Suxiao Jiuxin Pill, Ligusticum chuanxiong Hort. [Apiaceae, Chuanxiong Rhizoma] extract and Bornolum syntheticum were provided by the Tianjin Darentang Group Co., Ltd., the sixth TCM factory (Tianjin, China). Voucher specimens were deposited in the Institute of Traditional Chinese Medicine, Tianjin University of Traditional Chinese Medicine. Scopoletin, senkyunolide H/I/A, sedanolide, ligustilide, angelicide, isochlorogenic acid C, vanillin, tetramethylpyrazine, isochlorogenic acid B, ferulic acid, caffeic acid, isochlorogenicacid A, cryptochlorogenic acid, protocatechuic acid, vanillic acid, neochlorogenic acid, succinic acid, chlorogenic acid, and isoborneol, (Z)-3-butylidenephthalide were purchased from Yuanye Bio-technology Co. Ltd. (Shanghai, China). Acetylcholine chloride (Ach), 9,11-dideoxy-11α, 9α-epoxy-methanoprostaglandin F_2α (U46619), N^G-nitro-l-arginine methyl ester (L-NAME), indomethacin (INDO), methylene blue (MB), glibenclamide [Gli, ATP-sensitive K⁺ channels (K_ATP) inhibitor], 4-aminopyridine [4-AP, voltage-dependent K⁺ channels (K_V) inhibitor], tetraethylammonium [TEA, Ca²⁺-activated K⁺ channels (K_Ca2+) inhibitor], BaCl₂ [inward rectifier K+ channels (K_ir)inhibitor], diltiazem (L-type calcium channel inhibitor), and dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich (St. Louis, MO). HPLC grade acetonitrile was obtained from Fisher Scientific (Fair Lawn, NJ, USA). Potassium chloride (KCl) was bought from Guang Fu Technology Development Co. Ltd. (Tianjin, China). Uprosertib and 1H- [1,2,4] oxadiazolo [4,2-α] quinoxalin-1-one (ODQ) were acquired from MedChemExpress Co. Ltd. (Shanghai, China). All the chemicals were of analytical grade.

Male Sprague-Dawley rats (200–250°g), 8–10°weeks old were purchased from SPF Biotechnology Company Ltd. (Beijing, China). All animals were kept in the Animal Center of Tianjin University of Traditional Chinese Medicine under standard conditions of 24°C–27°C, 60–70% relative humidity, and a 12 h light-dark cycle. All animals were fed a regular chow diet and water ad libitum. The Tianjin University of Traditional Chinese Medicine Institutional Animal Use and Care Committee approved all experimental procedures in this study with the following approval: No. TCM-LAEC2020079.

The rat AMI model was established by LAD ligation surgery (Li, J. et al., 2021). Rats were randomly divided into 5 groups: Sham, AMI model, SJP-IG group (54 mg/kg, intragastric gavage), SJP-IP group (54 mg/kg, intraperitoneal injection), and nitroglycerin group (0.0071 mg/kg, intragastric gavage). The SJP and nitroglycerin doses are the human equivalent doses calculated based on body surface area. SJP and nitroglycerin were administered immediately after ligation. Rats in both the sham and AMI groups received the same volume of saline.

Transthoracic echocardiography was performed using an ultrasound system (Vinno, Suzhou, China) with a 22 MHz transducer at 5, 10, and 20 min after SJP and nitroglycerin treatment. The left ventricular internal dimension diastole (LVIDd), left ventricular internal dimension systole (LVIDs), left ventricular ejection fraction (EF), and fractional shortening (FS) were obtained with the heart rate controlled between 400 and 450 bpm. Three consecutive cardiac cycles were averaged to measure cardiac function.

SD rats were anesthetized and kept in the supine position. Needle electrodes were inserted subcutaneously according to lead II (right front leg, left hind leg, and left front leg) per the ECG protocol. ECG was recorded using PowerLab connected to BioAmp and analyzed by LabChart 8 software (AD instrument, Australia). The ECG signal was continuously monitored before ligation and for 20 min after drug administration. Changes in ECG patterns (ST segment) were analyzed.

Gene targets of 14 identified serum compounds were summarized from TCMSP (https://old.tcmspe.com/tcmsp.php), BATMAN-TCM (http://bionet.ncpsb.org.cn/batman-tcm/index.php), and Swiss Target Prediction (http://www.swisstargetprediction.ch/). AMI-associated targets were retrieved from OMIM (https://omim.org/), GeneCards (https://www.genecards.org/), and CTD (http://ctdbase.org/). Cytoscape 3.7.2 and DAVID (https://david.ncifcrf.gov/) were used to filter key targets and search for key compounds and pathways. The compounds-targets-pathway network was visualized through Sankey Tools (https://www.zxgj.cn/g/sankey).

The fresh heart was removed from SD rats and immediately immersed in a petri dish containing a K-H solution composed of NaCl 118 mmol/L, KCl 4.7 mmol/L, NaHCO₃ 25 mmol/L, KH₂PO₄ 1.2 mmol/L, MgSO₄ 1.2 mmol/L, CaCl₂ 1.3 mmol/L, and d-glucose 10 mmol/L. The rat coronary arteries were carefully separated and cut into 2 mm rings. The coronary artery from one rat was cut into 2 segments for each sample measurement. The arterial ring was suspended in an organ bath of wire myograph (DMT A/S, Model 630MA, Denmark) containing K-H solution with an aerating mixture of 95% O₂ and 5% CO₂ at 37 °C.

After equilibration, arterial rings were constricted with U46619 (100 nmol/L) to the maximal contraction, and endothelial integrity was evaluated using cumulative concentrations of Ach (10-9–10–5 mol/L). The endothelium was considered intact when the relaxation rate was more than 80%. Endothelium-denuded rings were obtained by rubbing the inner surface of the artery several times with rat whiskers. Finally, the cumulative concentration-response curve for the SJP and its components was investigated. The EC₅₀ and R_max (maximal vasorelaxation) values were then calculated.

Human umbilical vein endothelial cells (HUVECs) were purchased from AllCells Biotech Co., Ltd. (Shanghai, China). The cells were cultured in an Endothelial Cell Complete Culture Medium (Procell, Wuhan, China) containing 10% FBS and 1% penicillin/streptomycin at 37 °C in a humidified atmosphere of 5% CO₂ and 95% air. The cells within 3-8 passages were used in the following experiments. The cells were serum-starved for 6 h, then treated with senkyunolide A or scopoletin for 15 min. Western blot analysis was performed using anti-eNOS (1:1000), anti-p-eNOS (1:1000, Ser1177), anti-Akt (1:1000), anti-p-Akt (1:1000, Thr308), and anti-GAPDH antibodies (1:1000) (CST, MA, USA).

The X-ray 3D crystal structure of the human Akt1 pleckstrin homology (PH) domain, PDB ID: 1H10 (Thomas et al., 2002), was downloaded from the protein data bank repository (http://www.rcsb.org/). The 3D chemical structure of senkyunolide A, or scopoletin, was obtained from the PubChem database. The X-ray crystal protein preparation and the molecular docking were done using Discovery Studio Client (2019) and Autodock Vina, as described in our previous study (Fang et al., 2022), and the results were visualized with PyMOL.

To evaluate the binding affinity of senkyunolide A or scopoletin to Akt, an SPR assay was performed using the Biacore T200 (Cytiva Corp., MA, USA). The recombinant human Akt1 protein (Bio-techne, Shanghai, China) was immobilized on an S CM5 sensor chip using standard amine-coupling chemistry. A serial dilution of the senkyunolide A (62.5, 125, 250, 500, and 1000 μmol/L) or scopoletin (5, 10, 20, 40, and 80 μmol/L) in running buffer was injected over the human Akt1 protein (1775-KS-010, Bio-Techne). With Biacore T200 evaluation software (Cytiva), the data fit into a 1:1 steady binding model to calculate the binding affinity KD.

SPSS version 21.0 for Windows (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. All values are given as means ± standard error of the mean (S.E.M). Statistical significance was analyzed using one-way or two-way analysis of variance, followed by the Bonferroni post hoc analysis, and p < 0.05 was considered to be statistically significant.

To assess the therapeutic effect of SJP on angina, AMI was induced in rats by LAD ligation, after which cardiac function was examined. In the rat AMI model, EF and FS were significantly reduced, with an increase in LVIDs and LVIDd values after LAD ligation. SJP treatment (54 mg/kg) significantly reduced EF, FS, and LVID values 5–20 min after intraperitoneal injection and 10–20 min after intragastric gavage administration. Nitroglycerin, as a positive control, exhibited a similar effect on cardiac function (Figure 1A–E). The ST segment elevation in ECG was observed in the AMI model group, compared with sham control rats. However, SJP treatment significantly reduced the ST segment elevation caused by LAD ligation (Figures 1F,G).

To identify the active ingredients of SJP, LC-MS, and GC-MS analysis were performed on serum from SJP-treated rats and Agilent Masshunter Qualitative Analysis was used to determine the identities of the various circulating compounds. The base peak chromatogram of the non-volatile compounds for the positive and negative ion modes is shown in Figure 2A. By comparing their retention time, mass-to-charge ratio, and secondary mass spectrum to their respective standard chemical compounds and previous phytochemical investigation reports, a total of 28 compounds were identified (Table 1). In addition, a total of 11 volatile circulating compounds were detected when standard chemical compounds were compared to the NIST database. The GC-MS TIC traces chromatogram is shown in Figure 2B and Table 2 lists the GC-MS profiling of the volatile circulating compounds.

Network pharmacology was used to identify potential pharmacological targets of SJP in the treatment of AMI. A total of 14 compounds found in the serum of SJP-treated rats were mapped to databases of cellular drug targets, and a total of 20 genes, and 15 pathways were identified as a result. To better understand the differential relationship between SJP and AMI, a PPI network was constructed, which identified PTGS2, CASP3, MMP9, and NOS3 genes as the leading targets (Figure 2C). In addition to understanding the cellular connection between SJP and AMI, a pathway enrichment analysis was conducted on the key targets. The cellular pathways through which SJP may exert its beneficial therapeutic effect were identified as lipid and atherosclerosis, fluid shear stress, atherosclerosis, and AGE-RAGE signaling (Figures 2D,E). It was also uncovered that ferulic acid and senkyunolide A were all closely associated with the top two cellular drug targets, PTGS2 and NOS3 (Figure 2F). These findings, combined with the previous observations indicate that PTGS2 and NOS3 play a direct role in the regulation of endothelial function and vascular tone, reveal that SJP improves AMI by regulating artery relaxation.

A test was conducted to ascertain whether SJP could dilate coronary artery rings. Results revealed that SJP has a concentration-dependent vasodilatory effect on the endothelium-intact coronary artery rings (EC₅₀ = 136.1 μg/mL; R_max = 80.72 ± 10.27%). Nitroglycerin, as a positive control, relaxed the coronary artery with an EC₅₀ of 22.1 μg/mL (Figures 3A–C).

In line with the network pharmacology findings, investigations into whether the NOS3 inhibitor L-NAME or PTGS2 inhibitor INDO could prevent SJP-induced coronary artery relaxation revealed that the relaxation effect of SJP was significantly suppressed by L-NAME treatment (R_max dropped from 80.72% to 65.96%) but not by INDO (Figures 3D,E), indicating that SJP modulates the eNOS-NO pathway to induce vasorelaxation.

Extracts of the two active ingredients in SJP (Ligusticum chuanxiong Hort. and Borneolum syntheticum) had potent vasodilatory effects on the endothelium-intact coronary artery rings (Figure 3F). The findings inform that the active compound(s) of SJP is responsible for SJP’s vasorelaxation action. It was therefore tested to determine whether the compounds identified in the serum of SJP-treated mice, as well as other chemical compounds isolated from SJP (Lei et al., 2018; Li, S. L. et al., 2003; Liang et al., 2005), could respond to constricted coronary arteries in the vasoreactivity assay. All the compounds tested exhibited concentration-dependent vasodilation effects on coronary artery rings (Table 3). Scopoletin, sedanolide, butylphthalide, senkyunolide A, and borneol particularly had potent relaxation effects.

Senkyunolide A, a chemical constituent of Ligusticum chuanxiong Hort., was found to relax isolated rat aortas (Chan et al., 2007). In line with the previous finding, this study also shows that senkyunolide A (1 mmol/L) significantly relaxes constricted endothelium-intact coronary artery rings with a R_max of 101.77%. However, in the endothelium-denuded coronary artery rings, senkyunolide A R_max value decreased to 64.43% (Figures 4A,B). Senkyunolide A relaxation curve and R_max value were significantly reduced in coronary artery rings pretreated with L-NAME but not INDO (Figures 4C,D). Furthermore, in the presence of sGC inhibitor ODQ or cGMP inhibitor MB, the relaxation curve and Rmax of senkyunolide A were significantly reduced compared to the control (Figures 4E,F), indicating that senkyunolide A promotes endothelial-dependent vasorelaxation mainly via the eNOS-NO-sGC-cGMP pathway. Coronary arteries are mainly through endothelium-dependent and non-endothelium-dependent dilation. This study revealed that eNOS, sGC, and cGMP inhibitors have limited effect. Senkyunolide A may also play a role in vasodilating through other pathways. Scopoletin, like senkyunolide A, had a potent endothelium-dependent vasorelaxant effect. The R_max of scopoletin in endothelium-intact coronary arteries dropped from 96.58% to 75.98% in endothelium-denuded coronary arteries (Figures 5A,B). Scopoletin relaxation curves decreased significantly in the presence of L-NAME and INDO compared to the control (Figures 5C,D). Furthermore, in the presence of ODQ and MB, the relaxation curves for scopoletin were significantly reduced (Figures 5E,F). These findings indicate that scopoletin induces endothelial-dependent relaxation via both the NO and PGI2 pathways.

The Akt kinase is needed to control the phosphorylation of eNOS and, in turn, NO production (Huang et al., 2020). Thus, the Akt-eNOS-NO endothelium-dependent pathway was studied to determine how senkyunolide A and scopoletin induce endothelium-dependent vasorelaxation. In HUVECs, senkyunolide A or scopoletin treatment increased Akt phosphorylation at Thr308 and eNOS phosphorylation at Ser1177 (Figure 6A–B, Supplementary Figure S1). Following that, in silico molecular docking was used to ascertain whether senkyunolide A or scopoletin could bind to critical regulatory sites on Akt, thereby promoting its phosphorylation.

For PDK1 or mTOR2 to phosphorylate Akt, the pleckstrin homology (PH) of Akt must first bind to the lipid second messengers PI (3,4,5) P3 and PI (3,4) P2 (Truebestein et al., 2021). Senkyunolide A and scopoletin both exhibited a strong binding affinity for the Akt1 PH domain (PDB ID: 1H10), with the former having a binding energy of -4.8 kcal/mol and the latter having a binding energy of -5.3 kcal/mol. Additionally, protein-ligand interaction analyses showed that senkyunolide A forms three hydrogen bonds with GLU17, ILE19, and ARG23, while scopoletin forms six hydrogen bonds with LYS14, GLU17, TYR18, ILE19, ARG23, and ASN53 (Figures 6C,D). This indicates that senkyunolide A and scopoletin may exert a regulatory activity on Akt phosphorylation through their interaction with key binding site residues in the PH domain. The SPR biophysical assay was carried out with recombinant Akt protein to investigate the direct interaction between senkyunolide A/scopoletin and Akt. The SPR assay consistently demonstrated that senkyunolide A or scopoletin interacts with Akt in a dose-dependent manner (Figures 6E,F). The value of the equilibrium dissociation constant (KD) for scopoletin was determined to be 11.09 μmol/L, while the value for senkyunolide A was 752 μmol/L.

Uprosertib, an inhibitor of Akt, decreased the expression of p-eNOS in HUVECs that had been treated with senkyunolide A and scopoletin (Figure 6G, Supplementary Figure S2). Consistently, in ex-vivo, coronary artery rings pretreated with 10 μmol/L of uprosertib, the relaxation curves of senkyunolide A and scopoletin were significantly decreased compared to the negative control (Figures 6H,I), indicating that senkyunolide A and scopoletin trigger endothelium-dependent vasodilation via the activation of the Akt-eNOS-NO signaling pathway.

The EC₅₀ value for borneol-induced concentration-dependent coronary artery ring relaxation was 119.7 μmol/L, and the R_max value was 83.63% (Table 3). However, the R_max value did not decrease in endothelium-deficient coronary arteries, which shows that the vasorelaxation action of borneol was caused by pathways that do not depend on endothelium (Figures 7A,B). The vasorelaxation effect of borneol was blocked in the presence of potassium channel inhibitors like BaCl₂, TEA, or 4-AP (Figures 7C–E). However, the vasodilation activity of borneol was unaffected by the K_ATP inhibitor glibenclamide or the L-type calcium channel inhibitor diltiazem (Figures 7F,G), indicating that borneol likely stimulates the opening of K_ir, K_Ca2+, and K_v channels in vascular smooth muscle cells to induce vasorelaxation.