2.1 Reagents and chemicals

Hydrochloric acid was purchased from Tianjin Damao Chemical Reagent Factory (Tianjin, China). Glacial acetic acid, sodium bicarbonate, and sodium acetate anhydrous were bought from Tianjin Fengchuan Chemical Reagent Co., Ltd. (Tianjin, China). Phenyl isothiocyanate (PITC), triethylamine, and formic acid were supplied by Shanghai Aladdin Bio-Chem Technology Co., Ltd. (Shanghai, China). n-Hexane was bought from Tianjin Concord Technology Co., Ltd. (Tianjin, China). Reference standards including geniposide (Gen), Gep, Cat, aspartate (Asp), glutamate (Glu), Lys, Arg, histidine (His), alanine (Ala), valine (Val), glutamine (Gln), proline (Pro), leucine (Leu), isoleucine (Ile), tryptophan (Trp), phenylalanine (Phe), Fru, Glc, sucrose (Suc), melibiose (Mel), raffinose (Raf), manninotriose (Man), and Sta were obtained from Shanghai Yuanye Biotech Co., Ltd. (Shanghai, China). Leo was provided by Chengdu Ruifen Si Biotech Co., Ltd. (Chengdu, China). The purity of all chemicals was greater than 98 %. β-Glucosidase, acetonitrile (HPLC grade), and methanol (HPLC grade) were obtained from Sigma-Aldrich Co., Ltd. (St. Louis, MO, USA). Column chromatography (CC) was carried out using MCI Gel (CHP20/P120, Mitsubishi Chemical Corporation, Tokyo, Japan). Water used in the experiment was purified by a Milli-Q system (Millipore, Milford, MA, USA).

The sodium acetate buffer solutions at pH 4.8 and 6.0 were prepared according to the guidelines released by Chinese Pharmacopoeia, respectively.

FRR samples, collected in Jiaozuo city (Henan, China) in November 2020, were authenticated by Prof. Tianxiang Li of Tianjin University of Traditional Chinese Medicine (Tianjin, China) based on original, characteristic, microscopical, and chemical identifications specified by Chinese Pharmacopoeia.

2.2 Preparation of RRR and PRR

To acquire RRR, three batches (2500 g/batch) of FRR with a uniform size (weight of 110 ± 15 g, diameter of 58 ± 10 mm) were selected after removing their fibrous roots. The FRR samples were then dried at 50 °C for 10 days (10 h per day) until the moisture was less than 15 %.

PRR samples were obtained through the following procedure: Three batches of RRR samples were mixed with RW at a weight ratio of 10:4, soaked for 24 h, and then steamed for 12 h until softened. The softened samples were taken out, dried, blended with resulting steamed liquid, and continued to be steamed for 6 h. Afterwards, they were dried at 60 °C in the oven until the texture became viscous and dark-black colored, and then cut into thick slices (2-4 mm).

2.3 Preparation of standard and sample solutions

2.4 The enrichment of crosslinking coloring reaction products (CCRPs) and Maillard reaction products (MRPs) from processed products of RR

RRR sample powder (2.5 g) was ultrasonically extracted (144 W, 35 kHz) with 75 mL ultrapure water for 30 min. After centrifugation at 13,322 g for 10 min, the supernatant was collected and concentrated to 30 mL. The concentrated solution was subjected to MCI Gel CC by elution with MeOH-H₂O (0:100, 50:50, 75:25, v/v) and 150 mL of each gradient elution was collected to afford fractions with CCRPs. Each fraction was then concentrated and analyzed by UV–vis. The enriched CCRPs from PRR samples were analyzed by the established method.

RRR sample powder (0.25 g) was ultrasonically extracted (144 W, 35 kHz) with 25 mL 75 % methanol aqueous solution for 30 min. After centrifugation at 13,322 g for 10 min, the supernatant was collected. Then, 10 mL of the supernatants was taken into the sealed dialysis bag (molecular weight cutoff 3500 Da) and immersed in ultrapure water for 24 h. During this period, the ultrapure water was refreshed three times. The enriched MRPs were analyzed via a UV–vis system. The successful method was undertaken to enrich MRPs from PRR samples.

2.5 UPLC-PDA analysis

The amino acids contents of different processed products of RR and RW were analyzed by an ACQUITY UPLC system (Waters, Milford, MA, USA) equipped with an ACQUITY UPLC® BEH C18 column (2.1 × 100 mm, 1.7 µm, Waters) maintained at 52 °C, using the method outlined by Zhang et al. (2016) with slight modifications. The mobile phase system consisted of pH 6.5 sodium acetate buffer solution (A) and acetonitrile (B), using the following gradient program: 0-2 min, 1 %-4% B; 2-4 min, 4 %-15 % B; 4-5 min, 15 %-17 % B; 5-6 min, 17 %-20 % B; 6-10 min, 20 %- 40 % B.

In order to explore the rule of crosslinking coloring reaction, UPLC analysis was performed on an ACQUITY UPLC® HSS T3 column (2.1 × 100 mm, 1.8 μm, Waters) maintained at 35 °C. The mobile phase consisted of 0.1 % formic acid solution (A) and acetonitrile (B), using a following gradient program: 0-6 min, 1 %-40 % B. Both mobile phase systems were maintained at flow rate of 0.3 mL/min. The sample injection was 2 µL, and the monitoring wavelengths were set at 254 and 203 nm.

2.6 HPLC-ELSD analysis

Using the method described in a previous study (Hu et al., 2013), the analysis of oligosaccharides and monosaccharides in different processed RR products and the exploration of oligosaccharides hydrolysis were performed on a Waters 2695 HPLC system coupled with ELSD. The separation was achieved on a Prevail Carbohydrate ES column (4.6 mm × 250 mm, 5 μm, Shanghai Chenqiao Biotech Co., Ltd., China) operated at 30 °C. The mobile phase was composed of acetonitrile (A) and water (B), using following elution condition: 0-20 min, 28 %-36 % B. The sample injection was 5 µL. The evaporator temperature for the ELSD was set at 100 °C with a nitrogen gas flow at 1.0 L/min.

2.7 UV–vis analysis

UV–vis analysis was performed on a UV-2600 system (SHIMADZU, Tokyo, Japan) equipped with quartz cells with 1 mm path length. The absorbance spectra were recorded ranging from 200 to 800 nm.

2.8 Statistical analysis

The tested values were represented as mean ± standard deviation (SD). Excel 2016 and SPSS Statistics 24.0 were used for data analysis. Unpaired t-test was used for multiple comparisons (P<0.05 for significant difference) and plotted with Origin 2019b.

3.1 Influence of processing on the physicochemical properties of RR

By taking the weight, water loss rate, buffering capacity, and pH into account, the variations of key physicochemical properties were comprehensively and systematically investigated during the processing of RR. As shown in Fig. 1(A-D), a notable transition from FRR to PRR was observed, characterized by darkening color, sweetening flavor, and stickier texture, which indicated the change of internal physicochemical environment of RR. The lab values about color changes from FRR to PRR are provided in Fig. S1. Obvious decrease in the weight of RR was recorded, with moisture reducing by 71.5 % after extended heating. To evaluate buffering capacity, different volumes of 0.5 % HCl (0 ∼ 1000 µL) and 0.5 % NaOH (0 ∼ 1000 µL) were added into 50 mL of ultrapure water and FRR juice, respectively. Contrary to water, whose pH value fluctuated between 2.49 and 11.40, the pH value of FRR juice consistently stabilized around 6.5, demonstrating its incredible buffering ability. Besides, the pH value showed slight decrease from FRR to RRR, followed by a distinct drop from RRR to PRR, with pH value falling from 6.5 to 4.8. This decline could be ascribed to the decomposition of the proteins into acidic peptides or free amino acids released under high temperature and long-term heating during the processing of PRR (Ni et al., 1989).

Close

Fig. 1

(A) Variations in the appearance during the processing from FRR to PRR. (B) Variations of the weight and water loss rate during the processing of RRR. (C) The buffering capacity of FRR. (D) The pH values in different processed RR products. *P＜0.05; **P＜0.01; ***P＜0.001.

Export to PPT

3.2 Unveiling the transformation mechanism of saccharides hydrolysis for “sweetness as candy”

Five oligosaccharides (Suc, Mel, Raf, Man, and Sta) and two free monosaccharides (Glc and Fru) were determined in different processed RR products, whose content variations are illustrated in Fig. S2. It was observed that Sta was the most abundant oligosaccharide, reaching up to 562.4 mg/g in FRR, yet dramatically decreased to 42.93 mg/g in PRR induced by processing. On the contrary, Man, Mel, Glc, and Fru were barely undetectable in FRR, yet witnessed to climb in content especially from RRR to PRR, with concentrations reaching 329.3 mg/g, 51.39 mg/g, 101.2 mg/g, and 163.2 mg/g in PRR, respectively. The levels of Suc and Raf were highest in RRR, followed by FRR, and were lowest in PRR, suggesting the hydrolysis of oligosaccharides during processing.

To further explore the impact of processing on the changes of saccharides contents in RR, the mechanisms were discussed under simulated processing condition of PRR (100 °C, pH 4.8) for 18h. As shown in Fig. 2A, Suc, composed of Glc and Fru, was hydrolyzed into the corresponding monosaccharides after 18 h of heating, revealing the increase of Glc and Fru and the decrease in Suc levels in PRR. Similarly, Raf, which is formed by Fru and Mel, broke down into corresponding Fru and Mel, leading to a decline in Raf and an increase in Mel levels. Furthermore, the increase of Man and Fru can be attributed to the release of fructofuranosyl in Sta. However, Mel and Man exhibited no significant change in content after 18 h of heating, suggesting that the glycosidic bond between Fru and galactose (Gal) is stable and may require more severe conditions for hydrolysis. The processing-induced chemical transformation pathways of saccharides in RR are presented in Fig. 2B, from which the tasteless saccharides such as Sta and Raf are gradually converted into sweeter monosaccharides such as Fru, contributing to a “sweetness as candy” flavor in PRR.

Close

Fig. 2

(A) Representative HPLC-ELSD chromatograms of reacted solutions of saccharides hydrolysis at 0 h and 18 h (1. Fru; 2. Glc; 3. Suc; 4. Mel; 5. Raf; 6. Man; 7. Sta). (B) The processing-induced chemical transformation pathways of saccharides in RR (Solid arrows: validated degradation pathways; dotted arrows: possible degradation pathways).

Export to PPT

3.3 Exploration on the CCRPs in different processed RR products to explain the mechanism of crosslinking reaction for“black as lacquer”

Gen, driven by β-glucosidase secreted by medicinal plants, is transformed into Gep and subsequently reacts with methylamine to produce blue pigment (Bentes et al., 2015). Researchers have documented the presence of β-glucosidase in FRR (Zhao et al., 2006). Our previous study has demonstrated that iridoid glycosides, especially Cat, Leo, and Gen, presented declining tendency from RRR to PRR (Yang et al., 2024). In this study, the dynamic variations in amino acids contents in different processed RR products were investigated using UPLC-PDA analysis. As shown in Fig. S3, the contents of alkaline amino acids (Lys, His, and Arg), acidic amino acids (Asp and Glu), and neutral amino acids (Trp and Gln) decreased by more than 78 % during the process from FRR to PRR. Noteworthily, RW also contains abundant amino acids, significantly influencing the contents of amino acids in PRR. Therefore, it is speculated that iridoid glycosides, catalyzed by β-glucosidase, can react with various amino acids to form CCRPs in PRR.

To substantiate our hypothesis, an MCI CC method was performed to enrich the CCRPs in different processed RR products. The ratio of material/solvent (1:10, 1:20, 1:30, and 1:40) and extraction time (20, 30, and 40 min) were optimized to maximize extraction efficiency and satisfactorily yield CCRPs. As shown in Fig. 3, the absorbance measurements of enriched fractions from FRR, RRR, and PRR revealed a characteristic absorption peak at 540 nm in RRR and PRR sample solutions, with the peak being particularly noticeable in the 50 % MeOH fraction of PRR. However, there was no absorption at 540 nm in FRR. This phenomenon offered stronger evidence that crosslinking coloring reaction does occur during the processing of both RRR and PRR, with more obvious effects observed in PRR.

Close

Fig. 3

UV–vis absorption curves of CCRPs in (A) FRR, (B) RRR, and (C) PRR. (D) The absorbance of different enriched fractions from FRR, RRR, and PRR at 540 nm (n = 3). *P＜0.05; **P＜0.01; ***P＜0.001.

Export to PPT

To elucidate the formation mechanisms of CCRPs in RRR, Cat, Leo, and Gen, respectively represented by the epoxy ether[c]pyran, cyclopentano[c]pyran, and cyclopentenyl[c]pyran ring systems, were employed for reaction with Ala, Arg, and Glu (neutral, alkaline, and acidic amino acids) seperately under simulated processing conditions of RRR (50 °C, pH 6.0) with the presence of β-glucosidase. As depicted in Fig. 4, in 12 h, the content of Gen consistently decreased, whereas the content of its aglycone, Gep, initially increased within the first 6 h and then approached an equilibrium state, indicating that Gen is metabolized into Gep by β-glucosidase. Concurrently, the absorption values of CCRPs formed from reactions between Gen and different amino acids gradually raised in the presence of β-glucosidase, manifesting that Gep can react with amino acids to produce CCRPs. Notably, characteristic absorption peaks at 590-600 nm were detected in the UV–vis spectra at 12 h. In contrast, Cat and Leo fail to form CCRPs (Figs. S4 and S5) due to the inability of β-glucosidase to transform these compounds into their corresponding aglycones.

Close

Fig. 4

UV–vis spectra of CCRPs from reactions between Gen and (A₁) Ala, (A₂) Arg, or (A₃) Glu in the presence of β-glucosidase. Dynamic variations on the contents of Gen and Gep, and the absorbance of CCRPs from reactions between Gen and (B₁) Ala, (B₂) Arg, or (B₃) Glu in the presence of β-glucosidase in 12 h.

Export to PPT

The formation of CCRPs from Gep is happened through nucleophilic attack by primary amine groups on the C-3 atom of Gep, subsequently embedding tertiary nitrogen in the six-member ring in place of an oxygen atom (Mu et al., 2013). Gardenia blue, the final product of a crosslinking coloring reaction, is supramolecular assemblies, which is formed via the covalent polymerization of basic units into dimers, followed by non-covalent self-assembly. The development of color is influenced by various factors, such as pH, temperature, type of amino acid, and genipin-to-amino acid ratio (Brauch, 2016; Zhang et al., 2023). Crosslinking coloring reaction derived from reactions between Gep and amino acids provides a plausible basis for the darkening color of PRR.

3.4 Exploration on the MRPs in different processed RR products to explain the mechanism of Maillard and caramelization reactions for“black as lacquer”

Maillard reaction not only generates volatile small molecules and brown macromolecules that are crucial for the food and nutrition industries, but also plays an essential role in the processing of TCMs (Cao et al., 2009; Liu et al., 2020). MRPs are formed in the early, intermediate, and final stages. As significant Maillard reaction intermediates, Amadori rearrangement products are key flavor and browning precursors (Chen et al., 2022), which can be monitored through absorbance at 294 nm. The increase in absorbance at 420 nm serves as an indicator for browning development in the final stage of the reaction (Zhou et al., 2016b; Sun et al., 2023). The MRPs formation may be significantly affected by the type and concentration of precursors, pH, time, and temperature (Gong et al., 2019b). Along with the Maillard reaction, caramelization of saccharides also occurs, which is a group of reactions that take place when carbohydrates are exposed to high temperatures in the absence of amino groups (Henning and Glomb, 2016; Wang et al., 2022). The two reactions take place simultaneously, influencing each other and enhancing chemodiversity in the processing (Eliodório et al., 2023). Therefore, MRPs are likely presented in PRR which is evidenced by the gradually deepened color during the processing of RR.

The preparation method of MRPs was optimized by investigating extraction solvent (H₂O, 25 %, 50 %, and 75 % methanol aqueous solution), the ratio of material/solvent (1:50, 1:100, and 1:200), and extraction time (20, 30, and 40 min) to achieve the highest extracting efficiency. To minimize the impact of extraneous components, dialysis was employed to further enrich MRPs and remove interfering substances. The dialysis results confirmed that MRPs exist in different processed products of RR. As shown in Fig. 5, distinct absorption peaks at 440 nm were observed in FRR, RRR, and PRR, showing the highest absorption in PRR, which indicates that the processing conditions for PRR are more favourable of producing MRPs. However, the formation mechanisms of the MRPs in PRR remain uncertain and deserve further exploration.

Close

Fig. 5

(A) UV–vis absorption curves of MRPs in different processed RR products. (B) The absorbance value of MRPs at 440 nm (n = 3). *P＜0.05; **P＜0.01; ***P＜0.001.

Export to PPT

The rules of Maillard reaction occurring between saccharides and representative amino acids were explored under simulated preparation condition of PRR (100 °C, pH 4.8), whose results are presented in Fig. 6. The UV–vis absorbance at 450 nm was applied to assess changes on the MRPs from the final stage in Fru, Fru-Ala, Fru-Arg, and Fru-Asp reaction systems. From the UV–vis absorption spectra, obvious characteristic absorption peak at 450 nm can be witnessed in Fru, Fru-Ala, Fru-Arg, and Fru-Asp reaction systems. Similarly, the Sta, Sta-Ala, Sta-Arg, and Sta-Asp reaction systems exhibited strong absorption peak at the focused wavelength. In contrast, Glc, Suc, Raf, and their combinations with Ala, Arg, and Asp showed no significant absorbance at 450 nm. This suggests that Fru and Sta are primary saccharides contributing to the black color of PRR, thus explaining the mechanism behind the “black as lacquer” appearance of PRR.

Close

Fig. 6

UV–vis spectra of MRPs from reactions between (A) Fru and different amino acids (Ala, Arg, and Asp), (B) Glc and different amino acids (Ala, Arg, and Asp), (C) Suc and different amino acids (Ala, Arg, and Asp), (D) Raf and different amino acids (Ala, Arg, and Asp), and (E) Sta and different amino acids (Ala, Arg, and Asp), respectively. (F) UV–vis spectra obtained by reactions of amino acids themselves.

Export to PPT

The products from Maillard, crosslinking coloring, caramelization, and saccharides hydrolysis reactions are closely related to processing conditions. With the extended steaming time and increased steaming cycles, the PRR became darker in surface color and sweeter in taste. This is due to the fact that the extended steaming time and increased steaming cycles promote the formation of black materials and degradation of saccharides during the processing of PRR. Shennong’s Classic of Materia Medica also have documented that RR must undergo multiple cycles of steaming and sun-drying, even to the extent of being “nine times steamed and nine times sun-dried,” to ensure it achieves a quality “black as lacquer and sweetness as candy”. Therefore, the steaming time and temperature should be strictly controlled to obtain PRR with high quality and efficacy. Furthermore, based on a deeper understanding of the scientific principles underlying the processing of PRR, our study suggested that the level of monosaccharides and black materials are the important quality indicators for PRR.