Synergistic Effect of Ferulic Acid and Z-Ligustilide, Major Components of A. sinensis, on Regulating Cold-Sensing Protein TRPM8 and TPRA1 In Vitro

Angelica sinensis has been used to attenuate cold-induced cutaneous vasospasm syndrome, such as Raynaud's disease and frostbite, in China for many years. Ferulic acid (PubChem CID: 445858) and Z-ligustilide (PubChem CID: 529865), two major components extracted from Angelica sinensis, had been reported to inhibit vasoconstriction induced by vasoconstrictors. In this study, the pharmacological interaction in regulating cold-induced vascular smooth muscle cell contraction via cold-sensing protein TRPM8 and TRPA1 was analyzed between ferulic acid and Z-ligustilide. Pharmacological interaction on inhibiting [Ca2+]i influx evoked by TRPM8 agonist WS-12 or TRPA1 agonist ASP 7663 as well as cold-induced upregulation of TRPM8 was determined using isobolographic analysis. The isobolograms demonstrated that the combinations investigated in this study produced a synergistic interaction. Combination effect of two components in inhibiting RhoA activation and phosphorylation of MLC20 induced by WS-12 or ASP 7663 was also being quantified. These findings suggest that the therapeutic effect of Angelica sinensis on cold-induced vasospasm may be partially attributed to combinational effect, via TRPM8 and TPRA1 way, between ferulic acid and Z-ligustilide.


Introduction
The roots of Angelica sinensis (RAS) are a well-known traditional Chinese medicine which had been used for thousands of years in China. RAS is widely accepted for its anticancer, memory, neuroprotective, and immunoregulatory effects [1]. It has also been used to cure some cold-induced cutaneous vasospasm diseases such as Raynaud's phenomenon and frostbite [2]. Ferulic acid and Z-ligustilide, which proved to be effective in cardiovascular disease, are two major components extracted from Angelica sinensis [3,4]. However, mechanism of RAS in regulating cold-induced vasospasm is still unknown.
Ion channels, especially Ca 2+ -permeable channels, play an important role in regulating vascular tone [5]. Members of transient receptor potential (TRP) family that are considered as novel nonvoltage Ca 2+ permeable cation channels attract people's attention in vascular research. Several members of TRP family exhibit temperature-sensitive characters [6]. TRP channels, expressed in sensory neurons of dorsal root ganglia (DRG) and trigeminal ganglia (TG), are the primary detectors for sensing environmental change and stimuli [7]. Among them, TRPM8 and TRPA1 are sensors for noxious cold and innocuous cool temperature, respectively [6,8,9]. TRPM8 can be activated by some cooling compounds, such as menthol or icilin, while TRPA1 can be activated by mustard oil or AITC. Recently, several studies proved that TRPM8 and TRPA1 not only existed in sensory neurons but also could be detected in rat aorta and pulmonary artery [10]. Both of them are involved in the regulation of vascular tone, especially response to cold environment [5,11,12].
Local cold exposure leads to an initial vasoconstriction which protects against heat loss, followed by a vasodilation which prevent local area from cold-induced injuries such 2 Evidence-Based Complementary and Alternative Medicine as frostbite [13]. The complex regulator mechanisms involve a combination of neurotransmitter synthesis and release, Ca 2+ homeostasis, adrenergic receptor function, or VSMC contractile ability [14]. Sympathetic nerve, sensory nerve, and nonneuronal factors, such as NOS system, contributed to the cutaneous vasoconstrictor response to local cooling [14,15]. Contraction of vascular smooth muscle cell (VSMC), influenced by systems mentioned above, controlled vascular tone directly. Constriction of VSMC can be achieved by an increase of intracellular Ca 2+ influx, which is recognized as "calcium-dependent," as well as an increase of Ca 2+ sensitivity, which is recognized as "calcium sensitization" [16]. TRPM8 and TRPA1 could regulate vascular tone through sensory nerve, sympathetic nerve, and NOS system [12,17].
In order to discuss TRPM8 and TRPA1 direct effect on vascular smooth muscle cell contraction, TRPM8 specific agonist WS-12 and TRPA1 specific agonist ASP 7663 were used. Both "calcium-dependent" pathway and "calciumsensitization" pathway were explored. Since cold temperature can activate TRPM8 and TRPA1 ion channel, its regulation on TRPM8 and TRPA1 expression was discussed as well.
Since prolong exposure to cold environment leads to some local area cold-induced injuries which could be attenuated by application of Angelica sinensis, synergistic effect of ferulic acid and Z-ligustilide on regulating TRPM8/TRPA1 function and expression would be discussed.

Cell
Culture. The human aortic smooth muscle cell (Sci-enCell, Carlsbad, CA, USA) was cultured in smooth muscle cell medium (ScienCell, Carlsbad, CA, USA) containing 10% FBS and antibiotics (100 U/mL penicillin and 100 g/mL streptomycin) at 37 ∘ C in a humidified incubator with 5% CO 2 .

Calcium Imaging.
Human aortic smooth muscle cells on coverslip were loaded with 5 M fluo-4 and 0.01% pluronic acid in HBSS for 60 min at 37 ∘ . Coverslip was placed on fluorescence microscope (Olympus IX71, Xuhui District, Shanghai). Fluorescence signal was measured at 561 nm. Images were captured every 4 s for 15 min with a ×5 objective lens and analyzed using ImagJ software. For each coverslip at least 50 cells in each image were included as a region of interest (ROI). The average intensity of ROIs was calculated at each time-point. The highest fluorescence ( max ) was recognized as the maximal response to agonist WS-12 or ASP 7663 (50 M; set at 100%) whereas the baseline fluorescence ( 0 ) was the weakest and stable fluorescence before agonist added. Change of fluorescence max / 0 was recorded.

RNA Extraction and RT-qPCR Analysis.
Cells were seeded at a density of 2 × 10 5 cells/dish and incubated with ferulic acid, Z-ligustilide, or their combination in the presence of cold stimulation. RNA was extracted from HASMC using Trizol reagent (Invitrogen Life Technologies, Darmstadt, Germany) under manufacturer's instructions. The RNA quantity was estimated by a nanodrop spectrophotometer (Thermo Scientific, Rockford, IL). cDNA was made from 500 ng of total RNA using PrimeScript6 RT Master Mix (TAKARA, Kyoto, Japan). The resulting cDNA was used to perform real-time fluorescence quantitative PCR reactions in triplicate using SsoFast6 EvaGreen5 Supermixes (Bio-Rad, California, USA) and CFX Connect6 Real-Time PCR Detection System (Bio-Rad, California, USA) operated by CFX Manager Software. RTFQ PCR was initiated with 1 × 98 ∘ C for 2 minutes, followed by 39 cycles of 98 ∘ C for 2 seconds and 58 ∘ C for 5 seconds, and finalized with melt curve for 75 ∘ C to 95 ∘ C for 10 seconds. Forward and reverse primers were designed and ordered from Sangon Biotech (Songjiang, Shanghai, China). -actin was used as reference gene. Primer sequences were as follows: TRPM8 forward CAG CAC TGG CAC CTG AAA AC; TRPM8 reverse GGA CTG CGC GAT GTA GAT GA; TRPA1 forward GCT ACT CTC TAA AGG TGC CCA AG; TPRA1 reverse CGT TGT CTT CAT CCA TTA CCA G; -actin forward GGG AAA TCG TGC GTG ACA TTA AGG; -actin reverse CAG GAA GGA AGG CTG GAA GAG TG. The expression levels of the target genes were normalized to -actin.

Western
Blotting. Lysis of HASMC was made in RIPA including protease inhibitor purchased from Beyotime Biotechnology (Songjiang, Shanghai, China). Lysate concentration was determined using BCA assay and a standard curve. Equal amounts were loaded on 8%-12% Bis-Tris gels, subjected to electrophoresis, and transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were probed with appropriate primary antibodies and then incubated with the horse-radish peroxidase-conjugated secondary antibodies. Immunoreactive bands were visualized using the chemiluminescence detection kit (CST, Beverly, MA, USA).

Detection of Total and GTP-Bound
RhoA. HSMAC were seeded in 60 mm tissue culture dish and grow to 30% confluence. Then they were made quiescent by serum starvation (0% FBS) for 24 hours. Active and GTP-bound RhoA were measure from cell lysates using Rhotekin-RBD bead pulldown assay according to the manufacturer's instruction. Total RhoA was determined by Western blot with an anti-RhoA and was used to normalize GTP-RhoA expression. Activated RhoA in stimulated samples was compared to the level of GTP-RhoA in unstimulated samples.

Analysis of Combined Drug Effects.
We used the Calcusyn software (Biosoft, Version 2.1) to determine drug synergy by the isobolograms and combination index originated from the median effect principle of Chou and Talalay [18]. The isobolograms method, which is formed by selecting a desired fractional affect (Fa), represents the pharmacologic interaction. A straight line was drawn to connect the Fa points plotted with fixed ration between ferulic acid and Z-ligustilide on -and -axes to generate isobolograms. Combination data points falling on the line represented an additive drug-drug interaction while points falling below or above the line represented synergism or antagonism. The combination index (CI) was calculated by the following formula [19]: were the concentrations of drug A and drug B, used in combination to achieve % drug effect. ( ) A and ( ) B were the concentrations of individual components to achieve the same effect. CI values were generated over a range of Fa levels from 0.05 to 0.90 (5%-90% inhibition). CI of 1 indicated additive effect between the two components whereas CI < 1 or CI > 1 indicated synergism or antagonism separately.

Statistical Analysis.
Each experiment was performed using 3 individual coverslips of cells and data are presented as mean ± SEM. Data statistical analysis was performed using two-tailed Student's -test or one-way analysis of variance followed by the SNK post hoc test. values of <0.05 were considered statistically significant.

Responses to TRPM8 and TRPA1 Agonist-Evoked
Ca 2+ Signal 3.1.1. Contribution of Ca 2+ Store to TRPM8 and TRPA1 Agonist-Evoked Ca 2+ Signal. We first examined the effects of TRPM8 specific agonist WS-12 50 M ( max / 0 = 7.186 ± 0.8334) and TRPA1 specific agonist ASP 7663 50 M ( max / 0 = 5.153 ± 0.9831). As illustrated in Figure 1(a), both of them could induce significant change of fluorescence. TRPM8 and TRPA1 were Ca 2+ permeable cation channels which could activate calcium induced calcium release (CICR) through either ryanodine receptor (RYR) or inositol trisphosphate receptor (IP 3 R). To assess if Ca 2+ release via RYR and IP 3 R contributes to TRPM8 and TRPA1 agonistevoked Ca 2+ signal, experiments were carried out with the presence of RYR inhibitor or IP 3 R inhibitor. As shown in Figure 1(b), application of a high concentration of RYR inhibitor ryanodine (30 M) could significantly reduce both ASP 7663 ( max / 0 = 3.384 ± 0.8142) and WS-12 ( max / 0 = 5.066 ± 0.6962) evoked [Ca 2+ ] i , while application of IP 3 R inhibitor 2-APB (100 M) or xestospongin C (5 M) could only reduce ASP 7663 ( max / 0 = 3.499 ± 0.4609 for 2-APB and max / 0 = 3.122 ± 0.6891 for xestospongin C) but not WS-12 ( max / 0 = 6.395 ± 1.030 for 2-APB and max / 0 = 6.76 ± 0.6533 for xestospongin C) evoked [Ca 2+ ] i . These data indicated that Ca 2+ release via RYR contributes to TRPM8 agonist-evoked Ca 2+ signal, while Ca 2+ release via both RYR and IP 3 R contributes to TRPA1 agonist-evoked Ca 2+ signal.   Figure 3 and    (Figure 4(c)). Figure 5(a) showed higher gene expression of TRPM8 from HASMC incubating at 10 ∘ C (5% CO 2 ) for 4 hours, when compared to control group. TRPA1 gene expression was not influenced by 10 ∘ (Data not shown). Different concentration of ferulic acid and Z-ligustilide could inhibit upregulation of TRPM8 mRNA expression (Figure 5(a)). The individual IC 50 values ( Figure 5(b)) of ferulic acid (3.416 ± 1.04 M) and Z-ligustilide (2.325 ± 1.05 M) were used to define four concentrations (Table 3) of the drugs in the combination at a fixed ratio for response to cold-induced  upregulation of TRPM8. Isobolograms ( Figure 5(c)) and CI values (Table 4) represented that ferulic acid and Zligustilide synergistically inhibited cold-induced upregulation of TRPM8 expression. TRPM8 and TRPA1 ion channel have been improved to play an important role in processing pain, but the role in cardiovascular system is still controversial. Role of TRPM8 in mediating vasoconstriction and vascular tone is full of paradox. On the one hand, cold-induced vasoconstriction in hind paw was reduced in TRPM8 KO mice or in WT mice pretreated with TRPM8 antagonist AMTB [5]; meanwhile, cold-induced contraction of rat gastric fundus can be attenuated by TRPM8 antagonist as well as ROCK antagonist Y-27632 [20]. On the other hand, Johnson et al. [12] proved that TRPM8 nonspecific agonist menthol caused moderate contraction in relaxed tail artery, but relaxation of vessels precontracted with KCL or phenylephrine. It is possible that TRPM8 channel in sympathetic terminals can cause release of neurotransmitters, resulting in vasoconstriction under relaxed condition. Besides, TRPM8 may be activated in sensory axon collateral nerves which can release vasodilator substances such as calcitonin gene-related peptide [21]. As far as TRPA1 was concerned, activation of TRPA1 could induced either vasoconstriction accomplished by norepinephrine and ROS [5] or vasodilation induced by vasodilator substance released from sensory nerve [17,22].  Our results showed that both TRPM8 specific agonist WS-12 and TRPA1 specific agonist ASP 7663 could extent phosphorylation of MLC 20 , that is, stimulated HASMC contraction directly. However, some differences of contraction mechanisms existed. TRPA1-induced contraction was achieved by calcium-dependent and calcium-sensitization way, while TRPM8-induced contraction was achieved mainly through calcium-dependent way. As far as CICR was concerned, activation of TRPM8 could induce calcium released from ryanodine receptor while TRPA1 could activate both ryanodine receptor and IP 3 receptor.

Response to Cold-Induced Upregulation of TRPM8 and TRPA1.
In the last part, we discussed cold-induced upregulation of TRPM8 and TRPA1. Both oxaliplatin-induced acute cold-hypersensitivity model and chronic construction injury (CCI) model of neuropathic pain in rat were accompanied with upregulation of TRPM8 and TRPA1 expression in dorsal root ganglia (DRG) [23][24][25]. Whether cold temperature can alter TRPM8 and TRPA1 expression in vascular system is still unknown. It is reported that phosphorylation of ERK and p38 in DRG appeared when the peripheral tissue was stimulated by cold temperature. Furthermore, immunohistochemistry proved that p-p38-IR neurons heavily colocalized with TRPA1, whereas the majority of p-ERK-IR neurons were positive for TRPM8 [26]. Cold-induced upregulation of p-ERK may contribute to upregulation of TRPM8. Our results indicated that cold stimulation could induce upregulation Evidence-Based Complementary and Alternative Medicine 9 of TRPM8 but not TRPA1. In previous experiment, we discovered that amount of TRPA1 was much lesser than TPRM8 in human aortic smooth muscle cell. Expression of TRPA1 may not enough to be detected by qRT-PCR. Further investigation is needed.
A. sinensis has been used as a medical plant in numerous traditional Chinese medicine prescriptions for thousands of years. Previous studies showed that the effective constituents of A. sinensis extract could be classified into water-soluble parts and essential oil. One of the major constituents of water-soluble parts is ferulic acid, while Z-ligustilide has been reported as one of the major essential oil components [3,4]. It is proved that Z-ligustilide can inhibit vasoconstriction induced by vasoconstrictors, such as norepinephrine and calcium chloride, on rat aortic segment. It may also protect ischemic brain injury in rats due to inhibiting the calcium influx through voltage-dependent calcium channel and receptor-mediated calcium [27]. Ferulic acid can induce concentration-dependence aorta vasodilation in chronic renal hypertensive rats [28]. It may also reduce carotid arterial pressure in anesthetized spontaneously hypertensive rats [29]. Synergistic inhibiting effect of these two components on calcium influx and combination inhibiting effect on RhoA-GTP induced by TRPM8 and TRPA1 agonist as well as synergistic inhibition on cold-induced upregulation of TRPM8 indicated that A. sinensis may attenuate, at least partially, vasoconstriction induced by cold-sensing channels TRPM8 and TRPA1.

Conclusion
Activation of TRPM8 and TRPA1 could induce contraction of human aortic smooth muscle cell via calcium influx and RhoA pathway. Cold temperature could upregulate TRPM8 expression. Synergistic effect on inhibiting calcium influx and cold-induced upregulation of TRPM8 by ferulic acid and Zligustilide was discovered.