2.1 Reagents and materials

HPLC-grade methanol and acetonitrile were purchased from Sigma-Aldrich Inc. (St. Louis, MO, USA). HPLC-grade formic acid was obtained from Shanghai Aladdin Biochemical Technology Co. Ltd (Shanghai, China). Water employed in the experiment was purified with Milli-Q water purification system (Millipore, Billerica, USA). PO (B12002008028), IPO (Y22802008028), Neo (X05711812017), and Bav (B07611804027) were purchased from Chengdu Herbpurify Co., Ltd (Sichuan, China). P (ST01250120), IP (ST03150120), Iso (ST04850120), and Pso (ST03720120) were obtained from Shanghai Standard Biotech Co., Ltd. (Shanghai, China). Bak was prepared in our laboratory. The purity of all standards was above 98% by LC-UV analysis.

Forty two batches FP, numbered B1–B42, were purchased from the different medicinal materials markets, whose detailed information is displayed in Table S1. The sample (B41) for the accelerated stability test was stored in the stability test chamber (40 °C ± 2 °C, RH 75%±5%) for 0, 1, 2, 3, and 6 months. All samples were deposited in State Key Laboratory of Compound-based Chinese Medicine, Tianjin University of Traditional Chinese Medicine (Tianjin, China).

Three parallel plants of P. corylifolia, raised in the medicinal garden of Hebei University of Chinese Medicine (Hebei, China), were randomly selected as representative samples and repeatedly collected at six-time points (t1–t6 were designated at Jul.15, Aug.15, Aug.30, Sep.15, Sep.30, and Oct.15 in 2020, respectively). The P. corylifolia samples collected on Sep.15 were selected and divided into different grades according to the specifications of different plant parts (stems, leaves, flowers, and fruits). The stems were cut into about length of 10 cm and subsequently classified into five specifications by measuring the diameter, <1mm (0.83 ± 0.06 mm), 1–3 mm (2.64 ± 0.02 mm), 3–7 mm (5.46 ± 0.65 mm), 7–10 mm (8.67 ± 0.59 mm), and > 10 mm (10.91 ± 0.47 mm). The leaves were divided into four levels according to their width, namely 2–3 cm (2.54 ± 0.30 cm), 3–4 cm (3.41 ± 0.24 cm), 4–5 cm (4.3 ± 0.26 cm), and 5–6 cm (5.53 ± 0.30 cm). The flowers were grouped into initial flowers, blooming flowers, and terminal flowers according to the state of flowers' opening. The fruits were divided into green and mature fruits based on the organoleptic status. All P. corylifolia samples were dried by freeze dryer (FDU-2110, Tokyo Rikakikai Co., Ltd., Japan).

The sample (B42) was toasted with blast drier (101-2AB, Tianjin Test Instrument Co., Ltd., China) at different temperatures (40, 60, 90, 100, 110, 120, 130, 150, 180, 190, 200, and 210 °C) for 2 h, respectively. All three parallel samples were subjected to each temperature.

2.2 Preparation of reference and sample solutions

Standards were accurately weighed and dissolved with methanol to obtain respective stock solutions at the concentration of 2 mg/mL for PO and IPO; 1 mg/mL for P, IP, Neo, Iso, and Bav; 0.5 mg/mL for Pso; and 16 mg/mL for Bak. Afterward, the mixed stock standard solution was prepared by employing stock standard solution of nine compounds to reach the final concentrations: PO, 147.7 μg/mL; IPO, 60.92 μg/mL; P, 35.60 μg/mL; IP, 34.69 μg/mL; Neo, 29.79 μg/mL; Iso, 34.76 μg/mL; Pso, 20.12 μg/mL; Bav, 49.15 μg/mL; and Bak, 720.9 μg/mL. Subsequently, the working solutions at the different concentrations were diluted for constructing calibration curves.

The accurately weighed sample powder (0.2 g) was transferred into 50 mL volumetric flask and ultrasonically extracted at 60 °C by methanol for 20 min, then cooled to ambient temperature and diluted to scale by adding methanol. The extracted solution was centrifuged at 12,000 rpm for 10 min to obtain the supernatant and diluted with ultrapure water (v:v, 1:1) to obtain the sample solution.

2.3 UHPLC-PDA conditions

ACQUITY UPLC H-class plus system (Waters Corporation, Milford, MA, USA) was used to perform the chromatographic analysis by using ACQUITY UPLC BEH C18 column (100 mm × 2.1 mm, 1.7 μm) at 60 °C. The mobile phase was composed of 0.1% formic acid aqueous solution (A) and methanol (B) and implemented in the gradient elution as follows: 15–78% B in 0–7 min, 78–90% B in 7–11 min, 90% B in 11–13 min. The flow rate was 0.3 mL/min, and the sample injection was 2 μL. The detection wavelength was set at 246 nm.

2.4 Methodological validation

The analytical method established in this study was employed to validate linearity, the limit of detection (LOD), the limit of quantitation (LOQ), precision (intra- and inter-day), stability, and recovery test. The calibration curves were drawn using the peak area (y-axis) and the corresponding concentrations of the compounds (x-axis). The LOQ and LOD of the tested compounds were determined by gradually reducing the concentration of the standard solution until the S/N was about 10 and 3, respectively. In order to test the intra- and inter-day precisions, the sample solution was injected six times on the same day and three consecutive days, subsequently. Repeatability was confirmed by preparing and analyzing six sample solutions. The stability was studied by injecting sample solution at 0, 2, 4, 6, 8, 10, 12, and 24 h. The recovery test was carried out by adding standard solution to 0.1 g sample powder (n = 6), which was processed according to the method of sample preparation.

2.5 Data analysis

All figures were drawn by Origin 9.1 software (Originlab Corp., Northampton, MA, USA) and Adobe Illustrator CC2017 (Adobe Systems Incorporated, San Jose, CA, USA). The Origin 9.1 software and SPSS 21.0 software (IBM, Armonk, NY, USA) were used for statistical analysis of data.

3.1 Methodological validation of UHPLC-PDA analysis for quantitation of nine compounds

For the sake of the resolution and sensitivity of the tested compounds, gradient elution condition, column temperature, and detection wavelength were systematically optimized. Also, based on our previous extraction conditions regarding FP (Yang et al., 2018), the sample preparation method of nine compounds was optimized by systematically investigating extraction solvent, extraction time, material-liquid ratio, and dilution solvent. The typical chromatograms of sample and mixed standard solution are displayed in Fig. 1.

Close

Fig. 1

UHPLC-PDA chromatograms of FP sample solution (A) and the mixed standard solution (B). (1. PO, 2. IPO, 3. P, 4. IP, 5. Neo, 6. Iso, 7. Pso, 8. Bav, 9. Bak).

Export to PPT

As shown in Table 1, the simultaneous quantification of nine compounds in FP was validated. The calibration curves of nine compounds were established with the determination coefficient (r²) exceeding 0.9996, indicating a good linear correlation within the tested ranges. The LOD and LOQ were 0.017–0.139 μg/mL and 0.052–0.417 μg/mL, respectively. For the quantified compounds, the RSDs of intra-day and inter-day precisions were below 0.9% and 2.0%, respectively. The RSD of repeatability was below 1.0%. With RSD less than 0.5%, the stability test result showed that the determined compounds in the sample solution were stable for 24 h. The mean recovery was in the range of 95.43%–104.1%, with RSD less than 1.8%. Altogether, the results demonstrated that the established method could be successfully used in subsequent studies.

3.2 Characteristics of compounds' accumulation and distribution in different parts and different specifications of P. Corylifolia

Characteristics of compounds' accumulation and distribution in different parts of P. corylifolia were studied in the light of stems (1–3 mm), leaves (3–4 cm), flowers (blooming flowers), and fruits (mature fruits) collected on Sep.15. As displayed in Fig. 2, coumarins (PO, IPO, P, and IP) and monoterpene phenol (Bak) were found in all plant parts of P. corylifolia. The content of PO, IPO, P, and IP was higher in flowers and fruits than in stems and leaves. Bak was the most abundant compound in P. corylifolia, whose content was a descending order in flowers, fruits, leaves, and stems. Flavonoids (Neo, Iso, and Bav) were only detected in fruits but not in flowers, stems, and leaves, which may be related to the specific expression of key enzymes and genes for flavonoids biosynthesis (Huang et al., 2016; Moriguchi et al., 2002).

Close

Fig. 2

Histogram of compounds content in different parts of P. corylifolia.

Export to PPT

Characteristics of compounds' accumulation and distribution in stems, leaves, flowers, and fruits of different specifications were studied. As shown in Fig. 3A, the PO, IPO, P, IP, and Bak content were higher in young leaves (2–3 cm) and lower in mature leaves (3–6 cm). The results suggest that as the leaves develop, noticeably progressive increase is witnessed in width (larger surface area), the amount of each compound decreases correspondingly. As depicted in Fig. 3B, similar to leaves, the content of PO, IPO, P, IP, and Bak were higher in young stems (<1mm). The closer to the base of the plant the stem is, the larger the diameter of the stem is, the more mature the stem is, and the lower the content of each compound is. It is speculated that coumarins and monoterpene phenol are mainly distributed on the stem surface, suggesting that the more mature the stems is, the smaller the mass ratio of the surface layer to the vascular cylinder is, and the lower the content of each compound is. The content determination results in flowers at the different opening states is shown in Fig. 3C. From initial flowers to terminal flowers, the content of PO and IPO increased gradually, while the variation in the content of P and IP was opposite. However, Pso was not detected, and the amount of Bak was highest in terminal flowers, followed by initial flowers and blooming flowers. Flavonoids were detected only in terminal flowers close to the fruit-bearing stage. As shown in Fig. 3D, the content of flavonoids (Neo, Iso, and Bav) and monoterpene phenol (Bak) in young fruits was higher than that in mature fruits. As for coumarins, two differential trends were observed for their content distribution in fruits. On the one hand, the content of Pso in green fruits was significantly lower, which was about 0.12 times that in mature fruits. On the other hand, the PO, IPO, P, and IP content was almost same in green and mature fruits. The secretory cavities found in secretory tissues of seed-bearing plants are mainly involved in the biosynthesis and accumulation of secondary metabolites including coumarins, flavonoids and phenols (Zhang et al., 2020). Qin and Liu, (2007) reported that the density of secretory cavities in P. corylifolia is positively correlated with the content of P and IP. The number and volume of secretory cavities in stems and leaves are smaller than that in fruits, and the density of secretory cavities in young stems and leaves is greater than that in mature stems and leaves. Collectively, our study shows that the content of each compound in fruits or flowers is higher than that in leaves or stems, and the more mature the leaves or stems are, the lower the content of each compound is, proving that the content of coumarins, flavonoids, and monoterpene phenol may be closely related to the density coefficient of the secretory cavities.

Close

Fig. 3

The accumulation results of compounds in leaves (A), stems (B), flowers (C) and fruits (D) with different specifications from P. corylifolia.

Export to PPT

3.3 Dynamic variations of the tested compounds in P. Corylifolia from flowering to fruiting

Flowers of P. corylifolia were collected at different times, and the changes of their appearance and moisture content are illustrated in Fig. 4A and B. Also, the growth process was roughly divided into three periods according to the content results, as shown in Fig. 4C. Around t1–t2 was considered as the initial flowering period. During this period, mostly buds unopened. And, flavonoids and Pso were not detected while the content of PO, IPO P, IP, and Bak increased slowly. Around t3, most flowers were in a full-bloom stage. The content of PO, IPO, and Bak increased rapidly, the content of P and IP decreased gradually, while flavonoids and Pso were still not detected during this period. Around t4 was deemed as the final flowering period including a few blooming flowers, and many flowers beginning to bear fruit, flavonoids (Neo, Iso, and Bav) were detected, but not Pso. The above results were similar to that in flowers of different specifications.

Close

Fig. 4

The dynamic variations on appearance and the content of the detected compounds in fruits and flowers during P. corylifolia development. (A) The appearance of fruits and flowers at the different harvest times. (B) The change of moisture during fruits and flowers' growth. (C) The dynamic variations on the content of the detected compounds during flowers' growth. (D) The dynamic variations on the content of the detected compounds during fruits' growth. (E) The dynamic variations on the content of free and conjugated P and IP during fruits' growth.

Export to PPT

Fruits of P. corylifolia were collected at different times, and the changes of their appearance and moisture content are shown in Fig. 4A and B. From t1 to t5, fruit volume gradually expanded while the moisture content decreased from 72.1%±2.73% to 67.6%±3.79%. The volume gradually shrank from t5 to t6, and the moisture content rapidly decreased to 42.7%±3.92% as the fruit ripened. The dynamic variations of the tested compounds in fruits during P. corylifolia development are presented in Fig. 4D.

According to the dynamic variation rules in the content of flavonoids (Neo, Iso, and Bav) and monoterpene phenol (Bak), the growth and development stage of fruits could be divided into three periods, namely the growth, the stagnation, and the maturation period. t1–t4 was regarded as the growth period, during which the content of each compound gradually increased and reached summit at t4. t4–t5 was deemed as the stagnation period, during which the content of each compound decreased rapidly. t5–t6 was taken as the maturation period, during which the content of compounds was relatively stable as the fruits basically stopped growing. It was worth mentioning that coumarins presented two characteristics. That is, the content of (I)PO increased from t1 to t4, gradually decreased from t4 to t5, and slowly increased from t5 to t6, while (I)P and Pso increased from t1 to t2 and decreased from t2 to t6.

A preliminary study showed that (I)PO could be transformed into (I)P under the action of β-glucosidase (Yang et al., 2018). Given this, the total content of conjugated (I)P (Eq. (1)) and free (I)P (Eq. (2)) respectively defined as the total content of (I)PO and (I)P was investigated to determine the dynamic variations of (I)PO and (I)P during fruits growth. It can be deduced from Fig. 4E that the variation in the total content of conjugated and free (I)P was consistent with that of flavonoids and monoterpene phenol. However, the different proportions of conjugated and free forms of (I)P were observed during the different growth periods, suggesting that the dynamic transformation between (I)PO and (I)P exists throughout fruits' growth. From t1 to t2, the content of conjugated and free (I)P were both increased, speculating that the synthesis of (I)P is dominant and accompanied by synthesized (I)PO under the influence of glycosyltransferases. From t2 to t4, according to the increase in the content of (I)PO and the decrease in the content of (I)P, it is inferred that the consumption rate of (I)P is greater than that of synthesis rate, that is, a large amount of (I)P is synthesized into (I)PO by glycosyltransferases. Moreover, from t5 to t6, the total content of conjugated and free (I)P remained unchanged, despited that the slight increase and decrease observed for (I)PO and (I)P, respectively. This observation implies that the synthesis of (I)P is basically stopped and a few (I)P is still devoted to the synthesis of (I)PO at this stage.

3.4 Clarification of the correlation between storage time and the content of the focused compounds during the storage of FP

Forty batches FP were sampled from the different medicinal materials markets. The appearance of samples is shown in Fig. 5A. Among them, B12 was taupe, which was significantly different from the other samples. Based on the established UHPLC-PDA method, nine tested compounds were investigated in 40 batches FP. As shown in Fig. 5B, the content of coumarins, flavonoids, and monoterpene phenol in B12 is noticeably distinct from that of other samples, with PO, IPO, P, IP, Neo, Iso, Pso, Bav, and Bak quantified to be 0.1901, 0.1147, 14.80, 15.38, 0.2758, 0.5241, 0.09823, 0.3015 and 16.78 mg/g, respectively. It can be seen that B12 is an abnormal sample with lower content of key compounds except P and IP. Except for B12, the range of compounds' content in remaining 39 batches FP were as follows: PO 6.281–34.17 mg/g (RSD, 34.0%), IPO 4.039–29.35 mg/g (RSD, 37.6%), P 0.7861–7.738 mg/g (RSD, 45.1%), IP 0.5312–7.506 mg/g (RSD, 49.5%), Neo 2.359–5.436 mg/g (RSD, 18.1%), Iso 1.306–6.791 mg/g (RSD, 31.4%), Pso 1.817–3.843 mg/g (RSD, 17.5%), Bav 4.401–10.36 mg/g (RSD, 18.1%), and Bak 60.94–129.3 mg/g (RSD, 16.5%). Among the quantified compounds, the content of PO, IPO, P, and IP fluctuated obviously in samples.

Close

Fig. 5

The appearance of 40 batches FP (A) and the heat map of the tested compounds' content in 40 batches FP (B).

Export to PPT

The Chinese Pharmacopoeia (ChP) stipulates that the total content of P and IP in FP must not be less than 0.70% (National Commission of Chinese Pharmacopoeia, 2020). Moreover, (I)PO could be converted into (I)P under certain conditions (Yang et al., 2018). The detailed content of (I)PO as conjugated (I)P and (I)P as free form for 40 batches FP are shown in Fig. 6A. Interestingly, in B12, the total content of free (I)P complied with the ChP criterion, and the sum of conjugated and free (I)P was close to the average value of 40 batches FP. However, the total content of conjugated (I)P in B12 was extremely lower. Combined with the market survey, it is possibly concluded that almost all (I)PO are converted into (I)P under the action of β-glucosidase since the storage time of B12 is more than ten years. The total content of conjugated and free (I)P in 39 batches FP except for B12 was 5.245–31.95 mg/g (RSD, 35.4%) and 1.323–15.12 mg/g (RSD, 47.0%), respectively, with large fluctuations. Nevertheless, the sum of conjugated and free (I)P between 20.00 and 35.05 mg/g (RSD, 14.4%) is distributed in a narrow range.

Close

Fig. 6

The changes on the total content of the detected compounds in samples from different origins. (A) Histogram of the content of conjugated and free P and IP in 40 batches FP. (B) PCA analysis of 40 batches FP based on the content of conjugated and free P and IP. (C) Scatter plot of 39 batches FP based on the content of conjugated and free P and IP. (D) Histogram of the content of the tested compounds for FP obtained from the accelerated stability test.

Export to PPT

Correlation analysis is a statistical method used to evaluate the associations between variables (Kim et al., 2013). To determine the detailed relationship between the content of (I)PO and (I)P, the correlation between the total content of free (I)P and that of conjugated (I)P, and the correlation between the total content of free (I)P and that of conjugated and free (I)P were analyzed, respectively. As shown in Fig. 6B, PCA analysis showed that B12 was significantly different from other batches, which was consistent with the above results. Also, as shown in Fig. 6C, by Pearson's correlation analysis, a strong correlation existed between the total content of free (I)P and that of conjugated (I)P, also between the total content of free (I)P and that of free and conjugated (I)P in 39 batches samples by excluding B12. The total content of free (I)P was negatively correlated with that of conjugated (I)P (r = –0.896, p < 0.001) and that of free and conjugated (I)P (r = –0.629, p < 0.001). It means that the total content of conjugated (I)P and that of free and conjugated (I)P could be reduced as the total content of free (I)P increased. Therefore, it is conjectured that with the extension of storage time for FP, (I)PO will be constantly transformed into (I)P under the action of β-glucosidase, resulting in a decrease in the total content of (I)PO and an increase in the total content of (I)P, while (I)P will be slowly degraded, resulting in a decrease in the total content of (I)PO and (I)P.

This phenomenon was witnessed in the accelerated stability test designed to simulate changes of compounds' content during the storage of FP, as displayed in Fig. 6D. The determination result of content for FP treated for 6 months was compared with that of untreated FP (0 month). The content of P and IP increased by 513.9% and 648.0%, respectively; the content of PO and IPO decreased by 24.9% and 25.7%, respectively. Also, the content of Neo, Iso, Pso, Bav, and Bak decreased by 43.8%, 53.5%, 6.2%, 35.7%, and 30.6%, respectively. It demonstrates that compounds' content is closely related to storage time, that is, with prolonged storage time, (I)PO could be converted into (I)P driven by β-glucosidase, and other compounds gradually degrade to a certain extent. However, we did not observe a decrease in the total content of free and conjugated (I)P due to the short storage time, but it is speculated that further extension of the storage time would aggravate (I)P degradation.

3.5 Compounds' transformation actuated by enzyme and temperature in FP exposed to the different processing temperatures

The complex changes of compounds after processing of Chinese materia medica are important reasons for its efficacy and toxicity. The processing temperature is one of the important parameters that trigger changes of compounds. In this study, we focused on the effect of different processing temperatures on compounds' transformation driven by enzyme and temperature. From our previous study, we observed that (I)PO extracted from FP with 50% methanol was easily converted into (I)P under the action of β-glucosidase. However, the extraction efficiency of (I)PO and (I)P in roasted FP with inactived β-glucosidase was insignificantly different between 50% methanol and methanol (Yang et al., 2018). In order to unravel the effect of processing temperature on enzyme activity and the stability of compounds in FP, the samples were prepared by placing FP at different temperatures (40–210 °C) for 2 h, and afterward extracted with 50% methanol and methanol, respectively.

The quantitative results of the tested samples prepared by methanol as extraction solvent represented the reference value of compounds. And the content of free and conjugated (I)P for FP extracted with 50% methanol was compared with the reference value to study enzyme activity in the processed FP, as shown in Fig. 7A. When the temperature was lower than 90 °C, the total content of free and conjugated (I)P remained basically unchanged for FP extracted with methanol. However, for FP extracted with 50% methanol, the content of free (I)P decreased gradually while the conjugated (I)P content increased. It suggests that only enzyme-driven transformation of compounds occurred within this temperature range. From 90 to 120 °C, where the total content of free and conjugated (I)P for FP extracted with methanol decreased slightly, but the content of conjugated (I)P increased rapidly and the content of free (I)P decreased obviously in samples extracted with 50% methanol along with the processing temperature elevating. In this temperature range, β-glucosidase activity decreases fleetly, the focused compounds begin to degrade, and compounds' transformation is mainly driven by β-glucosidase. When the temperature was above 120 °C, the total content of free and conjugated (I)P decreased rapidly for FP extracted with methanol, the content of corresponding compounds was insignificantly different between FP extracted with 50% methanol and methanol. It demonstrates that the compounds are rapidly degraded, and β-glucosidase completely inactive when the temperature exceeded 120 °C. And compounds' transformation is only driven by high temperature.

Close

Fig. 7

The effect of temperature on the stability of compounds and enzyme activity in FP. (A) Line chart of the content of the focused compounds extracted with 50% methanol and methanol in FP exposed to the different processing temperatures; (B) Line chart of the content of conjugated and free P extracted with methanol in FP exposed to the different processing temperatures; (C) Line chart of the content of conjugated and free IP extracted with methanol in FP exposed to the different processing temperatures; (D) Line chart of the content of flavonoids and bakuchiol extracted with methanol in FP exposed to the different processing temperatures.

Export to PPT

In addition, the effect of the different processing temperatures on the stability of compounds in FP was investigated in detail based on the quantitative results of samples extracted with methanol. As shown in Fig. 7B and C, the content of four compounds, conjugated and free (I)P, remained basically stable at the temperature below 90 °C. Ranging from 90 to 120 °C, the content of free P was basically stable while the content of conjugated P decreased slightly. And, the decrease in the content of conjugated IP was basically consistent with the increase in the content of free IP. Also, from 120 to 180 °C, the content of free P and IP increased significantly, and conjugated P and IP content decreased markedly, both the total content of free and conjugated P and that of free and conjugated IP decreased noticeably, which indicates that the decreased degree of conjugated P and IP is greater than the generation degree of free P and IP, respectively. It is speculated that partial conjugated P and IP are degraded in this temperature range. Between 180 and 200 °C, the decreased degree of conjugated P was approximately equal to the production degree of free P, the conjugated IP was completely degraded while the content of free IP was basically unchanged. And, the conjugated P was not detected at 210 °C. This shows that free P and IP are relatively stable when the temperature is below 210 °C, and the reduction of the total content of four compounds is mainly due to the degradation of conjugated P and IP. The sensitivity to temperature for other compounds in FP is shown in Fig. 7D, from which Iso, Bav, Bak, Neo and Pso began to degrade at 120, 120, 130, 150, and 180 °C, respectively.