EFCS-induced vasorelaxation, suggesting the involvement at least of these 3 K + channels in the vascular efect. When the MAB was incubated simultaneously with L-NAME, Indomethacin, and the latter blockers of the three types of K + channels, the EFCS's efect was completely abolished, confrming that the observed relaxation is totally endothelium dependent and mediated at least by NO production and K + channels opening corresponding to EDHF.
Previous studies already reported that nonpsychoactive cannabinoid induces vasorelaxation in diferent vascular beds, through cannabinoid receptors and endothelium activation [20,21]. In this study, we found that EFCS induces arterial dilation via NO and EDHF, but the frst target of our extract seems to be muscarinic receptor since atropine abolished this vasorelaxation. Tis result is in concordance with our previous study [7] and others [22,23] who showed that arterial relaxation by medicinal plant extracts can be mediated by the activation of muscarinic receptors.
In this study, we also confrmed that the increased level of NO release by aortic rings when incubated with EFCS is similar to that of ACh incubation, suggesting the implication of eNOS activation and muscarinic receptors in the efect also in conductance arteries. Tis fnding suggests that the extract is potentially able to improve endothelial function in both resistance and conductance arteries.
Te phytochemical analysis initiated on our extract showed the presence of compounds pertaining to diferent families such as alkanes/hydrocarbons, fatty acids derivatives, sterols, and noncannabinoids phenolic compounds among others, which were previously reported in C. sativa extracts [9,12,13,16]. Diferences and discrepancies which may be evidenced between the phytochemical profle observed in this work and those previously reported may be due to diferent conditions such as climate, soil composition, date and place of harvest, and orientation.
Te GC-MS analysis of EFCS is also notably signifcant because it confrms that our extract does not almost contain the more known cannabinoids as CBD or CBG, even not THC, but shows that it contains high levels of undecane, neophytadiene, squalene, β-sitosterol, heptacosane, and 17pentatriacontene. β-sitosterol and squalene have been previously reported in Cannabis species [39,40].
Te HPLC analysis confrmed the presence of phenolic compounds in EFCS. Among the nine compounds detected, fve have been tentatively identifed from their UV absorbance and mass spectral characteristics. Te fve compounds were concluded to pertain to the favonoids' family [41][42][43][44] as it was clearly observed on the chromatographic profle recorded at 350 nm where the area of these fve peaks washigher than those recorded at 280 nm. Te fve compounds were tentatively identifed as apigenin and derivatives of luteolin skeleton.
In addition to these compounds, we tried to check for other metabolites such as condensed tannins. Teir occurrence was assayed through phloroglucinolysis reaction [41] which demonstrated the absence of such derivatives in the investigated extract.
Apigenin and luteolin have been already detected and isolated from Cannabis [45]. In addition to their welldocumented antioxidant efect, those favonoids have been described to exhibit endothelium-dependent vasorelaxant efect [22,23]. Also, our results are in concordance with those of Abdallah et al. [6] who show that β-sitosterol induces signifcant vasodilation activities that were blocked by either endothelial denudation or L-NAME (NOS inhibitor), pointing towards a role of endothelial NO in their activities.
Tus, the presence of neophytadiene, squalene, β-sitosterol, apigenin, and luteolin derivatives in high amounts in EFCS could be responsible for the antioxidant and the vasorelaxant efects of EFCS.
Te summary of our study on precontracted isolated rat MAB shows that the efect of EFCS was drastically diminished by NO-synthase inhibition and potassium channels blockers, and totally abolished by endothelium removal. Tese results are relevant for the better understanding of C. sativa threshing compounds benefcial actions on the cardiovascular system, and could suggest the use of the extract to improve endothelial functions. Since hypertension can be due to increased peripheral resistance and endothelial dysfunction, and since vasodilator and antioxidant activities of EFCS were proven only ex vivo, a follow up in vivo studies on hypertensive experimental animal models are required in Evidence-Based Complementary and Alternative Medicine order to assess the in vivo efect of EFCS's compounds and to prove their potential therapeutic use against high-bloodpressure-related cardiovascular diseases.

Data Availability
Data in supplementary information fle concern the repeated individual measurements of each parameter investigated in this study. Data will be made available upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.