Development of Global Chemical Profiling for Quality Assessment of Ganoderma Species by ChemPattern Software

Triterpenoids are the major secondary metabolites and active substances in Ganoderma, considered as the “marker compounds” for the chemical evaluation or standardization of Ganoderma. A response surface methodology was used to optimize the ultrasonic-assisted extraction of triterpenoids. The extraction rate was 7.338 ± 0.150 mg/g under the optimum conditions: 87% ethanol, ratio of solid to liquid (w : v) 1 : 28, and ultrasound extraction time 36 min. Based on the high sensitivity and selectivity of HPLC-LTQ-Orbitrap-MSn, 24 components of triterpenoids were tentatively identified in the negative mode. Then, the global chemical profiling consisting of HPLC and TLC fingerprints generated by ChemPattern™ software was developed for evaluation of Ganoderma species. For fingerprint analysis, 11 peaks of triterpenoids were selected as the characteristic peaks to evaluate the similarities of different samples. The correlation coefficients of similarity were greater than 0.830. The cluster analysis showed a clear separation of three groups, and 11 peaks played key roles in differentiating these samples. The developed global chemical profiling method could be applied for rapid evaluation, quality control, and authenticity identification of Ganoderma and other herbal medicines.


Introduction
Ganoderma, a popular edible and medicinal mushroom, is commonly used as dietary supplements and touted as a remedy to promote health and longevity [1]. So far, more than 200 species of Ganoderma have been found in the world, and Ganoderma lucidum (Leyss. ex Fr.) Karst. and Ganoderma sinense Zhao, Xu et Zhang are o cially recorded in Chinese pharmacopoeia [2]. Previous studies have demonstrated that Ganoderma possesses various biological properties, such as antitumor, antiaging, antioxidant, hyperglycemic, and regulating immunity [3][4][5]. Owing to its satisfactory clinical e ects, more and more Ganoderma products as health foods or medicines have appeared at the market. However, it is hard to say whether its quality is good or bad, and it is also di cult to identify whether the raw materials are authentic or adulterant.
For Ganoderma, triterpenoids are the major secondary metabolites and active substances [6]. e pioneering has isolated and identi ed more than 300 triterpenoids from the spores, fruiting bodies, and cultured mycelia of Ganoderma [7][8][9]. Triterpenoids could be considered as the "marker compounds" for the chemical evaluation or standardization of Ganoderma. Professor Guo and his team's researches concluded that the content and composition of triterpenoids vary signi cantly due to di erence in the strain, geographic origin, cultivation method, extraction process, and other factors [10,11]. Besides, scholars have made a lot of exploratory work on the chromatographic ngerprint of Ganoderma products [12,13]. However, the traditional ngerprints were always processed by the ngerprint similarity software (2004 or 2012 version), only limited to HPLC pro les without TLC. After that, researchers always apply other statistical software (such as SPSS and SAS) to process the data for cluster analysis or principal component analysis, which is relatively cumbersome and time-consuming. erefore, it is essential to develop a global chemical pro ling method for rapidly evaluating the quality of Ganoderma to ensure the e cacy.
Chemical pro le re ects the totality of intrinsic chemical compounds of herbal medicines and emphasizes the integral characterization of a complex system [14,15]. Here, the global chemical pro ling method contained the establishment of HPLC and TLC ngerprints, the characterization of common peaks and statistical analysis. For processing the large scale of physical properties characteristic in the complex herbal extract, an advanced chemometric and chemical ngerprinting software was developed by Chemmind Technology Co., Ltd. ChemPattern is an advanced chemometric and chemical ngerprinting software, which endeavors to provide solutions for qualitative and quantitative quality evaluation and characteristic analysis. e software was employed to calculate the correlation coe cients between di erent chromatographic pro les, as well as to generate the representative standard ngerprint by mean simulation.
In the present work, triterpenoids were extracted from di erent Ganoderma samples under the ultrasonic-assisted condition optimized by response surface methodology (RSM) with the Box-Behnken design (BBD) [16]. e extract was analyzed by high-performance liquid chromatography coupled with linear ion trap-Orbitrap mass spectrometry (HPLC-LTQ-Orbitrap-MS n ) for a comprehensive study of the multiple chemical constituents. In addition, the chromatographic data of HPLC and TLC were submitted into the professional Chem-Pattern software for establishing global chemical pro ling and evaluating similarity. Furthermore, Ganoderma samples from di erent regions could be distinguished by clustering analysis and principal component analysis of ngerprint data. Figure 1. Samples were pulverized into powder and kept in a vacuum dryer. e standard sample of Ganoderma was identi ed by National Institutes for Food and Drug Control. Cellulase and oleanolic acid were obtained from Aladdin. Ganoderic acid A was purchased from Chengdu Must Bio-technology Co., Ltd. Acetonitrile and methanol were of HPLC grade and obtained from Anhui Tedia High Purity Solvents Co., Ltd.

Materials and Reagents. Ganoderma collected from di erent regions is shown in
Ethanol, vanillin, sulfuric acid, acetic acid, perchloric acid, phosphoric acid, formic acid, petroleum ether, ethyl acetate, and other chemicals were of analytical grade.

Heating Re ux Extraction.
Sample powder (2.0 g) and 95% ethanol (30 mL) were extracted under re ux at 75°C for 30 min. en, the extraction solution was ltered through a lter paper and evaporated to dryness at 60°C.

Ultrasonic Extraction of Total Triterpenoids.
Sample powder (2.0 g) and 95% ethanol (30 mL) were placed in an ultrasonic bath (300 W) at 75°C for 30 min. e suspension was cooled to room temperature and ltered. e ltrate was vacuum-dried at 60°C.

Determination of the Total Content of Triterpenoids.
e total content of triterpenoids was determined according to the method of Hou with some modi cations [17]. e oleanolic acid (2.0 mg) was dissolved in methanol (10 mL) to produce a standard solution. 0 μL, 100 μL, 200 μL, 300 μL, 400 μL, 500 μL, and 600 μL standard solutions were added into a test tube, respectively, and evaporated in a water bath. 5% vanillin-acetic acid reagent (400 μL) and perchloric acid (1000 μL) were added, and the tube was placed in a water bath for 30 min at 65°C. When the reaction solution was cooled, acetic acid (5 mL) was added. e absorbance was determined at 546 nm using a microplate reader (In nite ® 200 Pro NanoQuant, Tecan, Switzerland). e sample (100 μL) was determined following the abovementioned method.
en, the weight of triterpenoids was calculated according to the standard curve. e extraction yield of triterpenoids was calculated as follows: yield (mg/g, w/w) � weight of triterpenoids/weight of raw materials. All determinations were performed in triplicates. e standard curve was y � 223.2x + 4.473, R 2 � 0.9950. e oleanolic acid weight in 20-120 μg range showed a good linear relationship.

Experimental Design of RSM for Ultrasonic Extraction.
ree independent factors of ultrasonic extraction were  Table S1. e three levels were designated as −1, 0, and +1 for low, intermediate, and high values. In order to predict the conditions of ultrasound extraction, experimental data were analyzed using the software Design-Expert version 8.06 and explained using the following nonlinear computer-generated quadratic model [18]: where β 0 is the constant coe cient, β i , β ii , and β ij are the coe cients for the linear, quadratic, and interaction e ect, x i and x j are the independent variables, and ε is the error. e adequacy of the model was tested through analysis of variance (ANOVA). e coe cients of determination R 2 and adj R 2 expressed the quality of t of the resultant polynomial model, and the statistical signi cance was checked by F-value and lack of t [19].  e DAD was set at 254 nm. e injection volume was 10 μL, and the ow rate was 0.6 mL/min. e operation parameters of mass spectrometry were as follows: source voltage, 4.0 kV; sheath gas, 20 (arbitrary units); auxiliary gas, 12 (arbitrary units); sweep gas, 2 (arbitrary units); and capillary temperature, 350°C. Default values were used for most other acquisition parameters: Fourier transformation (FT) automatic gain control (AGC) target 5 × 10 5 for the MS mode and 5 × 10 4 for the MS n mode. Perfusion samples were analyzed in the data-dependent scan mode at a resolving power of 60,000 at m/z 400. e most intense ions were selected, and parent ions were fragmented by high-energy C-trap dissociation (HCD) with a normalized collision energy of 45% and an activation time of 100 ms. e maximum injection time was set to 100 ms with two microscans for the MS mode and to 1000 ms with one microscan for the MS n mode. e mass range was from m/z 100 to 1500. Each sample was analyzed both in negative and positive modes. Data were analyzed using Xcalibur software version 2.2 ( ermo Fisher Scienti c).

HPLC Chromatographic Fingerprint Analysis
Conditions. HPLC chromatographic ngerprint analysis was conducted on a liquid chromatography system (1260, Agilent, America) equipped with a quaternary solvent deliver system, an autosampler, and a DAD (Agilent Technologies). e mobile phase consisted of acetonitrile (A) and 0.03% phosphoric acid solution (B) using a gradient elution of 75-68% B at 0-40 min, 68-60% B at 40-60 min, 60-40% B at 80-120 min, 40-0% B at 80-120 min, and 0% B at 120-125 min. Chromatographic separation was carried out at an Agilent Zorbax Extend-C 18 column (4.6 mm × 250 mm, 5 μm) with a solvent ow rate of 1.0 mL/min at a temperature of 35°C. e wavelength was set at 254 nm. e injection volume was 10 μL.

TLC Chromatographic Conditions.
Ganoderic acid A solution and samples a to k (5 μL) were spotted using a microinjector on a 20 × 20 cm silica gel plate (GF254, Qingdao, China). e silica gel was activated in an oven at 80°C for 30 min before use. e mobile phase consisting of petroleum ether : ethyl acetate : formic acid (1 :1 : 0.02, v/v/v) was added into a twin-trough chamber and saturated for 10 min. e plate in the chamber was developed upward over a path of 15 cm and sprayed with 1% vanillin-sulfuric acid solution. e plate was placed in an oven at 80°C for 10 min until the color of the triterpenoid spots was distinct. e image of TLC was reverse-phase processed, and the information of TLC spots was turned into the gray curve by software.

Statistical Analysis. All experiments were performed at least in triplicate.
e values were expressed as means ± standard deviation (SD).

Comparison of Ultrasonic-Assisted Extraction and Re ux
Extraction. Comparing heating re ux extraction (6.404 mg/g) with ultrasonic-assisted extraction (6.869 mg/g), the extraction yield of triterpenoids was increased signi cantly by ultrasonic-assisted extraction in the same extraction time with simple operation. e ultrasonic wave produced a strong cavitation, mechanical crushing, and thermal e ect, which dissolved the active ingredients into the solvent more adequately and saved more energy. So, the ultrasonic-assisted extraction was selected for further optimization.

RSM for Optimization of Ultrasonic-Assisted
Extraction.
e main factors which a ected the ultrasonic extraction yield of triterpenoids were investigated by single-factor experiments, including the temperature, liquid-solid ratio, extraction time, and ethanol concentration. When the procedures were conducted at 30 min with a concentration of 95% ethanol, the maximum yield of triterpenoids was 0.6906% at the liquid-solid ratio 25 mL/g (Figure 2(a)). e liquid-solid ratio and concentration of ethanol were xed at 25 mL/g and 95%.
e result showed that the extraction e ciency increased to the maximum amount of 0.7097% at 45 min ( Figure 2(b)). e yield of triterpenoids was signi cantly increased with the ethanol concentration varying from 15% to 95%, and the optimal ethanol concentration was from 75% to 95% (Figure 2(c)). erefore, the liquid-solid ratio 25 mL/g, extraction time 45 min, and 85% ethanol were selected as the center points for each factor in the RSM experiments. e single-factor experiment of temperature showed that the extraction rate of triterpenoids was almost unchanged with increasing temperature after 50°C. So, the extraction temperature was not considered as the experimental factor of BBD and set at 50°C. An experimental program for optimizing the extraction of triterpenoids using RSM with BBD is shown in Table S2. e predicted values were obtained from the model tting technique using the software Design-Expert version 8.06.
ANOVA was applied to optimize the extraction conditions of ultrasonic-assisted extraction for the triterpenoid yield and evaluate the relationship between response and variables. ANOVA for the response surface quadratic regression model showed that the F-value of model was 30.56 and the P value of model was smaller than 0.0001 (Table  S3), suggesting the model was signi cant [16]. e coe cient of determination (R 2 ) of the model was 0.9752, and the adjusted determination coe cient (adj R 2 ) was 0.9433, which indicated good agreement between the experimental parameters and the predicted values of triterpenoids. e sequence of three factors in uencing the triterpenoid yield was the ethanol concentration (C), extraction time (B), and liquid-solid ratio (A). By statistically processing, the multiple second-order equation for the extraction yield of triterpenoids was obtained as follows: where Y is the extraction yield of triterpenoids, A, B, and C are the coded values of the liquid-solid ratio, extraction time, and ethanol concentration, respectively.   e extraction yield and the interaction of di erent variables could be predicted from the three-dimensional (3D) response surface (Figures 2(d)-2(f)). e steeper slope represented that the factor had a more signi cant e ect on the extraction yield. e steeper slope of ethanol concentration implied that it had greater e ect on the extraction yield of triterpenoids than on the extraction time and liquid-solid ratio.
According to the regression equation, the optimal parameters were the liquid-solid ratio 28.32 mL/g, extraction time 35.64 min, and ethanol concentration 87.25%. e theoretical highest yield of triterpenoids was 7.3118 mg/g predicted by the model. In order to validate the adequacy of the model, verication experiments were carried out by slightly modi ed conditions: liquid-solid ratio 28 mL/g, extraction time 36 min, and ethanol concentration 87%. e yield of triterpenoids of 7.338 ± 0.150 mg/g could be attained, which was 6.83% higher than the previous ultrasonic extraction method.

Characterization and Identi cation of Triterpenoids by LC-LTQ-Orbitrap-MS n .
Qualitative analysis of triterpenoids was performed on the HPLC-LTQ-Orbitrap-MS n system. ESI-MS spectra in both negative and positive modes were examined in this study. Negative-mode ESI was found to be sensitive for triterpenoids. All triterpenoids gave [M -H] − ions in their negative ion mass spectra. e total ion chromatograms (TICs) of triterpenoids in the negative ion mode by LC-LTQ-Orbitrap-MS n are shown in Figure 3. e fragmentation pathway of triterpenoids is summarized by using ganoderic acid A as the standard compound. e mass spectrum of ganoderic acid A and its major fragmentation pathways are given in According to the nontarget compound identi cation strategy based on the accurate mass measurement (<5 ppm), MS/MS fragmentation patterns, diagnostic product ions, and di erent chromatographic behaviors, 24 compounds were unambiguously identi ed from triterpenoids [9,10]. Table 1 Figure S1. ese results provided the critical information for constructing chemical ngerprints of triterpenoids.

Validation of HPLC and TLC Methods.
Prior to the establishment of the HPLC ngerprint, the precision, repeatability, and stability were chosen to validate the reliability of HPLC, which were expressed by the relative standard deviations (RSDs) of the retention time (t R ) and peak area (Pa). For the precision test, the working solutions were analyzed in triplicate, and RSD values of t R and Pa were lower than 0.2% and 4% (Table 2). To con rm the repeatability, ve di erent working solutions prepared from   experiment, developing solvent, sample concentration, chromogenic reagent, chromogenic temperature, and chromogenic time were optimized to achieve the optimum e ect of separation and coloration. Owing to the large polarity of triterpenoids, a small amount of formic acid was added in the developing agent. en, 1% vanillin-sulfuric acid solution was chosen as a chromogenic agent to color the compounds of triterpenoids. In order to validate the reliability of the TLC method, precision, repeatability, and stability were determined. e operation methods of working solutions were the same as HPLC. As shown in Table 3, RSDs of the ow rate value (R f ) ranged from 0.33% to 2.48%, which proved that the TLC experiments were reliable. However, RSDs of the peak area value (Pa) of the main characteristic peaks were from 2.32% to 7.79% (Table 3), inferring that a certain amount of error was produced by manual spotting.

Establishment of Global Chemical Pro le by ChemPattern
Software. Both common patterns of HPLC and TLC ngerprints for triterpenoids were generated to represent the characteristic peaks of this authenticated herbal medicine by ChemPattern software (Chemmind Technologies, Beijing, China). e similarity and clustering analyses on di erent kinds of chromatographic ngerprint data comprehensively evaluated the types and quantities of triterpenoids by ChemPattern software. e HPLC common pattern of 10 batches of samples is shown in Figures 5(a) and 5(b). Sample c from America was not suitable to establish the common model of the HPLC ngerprint because of its signi cant di erences in chemical composition. 11 peaks existing in 10 batches of samples were found as common peaks. According to the compound database of triterpenoids obtained from HPLC-LTQ-Orbitrap-MS n , peaks 1 to 11 were identi ed as lucidenic acid LM 1 , ganoderic Journal of Analytical Methods in Chemistry acid G, ganoderic acid B, lucidenic acid E, ganoderenic acid A, ganoderic acid A, lucidenic acid A, ganoderenic acid D, ganoderic acid D, lucidenic acid D, and ganoderic acid F, respectively. e relative retention time (t ′ R � retention time of the characteristic peak/retention time of the marker peak) and relative peak area (RPA � peak area of the characteristic peak/peak area of the marker peak) of the common peaks are shown in Tables S4 and S5. Peak 6, identi ed as ganoderic acid A, was selected as the reference peak. e results indicated that t ′ R of 11 common peaks (between 0.13% and 0.57%) was invariable between samples, and t ′ R was a valid parameter for constituent identi cation. However, the RPA (from 29.19% to 135.20%) showed signi cant di erences between 10 batches of samples, indicating that the content of triterpenoids from various sources was di erent. e similarity and clustering analyses of HPLC ngerprints were also analyzed by ChemPattern software. e advanced chemical pattern recognition module of ChemPattern could forecast the classi cation of complex samples. e similarities of samples a, b, d, e, f, g, h, and i were greater than 0.830 ( Figure 5(c)). However, the pro les of samples j and k were di erent from the common pattern, suggesting that these two samples belonged to Ganoderma sinense. is method could distinguish Ganoderma sinense and Ganoderma lucidum quickly. When the clustering distance was extended to 1.0, the samples from di erent regions could be divided into three groups. As shown in Figure 5(d), samples f, d, h, b, and a belonged to class I. ese ve samples were collected from Anhui Province. Samples k, j, g, and e were classi ed into class II, for the reason that the triterpenoid content and species of these samples were similar. Sample i was classi ed into class III individually, due to the di erent categories of triterpenoids. ese results illustrated that the origins of Ganoderma could be distinguished via clustering analysis. e HPLC ngerprints could not only re ect the chemistry information of Ganoderma but also distinguish the Ganoderma species from di erent geographical origins. e image of TLC taken by a digital camera was reversephase processed before imported into ChemPattern software ( Figure S2). Twelve tracks were set manually, and the information of TLC spots was turned into the gray curve. TLC ngerprints of di erent triterpenoid extracts and the common model were obtained as shown in Figures 6(a) and 6(b). e   ngerprints exhibited that nine common peaks were likely to represent the major constituents of triterpenoids. e similarity and clustering analyses of TLC by ChemPattern software are shown in Figures 6(c) and 6(d). e similarities of the samples a, b, d, e, f, g, h, i, j, and k were all greater than 0.900, while samples j and k were relatively low, which was consistent with the result of the HPLC ngerprint ( Figure 6(c)).
e clustering analysis showed that the samples could be divided into three clusters if the Euclidean distance was equal to 25 ( Figure 6(d)). e samples e, d, b, f, i, h, and a belonged to class I. Among these samples, only samples e and h were not collected from Anhui. Sample k was classi ed into class II individually, while samples j and g were classi ed into class III. e cluster analysis of 10 batches of Ganoderma showed a clear separation of the three groups, and common peaks played key roles in di erentiating these samples. Both HPLC and TLC ngerprint classi cations could provide a simple reference standard for quality identi cation of Ganoderma.
3.6. Discussion. In this article, RSM with the BBD method was successfully applied to optimize the factors for the ultrasonic  extraction of triterpenoids. e suitable conditions were as follows: liquid-solid ratio 28 mL/g, ethanol concentration 87%, and extraction time 36 min at 50°C. e ultrasonic extraction technology has the advantage of accelerating the extraction time, lowering the temperature, causing less damage to the structure of plant materials, and increasing the extraction yield. It is more suitable for the extraction of triterpenoids. RSM is an e ective statistical method useful for optimizing a complex process and evaluating the interaction between multiple parameters [21]. e yield of triterpenoids of 7.338 ± 0.150 mg/g could be attained, which was consistent with the theoretical predicted value (7.3118 mg/g). erefore, the model was considered to be reliable, and RSM could be used for predicting the ultrasonic extraction yield of triterpenoids.
HPLC-LTQ-Orbitrap-MS n technique was applied to identify the chemical structure of triterpenoids in Ganoderma with higher sensitivity. It provided high resolution and abundant structural information for not only the pseudomolecular ions but also the fragment species. e application of HPLC-LTQ-Orbitrap-MS n could provide a large amount of information related to the chemical structure and make up the de ciency of UV and DAD detectors [8]. Based on the accurate mass measurement, MS/MS fragmentation patterns, and diagnostic product ions provided by HPLC-LTQ-Orbitrap-MS n and literatures, 24 compounds were identi ed, which was one of the major tools for the study of the chemical substance of Ganoderma.
To investigate triterpenoids further, the chromatographic ngerprint method was applied to analyze the complex composition of Ganoderma. e ngerprint analysis technology is di erent from traditional analysis methods because it analyzes objects from the perspective of whole component information. e chromatographic ngerprint method was established through importing HPLC and TLC pro les into ChemPattern software for the comprehensive quality control of Ganoderma.
e common pattern of ngerprints provided the critical information for constructing chemical ngerprints of triterpenoids. In addition, the ngerprint method combined with the stoichiometry could be applied to identify the species and origins of Ganoderma. e results obtained by clustering analysis of HPLC and TLC ngerprints were slightly di erent due to the fact that HPLC possesses high resolution and reproducibility. e correct classi cation percentages of the pattern recognition method of HPLC ngerprint identication were higher than those of TLC. However, in contrast with the HPLC ngerprint, the TLC ngerprint was more cost-e ective and provided a vivid colorful image for parallel comparison. In summary, the chemical ngerprint method could be adopted as a reliable tool for the authentication and quality control of Ganoderma.

Conclusion
In the present study, the chemical composition of triterpenoids was clari ed by HPLC-LTQ-Orbitrap-MS n , and the global chemical pro le consisting of HPLC and TLC ngerprints was established. Eleven triterpenoid peaks which di ered signi cantly in all the analyzed samples were used as markers for origin identi cation and authenticity establishment of Ganoderma.
is work suggested that the developed global chemical pro ling method could provide a convenient approach, which might be applied for rapid evaluation, quality control, and authenticity establishment of Ganoderma products. In the future, the chemical constituents and pharmacological activities of Ganoderma would be explored in depth through the multidimensional ngerprints combined with chemometric methods, molecular biology, and pattern recognition techniques.

Conflicts of Interest
e authors declare that they have no con icts of interest. Table S1: Factors and levels in extraction experiments. Table  S2: Box-Behnken design of extraction parameters. Table S3: ANOVA for the response surface quadratic model of extraction parameters. Table S4: e t ′ R a of the common peak in di erent Ganoderma species. Table S5: e RPA b of the common peak in di erent Ganoderma species. Figure S1: e structures of compounds identi ed in triterpenoids from Ganoderma. Figure S2: TLC image of triterpenoids (a) and the reverse-phase processing image (b). Twelve tracks from left to right were ganoderic acid A and samples a, b, c, d, e, f, g, h, i, j, and k, respectively. (Supplementary Materials)