2.1 Reagents, chemicals, instruments, and materials

The crude FC, purchased from Panan (Zhejiang Province, P.R. China) and authenticated by Prof. Jianwei Chen, is the dried fruits of Cornus officinalis Sieb et Zucc. The steamed samples were processed at 100 ℃ for 1 h, while wine steamed FC were treated at 100 ℃ for 2 h after 30 % wine infiltration for 0.5 h. High-pressure steaming conditions were at 125 ℃ for 1 h under high pressure, while high-pressure wine steaming conditions were at 115 ℃ for 1 h under high pressure after 1 h immersion with 25 % wine. Wine immersion is to soak the medicinal materials in wine only for 1 h, and wine stir-frying is to fry the medicinal materials dried with after soaking with wine for 1 h. The details of processing methods were followed the previous investigations (Han et al., 2022). The voucher specimens were deposited in Jiangsu Key Laboratory of Chinese Medicine Processing, Nanjing University of Chinese Medicine (Nanjing, P.R. China).

UHPLC (Shimadzu 30A, Japan); AB Sciex Triple TOF 5600⁺ system (Canada) equipped with Analyst TF1.6, PeakView^TM 1.2.1, and MarkerView^TM 1.2.1 softwares; HPLC (Waters e2695, USA) allocated with UV detector (Waters 2489, USA) and ELSD detector (600ES-1104010, Alltech, USA); semi microbalance (0.01 mg, Quintix 35–1, Sartorius, P.R. China); ultra-pure water purification system (Millipore, Integral-3, USA); ultrasonic cleaner (KQ5300DC, Kunshan, P.R. China); SPE solid-phase extraction column (CNWBOND LC-C₁₈ SPE Cartridge, Shanghai, P.R. China); rotary evaporator (Buchi, Switzerland); electric heating adaptor (Xiangru, Nanjing, P.R. China).

Methanol, acetonitrile (chromatographic grade, TEDIA, USA; MS grade, E. Merck, Germany); phosphoric acid, ammonium acetate, trifluoroacetic acid (TFA), ammonium hydroxide (chromatographic grade, Aladdin, P.R. China); 1-phenyl-3-methyl-5-pyrazolone (PMP), glacial acetic acid, ethanol (Macklin, P.R. China); chloroform (20190920, Sinopharm, P.R. China); formic acid (75C1812LP, ACS, USA).

The reference substances including 5-hydroxymethyl-2-furfural (5-HMF, Q-042–180112), morroniside (M−027−190304), loganin (M−010−180118), cornuside (S-125–180525), and glucose (Glu) were prepared from Chengdu Herbpurity Co., ltd (Chengdu, P.R. China). Dextrans with known molecular sizes (2.70–2000.00 kDa), galacturonic acid (Gala), glucose (Glu), arabinose (Ara), mannose (Man), rhamnose (Rha), galactose (Gal), fucose (Fuc), fructose (Fru), and swertiamarin (111742–200501) were purchased from National Institute for Control of Pharmaceutical and Biological Products (P.R. China), while raffinose (Raf) was obtained from Aladdin (Shanghai, P.R. China).

2.2 Sample preparation for secondary metabolites

The reference substances (10 mg) including 5-HMF, morroniside, loganin, cornuside, and swertiamarin were weighed accurately and dissolved ultrasonically with 10 mL methanol to prepare for the mixture reference solution. The smashed and sieved (60 mesh) herbal materials were weighed (5 mg) and extracted with 90 % (v/v) and 50 % (v/v) ethanol for 1.5 h, respectively. The combined filter liquor was concentrated by a rotary evaporator in vacuum to 20 mL. The concentrated solution (1 mL) was diluted to 4 mL with methanol and centrifuged for 5 min at the speed of 4000 rpm. The elution program of the supernatant (2 mL) through SPE column to purify and dilute was set as follows: 5 %, 25 %, 50 %, 75 %, and 100 % (v/v) methanol for 3 mL, 2 mL, 2 mL, 1 mL, and 1 mL, separately.

2.3 Sample preparation for glycomics

Saccharides were prepared with the related literature (Zhou et al., 2018). Dried sample powder (40 mg) was extracted with water (400 mL × 1.5 h × two times). Ethanol was added to the cooling combined filtrate from filtration to make a final concentration of 80 % (v/v) and left overnight (24 h) under 4 ℃. Set aside the concentrated liquor (100 mL) prepared from supernatant filtrated again, while the sediment was placed in the ventilated area after being washed with ethanol three times.

1 mg prepared sediment was firstly dissolved with 15 mL ultra-pure water, and the liquor was loaded into the 3500 Da dialysis bag overnight. During the period, the water was changed at the 4th hour and again at the 12th hour, and the last change continued for 2 h. The dialysate lyophilized was the polysaccharides that were prepared for analysis.

2.4 Preparation of reference substances of saccharides

10.00 mg of dextrans with different molecular weights were weighed precisely and dissolved with 1 mL ultra-pure water, respectively.

2.00 mg of monosaccharides were weighed accurately and dissolved with 1 mL ultra-pure water, respectively. 200 μL hydrolysate was loaded into the test tube with a stopper and 200 μL ammonium hydroxide was added, respectively. The subsequent operations were identical to section 2.3.2.

5.00 mg of oligosaccharides were weighed precisely and dissolved with 1 mL ultra-pure water, respectively. Furthermore, the preparation of fructose followed the same disposed principles.

2.5 Chromatographic analysis

2.6 Method validation

The quantitative method was validated in terms of linearity, precision, accuracy and stability.

2.7 Component identification by UHPLC-Q-TOF-MS/MS analysis

Firstly, an in-house database was built by searching literature and database covering CNKI (China National Knowledge Infrastructure), Scifinder, Pubmed, and ScienceDirect comprehensively to collect the components information containing chemical name, molecular mass, molecular formula, and molecular structure of FC, and the exact mass to charge ratios (m/z) of compounds in positive and negative ion modes were computed. Subsequently, the Analyst TF 1.6 software was employed for collecting the chemical information of FC. And then, the obtained total ion chromatographic information was imported into the PeakView^TM 1.2.1 software, and the calculation function of the XIC Manager module was employed for preliminary analysis and calculation according to the retention time (error < 0.2 min) and mass-to-charge ratio (error < ± 10 ppm). The relevant parameters were set as follows: XIC intensity > 50 counts, S:N > 10, mass error ≤ 10, and isotope ratio% difference ≤ 20. Finally, the ion peak attribution could be determined when the matching degrees of molecular structure and the secondary fragment information were>75 %.

2.8 Simulation processing assay of morroniside

The aqueous solutions of morroniside (2 mg·mL⁻¹) were prepared, respectively, and the pH value was adjusted to 4.4 with diluted hydrochloric acid solution later. The mixture was heated at 125 °C for 1 h and centrifuged at 13,000 rpm for 5 min twice.

2.9 Data processing

MarkerView^TM 1.2.1 software from AB Sciex company was applied for screening the substances with significant differences in the crude and processed FC. The original data were imported into the software to operate the principal component analysis (PCA) and statistical analysis (t-test). The raw data collected and processed by Analyst TF 1.6 software were imported into MarkerView^TM 1.2.1 software. Subsequently, peak alignment, noise filtering and normalization were performed in the software, and the ions with significant differences were found by t-test analysis (P < 0.05 means statistical difference. t-value > 0 means the content increasing of the components). PCA of content information of saccharides from the crude and processed FC were performed by software SIMCA 14.1. Data were presented as mean ± SD of the determinations. Data sets performed using one-way analysis of variance (ANOVA) and values of P < 0.05 were considered significant. A defined variable (V_p) of polysaccharides was calculated as following equation: $$\mathit{Vp}=(Dp/Dc)\times C$$

(D_c: the D value of crude samples; D_p: the D value of processed samples; C: the total content of polysaccharides).

3.1 Secondary metabolites

85 components including iridoids, flavones, organic acids, tannins, etc. of the crude and processed FC were recognized in the positive and negative ion modes (Table 1). Representative total ion chromatograms were shown in Fig. 2. Iridoids and their glycosides, the representative monoterpenes of cyclopentane pyranoid contained in FC, possess extensive biological activities such as anti-cancer, liver protection, anti-inflammatory effects, and so on (Wang et al., 2020). Iridoids in FC identified by liquid chromatography-mass spectrometry were loganin, loganin acid, swertiamarin, hastatoside, morroniside, cornin, geniposide, dehydrologanin, secologanin, monotropein, secologanoside, dihydrocornin, secoxyloganin, sweroside, 7β-O-methyl-morroniside, 7α-O-methyl-morroniside, and cornuside. Loganin appeared at the retention time of 6.39 min and exhibited a parent ion at m/z 435, generating a fragment ion at m/z 227 by loss of a glucose fragment (162 Da) and a carboxyl and continuously removing a C₇H₁₀O₂ fragment (106 Da) as shown in Fig. 3a. Organic acids and furans are also the typical active compounds contained in FC. The fragments of 5-HMF under the experimental conditions were consistent with the report (Serra-Cayuela et al., 2013). The main ion fragments were exhibited at m/z 109, 81, and 53 (Fig. 3b). The fragmentation was elucidated by loss of a neutral fragment (H₂O) and CO group.

Close

Fig. 2

Representative total ion chromatograms in positive and negative ion modes. S: crude FC; B: wine steamed FC; C: high-pressure wine steamed FC.

Export to PPT

Close

Fig. 3

Fragmentation patterns of loganin and 5-HMF. a: loganin; b: 5-HMF.

Export to PPT

The PCA score plots of the crude and differently processed FC in the positive and negative ion modes were presented in Fig. 4, from which an interesting finding could be cleared that there were differences in the composition of FC after processing due to the separation among groups, especially in the negative ion mode. To describe the difference between the crude and processed FC deeply, SIMCA software was used for analysis. All original data was imported into software for Par standardization to perform OPLS-DA, a supervised recognition pattern to screen variable markers. In the positive ion mode, R²X (cum) = 0.759, R²Y (cum) = 0.953, Q² (cum) = 0.680, while in the negative ion mode, R²X (cum) = 0.8639, R²Y (cum) = 0.938, Q² (cum) = 0.724, indicating that the model had a decent representativity, stability, and predictivity. OPLS-DA score plots intuitively expressed the dispersion degrees of different samples and revealed quite a gap among the wine steamed, high-pressure wine steamed, and crude materials. That is the reason why the wine steamed and high-pressure wine steamed samples were selected to compare different components with the crude one. Interestingly, the spots of group E (wine stir-fried samples) in PCA score plots didn’t present a favorable aggregation since not only the method of stir-frying was difficult to control the temperature and time in practical operation, but also the subjective judgment contributed to the products. Different processing methods related to processing time, temperature, and procedure contributed to the composition of products (Jia et al., 2017). Additionally; in the PCA score plots, we found that the points in the high-pressure wine steaming group and the wine steaming group were close together, indicating that these two methods had similar chemical compositions on secondary metabolites. The Chinese pharmacopoeia stated that FC should be wine steamed for clinical application, while high-pressure wine steamed FC usually applied in clinic to save the costs. From the results of PCA, it was found that the types and contents of secondary metabolites of FC obtained from both methods were similar. From this perspective, it could be assumed that both methods could be applied clinically. The chemical informations of the differential ions (P < 0.05) screened by the MarkerView^TM 1.2.1 software were compared to the components identified in PeakView^TM 1.2.1 software. t > 0 was considered as the content increase of the components after processing. 32 and 31 components were selected to compare wine steamed and high-pressure wine steamed FC with crude one, respectively. From Table 2 and Table 3, we could know that the main components including iridoid glycosides decreased while triterpenoid acids and 5-HMF increased after wine steaming and high-pressure wine steaming. It was speculated that the processing methods of wine as an adjuvant could increase the solubility of acids and on the other hand, the iridoid glycosides were converted and degraded under the high temperature for a long time.

Close

Fig. 4

PCA and OPLS-DA score plots of crude and processed Fructus Corni (FC) in positive and negative ion modes. a: PCA (ESI⁺); b: PCA (ESI^-); c: OPLS-DA (ESI⁺); d: OPLS-DA (ESI^-). S: crude FC; Q: steamed FC; A: high-pressure steamed FC; B: wine steamed FC; C: high-pressure wine steamed FC; D: wine immersed FC; E: wine stir-fried FC.

Export to PPT

3.2 Molecular weight distribution of polysaccharides in crude and processed FC

Broad pharmacological investigations conducted by the scientific community indicated that polysaccharides from Fructus Corni have a series of bioactivities involving immunomodulation, anticancer, antioxidation, anti-aging, etc (Dong et al., 2018). Oligosaccharides could be a neotype given immunomodulatory and anti-inflammatory function (Coleman and Ferreira, 2020), and raffinose as a natural oligosaccharide from the botany could be a promising alternative to inhibit bacterial biofilm formation (Kim et al., 2016; Kim et al., 2021). However, the significances of saccharides as mentioned above were ignored in a large number of investigations on effective components in FC.

Multiple chromatographies such as HPLC (C₁₈)-PDA, HPLC (NH₂)-ELSD, and HPGPC-ELSD were applied to depict the saccharides of FC generally. It is widely confirmed that the bioactivity of polysaccharides is relative to molecular weight distributions and monosaccharide compositions.

The method for determining the molecular weights was high-performance gel permeation chromatography (HPGPC), the typical chromatograms were shown in Fig. 5. Taking the retention time of dextran as the abscissa and the logarithm of molecular weight as the ordinate, the standard curve was in line with the equation: Y = -0.3564X + 9.957 (R² = 0.9950). The properties were associated with molecular weights, which were the primary characterization of polysaccharides. The fluctuation of molecular weight distributions of FC polysaccharides indicated that processing had vital significance on the structure of polysaccharides. A novel parameter (D) to quantize the molecular weights of polysaccharides was introduced to evaluate the molecular weight distributions (Table 5). The equations were as follows: $$D=\frac{\sum (RIiMi)/\sum RIi}{\sum RIi/\sum (RIiMi)}$$

Close

Fig. 5

HPGPC chromatograms of polysaccharides from crude and processed FC. S: crude FC; Q: steamed FC; A: high-pressure steamed FC; B: wine steamed FC; C: high-pressure wine steamed FC; D: wine immersed FC; E: wine stir-fried FC.

Export to PPT

in which RIi and Mi, meant the peak height and molecular weight for each polysaccharide peak in the FC samples, respectively. The parameter (D) showed that for the molecular weight compositions of polysaccharides, there were little differences among the high-pressure wine steamed, wine steamed, wine stir-fried, and crude products.

3.3 Monosaccharides and oligosaccharides in crude and processed FC

3.3.2

3.3.2 Total monosaccharides compositions of polysaccharides

The monosaccharide compositions of polysaccharides were performed by HPLC-PDA with the crude and high-pressure wine steamed samples as references (Fig. 6). What could be believed was that both crude and processed samples contain polysaccharides made up of Man, Rha, Gala, Gal, Ara, Fuc, and Glu. It was clearly showed that the total contents of monosaccharides constituting polysaccharides had significant differences (Table 7, Fig. 7a), particularly in the wine stir-fried samples. The compositions of hydrolyzed and derivatized monosaccharides from polysaccharides were identified that the content of Gala was the highest in each sample, except for the wine stir-fried product, reaching 0.620 ± 0.034 mg·g⁻¹ and 0.643 ± 0.023 mg·g⁻¹ in the crude and high-pressure wine steamed products. The contents of Man, Ara, Fuc, and Gal clearly decreased after processing, while the content of Rha increased to 0.212 ± 0.009 mg·g⁻¹ after wine steaming, and the content of Glu increased to 0.460 ± 0.018 mg·g⁻¹ under wine immersion. Overall, the results revealed that the total contents of monosaccharides constituting polysaccharides from the processed FC samples decreased compared to the crude one due to the degradation of polysaccharides into free monosaccharides and oligosaccharides or the structural transformation. An investigation on the different effects of polysaccharides between crude and processed Astragalus on colitis mice revealed that structure could be the basis of the effect (Wu et al., 2021). These results indicated comprehensively that structure might play an important role in pharmacodynamics of polysaccharides.

Close

Fig. 6

HPLC-PDA chromatograms of monosaccharide compositions of polysaccharides. a: crude FC; b: high-pressure wine steamed FC. 1: Man; 2: Rha; 3: Gala; 4: Glu; 5: Gal; 6: Ara; 7: Fuc.

Export to PPT

Close

Fig. 7

Contents of monosaccharides and oligosaccharides from crude and processed FC. a: monosaccharide compositions of polysaccharides; b: free monosaccharides; c: raffinose; d: fructose; e: chromatograms of free monosaccharides: (1) crude samples, and (2) high-pressure wine steamed samples. 1: Man; 2: Rha; 3: Glu; 4: Gal; 5: Ara; 6: Fuc. **P＜0.01, ***P＜0.001, compared to crude samples (n = 6).

Export to PPT

Based on the characterization of the property and quantity of polysaccharides systematically, the total contents of monosaccharides were introduced and regarded as the contents of polysaccharides with a descriptive defined viable parameter (V_p = polysaccharides contents × D, Table 6) to evaluate the content changes of polysaccharides for carrying out the experiments of glycomics. The results showed that the high-pressure wine steamed and wine immersed samples differed less from the crude one. The small difference between the steamed and high-pressure steamed samples suggested that the treatments of both had less effects on the property of polysaccharides. For the wine stir-fried FC sample, the V_p was only 1.0372, less than half of the crude FC sample (2.2612), indicating a greater difference with the crude FC sample. The greater effect of wine stir-frying on polysaccharides might be related to the non-standard operation.

Table 6 Redefined polysaccharide variable (V_p) in crude and processed FC.

Group	Crude	Steaming	High-pressure steaming	Wine steaming	High-pressure wine steaming	Wine immersion	Wine stir-frying
Vp	2.2612	1.5411	1.5947	1.4409	1.8727	1.8591	1.0372

Table 7 Variability in monosaccharide contents of polysaccharides from crude and processed FC.

Group Content / mg·g−1	Crude	Steaming	High-pressure steaming	Wine steaming	High-pressure wine steaming	Wine immersion	Wine stir-frying
Man	0.110 ± 0.002	0.080 ± 0.007**	0.056 ± 0.003**	0.068 ± 0.003**	0.074 ± 0.003**	0.078 ± 0.006**	0.052 ± 0.001**
Rha	0.193 ± 0.005	0.161 ± 0.009**	0.133 ± 0.002**	0.212 ± 0.009**	0.162 ± 0.003**	0.171 ± 0.008**	0.114 ± 0.005**
Gala	0.620 ± 0.034	0.370 ± 0.025*	0.399 ± 0.012**	0.328 ± 0.014**	0.643 ± 0.023	0.384 ± 0.019	0.229 ± 0.018**
Glu	0.388 ± 0.012	0.314 ± 0.017**	0.311 ± 0.006**	0.213 ± 0.005**	0.314 ± 0.010**	0.460 ± 0.018**	0.193 ± 0.012**
Gal	0.399 ± 0.015	0.298 ± 0.022**	0.317 ± 0.007**	0.301 ± 0.011**	0.362 ± 0.015**	0.351 ± 0.014**	0.186 ± 0.016**
Ara	0.443 ± 0.014	0.246 ± 0.013**	0.297 ± 0.006**	0.291 ± 0.009**	0.283 ± 0.010**	0.346 ± 0.013**	0.173 ± 0.012**
Fuc	0.109 ± 0.010	0.091 ± 0.011*	0.060 ± 0.010**	0.024 ± 0.002**	0.032 ± 0.002**	0.098 ± 0.011	0.089 ± 0.009**

*P < 0.05, **P < 0.01, compared to crude samples.

3.4 Multivariate statistical analysis of glycomics

SIMCA 14.1 was employed to conduct principal component analysis (PCA), an unsupervised recognition pattern to analyze the contents of polysaccharides, raffinose, and free monosaccharides comprehensively (Fig. 8a-8b). The results indicated that the crude and processed products could be separated entirely, implicating deeply that processing exerted influence on the saccharides of FC. The primary components were separated from the samples and the values of R²X (cum) and Q² (cum) were 1.000 and 0.986 respectively, implying that a decent PCA model was built. Remarkably, the crude and high-pressure wine steamed samples could be assorted as a category, while the wine steamed sample differed considerably from the crude one, which were in contrast to PCA score plots of secondary metabolites. In our preliminary stage of the experiments, it was found by orthogonal design experiments that the crude and processed products of FC had good effects on TGF-β1-induced fibrotic HK-2 and HSC-T6 cells, and high-pressure wine steamed FC were found to be the most promising agent according to the index scores (Han et al., 2022). Combining the results of the secondary metabolomics, glycomics, and preliminary experiments, high-pressure wine steaming might be an alternative to wine steaming in the clinical application. However, further animal experiments were needed for an in-depth comparative study of their efficacies.

Close

Fig. 8

PCA score plots and transformation mechanism of saccharides. a: 2D PCA score plots of saccharides from crude and processed FC; b: 3D PCA score plots of saccharides from crude and processed FC. S: crude FC; Q: steamed FC; A: high-pressure steamed FC; B: wine steamed FC; C: high-pressure wine steamed FC; D: wine immersed FC; E: wine stir-fried FC.

Export to PPT

3.5 Simulation processing assay

The fructofuranose group, a key part of raffinose, tends to degrade in acidic and heating situations. Nevertheless, there was no increase in the content of fructose in the free monosaccharides on account of transformation. 5-HMF, produced by dehydration of hexose under high temperature and acidic conditions, is the distinctive product of Maillard reaction (Fallico et al., 2008). FC is a medicinal herb and functional food rich in saccharides and organic acids, thus it is possible to generate 5-HMF when heating for medication use. Morroniside, the characteristic component of iridoid glycosides containing glucose, could be a simulation processing object. Given the above, a simulation processing assay was introduced to clarify the transformation mechanism of characteristic constituents of FC. 5-HMF could be detected in the processed products of polysaccharides, raffinose, glucose, morroniside, and fructose (Fig. S1). The results were consistent with the literature reported that 5-HMF is a by-products existing in saccharides (Wu et al., 2012). Polysaccharides, raffinose, and morroniside could produce 5-HMF through glycosylation of 5-HMF glycosides (5-GMF). The results of simulation processing assay indicated that the fructose from raffinose could be hydrolyzed into 5-HMF, which was the potential transformation mechanism of the decreased content of fructose in processed FC and explained by the increase in 5-HMF content after processing. In the processed products of morroniside (Fig. 9b), glucose, 5-HMF, and aglycone were detected based on the TOF MS and TOF MS/MS spectra, indicating that in the acidic and heating conditions, the loss of glucose in the morroniside was to transform into 5-HMF. That was the reason why the decrease of morroniside in the processed FC compared to crude one. Furtherly, the intensity of 5-HMF in all samples was determined to investigate the influence of processing conditons on FC (Fig. 9a). Interesting enough, the intensity of 5-HMF exhibited the different variations. The crude and wine-immersed FC had the lower level of 5-HMF in all samples, revealing that the heating was the essential condition on transformation into 5-HMF due to the fact that these two samples were not heated during the pretreatment process. As many investigations reported, 5-HMF is the cure for confronting fibrosis and oxidative injury of the liver (Han et al., 2017) as well as it is toxic to the nervous system and able to induce anaphylactoid reactions (Li et al., 2020; Wang et al., 2020). Briefly, there is no denying that 5-HMF is an underlying chemical marker of processed FC and an indicator to estimate the safety of processed herbs. Traditional Chinese medicine has its distinctive pattern of processing, and wine steaming is a commonly used method for FC under clinical regulations. Therefore, the paper further investigated the diverse processing methods with and without wine on how to affect the FC from the perspective of saccharide compositions. Finally, it is a necessity to standardize the processing methods that could influence the variation of these components.

Close

Fig. 9

Intensity of 5-HMF in different samples and transformation mechanism in morroniside. a: intensity of 5-HMF in crude and different processed samples; b: mechanism on transformation of morroniside to 5-HMF. Compared to the crude FC, *P < 0.05,^**P < 0.01, ^***P < 0.001.

Export to PPT