1H NMR based metabolite profiling for optimizing the ethanol extraction of Wolfiporia cocos.

Metabolite profiling of Wolfiporia cocos (family: Polyporaceae) had been much advancement in recent days, and its analysis by nuclear magnetic resonance (NMR) spectroscopy has become well established. However, the highly important trait of W. cocos still needs advanced protocols despite some standardization. Partial least squares discriminant analysis (PLS-DA) was used as the multivariate statistical analysis of the 1H NMR data set. The PLS-DA model was validated, and the key metabolites contributing to the separation in the score plots of different ethanol W. cocos extract. 1H NMR spectroscopy of W. cocos identified 33 chemically diverse metabolites in D2O, consisting of 13 amino acids, 11 organic acids 2 sugars, 3 sugar alcohols, 1 nucleoside, and 3 others. Among these metabolites, the levels of tyrosine, proline, methionine, sarcosine, choline, acetoacetate, citrate, 4-aminobutyrate, aspartate, maltose, malate, lysine, xylitol, lactate threonine, leucine, valine, isoleucine, uridine, guanidoacetate, arabitol, mannitol, glucose, and betaine were increased in the 95% ethanol extraction sample compared with the levels in other samples, whereas level of acetate, phenylalanine, alanine, succinate, and fumarate were significantly increased in the 0% ethanol extraction sample. A biological triterpenoid, namely pachymic acid, was detected from different ethanol P. cocos extract using 1H-NMR spectra were found in CDCl3. This is the first report to perform the metabolomics profiling of different ethanol W. cocos extract. These researches suggest that W. cocos can be used to obtain substantial amounts of bioactive ingredients for use as potential pharmacological and nutraceuticals agents.

Introduction

Wolfiporia cocos is a fungus in the family Polyporaceae. It is a wood-decay fungus but has a subterranean growth habit. It is notable in the development of a large, long-lasting underground sclerotium that resembles a small coconut. This sclerotium called “(Chinese) Tuckahoe” or fu-ling, is not the same as the true tuckahoe used as Indian bread by Native Americans, which is the arrow arum, Peltandra virginica, a flowering tuberous plant in the arum family. W. cocos is also used extensively as a medicinal mushroom in Chinese medicine (Esteban, 2009, Wu et al., 2018, Liu et al., 2018).

Indications for use in the traditional Chinese medicine include promoting urination, invigorating the spleen function (i.e., digestive function) and calming the mind (Shah et al., 2014). Alcoholic extracts of W. cocos have been reported to contain various lanostane-type triterpenoids (Akihisa et al., 2007, Wang et al., 2018, Zhu et al., 2018, Chen et al., 2017). W. cocos also possesses abundant medicinal compounds including polysaccharides and triterpenoids (Feng et al., 2013). These compounds have been used to treat many diseases such as gastritis, nephrosis, edema, dizziness, nausea, and emesis. In addition, the surface layer of W. cocos has known to be functional in significant diuretic effects (Zhao et al., 2012, Shi et al., 2017, Hu et al., 2017, Lee et al., 2017) and famous for its biological efficacy such as anti-tumor effect (Kanayama et al., 1983, Jin et al., 2003, Li et al., 2017).

Until now, the metabolomic profiling using 1H NMR and multivariate statistical analysis of W. cocos has not been reported. The W. cocos extract according to different ethanol extraction is mainly performed by visual inspection. Therefore, such different ethanol extraction has been rather subjective and relies on a few experts in the experiment. Nowadays, metabolomics techniques combining spectrometric methods and multivariate statistical analysis such as principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), and hierarchical cluster analysis (HCA) (Eriksson et al., 2006). Additionally, the use of PLS makes it possible to estimate the important activities from multivariate data sets. These techniques are the rapid and reliable identification of different ethanol extract and will require all the traditional approaches of natural products chemistry and metabolomics as well as improved analytical methods and statistical tools. The multivariate statistical analysis techniques coupled with 1H NMR analysis using various selection protocols were used for metabolic profiling and trait of various kinds of plants, plant-derived preparations, foods, and tissues (Kim et al., 2010, Sekiyama et al., 2010, Wishart, 2008). We report the first identification and quantification of pachymic acid by 1H NMR and our hypothesis was that the metabolic profiles of compounds of W. cocos might change during different ethanol extracts. In this study, we first described 1H NMR spectroscopy followed by PLS-DA in metabolomic analysis of different ethanol W. cocos extracts.

Materials and methods

Solvents and chemicals

The following chemicals were obtained commercially: Monopotassium phosphate (KH2PO4), 3-(trimethylsilyl)-propionic-2,2,3,3-d4 acid sodium salt (TSP), Ethyl alcohol, Deuterated chloroform (CDCl3) and deuterium oxide (D2O) 99.8%, were purchased from Sigma-AldrichSigma Aldrich (St. Louis, MO, USA). NMR tubes were obtained from Optima (Tokyo, Japan).

P. cocos microwave-assisted extraction

The microwave-assisted extraction method used for W. cocos samples had as follows: Powdered W. cocos (2 g) were placed into a 250 mL in an extraction vessel with 40 mL each solvent (0, 25, 50, 75, and 95% ethanol). Each extraction vessel was inserted to the microwave oven for 50 min at 85 °C (960 W) (Transform 800. AR0800-MW-1800, Aurora instruments Ltd, Vancouver, B.C., Canada). First extraction was transferred to new flask and then the residue was re-extracted twice for 50 min at 85 °C (960 W). The extracts were evaporated, freeze-dried and then stored at -70 °C until analysis. The extraction samples were used for analysis of 1H NMR.

1H NMR analysis of W. cocos metabolomic profiling

To analyzing the hydrophilic substance, 100 mg of W. cocos extracts were dissolved in 1 mL phosphate buffer (90 mM, pH 7.0–7.4) in D2O containing 0.01% sodium salt (TSP) as an internal standard using an ultra-sonication for 60 min (Lab companion, Daejeon, Korea) to extract intracellular metabolites. Analyzing for pachymic acid, 100 mg of W. cocos extracts were dissolved in 1 mL CDCl3. After metabolite extraction, supernatants were clarified by centrifuging at 12,000 rpm for 20 min at room temperature (Labogene, Seoul, Korea), filtered using an Amicon Ultra 0.5 mL centrifugal filters (Millipore, Darmstadt, Germany), and collected into 1.5 mL tube (Eppendorf, Hamburg, Germany). 700 μL of each filtered extracts were loaded into 5 mm NMR tubes (n = 3). 1H NMR spectra were acquired at 300 K on a 600.13-MHz Bruker Advance spectrometer (Rheinstetten, Germany) using the ZGPR pulse sequence with water pre-saturation. In total, 128 transients were gathered into 32 K data points with a relaxation delay of 2 s with an acquisition time per scan of 1.70 s and a spectra width of 10.0 ppm. The NMR spectra were analyzed using Chenomx NMR suite software (version 8.2, Chenomx Inc., Edmonton, Alberta).

Data preprocessing and multivariate statistical analysis of 1H NMR data

1H NMR data processing and assigned were performed through Chenomx NMR suite software (version 8.2, Chenomx Inc., Edmonton, Alberta). Multivariate statistical analyses were performed by one-way ANOVA using PASW Statistics 22 software (IBM, Somers, NY, USA) following which a Tukey’s significant difference test. Significance was determined with a P-value threshold (<0.05). Metabolites levels were normalized using log2 function and then, mean centering and UV scaling was applied for all principal component analysis (PCA), and partial least squares discriminant analysis (PLS-DA) usingSimca-P+11.0 (Umetrics, Umeå, Sweden).

Results and discussion

Metabolic assignments of W. cocos different ethanol extractions

Table 1 and Fig. 1 indicate that 33 compounds were identified in D2O dissolved of W. cocos samples at the 5 different ethanol extractions. 11 amino acids, including leucine, valine, isoleucine, threonine, alanine, proline, methionine, aspartate, lysine, tyrosine, and phenylalanine, 2 amino acid derivatives, including sarcosine, and guanidoactate, 5 organic acid, including lactate, acetoacetate, acetate, citrate, and malate, 6 organic acid derivative, including malonate, fumarate, formate, 4-aminobutyrate (GABA), succinate, betaine, 2 sugars glucose and maltose, 3 sugar alcohols, including arabitol, xylitol, and mannitol, 1 nucleoside, including uridine, and 3 others, including trehalose, xanthine, choline were detected (Fig. 1A). In addition, pachymic acid was identified CDCl3 dissolved of W. cocos samples at the 5 different ethanol extractions. Fig. 1B shows a presentive NMR specrum of the CDCl3 extracts of W. cocos samples. The peak of pachymic acid were assigned by comparisons with the previously study (Wang et al., 2012).

Table 1

Metabolite assignments and chemical shifts of distinguishable peaks.

Assigned No	Metabolite compound	Chemical shift (ppm)
1	Leucine	0.94(t, J = 6.5), 1.70(m)
2	Valine	1.02(d, J = 6.0), 0.98(d, J = 6.87)
3	Isoleucine	1.02(t, J = 6.5), 1.06(d, J = 6.9)
4	Lactate	1.30(d, J = 6.91)
5	Threonine	1.34(d, J = 6.58)
6	Alanine	1.46(d, J = 8.31)
7	4-Aminobutyrate (GABA)	1.90(m), 2.30(t, J = 7.36), 3.02(t, J = 7.58)
8	Acetate	1.90(s)
9	Proline	1.98(m), 2.02(m)
10	Methionine	2.14(s), 2.22(m)
11	Acetoacetate	2.30(s)
12	Succinate	2.38(s)
13	Citrate	2.50(d, J = 15.92)
14	Malate	2.66(dd, J1 = 15.36, J2 = 2.9)
15	Aspartate	2.66(m), 2.82(dd, J1 = 3.8, J2 = 3.7)
16	Sarcosine	2.70(s), 3.58(s)
17	Malonate	3.10(s)
18	Choline	3.18(s), 4.06(m)
19	Glucose	3.22(t, J = 8.9), 3.46(m), 4.62(m),5.22(d, J = 4.07)
20	Betaine	3.22(s), 3.86(s)
21	Arabitol	3.54(dd, J1 = 8.3, J2 = 1.62), 3.90(m)
22	Xylitol	3.66(m)
23	Lysine	3.74(t, J = 6.09)
24	Guanidoacetate	3.78(s)
25	Mannitol	3.79(m), 3.90(m)
26	Trehalose	5.18(d, J = 3.8)
27	Maltose	5.34(d, J = 3.8), 5.42(d, J = 3.89)
28	Uridine	5.9(d, J=), 7.86(d, J=)
29	Fumarate	6.5(s)
30	Tyrosine	6.88(d, J = 8.45), 7.18(d, J = 8.41)
31	Phenylalanine	7.34(m), 7.42(m)
32	Xanthine	8.26(s)
33	Formate	8.46(s)
34	Pachymic acid	0.97–1.01(m), 1.54(s), 4.54–4.78(m)

Fig. 1

Representative 1H NMR spectrum (a) NMR spectra of W. cocos sample analyzed with D2O as an NMR dissolution solvent after 95% methanol extraction. (b) pachymic acid analyzed with CDCl3 as an NMR dissolution solvent after 95% ethanol extraction.

Multivariate data analysis and metabolic characterization of W. cocos

The NMR data sets that included 33 aqueous metabolite data from the W. cocos of different ethanol extraction either the 0, 25, 50, 75, and 95 % were assessed using PCA and PLS-DA (Fig. 2A and B) score plotting. As an unsupervised method, PCA enables recognition of any inherent sample clustering without any bias because it does not require any information on the data sets (Lu et al., 2008, Woo et al., 2009, Carraro et al., 2009). Therefore, it was applied to confirm the clustering prior to analysis by supervised methods such as PLS-DA. A supervised PLS-DA is one of the classification methods where the response variable is a ‘dummy’ Y matrix expressing an orthogonal unit vector of each class (Barker and Rayens, 2003).

Fig. 2

Validation plots of the PLS-DA model using 5 different ethanol extracts of W. cocos samples. PCA model (A), PLS-DA model (B), Metabolites in PLS1 (C), R2 and Q2 intercept values (D).

Our data show that significant differences were observed among the five groups. The 0% and 95% extraction samples were clearly separated from the 25%, 50% and 75% extraction samples (PCA and PLS-DA score plots). PLS component 1 versus PLS component 2 explained separation among W. cocos samples of 3 groups. Combining PLS component 1 and PLS component 2 explained 75.4% of total variance (54.1% and 21.3%, respectively) (Fig. 2 B). When internal cross-validation on PLS-DA was performed, component 7 and parameters, such as R2X = 0.95, R2Y = 0.99, and Q2Y = 0.83, which showed its predictive power and degree of fit to the data, were obtained. The major metabolites involved in the separation along PLS1 (positive: 95% ethanol extraction sample) were valine, isoleucine, leucine, lactate threonine, 4-aminobutlyate, proline, methionine, acetoacetate, citrate, malate, aspartate, sarcosine, malonate, choline, glucose, betaine, arabitol, xylitol, lysine, guanidoacetate, mannitol, trehalose, maltose, uridine, tyrosine, and xanthine (Fig. 2C), while alanine, acetate, succinate, fumarate, phenylalanine, and formate were the major compounds contributing to the separation along PLS1 (negative: 0% ethanol extraction sample) (Fig. 2C). A permutation test was processing in order to further validate the PLS-DA models (Bijlsma et al., 2006). The R2 and Q2 intercept values were 0.713 and −0.504 after 20 permutations (Fig. 2D). Table 2 shows the PLS-DA-derived VIP values of the major compounds contributing to the separation of each W. cocos extraction sample in the PLS-DA model. Variable importance in the projection (VIP) value is an weighted sum of squares of the PLS-DA weight, both with respect to Y as the correlation to all the responses and X as its projection, picking components that play important roles in the separation. It has been indicated that cutting-off for VIP around 1.0 worked well for variable selection even though the variables with larger than 1.0 were most influential for model (Lu et al., 2008, Woo et al., 2009). As shown in Table 2. The VIP values of the major contributing compounds for separation in the score plots derived from PLS-DA. The metabolites with VIP values over 1.0, such as proline, aspartate, 4-aminobutyrate, acetate, xanthine, methionine, alanine, malate, choline, malonate, uridine, formate, acetoacetate, sarcosine, phenylalanine, fumarate, glucose, succinate, and threonine were selected for further comparison by one-way analysis of variance (ANOVA). The relative changes in these metabolites and significant differences between the five groups were determined by ANOVA test. Among these metabolites, the levels of tyrosine, proline, methionine, sarcosine, choline, acetoacetate, citrate, 4-aminobutyrate, aspartate, maltose, malate, lysine, xylitol, lactate threonine, leucine, valine, isoleucine, uridine, guanidoacetate, arabitol, mannitol, glucose, and betaine were increased in the 95% ethanol extraction sample compared with the levels in other samples, whereas level of acetate, phenylalanine, alanine, succinate, and fumarate were significantly increased in the 0% ethanol extraction sample (Table 3). Main variations are summarized in the form of a heatmap shown in Fig. 3. In this figure the red color indicates increased relative concentration of the metabolites within the different ethanol W. cocos, while green color indicates a decreased relative concentration (or absence) of metabolites.

Table 2

The VIP values of the major metabolites for the separation of W. cocos samples in the PLS-DA derived score plots.

No.	Compounds	VIP
1	Proline	1.333
2	Asparate	1.276
3	4-Aminobutyrate	1.255
4	Acetate	1.204
5	Xanthine	1.168
6	Methionine	1.163
7	Alanine	1.145
8	Malate	1.119
9	Choline	1.069
10	Malonate	1.056
11	Uridine	1.052
12	Formate	1.050
13	Acetoacetate	1.047
14	Sarcosine	1.044
15	Phenylalanine	1.043
16	Fumarate	1.031
17	Glucose	1.020
18	Succinate	1.017
19	Threonine	1.007

Table 3

Relative levels of 34 metabolites in different ethanol extraction sample of W. cocos.

Compounds	Poria cocos ethanol extracts
	0%	25%	50%	75%	95%
Valine	0.148 ± 0.0284a	0.134 ± 0.0037a	0.099 ± 0.0531a	0.352 ± 0.0215b	0.680 ± 0.0321c
Isoleucine	0.120 ± 0.0247a	0.140 ± 0.0166a	0.094 ± 0.0418a	0.335 ± 0.0329b	0.675 ± 0.0396c
Leucine	0.213 ± 0.0326a	0.127 ± 0.0322a	0.125 ± 0.0385a	0.452 ± 0.0281b	0.824 ± 0.0528c
Lactate	0.340 ± 0.0207a	0.285 ± 0.0197a	0.379 ± 0.0679a	0.533 ± 0.0337b	1.174 ± 0.0615c
Threonine	0.279 ± 0.0128ab	0.249 ± 0.0061a	0.275 ± 0.0328ab	0.372 ± 0.0516b	0.900 ± 0.0316c
Alanine	0.479 ± 0.0100d	0.293 ± 0.0106bc	0.237 ± 0.0077a	0.257 ± 0.0191ab	0.324 ± 0.0129c
4-Aminobutyrate	0.471 ± 0.0279abc	0.729 ± 0.0482c	0.415 ± 0.1185a	0.451 ± 0.0356ab	0.697 ± 0.1118bc
Acetate	0.234 ± 0.0061b	0.174 ± 0.0181b	0.083 ± 0.0382a	0.069 ± 0.0045a	0.082 ± 0.0222a
Proline	0.143 ± 0.0303a	0.134 ± 0.0344a	0.114 ± 0.0194a	0.097 ± 0.0023a	0.155 ± 0.0522a
Methionine	0.200 ± 0.0310a	0.236 ± 0.0583a	0.167 ± 0.0261a	0.152 ± 0.0155a	0.239 ± 0.0669a
Acetoacetate	0.053 ± 0.0115a	0.116 ± 0.0368a	0.083 ± 0.0201ab	0.074 ± 0.0061a	0.168 ± 0.0300b
Succinate	0.775 ± 0.0067b	0.200 ± 0.0299a	0.174 ± 0.0330a	0.167 ± 0.0419a	0.232 ± 0.0273a
Citrate	0.032 ± 0.0068a	0.088 ± 0.0367ab	0.054 ± 0.0098ab	0.063 ± 0.0274ab	0.118 ± 0.0130b
Malate	0.108 ± 0.0114a	0.130 ± 0.0134a	0.083 ± 0.0044a	0.087 ± 0.0074a	0.249 ± 0.0277b
Asparate	0.285 ± 0.0130a	0.424 ± 0.0618b	0.248 ± 0.0030a	0.220 ± 0.0155a	0.484 ± 0.0678b
Sarcosine	9.446 ± 0.8679bc	6.527 ± 0.2543a	7.308 ± 0.2788ab	9.019 ± 1.1207b	11.398 ± 0.5247c
Malonate	0.066 ± 0.0079	0.091 ± 0.0191	0.077 ± 0.0420	0.106 ± 0.0357	0.105 ± 0.0074
Choline	1.101 ± 0.1974ab	0.941 ± 0.0400b	0.812 ± 0.1059b	1.258 ± 0.2183ab	1.554 ± 0.0887b
Glucose	41.881 ± 4.2489a	85.977 ± 3.6322c	63.496 ± 1.8225b	88.431 ± 1.6366c	102.242 ± 3.1062d
Betaine	16.182 ± 1.7689a	25.322 ± 1.3454b	22.178 ± 1.0788b	26.971 ± 1.8695b	32.399 ± 1.7217c
Arabitol	20.125 ± 0.5451a	25.829 ± 2.0786bc	24.266 ± 0.3775bc	31.684 ± 4.1390bc	38.088 ± 2.4777c
Xylitol	26.176 ± 0.7212ab	31.144 ± 2.3288bc	21.976 ± 1.7453a	34.731 ± 3.3623c	56.220 ± 2.0529d
Lysine	8.552 ± 1.3995a	13.031 ± 2.5729ab	13.544 ± 2.3630ab	18.874 ± 2.9903b	27.290 ± 2.1021c
Guanidoacetate	3.366 ± 0.2804a	5.768 ± 0.4409bc	5.315 ± 0.7008b	7.104 ± 0.2362c	8.915 ± 0.6533d
Mannitol	10.925 ± 0.3843a	14.576 ± 1.4573a	14.439 ± 0.7339a	18.638 ± 1.3304b	23.531 ± 1.6786c
Trehalose	0.219 ± 0.0131a	2.966 ± 1.1683b	2.497 ± 0.3350b	3.437 ± 0.7295b	3.046 ± 0.0285b
Maltose	0.079 ± 0.0081a	0.206 ± 0.0715b	0.138 ± 0.0194ab	0.208 ± 0.0303b	0.328 ± 0.0101c
Uridine	0.031 ± 0.0056a	0.083 ± 0.0136b	0.060 ± 0.0057ab	0.090 ± 0.0134bc	0.099 ± 0.0051c
Fumarate	0.080 ± 0.0018d	0.037 ± 0.0020c	0.027 ± 0.0008b	0.021 ± 0.0007a	0.024 ± 0.0003ab
Phenylalanine	0.229 ± 0.0569b	0.098 ± 0.0277a	0.056 ± 0.0058a	0.065 ± 0.0081a	0.087 ± 0.0019a
Tyrosine	0.097 ± 0.0482	0.088 ± 0.0065	0.079 ± 0.0107	0.103 ± 0.0041	0.149 ± 0.0073
Xanthine	0.017 ± 0.0015a	0.050 ± 0.0062c	0.041 ± 0.0064bc	0.043 ± 0.0050bc	0.032 ± 0.0019b
Formate	0.037 ± 0.0089ab	0.048 ± 0.0040b	0.057 ± 0.0095b	0.041 ± 0.0041ab	0.024 ± 0.0034a
Pachymic acid	0.383 ± 0.0111b	0.354 ± 0.0070b	0.331 ± 0.0011 a	0.344 ± 0.0982b	0.426 ± 0.0313b

Tukey HSD a < b

Fig. 3

The heat map constructed based on the differential ethanol extract 34 metabolites of W. cocos relative concentration values. A red-blue color scale indicates normalized relative concentration level expression value. The red color indicate a Log2 Fold Change  ≥ 2 highest relative concentration value while the blue color means Log2 FC ≤ −2 lowest relative concentration value respectively.

In addition, pachymic acid contents of five different ethanol concentration W. cocos were analyzed and the results were summarized in Fig. 4. The pachymic acid content of 95% ethanol was significantly high value (0.426 ± 0.0313). Pachymic acid is a natural triterpenoid known to inhibit the phospholipase A2 family of arachidonic acid – producing enzymes. Pachymic acid has antioxidant activity, anti-inflammatory and anticancer properties, it can inhibit cell growth and modulate arachidonic acid metabolism in lung cancer and impair breast cancer cell invasion by suppressing nuclear factor kappa B-dependent metalloproteinase-9 expression (Gapter et al., 2005, Ling et al., 2010, Ling et al., 2011). W. cocos is used widely as traditional medicines and food in Korea, China, and Japan. W. cocos consumption is likely to increase in recently, due to its important bioactivities.

Fig. 4

Relative quantification of pachymic acid in different percentage of ethanol extracts of W. cocos.

Conclusion

In this study, we used a metabolomics approach based on 1H NMR spectroscopic analysis to show that the characteristic metabolic profiles of W. cocos. The results revealed that the metabolomics profiles different ethanol W. cocos extraction, making further investigations of the bioactivities of appropriated ethanol W. cocos extractions. This study can be offered that the combined use of NMR and PLS-DA is an efficient technique for the different ethanol W. cocos extractions and would be suitable for discriminating samples in industrial application.