Exploring the Active Components of Simotang Oral Liquid and Their Potential Mechanism of Action on Gastrointestinal Disorders by Integrating Ultrahigh-Pressure Liquid Chromatography Coupled with Linear Ion Trap-Orbitrap Analysis and Network Pharmacology.

Simotang oral liquid (SMT), a well-known traditional Chinese medicine formula composed of four medicinal and edible plants, has been extensively used for treating gastrointestinal disorders (GIDs) since ancient times. However, the major active constituents and the underlying molecular mechanism of SMT on GIDs are still partially understood. Herein, the preliminary chemical profile of SMT was first identified by ultrahigh-pressure liquid chromatography coupled with linear ion trap-Orbitrap tandem mass spectrometry (UHPLC-LTQ-Orbitrap). In total, 70 components were identified. Then, a network pharmacology approach integrating target prediction, pathway enrichment analysis, and network construction was adopted to explore the therapeutic mechanism of SMT. As a result, 170 main targets were screened out and considered as effective players in ameliorating GIDs. More importantly, the major hubs were found to be highly enriched in a calcium signaling pathway. Furthermore, 26 core SMT-related genes were identified, which may play key roles in ameliorating gastrointestinal motility. In conclusion, this work would provide valuable information for further development and clinical application of SMT.

Introduction

Gastrointestinal disorders (GIDs) are those that affect any part of the gastrointestinal tract, which directly impact on the intake, digestion, and absorption of nutrients.[1] GIDs mainly include functional gastrointestinal disorder, irritable bowel syndrome, inflammatory bowel disease, colorectal cancer, and malabsorptive syndromes. Among them, functional gastrointestinal disorders (FGIDs) are one of the most common health-related conditions in gastroenterology practice.[2] GIDs are highly prevalent and account for considerable health care utilization and spending in the world.[3] It has been reported that the total expenditures for GIDs were about $135.9 billion annually in the United States.[4] The pathogenesis of GIDs is quite complicated, with the current morbidity rate more than 41% worldwide.[5] GIDs could be induced by several factors including infection, genetic factor, smoking, and eating misbehavior.[6] Although many different types of drugs, such as dopamine receptor antagonists (metoclopramide and domperidone), 5-hydroxytryptamine receptor agonist (cisapride and mosapride), and motilin receptor agonist (erythromycin), have been used to alleviate some major symptoms of GIDs, patient outcomes still remain unsatisfying due to the unavoidable adverse effects after long-term use of these western medicines.[7] Therefore, novel and safety therapeutic strategies for GIDs are urgently required.

As an important part of the complementary and alternative medical system, TCM is universally acknowledged by its adoption of “multi-components” to take “multi-effects” on “multi-targets”.[8,9] It has been widely used in treating GIDs for thousands of years, especially the classical prescription called SMT. SMT was first recorded in YAN’s Jishengfang and approved by the Chinese Food and Drug Administration in the1980s.[10] It contains four Chinese herbs including Citrus aurantium L. (CL), Aucklandia costus Falc (AF), Areca catechu L (AL), and Lindera aggregata (Sims) Kosterm.(LK) at the weight ratio of 3:2:3:3. Modern pharmacological and clinical studies have revealed that SMT exerts significant therapeutic effects on functional dyspepsia,[11] the contraction of antral circular smooth muscle,[12] and constipation-predominant irritable bowel syndrome.[13] However, the underlying complex mechanisms of SMT acting on GIDs remain unclear.

With the rapid development of bioinformatics, network pharmacology has emerged as a systemic and promising tool for the mechanistic study of TCM.[14,15] It shifts the “one target, one drug” paradigm to the “network target, multi-component” strategy, which is in line with the holistic view of the TCM theory.[16] Until now, the network pharmacology approach has been successfully used to uncover the synergistic mechanisms of multiple components in TCM for treating specific diseases, such as the therapeutic mechanism of Carthamus tinctorius L. in the treatment of cardiovascular disease,[17] the action mechanism of Pulsatilla decoction (PD) on acute ulcerative colitis,[18] and so on.

In this work, an integrated strategy based on phytochemical analysis and network pharmacology was used to illuminate the complicated mechanism of action of SMT on GIDs. The flowchart of the experimental procedures of the current study is shown in Figure .

Figure 1

Whole framework of this study based on network pharmacology for deciphering pharmacological mechanisms of SMT acting on GIDs.

Results and Discussion

Characterization and Identification of Chemical Constituents Contained in SMT

By using the established UHPLC-LTQ-Orbitrap method, the chemical constituents of SMT in both positive and negative modes were well separated and detected (see Figure S1). In this study, for the compounds with available standards, the compounds were identified by comparing the retention time and high-resolution accurate mass with the reference compounds. Moreover, the MS fragmentation behaviors of the reference substances were summarized, which were really helpful for structural elucidation of their derivatives with the same basic skeleton. For those unavailable standard compounds, the structures were tentatively identified by comparing with previous reports, according to accurate mass, chromatographic behavior, MS/MS data, and fragmentation rules.[19] The mass errors for all the precursor ions of all identified constituents were within ±5 ppm. Overall, 70 components, including 26 flavonoids, 14 alkaloids, 4 coumarins, 5 amino acids, 6 organic acids, 13 terpenoids, and 2 other compounds in SMT were unambiguously or tentatively identified. Detailed information on these compounds is summarized in Tables and and Table S1.

Table 1

Identification of Compounds in SMT by UHPLC-LTQ-Orbitrap in Positive Ion Mode

peak no.	tR (min)	molecular formula	calculated mass (m/z)	measured mass [M + H]+	error (ppm)	MS/MS fragments	identification compound	source
1	1.27	C4H8N2O3	132.0529	133.0608	0.071	116.0343[M + H–NH3]+, 87.0553[M + H–CO–H2O]+, 70.0286[M + H–CO–H2O–NH3]+	asparagine	AF
2	1.41	C5H9NO2	115.0627	116.0704	–0.185	70.0650[M + H–H2O–CO]+	proline	AL
3	1.62	C6H9NO2	127.0627	128.0704	–0.175	110.0601[M + H–H2O]+, 109.0284[M + H–NH5]+, 99.0440[M + H–CH3N]+, 82.0650[M + H–H2O–CO]+, 81.0333[M + H–CH5NO]	guvacine	AL
4	1.69	C7H11NO2	141.0784	142.0862	–0.085	124.0758[M + H–H2O]+, 96.0807[M + H–H2O–CO]+, 81.0333[M + H–CH3OH–CH3N]+	arecaidine	AL
5	1.87	C5H11NO2	117.0784	118.0861	–0.145	72.0807[M + H–H2O–CO]+	valine	AL
6	2.09	C6H6O3	126.0311	127.0389	–0.101	109.0284[M + H–H2O]+, 99.0440[M + H–CO]+, 81.0333[M + H–CO–H2O]+	maltol	AL
7	2.35	C9H13NO2	167.0940	168.1012	–4.314	150.0916[M + H–H2O]+	synephrine	CL
8	2.85	C9H11NO3	181.0733	182.0811	–0.040	165.0551[M + H–NH3]+, 136.0762[M + H–H2O–CO]+	tyrosine	AL
9	3.16	C7H11NO2	141.0784	142.0862	–0.025	110.0602[M + H–CH4O]+, 113.0598[M + H–CH3N]+, 96.0807[M + H–CH2O2]+, 81.0334[M + H–C2H7NO]+	guvacoline	AL
10	3.47	C8H13NO2	155.0940	156.1018	–0.085	138.0553[M + H–CO]+, 124.0758[M + H–CH3OH]+, 113.0598[M + H–CH3–CH2N]+, 81.0334[M + H–CH3–CH3OH–CH2N]+	arecoline	AL
11	4.87	C6H6O3	126.0311	127.0389	–0.031	109.0286[M + H–H2O]+, 81.0335[M + H–H2O–CO]+	triacetate lactone	AL
12	4.93	C9H13NO	151.0991	152.1069	–0.071	121.0651[M + H–CH5N]+, 93.0698[M + H–CH5N–CO]+	N-methyltyramine	CL
13	6.32	C9H11NO2	165.0784	166.0862	–0.105	149.0599[M + H–NH3]+, 131.0493[M + H–NH3–H2O]+, 120.0810[M + H–H2O–CO]+, 80.0493[M + H–C4H6O2]+	phenylalanine	AL
17	11.09	C9H6O4	178.0260	179.0330	–4.721	161.0234[M + H–H2O]+, 151.0390[M + H–CO]+, 147.0442[M + H–O2]+, 137.0235[M + H–C2H2O]+, 123.0441[M + H–2CO]+, 111.0441[M + H–C3O2]+, 95.0487[M + H–C3O3]+	5,7-dihydroxycoumarin	CL
18	13.08	C27H30O15	594.1579	595.1655	–0.277	577.1582[M + H–H2O]+, 559.1478[M + H–2H2O]+, 529.1372[M + H–2H2O–HCHO]+, 511.1264[M + H–3H2O–HCHO]+	lonicerin	CL
19	13.12	C18H19NO4	313.1308	314.1387	0.045	300.1193[M + H-CH2]+, 299.1159[M + H–CH3]+, 285.1006[M + H–C2H5]+, 268.1340[M + H–CO–H2O]+, 151.0758[M + H–C9H9NO2]+	laurolitsine	LK
20	13.62	C21H30O8	410.1935	411.1987	–2.554	393.1896[M + H–H2O]+, 321.1677[M + H–C3H6O3]+, 291.1573[M + H–4CH2O]+, 249.1465[M + H–C6H10O5]+, 231.1359[M + H–C7H16O5]+, 203.0531[M + H–C13H20O2]+, 185.0424[M + H–C13H22O3]+	picriside B	AF
21	13.77	C20H29NO5	363.2040	364.2118	–0.089	318.2074[M + H–CO–H2O]+, 128.0710[M + H–C14H20O3]+	saussureanine D	AF
22	13.91	C9H6O3	162.0311	163.0389	–0.081	145.0286[M + H–H2O]+, 131.0494[M + H–O2]+, 119.0494[M + H–CO2]+, 107.0493[M + H–2CO]+, 91.0542[M + H–C2O3]	7-hydroxycoumarin	CL
23	14.22	C18H19NO4	313.1308	314.1387	0.045	297.1133[M + H–NH3]+, 265.0870[M + H–CH3OH–NH3]+	norisoboldine	LK
25	14.69	C19H21NO4	327.1465	328.1543	–0.005	313.1315[M + H–CH3]+, 297.1133[M + H–CH5N]+, 285.1134[M + H–COCH3]+, 265.0872[M + H–C2H7N–H2O]+, 253.0872[M + H–CH2O–C2H7N]+	boldine	LK
27	14.99	C19H21NO3	311.1515	312.1595	0.030	269.1181[M + H–C2H5N]+, 283.1336[M + H–CH3N]+, 206.1181[M + H–C7H6O]+	pronuciferine	LK
28	15.07	C15H16O4	260.1043	261.1120	–0.166	243.1013[M + H–H2O]+, 217.1203[M + H–CO2]+, 201.0914[M + H–2CH2O]+, 189.0551[M + H–C4H8O]+, 127.0004[M + H–C10H14]+	isomerancin	CL
30	15.35	C15H12O5	272.0679	273.0755	–0.250	153.0186[M + H–C8H8O]+, 147.0446[M + H–H2O–C6H4O2]+, 179.0343[M + H–C6H6O]+	naringenin	CL
32	15.39	C6H8O2	112.0518	113.0595	–0.186	95.0492[M + H–H2O]+, 85.0648[M + H–CO]+, 67.0541[M + H–CO–H2O]+	sorbic acid	CL
34	15.43	C19H23NO4	329.1621	330.1697	–0.305	192.1024[M + H–C8H10O2]+, 299.1286[M + H–CH5N]+	reticuline	LK
35	15.57	C15H12O5	272.0679	273.0756	–0.190	153.0186[M + H–C8H8O]+, 147.0445[M + H–H2O–C6H4O2]+, 179.0343[M + H–C6H6O]+	garbanzol	CL
36	15.60	C21H22O9	418.1258	419.1330	–0.689	404.1112[M + H–CH3]+, 389.0879[M + H–2CH3]+	natsudaidain	CL
38	15.81	C19H21NO4	327.1465	328.1543	–0.005	297.1130[M + H–CH3N]+, 265.0869[M + H–C2H9NO]+	laurotetanine	LK
40	15.93	C16H14O6	302.0784	303.0849	–4.569	179.0343[M + H–C7H8O2]+, 177.0553[M + H–C6H4O2–H2O]+, 153.0187[M + H–C9H10O2]+, 285.0767[M + H–H2O]+	hesperetin	CL
41	15.93	C22H24O10	448.1363	449.1442	–0.513	431.1353[M + H–H2O]+, 413.1249[M + H–2H2O]+, 369.0984[M + H–C2H8O3]+, 345.0984[M + H–C4H8O3]+, 303.0876[M + H–C6H10O4]+, 263.0561[M + H–C9H14O4]+	isosakuranin	CL
43	16.11	C17H16O4	284.1043	285.1097	–2.426	267.1000[M + H–H2O]+, 257.0815[M + H-C2H4]+, 241.1205[M + H–CO2]+, 211.0735[M + H–CH2O2]+, 201.0551[M + H–C5H8O]+	ethyllucidone	LK
44	16.38	C19H21NO4	327.1465	328.1540	–0.315	297.1127[M + H–CH3N]+, 265.0866[M + H–C2H9NO]+, 311.1284[M + H–NH3]+	morphine acetate	LK
45	16.49	C15H16O4	260.1043	261.1120	–0.136	243.1024[M + H-H2O]+, 189.0553[M + H–CH3–C3H5O]+, 177.0553[M + H–C5H8O]+, 217.0865[M + H–CO2]+	linderane	LK
46	16.52	C17H16O5	300.0992	301.1044	–2.650	283.0949[M + H–H2O]+, 243.1011[M + H–C2H2O2]+	pashanone	LK
47	17.13	C19H21NO4	327.1465	328.1542	–0.125	311.1290[M + H–NH3]+	naloxone	LK
49	17.32	C16H14O5	286.0835	287.0914	–0.030	245.0813[M + H–C2H2O]+, 179.0343[M + H–C7H8O]+, 161.0601[M + H–C6H6O3]+, 153.0186[M + H–C9H10O]+	isosakuranetin	CL
50	17.63	C19H18O6	342.1097	343.1154	–2.225	325.1050[M + H–H2O]+, 311.1284[M + H-O2]+, 283.0947[M + H–CO–CH3OH]+	4′,5,7,8-tetramethoxyflavone	CL
51	17.80	C10H16O	152.1195	153.1273	–0.072	135.1172[M + H–H2O]+, 109.1013[M + H–C2H4O]+, 107.0857[M + H–C2H6O]+, 95.0856[M + H–C3H6O]+, 93.0700[M + H–C3H8O]+, 81.0698[M + H–C4H8O]+	2-methyl-5-propan-5-ylcyclohex-2-en-1-one	AF
52	18.10	C20H27NO4	345.1934	346.2012	–0.085	328.3216[M + H–H2O]+, 300.1967[M + H–HCOOH]+, 128.0710[M + H–C14H18O2]+, 100.0758[M + H–C15H18O3]+	saussureanine B	AF
53	18.46	C15H20O2	232.1457	233.1536	0.034	215.1435[M + H–H2O]+, 205.1591[M + H–CO]+, 197.1329[M + H–2H2O]+, 187.1486[M + H–H2O–CO]+	costunolide	AF
54	18.77	C20H29NO4	347.2091	348.2170	0.045	330.2073[M + H–H2O]+,128.0710[M + H–C14H18O]+	saussureanine A	AF
55	18.81	C20H20O7	372.1203	373.1282	–0.009	358.1057[M + H–CH3]+, 343.0824[M + H–2CH3]+	tangeretin	CL
56	19.04	C20H20O8	388.1152	389.1232	–0.136	374.1009[M + H–CH3]+, 359.0774[M + H–2CH3]+, 328.3221[M + H–C2H5O2]+	demethylnobiletin	CL
57	19.12	C15H18O3	246.1250	247.1329	0.029	229.1229[M + H–H2O]+, 211.1123[M + H–2H2O]+, 201.1279[M + H–H2O–CO]+, 183.1173[M + H–CO–2H2O]+, 173.1329[M + H–H2O–2CO]+, 135.0808[M + H–C6H8O2]+, 107.0857[M + H–C7H7O3]+	isozaluzanin C	AF
58	19.37	C15H20O2	232.1457	233.1536	0.004	215.1436[M + H–H2O]+, 205.1593[M + H–CO]+, 197.1130[M + H–2H2O]+, 187.1487[M + H–H2O–CO]+	isoalantolactone	AF
59	19.53	C20H20O7	372.1203	373.1268	–3.617	358.1057[M + H–CH3]+, 343.0824[M + H–2CH3], 329.1031[M + H–C2H5O]+, 312.1002[M + H–2CH3–CH3O]+	sinensetin	CL
60	19.68	C15H16O4	260.1043	261.1124	0.234	243.1025[M + H–H20]+, 189.0554[M + H–C4H8O]+	meranzin	CL
61	19.94	C15H20O2	232.1457	233.1536	0.034	215.1436[M + H–H2O]+, 205.1593[M + H–CO]+, 197.1330[M + H–2H2O]+, 187.1487[M + H–H2O–CO]+	helenine	AF
62	20.17	C15H18O3	246.1250	247.1327	–0.181	229.1228[M + H–H2O]+, 201.1279[M + H–H2O–CO]+	zaluzanin C	AF
63	20.32	C21H22O8	402.1309	403.1383	–0.404	388.1164[M + H–CH3]+, 373.0928[M + H–2CH3]+, 355.0822[M + H–2CH3–H2O]+, 342.1107[M + H–C2H5O2]+	nobiletin	CL
64	20.47	C19H18O6	342.1097	343.1177	0.095	328.0952[M + H–CH3]+, 310.0847[M + H–CH3–H2O]+, 299.0924[M + H–C2H4O]+, 282.0897[M + H–CO–H2O–CH3]+	3′,4′7,8-tetramethoxyflavone	CL
65	20.83	C15H18O3	246.1250	247.1328	–0.041	229.1228[M + H–H2O]+, 201.1279[M + H–H2O–CO]+	irofulven	AF
66	20.86	C22H24O9	432.1414	433.1493	–0.049	418.1273[M + H–CH3]+, 403.1039[M + H–2CH3]+, 385.0933[M + H–2CH3–H2O]+	3,3′,4′,5,6,7,8-heptamethoxyflavone	CL
67	21.26	C20H20O7	372.1203	373.1268	–3.617	358.1060[M + H–CH3]+, 343.0826[M + H–2CH3]+	auranetin	CL
68	21.33	C10H16	136.1246	137.1326	0.093	109.1013[M + H–C2H4]+, 95.0856[M + H–C3H6]+, 81.0699[M + H–C4H8]+	α-terpinene	AF
69	21.85	C20H20O8	388.1152	389.1230	–0.074	356.0905[M + H–CH5O]+, 341.0667[M + H–C2H8O]+, 374.1006[M + H–CH3]+, 359.0772[M + H–2CH3]+, 328.3221[M + H–C2H5O2]+	uumuhengerin	CL
70	23.28	C15H18O2	230.1301	231.1378	–0.136	213.1280[M + H–H2O]+, 195.1174[M + H–2H2O]+, 185.1331[M + H–H2O–CO]+, 145.1016[M + H–C4H6O2]+, 131.0860[M + H–C5H8O2]+	dehydrocostus lactone	AF

Note: t: retention time; AF: Aucklandia costus Falc; LK: Lindera aggregata (Sims) Kosterm.; CL: Citrus aurantium L.; AL: Areca catechu L.

Table 2

Identification of Compounds in SMT by UHPLC-LTQ-Orbitrap in Negative Ion Modea

peak no.	tR (min)	molecularformula	calculated mass (m/z)	measured mass [M – H]−	error(ppm)	MS/MS fragments	identification compound	source
14	7.95	C16H18O9	354.0945	353.0865	–0.189	191.0560[M – H–C9H6O3]−, 179.0352[M–H–C7H10O5]−, 173.0455[M – H–C9H6O3–H2O]−, 135.0453[M + H–C8H10O7]−	chlorogenic acid	AL
15	9.34	C16H18O9	354.0945	353.0863	–0.379	191.0559[M – H–C9H6O3]−	neochlorogenic acid	AL
16	9.87	C16H18O9	354.0945	353.0868	0.051	179.0560[M – H–C7H10O5]−, 173.0452[M – H–C9H6O3–H2O]−, 191.0560[M – H–C9H6O3]−, 135.0453[M + H–C8H10O7]−	4-dicaffeoylquinic acid	AL
24	14.55	C27H32O15	596.1735	595.1632	–2.537	475.1077[M – H–C8H8O]−, 459.1149[M – H–C8H8O2]−, 287.0563[M – H–C12H20O9]−, 271.0617[M – H–C12H20O10]−, 235.0251[M – H–C16H24O9]−	neoeriocitrin	CL
26	14.81	C27H32O15	596.1735	595.1652	–0.587	459.1149[M – H–C8H8O2]−, 287.0563[M – H–C12H20O9]−, 235.0251[M – H–C16H24O9]−	eriocitrin	CL
29	15.26	C27H32O14	580.1786	579.1705	–0.302	271.0608[M – H–C12H20O9]−, 459.1131[M – H–C8H8O]−	narirutin	CL
31	15.37	C27H30O14	578.1630	577.1536	–1.552	269.0455[M – H–C12H20O9]−, 413.0869[M – H–C6H12O5]−, 431.0979[M – H–C6H10O4]−	rhoifolin	CL
33	15.41	C27H32O14	580.1786	579.1704	–0.482	459.1141[M – H–C8H8O]−, 353.0822[M – H–C12H14O4]−, 339.0718[M – H–C12H16O5]−, 313.0715[M – H–C10H18O8]−, 271.0608[M – H–C12H20O9]−, 235.0250[M – H–C16H24O8]−	naringin	CL
37	15.64	C28H34O15	610.1892	609.1798	–1.587	489.1440[M – H–C4H8O4]−, 447.1294[M – H–C6H10O5]−, 403.1031[M – H–C8H14O6]−, 343.0826[M – H–C10H18O8]−, 325.0717[M – H–C10H20O9]−, 301.0720[M – H–C12H20O9]−, 286.0484[M – H–C13H23O9]−, 242.0587[M – H–C14H23O11]−	hesperidin	CL
39	15.86	C28H34O15	610.1892	609.1802	–1.157	489.1440[M – H–C4H8O4]−, 447.1294[M – H–C6H10O5]−, 403.1031[M – H–C8H14O6]−, 343.0826[M – H–C10H18O8]−, 325.0717[M – H–C10H20O9]−, 301.0720[M – H–C12H20O9]−, 286.0484[M – H–C13H23O9]−, 242.0587[M – H–C14H23O11]−	neohesperidin	CL
42	16.05	C22H24O11	464.1313	463.1233	–0.168	301.0718[M – H–C6H10O5]−	hesperetin 5-O-glucoside	CL
48	17.26	C28H34O14	594.1943	593.1868	0.348	473.1454[M + H–C4H10O4]−, 327.0875[M + H–C10H20O8]−, 285.0768[M + H–C12H22O9]−	neoponcirin	CL

Note: t retention time; AF:Aucklandia costus Falc; LK:Lindera aggregata (Sims) Kosterm.; CL:Citrus aurantium L.; AL:Areca catechu L.

Flavones and Their Glycosides

Flavonoids have a diphenylpropane skeleton bearing two benzene rings (A and B) connected by a pyran ring attached to the A ring, and are further classified into several subclasses (flavones, flavonols, flavanones, flavanonol, and so on).[20] In the present study, a total of 26 flavonoids were identified from SMT in the positive and negative mode in which the main source of flavonoids was CL. Among them, five standards, including naringin (peak 33), narirutin (peak 29), neohesperidin (peak 39), hesperetin (peak 40), and nobiletin (peak 63) were first identified and used to explore the fragmentation rules of flavonoids. Briefly, the MS/MS behaviors of aglycones were described by the retro-Diels–Alder (RDA) fragmentation pathway. Successive loss of CO from the ketone group, C-fragmentation and loss of radicals, such as CH3 and CHO, had been described.[21] For flavonoid O-glycosides, the neutral loss of 146, 162, and 308 Da were the characteristic fragment ions.[22] Taking compound 33 as an example, it gave the [M – H]− ion at m/z 579.1704 (C27H31O14). In the MS2 spectra, the main fragment ions were observed at m/z values of 459.1141[M – H–C8H8O]−, 339.0718[M – H–C12H16O5]−, 313.0715[M – H–C10H18O8]−, 271.0608[M – H–C12H20O9]−, and 235.0250[M – H–C16H24O8]−. Thus, compound 33 was unambiguously identified as naringin. Its mass spectrum and proposed fragmentation pathways in negative mode are shown in Figure S2.

Compound 37 and neohesperidin (compound 39) showed the same molecular formula and similar [M – H]− ions at m/z values of 609.1798(C28H35O15) and 609.1802(C28H35O15). Moreover, they exhibited the similar fragment ions of m/z values of 489.1440[M – H–C4H8O4]−, 447.1294[M – H–C6H10O5]−, 403.1031[M – H–C8H14O6]−, 343.0826[M – H–C10H18O8]−, 325.0717[M – H–C10H20O9]−, 301.0720[M – H–C12H20O9]−, 286.0484[M – H–C13H23O9]−, and 242.0587[M – H–C14H23O11]−, so it was tentatively identified as hesperidin.

Similarly, based on the chromatographic behavior and the similar fragmentation pathways, seven flavone glycosides (peaks 18, 24, 26, 31, 41, 42, and 48) were tentatively identified as lonicerin, neoeriocitrin, eriocitrin, rhoifolin, isosakuranin, hesperetin 7-O-β-d-glucuronide, and neoponcirin, respectively.[23−25] The remaining compounds (peaks 30, 35, 36, 46, 48, 49, 50, 55, 56, 59, 66, 67, and 69) were tentatively identified as flavonoids via comparing their exact molecular masses and MS/MS spectra with the literature data.[25] Detailed information on these compounds is summarized in Tables and .

Alkaloids

Alkaloids are a group of naturally occurring chemical compounds that mostly contain basic nitrogen atoms.[26] In the present study, the main source of alkaloids was AL and LK. Fourteen alkaloids were tentatively identified, including four piperidines (guvacoline, arecaidine, guvacine, and arecoline), eight isoquinolines (pronuciferine, laurolitsine, norisoboldine, boldine, laurotetanine morphine acetate, naloxone, and reticuline), and two organic amine alkaloids (N-methyltyramine and synephrine). Among them, two reference compounds, guvacoline (peak 9) and arecoline (peak 10) were first identified and used to characterize the fragmentation behaviors. Their fragment ions are summarized in Tables and , and the losses of methanol, the methyl group, methylamine (CH3NH), H2O, and CO were their main pathways.[27,28] Taking compound 9 as an example, it exhibited a [M + H]+ ion at an m/z of 142.0862(C7H12NO2) and in the positive ion spectrum, the main fragment ions at m/z values of 110.0602[M + H–CH4O]+, 113.0598[M + H–CH3N]+, 96.0807[M + H–CH2O2]+, and 81.0334[M + H–C2H7NO]+ were detected. Thus, compound 9 was inferred as guvacoline. The proposed fragmentation pathway of guvacoline was shown in Figure S3. Similarly, 13 alkaloids (peaks 3, 4, 7, 10, 12, 19, 23, 25, 27, 34, 38, 44, and 47) were identified as guvacine, arecaidine, synephrine, arecoline, N-methyltyramine, laurolitsine, boldine, norisoboldine, pronuciferine, reticuline, laurotetanine, morphine acetate, and naloxone, respectively, based on the MS information as well as the literature reports.[29−33]

Coumarins

In the present study, the main source of coumarins was CL. Four coumarins (7-hydroxycoumarin, 5,7-dihydroxycoumarin, isomerancin, and meranzin) were identified from SMT, and the structures could be postulated based on high-accuracy quasi-molecular ion and MS mass spectrometry. Among them, one standard compound (7-hydroxycoumarin) was first used to characterize the fragmentation behaviors and explore the fragmentation rules. For most of coumarins, the primary fragmentation mechanism was first a neutral loss of side chain then a loss of CO and CO2 in MS/MS mode.[34] Compound 22 gave a [M + H]+ ion at an m/z of 163.0389(C9H7O3) and yielded fragment ions at m/z values of 145.0286[M + H–H2O]+, 119.0494[M + H–CO2]+, and 107.0493[M + H–2CO]+, which suggested that it was 7-hydroxycoumarin. The specific cleavage process is shown in Figure S4. Similarly, compounds 17, 28, and 60 were identified as 5,7-dihydroxycoumarin, isomerancin, and meranzin, respectively.[35]

Amino Acids

Amino acids are organic compounds containing amine (−NH2) and carboxyl (−COOH) functional groups, along with a side chain (R group) specific to each amino acid.[36] In the present study, the main source of amino acids was AL. Four amino acids and one derivative of aspartic acid were identified. Among them, two standard compounds, phenylalanine (peak 13) and valine (peak 5), were first identified and used to characterize the fragmentation behaviors and explore the fragmentation rules for amino acids. Their fragmentation mechanisms have been extensively studied in positive ion mode. Because of the presence of amino and acid groups in amino acid structure, fragment ions were usually generated by the loss of NH3 (17 Da), H2O (18 Da), CO2 (44 Da), and HCOOH (46 Da).[28] Taking compound 13 as an example, the precise molecular weight was 166.0862 (C9H12NO2), and in the positive ion spectrum, the main fragment ions were observed at m/z values of 149.0599[M + H–NH3]+, 131.0493[M + H–NH3–H2O]+, 120.0810[M + H–H2O–CO]+, and 80.0493[M + H–C4H6O2]+. Therefore, compound 13 was unambiguously identified as phenylalanine. The detailed fragmentation pathways of compound 13 are shown in Figure S5. Peak 1 exhibited a [M + H]+ ion at an m/z of 133.0608 and gave the fragment ions at m/z values of 116.0343[M + H–NH3]+, 87.0553[M + H–CO–H2O]+, 70.0286[M + H–CO–H2O–NH3]+. Thus, it was identified as the derivative of aspartic acid, asparagine. Similarly, peaks 2, 5, and 8 were identified or tentatively characterized as proline, valine, and tyrosine, respectively, from SMT (Tables and ) based on the MS information and reported data.[37−39]

Sesquiterpenoid and Derivatives

In the present study, the main source of sesquiterpenoids was AF and LK. A total of 13 sesquiterpenoids were found in SMT. Among them, four standard compounds, linderane (peak 45), costunolide (peak 61), isoalantolactone (peak 58), and helenine (peak 53) were first identified and used to characterize the fragmentation behaviors. The details of the identified components are summarized in Tables and . Their MS/MS behaviors can be described as the loss of radicals, such as H2O, CH3, and CO.[40] For example, compound 45 showed a [M + H]+ ion at an m/z of 261.1120 with an elemental composition of C15H17O4. The MS/MS spectrum of [M + H]+ exhibited fragment ions at m/z values of 243[M + H–H2O]+, 189[M + H–CH3–C3H5O]+, 177[M + H–C5H8O]+, and 217[M + H–CO2]+. Thus, compound 52 was identified as linderane. The detailed fragmentation pathways of compound 45 are shown in Figure S6. Based on these similar fragmentation patterns, nine sesquiterpenoids (peaks 20, 21, 43, 52, 54, 57, 62, 65, and 70) were identified as picriside B, saussureanine D, ethyllucidone, saussureanine B, saussureanine A, isozaluzanin C, zaluzanin C, garbanzol, and dehydrocostus lactone, respectively.[40−44]

Organic Acids

In the present study, the main source of organic acids was AL and CL, and six organic acids were identified from SMT. Peak 32 gave a [M + H]+ ion at an m/z of 113.0595(C6H8O2). The fragment ions at m/z values of 95.0492[M + H–H2O]+, 85.0648[M + H–CO]+, 67.0541[M + H–CO–H2O]+ suggested the structure of carboxyl. Based on the fragmentation information, peak 32 was tentatively identified as sorbic acid. Peaks 6 and 11 were isomers and yielded the fragment ions at m/z 127.0389. They also exhibited similar fragmentation ions. According to their MS behavior and reported data, peaks 6 and 11 were tentatively identified as maltol and triacetate lactone, respectively.[45,46] Peaks 14, 15, and 16 were isomers of chlorogenic acid, which gave [M – H]− ions at an m/z of 353.0865 (C16H19O9). The characteristic fragment ions at m/z values of 191.0560[M – H–C9H6O3]−, 179.0352[M – H–C7H10O5]−, 173.0455[M – H–C9H6O3–H2O]−, and 135.0453[M + H–C8H10O7]− suggest that these compounds may be the ester of caffeic acid and quinic acid.[47]

Miscellaneous

Another two obvious peaks in the extracted ion chromatogram of SMT were identified. Among them, compound 51 was tentatively assigned as 2-methyl-5-propan-5-ylcyclohex-2-en-1-one. Compound 68, with [M + H]+ at an m/z of 137.1326(C10H17), yielded ions at m/z values of 109.1013[M + H–C2H4]+, 95.0856[M + H–C3H6]+, and 81.0699[M + H–C4H8]+. It was plausibly identified as α-terpinene.[48]

Putative Targets of SMT

A total of 461 targets of SMT were predicted by MedChem Studio, as given in Table S2. The results showed that the candidate compounds could act on multiple targets, and one target could also be linked to multiple components.

Known Therapeutic Targets Acting on GIDs

We collected 315 known therapeutic targets of GIDs from three databases after eliminating redundancy, of which six were from Therapeutic Target Database (TTD), 19 were from DrugBank and 290 were from Online Mendelian Inheritance in Man (OMIM). Detailed information about putative targets of GIDs is provided in Table S3.

Network and Pathway Analysis

To illustrate the potential relationships between SMT and GIDs, a protein–protein interaction (PPI) network of interactions between SMT-related targets and GID-related targets was established. The network contains 515 nodes and 5957 edges. Detailed information about the PPI data is provided in Table S4. After calculating the values of the three topological features of all nodes in the network, 170 nodes were identified as major hubs because they satisfied the screening criteria (degree centrality (DC) > 20, betweenness centrality (BC) > 0.0015, closeness centrality (CC) > 0.3757). Among them, 152 hubs were putative targets of SMT and 16 hubs were known therapeutic targets of GIDs and 3 hubs were both putative targets of SMT and known therapeutic targets of GIDs. The details of these major hubs are shown in Table S5.

To facilitate scientific interpretation of the action mechanisms of SMT on GIDs, GO enrichment analysis was conducted on these major hubs. As depicted in Figure a–c, the top 10 significant GO entries of the biological process (BP), cellular components (CC), and molecular function (MF) were obtained after filtering by a parameter P-value cut-off of ≤0.05. The major biological processes included drug response, chemical synaptic transmission, response to hypoxia, phospholipase C (PLC)-activating G protein-coupled receptor (GPCR) signaling pathway, signal transduction, transport, calcium ion transmembrane transport, cellular amino acid biosynthetic process, cytosolic calcium ion concentration, cellular calcium ion homeostasis (Figure a). The majority of the protein responses were situated in a variety of cell components like postsynaptic membrane, plasma membrane, dendrite, neuronal cell body, cell junction, synapse, voltage-gated calcium channel complex, cytosol, cell surface, and mitochondrion (Figure b). Moreover, the associated molecular functions included ligand-gated ion channel activity, enzyme binding, drug binding, amino acid binding, dopamine binding, dopamine neurotransmitter receptor activity, heme binding, oxygen binding, voltage-gated calcium channel activity, and receptor binding (Figure c).

Figure 2

GO term performance and pathway enrichment analysis of major hubs. (a) GO term performance by a biological process (BP); (b) GO term performance by a molecular function (MF); (c) GO term performance by a cellular component (CC); (d) Pathway enrichment analysis by KEGG. The ordinate stands for GO terms or main pathways, the primary abscissa stands for minus log10(P).

To further explore the functional effects and systemic association of the major hubs, KEGG pathway enrichment analysis was carried out. As depicted in Figure d, the top 10 pathways were obtained, which could be classified into three major functional modules according to the functional annotation (Figure ). Maximum module was associated with neuro-immune regulation (neuroactive ligand–receptor interaction, cholinergic synapse, inflammatory regulation of TRP channels, HIF-1 signaling pathway, and biosynthesis of amino acids) and the second module was sorted as signal transduction (calcium signaling pathway, cAMP signaling pathway, and MAPK signaling pathway), while the minimum module was concentrated in cardiomyocyte contraction (vascular smooth muscle contraction and adrenergic signaling in cardiomyocytes). The network diagram composed of the interactions among the active components of SMT, major SMT-related targets, and major pathways was finally generated, as shown in Figure . It was worthy to note that the calcium signaling pathway was highly enriched in KEGG pathway analysis, which played an important role in smooth muscle contraction[49] and thus may mainly contribute to therapeutic effects of SMT on the improvement of gastrointestinal motility.[50]

Figure 3

SMT herbs–major hubs–main pathways network. Green diamonds represent each herb in SMT; round purple nodes represent putative targets of chemicals in SMT; round yellow nodes represent known therapeutic targets for GIDs; round red nodes represent both putative targets of components in SMT and known therapeutic targets for GIDs; orange rectangles represent top 10 pathways from enrichment analysis of major targets; edges represent interactions among SMT herbs, putative targets, known therapeutic targets for the treatment of GIDs, and main pathways.

SMT Attenuates the Gastrointestinal Dysmotility via Regulating the Calcium Signaling Pathway

Gastrointestinal dysmotility (GD), affecting the neuromuscular functions needed for propulsion of intraluminal contents[51] is closely associated with the pathophysiology of functional GIDs.[52] According to the network analysis (Table S6), the SMT putative targets associated with the calcium signaling pathway include cholinergic receptor muscarinic 1(CHRM1), CHRM2, dopamine receptor D1 (DRD1), DRD5, adrenoceptor alpha 1A (ADRA1A), ADRA1B, ADRA1D, adrenoceptor beta 2 (ADRB2), 5-hydroxytryptamine receptor 2A (HTR2A), HTR2B, HTR2C, protein kinase cAMP-activated catalytic subunit alpha (PRKACA), calcium voltage-gated channel subunit alpha1 A (CACNA1A), CACNA1B, CACNA1C, CACNA1D, CACNA1H, CACNA1S, cholinergic receptor nicotinic alpha 7 subunit (CHRNA7), glutamate ionotropic receptor NMDA type subunit 1(GRIN1), GRIN2A, GRIN2C, protein tyrosine kinase 2 beta (PTK2B), protein kinase C alpha (PRKCA), PRKCB, and nitric oxide synthase 1 (NOS1). Figure shows the calcium signaling pathway affected by major putative targets of SMT, and the details are discussed below.

Figure 4

Illustration of the calcium signaling pathway induced by major putative targets of SMT.

Ca2+ ions play key roles in the process known as excitation–contraction coupling.[53] Upon an appropriate autacoid or depolarizing stimulation, extracellular Ca2+ ions enter smooth muscle cells primarily through voltage-operated channels (VOCs) and receptor-operated channels (ROCs) of the plasma membrane, which in turn induces the release of Ca2+ ions from sarcoplasmic reticulum (SR), leading to a transient rise in intracellular Ca2+. Then, the contraction of gastrointestinal smooth muscles was initiated via the activation of Ca2+/calmodulin (CaLM)-dependent myosin light chain kinase (MLCK).[54−56] According to our predicted results, several alkaloid components of SMT were involved in Ca2+ influx and Ca2+ release from SR, such as boldine[57] and arecoline.[58] Therefore, SMT may modify the GD status by increasing intracellular Ca2+.

The PTK/PLC pathway has been recognized as an important upstream signaling leading to calcium release.[59] In response to growth factors, PTK activates phospholipase C (PLC) by phosphorylating the gamma 1 isoform of PLC.[60] PLC is a critical enzyme in the regulation of phosphatidylinositol (PIP2) metabolism, which catalyzes PIP2 to produce diacylglycerol (DAG) and inositol 1,4,5-triphosphate (IP3). IP3 acts on its receptors to release Ca2+ from intracellular stores. Also, DAG could trigger the activation of protein kinase C, which plays an important role in cell growth.[60,61] Accumulated evidences indicate that plant flavonoids exert significant inhibition effects on PKC, which may partially contribute to the anti-GID function of SMT.[62,63]

GPCR signaling also involves in the regulation of luminal SR Ca2+ channel. In gastrointestinal smooth muscle cells, neurotransmitters or autacoids trigger the release of the α subunit of a guanine nucleotide-binding protein Gs. Then, Gs binds to adenylyl cyclase (ADCY), which catalyzes the production of cyclic adenosine monophosphate (cAMP). Subsequently, cAMP induces the activation of PKA and then leads to the phosphorylation of phospholamban (PLN), which in turn alleviates the suppression effects of PLN on Ca2+-transport ATPases (SERCA), thereby improving SR Ca2+ uptake and contributing to relaxation.[56] In addition, the Gq-PLC-IP3-Ca2+ pathway also plays an important role in maintaining calcium homeostasis.[64] It has been evident that many components from SMT could ameliorate GD via regulating the GPCR pathway, such as N-methyltyramin[65] and naringenin.[66]

Other downstream targets such as NOS, PTK2B, and PKC could also be regulated by Ca2+/(CaMKII) and thus involved in the pathological process of GD. Inducible NOS is responsible for the increased NO production, leading to smooth muscle relaxation. PTK2B, a cytoplasmic tyrosine kinase, could initiate multiple signaling pathways associated with several cellular functions such as cell migration and proliferation.[67,68] Thus, NOS, and PTK2B may also be the potential targets of SMT on GD.

Recently, the pharmacological activity of a single herb of SMT on gastrointestinal disorders has also been reported. AL could promote gastrointestinal motility by promoting the secretion of glucagon like peptide-1, bile, and cholesterol.[69] The flavonoids from CL also possess prominent gastrointestinal motility promoting efficacy. However, AF produced actually a spasmolytic effect on gastrointestinal motility via the suppression of muscarinic receptors, 5-hydroxytryptamine receptors, and Ca2+ influx.[70] In addition, LK has been reported to show a protective effect on the intestinal barrier, ameliorating gut microflora dysbiosis and attenuating inflammation.[71] According to the TCM theory, the combination use of these four herbs would achieve higher efficacy and fewer side effects compared with a single herb. Taken together, the various types of constituents contained in SMT exert their synergistic effects on GIDs through acting on multiple target genes/proteins on multiple pathways. However, this work has some inevitable limitations. First, and particularly, some identified constituents may not be absorbed into the blood circulatory system, and in many situations, it is the metabolites, not the parent compounds, which have therapeutic effects. Second, it is really hard to confirm the inhibitory or activated effects of the predicted targets. Lastly and importantly, the in silicon work may be influenced by possible biases to frequently studied biological processes.[17]

Conclusions

In the present study, UHPLC-LTQ-Orbitrap combined with a network pharmacology method was used to unveil the chemical basis and investigate the action mechanism of SMT for treating GIDs. The results showed that the pharmacological mechanism of SMT in the treatment of GIDs may be mainly associated with the calcium signaling pathway. Moreover, 26 core genes were identified as key active targets involved in this pathway, and they were mapped to 51 key active components. However, to enhance the reliability of the results, further experiments were demanded to validate these hypotheses.

Materials and Methods

Reagents and Chemicals

SMT was purchased from Hunan Hansen Pharmaceutical Co., Ltd. (Yiyang, Hunan Province, China), which was prepared according to the preparation process of Chinese Pharmacopoeia (2015 edition). HPLC grade acetonitrile, methanol, and formic acid used in this study were purchased from Fisher Scientific. High pure water was prepared by using a Millipore Milli Q plus purification system. All other reagents were analytical grade and commercially available. Reference standards of valine, guvacoline, arecoline, phenylalanine, 7-hydroxycoumarin, narirutin, naringin, neohesperetin, hesperetin, linderane, costunolide, isoalantolactone, helenine, and nobiletin were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). All standards were of at least 98% purity and were suitable for UHPLC-LTQ-Orbitrap analysis.

Sample Preparation

A volume of 1 mL SMT was filtered through a 0.22 μm nylon membrane filter before LC/MS analysis. Individual stock solutions of 14 compounds (500 μg/mL) were prepared by dissolving accurately weighted amount of reference compound in HPLC grade methanol, respectively. Then, all the 14 standards were mixed and serially diluted to 50 μg/mL. The solution was filtered through a filter (0.22 μm) and analyzed directly by LC/MS.

LC–MS/MS System

Sample analyses were performed on an ultimate 3000 LC system coupled to an LTQ Orbitrap mass spectrometer via an ESI interface. The chromatography system was composed of an autosampler, a column compartment, and two pumps. Xcalibur, Metworks, and Mass Frontier 7.0 software packages were used for data collection and data analysis. Chromatographic separations were performed on a Thermo Scientific BOS Hypersil C18 column (2.1 mm × 150 mm, 2.4 μm). The mobile phase was composed of water containing 0.1% formic acid (A) and acetonitrile (B). The LC gradient program [time (in min)/% mobile phase B] was described as follows: 0.01/2, 5/2, 5.1/10, 10/10, 25/80, 26/2, and 31/2. The chromatographic runs were performed at a flow rate of 0.300 mL/min, and the injection volume was 3 μL. The column was maintained at 35 °C.

The ESI source parameters were set as follows: the capillary temperature, 250 °C; source voltage and spray voltage, 5 kV; sheath gas (N2) flow, 35 psi. The ESI source was operated in both positive and negative ionization modes. In the Fourier transform (FT) cell, full MS scans were acquired in the range of m/z 50–2000. The MS/MS experiments were set as data-dependent scans.

Prediction of Putative Targets of SMT

MedChem Studio (MedChem Studio, 3.0; Simulations Plus, Inc., Lancaster, CA, USA, 2012) was used to predict the targets associated with the identified SMT components based on structure similarity. The similarity threshold was set at 0.7.

Known Therapeutic Targets of GIDs

Known human therapeutic targets of GIDs were retrieved from three sources with the query “gastrointestinal disorders”. The first one was TTD (http://db.idrblab.net/ttd/), which could offer abundant information about the known therapeutic protein and nucleic acid targets.[72] The second one was DrugBank (https://www.drugbank.ca/), which is a unique bioinformatics and cheminformatics resource containing comprehensive molecular information about drugs and their corresponding targets.[73] The last one was OMIM (http://www.omim.org/), which is an authoritative database of human genes and genetic disorders.[74]

Protein–Protein Interaction Network

The PPI network of SMT-related targets and the targets of GIDs was constructed by using STRING database (https://string-db.org/), with the species limited to Homo sapiens. STRING is a web-enabled database that provides extensive information about known and predicted protein–protein interactions.[75] The cutoff value was set at 0.4.

Network Construction and Analysis

An interaction network of putative SMT target-known therapeutic targets of GIDs was constructed using the PPI data from STRING database, which was visualized by Cytoscape software 3.7.1 (https://cytoscape.org/). Cytoscape is a popular open-source software for biological network visualization and data integration.[76] Next, we used three topological properties, including DC, BC, and CC to evaluate the topological importance of each node. The median values of the three parameters were set as the cutoff values.[77]

Pathway Enrichment Analysis

To unveil the potential functions of the critical targets involved in the SMT-mediated treatment of GIDs, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and Gene Ontology (GO) enrichment analysis were carried out by the use of the Database for Annotation, Visualization, and Integrated Discovery (DAVID) system (http://david.abcc.ncifcrf.gov/home.jsp/, v6.7).[78]P values less than 0.05 were considered statistically significant.