Content determination results of DG
The components were identified and quantified by comparison of retention time and calculation of peak areas from the chromatograms with those of known standards. DG extract contained determine 6-gingerol, 8-gingerol, 6-shogaol and 10-gingerol were 6.95,0.99,0.42 and 1.32 mg/g, respectively (Fig. 1).
Pathological observation of vascular
The microstructures of the abdominal aorta in rats was observed. Vascular obstruction and a small amount of micro thrombosis were observed in the MG. Some endothelial cells fell off from the vascular wall, endothelial cells swelled, and intima thickened. Also, inflammatory cell infiltration was observed. The above pathological symptoms were alleviated in the administration group, especially in the GJH (C). The results are shown in Figure S1.
Effect of DG on blood viscosity
The impact of WBV is shown in Fig. 2 and Table S1. In the MG of BSS, WBV increased significantly at all shear rates. After administration, the WBV of each group was significantly reduced at high shear rates (p < 0.01 or 0.05). And the table also shows the effects of DG on plasma viscosity. The model rats had a significantly higher plasma viscosity than the controls. The plasma viscosity in the GJH, GJM, GJL was significantly decreased compared to the MG (p < 0.01 or 0.05).
Effect of DG on ESR, PCV, DI and EAI
The results of ESR, PCV, EAI, and DI for each group are shown in Table 1. All four indexes were significantly higher in the MG than in the NG. GJH, GJM are reduced ESR and PCV (P < 0.05). All DG dose groups decreased DI and EAI (P < 0.05 or P < 0.01).
Table 1
Effect of DG extract on ESR, PCV, DI, and EAI
Group | Dose(g/kg) | ESR(MM/H) | PCV(%) | DI (%) | EAI |
NG | — | 3.30 ± 1.15 | 31.67 ± 7.34 | 66.17 ± 7.63 | 2.23 ± 0.40 |
MG | — | 10.37 ± 2.85## | 51.83 ± 7.70## | 83.83 ± 7.65## | 5.24 ± 0.92## |
GJH | 2.10 | 6.28 ± 1.06* | 39.17 ± 8.11* | 71.33 ± 7.39* | 3.08 ± 0.65** |
GJM | 1.05 | 7.68 ± 0.43* | 42.33 ± 6.09* | 73.17 ± 8.04* | 3.49 ± 0.88** |
GJL | 0.53 | 7.90 ± 1.64 | 44.50 ± 6.89 | 74.33 ± 6.62* | 3.92 ± 0.83* |
Data represent mean ± S.D. n = 6. #Compared with NG, p < 0.05. ##Compared with NG, p < 0.01. *Compared with MG, p < 0.05. **Compared with MG, p < 0.01. |
Effect of DG on plasma coagulation parameters
The impacts of DG on blood coagulation were evaluated by assays of APTT, PT, TT, and FIB in the plasma. PT was decreased, FIB was increased, APTT and TT were significantly shortened in the model rats compared with the NG levels, as showed in Table 2. After administration, compared with the MG, the PT of GJH and GJM was significantly increased, and the FIB was significantly decreased (P < 0.05). In terms of TT and APTT, the DG group was significantly longer (P < 0.05).
Table 2
Effect of DG extract on Plasma Coagulation Parameters
Group | Dose(g/kg) | PT(INR) | FIB (g) | TT(S) | APTT(S) |
NG | — | 1.37 ± 0.15 | 2.11 ± 0.26 | 48.90 ± 4.70 | 32.33 ± 5.75 |
MG | — | 0.88 ± 0.21## | 6.38 ± 0.79## | 30.77 ± 6.30## | 20.67 ± 4.41## |
GJH | 2.10 | 1.21 ± 0.16* | 5.16 ± 1.02* | 44.24 ± 5.42** | 30.50 ± 8.55* |
GJM | 1.05 | 1.16 ± 0.21* | 5.33 ± 0.803* | 42.53 ± 645* | 29.33 ± 7.39* |
GJL | 0.53 | 1.00 ± 0.18 | 5.66 ± 1.08 | 39.82 ± 5.44* | 28.67 ± 7.01* |
Data represent mean ± S.D. n = 6. #Compared with NG, p < 0.05. ##Compared with NG, p < 0.01. *Compared with MG, p < 0.05. **Compared with MG, p < 0.01. |
Metabolic profiling analysis
To obtain the maximum possible information for each metabolite, the experiments were performed in both positive and negative ionization modes, and LC-MS-MS analyzed serum and urine samples under the same chromatographic conditions. The typical total ion chromatograms (TICs) of the serum and urine samples from the NG, MG, and GJH collected in the experiment are presented in Figure S2. To further analyze changes between complex sets of data, the multivariate data analysis techniques, including PCA-X, PLS-DA and OPLS-DA, were used to analyze LC-MS data.
PCA analysis was used to assess the difference in metabolite profiles between serum and urine samples of NG and MG. The apparent separation between them was obtained in the PCA scores plot (Fig. 3A, B), which indicated that the two groups had utterly different metabolic profiling. Then, the PLS-DA method was used to systematically evaluate the metabolomics of BSS rats (permutation number: 200). In PLS-DA, the NG was more distinct from the MG (Fig. 3C, D). The PLS-DA model parameters were as follows: R2 = 0.846 and Q2 = − 0.0669 in serum, and R2 = 0.831and Q2 = − 0.00977 in urine, which showed an excellent predictive power (Fig. 3E, F). To screen differential metabolites and maximize the discriminatory ability of serum and urine metabolites between the groups, orthogonal partial least squares discriminant analysis (OPLS-DA) was used. As showed in the score plot (Fig. 3G, H), the serum and urine samples in the MG were significantly different from those in the NG. The S-plots (Fig. 3I, J) showed differential metabolites between the two groups, and VIP was obtained based on OPLS-DA with a threshold than 1 would be viewed as potential biomarkers. Combined with the results of the S-plot and VIP-value plot together, the UPLC-Q-TOF/MS analysis platform provided the retention time, precise molecular mass, and MS/MS data for the structural identification of biomarkers. The same procedures were utilized to analyze the plasma samples derived from the NG, GJL, GJM and GJH.
Moreover, we investigated the differences in metabolic profiles between the MG and the GJL, GJM and GJH, using OPLS-DA analysis. Score 3D plots (Figure S3 A, B) from OPLS-DA were used to maximize the discrimination of metabolite differences among the five groups. The figure shows that the metabolites of serum and urine in GJL, GJM, and GJH gradually approach the NG. Meanwhile, the GJH was the closest to the NG, the GJL was the closest to the MG, and the GJM was between the GJH and the GJL. Therefore, the intervention of DG on endogenous metabolites in rats with BSS shows a significant dose-effect relationship.
Biomarker identification
To select potential biomarkers related to BSS, the first principal component of VIP was obtained. Firstly, the VIP values greater 1.0 are selected as changing metabolites. And then, the remaining variables were calculated by Student’s t-test (t-test), P < 0.05, variables are discarded between two comparison groups. Besides, several commercial databases, such as HMDB (http://www.hmdb.ca), KEGG (http://www.kegg.jp) and NIST (http://www.nist.gov/index.html), were used for searching the information of metabolites.
The retention time, precise molecular mass, and MS/MS data was provided for the structural identification of potential biomarkers using the analysis platform of EZ-info software. The accurate molecular mass was detected within 5 ppm in measurement errors by TOF. Meanwhile, the potential elemental composition, the degree of unsaturation and the fractional isotope abundance of compounds were obtained. The supposed molecular formula was searched in ChemSpider, the Human Metabolome Database, Mass Bank, and other relevant databases were used to identify the possible chemical constitutions, and MS/MS data were screened to determine the potential structures of the ions. Details of potential biomarkers were listed in Table 3.
Table 3
Potential biomarkers in serum and urine
| NO | RT | Formula | Metabolites | VIP | P | FC |
Serum | 1 | 0.6595 | C3H4O3 | Pyruvate | 1.0050 | 0.0024 | 0.6187 |
| 2 | 0.9963 | C15H12O | Chalcone | 2.4297 | 0.0031 | 2.9408 |
| 3 | 1.4248 | C29H50O4 | 6-Deoxohomodolichosterone | 1.4195 | 0.0091 | 2.0295 |
| 4 | 2.2860 | C18H29NO2 | Penbutolol | 1.2444 | 0.0190 | 1.9356 |
| 5 | 3.8942 | C18H39NO3 | Phytosphingosine | 1.2193 | 0.0367 | 1.6779 |
| 6 | 6.4867 | C20H41NO3 | Ethanolamine Oleate | 1.4903 | 0.0068 | 2.1090 |
| 7 | 7.6357 | C16H24N2O | Oxymetazoline | 1.2449 | 0.0004 | 1.7684 |
| 8 | 7.6965 | C19H38O4 | 1-Monopalmitin | 1.6229 | 0.0172 | 2.1786 |
| 9 | 8.5343 | C47H76O2 | α-calendic acid | 1.0186 | 0.0054 | 1.6551 |
| 10 | 14.0018 | C14H11NO | Phosphonoacetaldehyde | 1.7977 | 0.0004 | 2.2653 |
Urine | 1 | 0.7043 | C14H11NO | 3-Methyl-9H-carbazole-9-carboxaldehyde | 1.3875 | 0.0359 | 0.6445 |
| 2 | 2.7173 | C22H32N2O5 | Benzquinamide | 1.7277 | 0.0001 | 0.3362 |
| 3 | 0.6694 | C3H7NO2 | L-Carnitine | 1.4923 | 0.0291 | 0.2745 |
| 4 | 0.6896 | C4H9N3O2 | L-Alanine | 1.9506 | 0.0049 | 0.0529 |
| 5 | 0.6898 | C6H8N2O2 | Creatine | 1.9488 | 0.0063 | 0.0536 |
| 6 | 0.7301 | C9H13N3O3 | L-isoleucyl-L-proline | 1.5086 | 0.0078 | 0.3532 |
| 7 | 1.2144 | C5H7N3O | 5-Methylcytosine | 1.3817 | 0.0400 | 2.8193 |
| 8 | 1.7440 | C17H22N2O7 | Tetrahydropentoxyline | 1.5986 | 0.0180 | 1.6918 |
| 9 | 3.1393 | C11H19NO10S2 | Progoitrin | 1.5839 | 0.0082 | 2.6975 |
| 10 | 3.1389 | C11H19O13P | 1-Phosphatidyl-D-myo-inositol | 1.4146 | 0.0160 | 2.8472 |
| 11 | 9.6106 | C20H41NO3 | Ethanolamine Oleate | 1.8239 | 0.0291 | 2.8580 |
| 12 | 9.7938 | C24H40O6 | 3b,4b,7a,12a-Tetrahydroxy-5b-cholanoic acid | 1.8367 | 0.0104 | 2.1858 |
Metaboanalyst 4.0 (www.metaboanalyst.ca/) is used for pathway construction by employing the rice metabolic pathway databases as reference for the global test algorithm. Table 4 shows the identified biomarkers and their trends compared with the NG.
Table 4
Summary of intensity values of potential biomarkers in each group
Matrix | Biomarkers | NG | MG | GJH | GJM | GJL |
Serum | Pyruvate | 2054.53 ± 219.28 | 3540.26 ± 309.4## | 2898.62 ± 421.27 | 1941.29 ± 287.40** | 2406.15 ± 576.39** |
| Chalcone | 1190.27 ± 311.07 | 49.73 ± 166.64## | 933.49 ± 107.79 | 765.78 ± 125.21 | 655.97 ± 148.81 |
| 6-Deoxohomodolichosterone | 1459.10 ± 45.85 | 784.27 ± 132.91## | 1166.24 ± 315.19* | 1047.37 ± 371.41* | 903.75 ± 145.43 |
| Penbutolol | 1379.69 ± 485.53 | 783.49 ± 141.84## | 1170.86 ± 271.58 | 1066.08 ± 252.95 | 936.82 ± 164.42 |
| Phytosphingosine | 29605.10 ± 5830.21 | 19138.72 ± 4393.29# | 25570.68 ± 4553.94 | 23926.42 ± 3939.20 | 22274.01 ± 4349.25 |
| Ethanolamine Oleate | 4826.64 ± 1381.75 | 2523.15 ± 689.70# | 3780.75 ± 792.81 | 3368.81 ± 72.566 | 2757.25 ± 613.47 |
| Oxymetazoline | 2107.08 ± 130.59 | 1286.33 ± 265.67## | 1929.20 ± 245.88** | 1657.12 ± 477.21** | 1605.2 ± 221.02** |
| 1-Monopalmitin | 2564.01 ± 864.06 | 1280.57 ± 476.87## | 2093.21 ± 466.46 | 1977.15 ± 55.80 | 1518.93 ± 490.59 |
| α-calendic acid | 965.48 ± 145.78 | 643.90 ± 97.28## | 894.87 ± 103.88* | 714.86 ± 564.18 | 668.67 ± 59.73 |
| Phosphonoacetaldehyde | 7118.85 ± 675.99 | 3403.62 ± 1038.04## | 5999.82 ± 1260.77** | 5202.67 ± 1139.99* | 4360.61 ± 1204.03 |
Urine | 3-Methyl-9H-carbazole-9-carboxaldehyde | 434.90 ± 136.45 | 728.62 ± 191.06## | 455.14 ± 108.99** | 448.74 ± 84.66** | 539.19 ± 79.05** |
| Benzquinamide | 406.17 ± 153.89 | 1241.39 ± 162.91## | 583.35 ± 235.36** | 643.73 ± 334.58* | 717.24 ± 436.52** |
| L-Carnitine | 619.78 ± 189.29 | 2592.49 ± 1115.51# | 999.14 ± 250.12 | 1164.39 ± 514.70 | 1278.23 ± 478.98 |
| L-Alanine | 115.75 ± 57.41 | 2866.42 ± 501.03## | 1011.66 ± 151.37 | 1382.45 ± 603.93 | 1876.80 ± 939.28 |
| Creatine | 559.21 ± 208.48 | 13492.18 ± 5090.32## | 4305.16 ± 1336.07 | 5777.50 ± 1179.57 | 5987.95 ± 1430.89 |
| L-isoleucyl-L-proline | 1635.24 ± 725.09 | 5166.35 ± 1748.43# | 2102.71 ± 1215.84 | 3787.50 ± 1260.69 | 4485.61 ± 1489.07 |
| 5-Methylcytosine | 3626.66 ± 1238.58 | 1150.50 ± 609.22## | 3107.05 ± 944.98** | 2765.35 ± 1441.79* | 1767.78 ± 541.37** |
| Tetrahydropentoxyline | 1645.85 ± 737.43 | 944.80 ± 104.29# | 1772.12 ± 655.85 | 1247.53 ± 557.70 | 1020.69 ± 139.07 |
| Progoitrin | 1257.27 ± 413.71 | 482.18 ± 169.51# | 1056.89 ± 463.63 | 1132.06 ± 402.24** | 954.89 ± 334.78** |
| 1-Phosphatidyl-D-myo-inositol | 1017.94 ± 622.52 | 355.60 ± 143.10# | 977.23 ± 272.19* | 961.55 ± 557.54* | 636.25 ± 211.77 |
| Ethanolamine Oleate | 4809.81 ± 853.54 | 1997.75 ± 626.62## | 4761.77 ± 407.20** | 3986.53 ± 865.61** | 2888.81 ± 251.27** |
| 3b,4b,7a,12a-Tetrahydroxy-5b-cholanoic acid | 1438.69 ± 259.84 | 735.58 ± 115.77## | 1484.81 ± 774.85** | 1343.59 ± 575.07** | 1164.67 ± 374.06 |
#Compared with NG, p < 0.05. ##Compared with NG, p < 0.01. *Compared with MG, p < 0.05. **Compared with MG, p < 0.01. |
Metabolic Pathway Analysis
Possible ways to further explore the effects of BSS, using online MetaboAnalyst 4.0 software (www.metaboanalyst.ca), through the analysis of blood and urinary tract, found seven pathways affected (Fig. 4) in the BSS, these ways include arginine and proline metabolism, citrate cycle (TCA cycle), pyruvate metabolism, glycolysis/gluconeogenesis, phosphatidylinositol signaling system, inositol phosphate metabolism, and glycerophospholipid metabolism. Pathways with an impact value > 0.1 were considered to be the most important pathways. The results showed that pyruvate metabolism (Impact value = 0.21) and glycolysis/gluconeogenesis (Impact value = 0.10) and phosphatidylinositol signaling system (Impact value = 0.10) were the most important pathways for the development of BSS.
Correlation analysis between biomarkers and pharmacology Indicators
A correlation map of rat serum and urine metabolites of biomarkers and pharmacological indicators of BSS was conducted based on Pearson’s correlation coefficients. The correlation heatmap in Fig. 5, shows that the metabolites of Sm1 (serum, Pyruvate), Um1 (urine,3-Methyl-9H-carbazole-9-carboxaldehyde), Um2 (urine,Benzquinamide), Um3 (urine,L-Carnitine), Um4 (urine,L-Alanine), Um5 (urine,Creatine), Um6 (urine,L-isoleucyl-L-proline) were positive related to the level of WBV1, WBV5, WBV30, WBV50, WBV200,PV, ESR, PCV, DI, EAI, FIB, and they are negatively correlated with PT, TT and APTT. However, the correlation of pharmacological indicators corresponding to Sm2(serum, Chalcone), Sm3 (serum, 6-Deoxohomodolichosterone), Sm4 (serum, Penbutolol), Sm5(serum, Phytosphingosine), Sm6 (serum, Ethanolamine Oleate), Sm7 (serum, Oxymetazoline), Sm8 (serum, 1-Monopalmitin), Sm9(serum, α-calendic acid), Sm10(serum, Phosphonoacetaldehyde) Um7 (urine, 5-Methylcytosine), Um8 (urine, Tetrahydropentoxyline), Um9 (urine, Progoitrin), Um10 (urine, 1-Phosphatidyl-D-myo-inositol), Um11 (urine, Ethanolamine Oleate), Um12 (urine, 3b,4b,7a,12a-Tetrahydroxy-5b-cholanoic acid), is just the opposite. Among them Sm2 with TT, Sm8 with PT, Um4 with PCV, EAI, Um5 with DI are strongly positively correlated (r = 0.998, 0.998, 0.996, 0.996, 0.999, respectively), Sm2 with WBV1, WBV5, DI, Sm5 with PCV, Sm7 with WBV30, Um2, Um3 with APTT (r=-0.999, -0.999, -0.997, -0.998, -0.998, -0.997, -0.999, respectively), are strongly negatively correlated. These correlations may indicate that changes in metabolites are related to changes in metabolites and pharmacological indicators.
Correlation analysis of biomarkers in each group
The heat map (Fig. 6) shows the correlation of 22 potential biomarkers in different groups. These potential biomarkers were up-regulated and down-regulated to different degrees in NG and MG.In addition, the results of hierarchical clustering analysis provide visual visualization of each group, the contents of metabolites Sm1(serum, Pyruvate), Um1(urine, 3-Methyl-9H-carbazole-9-carboxaldehyde), Um2(urine, Benzquinamide), Um3(urine, L-Carnitine), Um4(urine, L-Alanine), Um5(urine, Creatine) and Um6 (urine, L-isoleucyl-L-proline) are decreased, while the contents of other metabolites are increased, while MG and GJL were opposite to NG. It can be seen that GJH and NG metabolism are the most similar, while GJL and MG metabolism are the closest (red: increased; blue: decreased).