*2.4. B. monnieri Compounds and Intracellular Ca2*<sup>+</sup> *Release*

The release of intracellular Ca2<sup>+</sup> from the sarcoplasmic reticulum is another important trigger of vascular contraction. Denuded arterial rings were pre-incubated in Ca2+-free Krebs' solution for 10 min and then 10 μM PE added thereby eliciting a transient contraction. Then the protocol was repeated with the same arterial ring in the presence of the test compounds (control, apigenin, luteolin, bacoside A and bacopaside I) producing reduced contractions (98.8 ± 1.2, 50.1 ± 8.5, 54.3 ± 14.9, 85.8 ± 7.2 and 66.2 ± 2.9%, respectively) (Figure 4). Luteolin, apigenin and bacopaside I caused significant decrease in PE-induced contraction compared to the vehicle control (*p* < 0.001, <0.01 and <0.001, respectively).

**Figure 4.** PE-induced contraction induced by Ca2<sup>+</sup> release from sarcoplasmic reticulum of endothelial denuded mesenteric arteries in the presence of DMSO (control), 10 μM of luteolin, apigenin, bacoside A and bacopaside I. The data is % contraction to 10 μM PE induced contraction compared to contractions produced by the initial protocol without test compound. Values are mean ± SEM of 5–6 individual arteries. \*\* *p* < 0.01, \*\*\* *p* < 0.001 each of the active compounds compared with control using unpaired Student's *t*-test (n = 5–6).

#### **3. Discussion**

This is the first study comparing the vasodilatory mechanisms elicited by saponins (particularly bacoside A and bacopaside I) and the principal flavonoids (luteolin and apigenin) were the most potent (EC50 4.4 and 8.9 μM) (Figure 1). However, these are present in BME at only about 1/20th the contents of the bacoside A saponins and bacopaside I (Figure S1 and Table S1) [42]. Thus in terms of the overall actions of the complete BME, the saponins would be expected to make a larger contribution to the vasorelaxation than the flavonoids.

However, higher potency of aglycone flavonoids compared to saponin glycosides may be due to sugar moieties interfering with the molecule interacting with the binding sites responsible for the vasorelaxation as suggested by previously, i.e., lipophilic groups in the ring skeleton of flavonoids increased their vasorelaxant activity [43]. This provides a basis for study of the molecular mechanisms of vasorelaxation of flavonoids.

We investigated the mechanisms of flavonoid- and saponin-induced relaxation by endothelial denudation in mesenteric arterial rings which impaired vasorelaxation (Figure 2). Role of NO was investigated using the eNOS inhibitor (L-NAME) with the test compounds. L-NAME increased EC50 and reduced Emax which imitated the effect of endothelial denudation, suggesting the relaxation was mainly medicated by NO. This accords with observations made by Jin et al. that a cyclooxygenase (COX) inhibitor did not affect the relaxation induced by apigenin [44], and consistent with our previous study of *B. monnieri* extract, where indomethacin had no effect on vasorelaxation [29]. There were some important concentration dependent differences between flavonoids and saponins. Firstly, denudation or blockade of eNOS reduced the effect of bacoside A more than bacopaside I, luteolin and apigenin. Perhaps this was a reflection of bacoside A being a mixture of saponins. However, curiously the responses of luteolin and apigenin to denudation and L-NAME where the latter had a greater effect.

Vascular smooth muscle express plasma membrane L-type Ca2<sup>+</sup> channels that allow depolarisation dependent Ca2<sup>+</sup> entry to trigger contraction. All three compounds (luteolin, apigenin and bacopaside I) tested in denuded vessels depressed this mechanism of contraction that can also explain in part, the vasorelaxant effect. But here, apigenin appeared to be more effective than luteolin while it was less effective in relaxation studies suggesting some heterogeneity in the mechanism of flavonoid action.

Ca2<sup>+</sup> release from intracellular stores also regulates contraction via inositol trisphosphate (IP3) or ryanodine receptors (RyR) associated channels in the SR membranes. IP3 associated channels are commonly activated by plasma membrane G-protein coupled receptors including α1-receptors which are activated by PE. RyR channels are activated by Ca2<sup>+</sup> itself. The three pure compounds also inhibited Ca2<sup>+</sup> released from stores which can account for at least some vasorelaxation of vessels precontracted by PE. However, the bacoside A was without clear effect again suggesting some heterogeneity between the four test substances. Other Ca2<sup>+</sup>-channels may also be involved, for example T-channels and TRP channels, especially TRPC4 which is activated by α1-receptor activation.

K<sup>+</sup> channels also play a role in regulation of vascular tone, i.e., voltage-dependent K<sup>+</sup> (Kv) channels open upon depolarization of the plasma membrane in vascular smooth muscle cells, and thus inhibits Ca2<sup>+</sup> influx through VOCCs, resulting in vasodilation [45]. Jiang et al. also reported that luteolin inhibited Ca2<sup>+</sup> channels, inhibited release of stored Ca2<sup>+</sup> while K<sup>+</sup> channels were activated, specifically via KATP, KCa, KV and KIR [40] therefore the effects of apigenin, bacoside A and bacopaside I involving K<sup>+</sup> channels deserve further investigation. Our findings support those of Si et al. that luteolin can directly act on vascular endothelial cells, by inducing eNOS phosphorylation at Ser1177, leading to NO production [41]. The flavonoids evoke relaxations and also protect endothelial dependent vasorelaxation against oxidative stress [44,46,47] and diabetes [48], however vasoprotective effects of saponins needs further comprehensive investigation.

#### **4. Materials and Methods**

### *4.1. General Information*

Tissues were from male Wistar rats (200–300 g) which were obtained from Nomura Siam International Co. Ltd. (Bangkok, Thailand). Experiments were approved by the Naresuan University Animal Care and Use Committee (NUACUC), protocol number NU-AE 600710. The rats were housed under the environmental conditions at 22 ± 1 ◦C, 12-h light and dark cycle, fed with standard rodent diet and tap water in Naresuan University Center for Animal Research (NUCAR) according to the guidelines for care and use of laboratory animals (Institute of Laboratory Animal Research, eighth edition 2011. Rats were anesthetized by intraperitoneal injection of thiopental sodium (100 mg/kg BW) and killed. The mesenteric arteries were excised, cleaned of surrounding loose connective tissue and cut into rings of 3–5 mm width. In some experiments, endothelial cells were mechanically removed by gently rubbing the lumen with a stainless steel wire. The mesenteric rings were mounted on a pair of intraluminal wires in organ chambers containing physiological Krebs' solution (mM): NaCl, 122; KCl, 5; [N-(2-hydroxyethyl) piperazine N'-(2-ethanesulfonic acid)] HEPES, 10; KH2PO4, 0.5; NaH2PO4, 0.5; MgCl2, 1; glucose, 11; and CaCl2, 1.8 (pH 7.3), at 37 ◦C and aerated [29,49–51]. The vessel segments were allowed to equilibrate for 1-h at a resting tension of 1–1.3 g during which the solution was replaced every 15 min. Changes in isometric tension were measured using force transducer lever (CB Sciences Inc., Milford, MA, USA) connected to a MacLab A/D converter (Chart V7; A.D. Instruments, Castle Hill, NSW, Australia), stored and displayed on a personal computer. Following stabilization, the arterial rings were tested for viability by the application of 10 μM PE. Upon development of a steady contraction, the endothelium status was tested with 10 μM ACh. The vessel was considered endothelial intact when the ACh induced >70% relaxation. After establishing the status of the endothelium, the rings were then rinsed with Krebs' solution for 30 min and one of the following protocols was initiated. Luteolin (lot 126M4061V) and apigenin (lot WE445301/1) were purchased from Sigma Aldrich (St. Louis, MO, USA). Bacoside A (lot 00002005-003) and bacopaside I (lot 00002002-T17H) were purchased from ChromaDex, Inc. (Irvine, CA, USA).
