Simultaneous detection of multiple mycotoxins in Radix Dipsaci and estimation of exposure risk for consumers

Abstract

Like many traditional Chinese herbal medicines, preparations from Radix Dipsaci are at risk of contamination by harmful mycotoxins; however, there have been no reports of actual contamination. In this study, we developed an analytical method to simultaneously detect eight mycotoxins in Radix Dipsaci and estimate the exposure risk for consumers. We have developed an analytical method utilizing ultra-high performance liquid chromatography and tandem mass spectrometry to accurately determine the levels of AFB1, AFB2, AFG1, AFG2, OTA, ZEN, T-2 and ST mycotoxins in 45 batches of Radix Dipsaci sourced from major medicinal herb markets across five regions in China. We also analyzed migration of mycotoxins from the raw herbs into water decoction. Based on these results and data on human consumption of the herbal medicine, we estimated risk of exposure and acceptable exposure limits to mycotoxins in the Radix Dipsaci using the “margin of exposure (MOE)” method. Of the 45 batches of Radix Dipsaci, 48.89% contained at least one of the eight mycotoxins, 24.44% contained one, 17.78% contained two and 6.67% contained three. The most frequent mycotoxins were aflatoxin B1, present in 35.56% of batches (at 0.25–34.84 μg/kg); aflatoxin G1, 15.56% (1.99–44.05 μg/kg); and ochratoxin A, 22.22% (16.11–143.38 μg/kg). These three mycotoxins transferred from the raw herb into water decoction at respective rates of 20.20%, 29.14%, and 24.80%. The 95th percentile values of the MOE risk factors for health effects of AFB1 were below 10,000 at high doses but above 10,000 at low doses of Radix Dipsaci long-term treatment. With the reduction in duration of exposure years, the MOE values of AFB1 and AFG1 gradually reverted to within the acceptable range. The mean, 50th, and 95th percentile values of the MOE risk factors for health effects of OTA exceeded 10,000 regardless of whether consumers received a low or high dose of Radix Dipsaci treatment for durations ranging from 1 to lifetime. Based on this exposure and a typical human diet, we have estimated the respective 20-year exposure limits for Radix Dipsaci to be 5.821 μg/kg, 4.035 μg/kg, and 56.073 μg/kg for the three mycotoxins under consideration. Contamination with multiple mycotoxins is frequently observed in Radix Dipsaci, and the three most prevalent contaminants have been found to leach into water decoctions, thereby posing a potential health hazard for individuals consuming this herbal preparation. This work highlights the need to monitor herbal medicines for mycotoxin contamination in order to protect consumers.

Introduction

Ensuring the quality and safety of Chinese herbal medicine is crucial for addressing public health demands and advancing traditional medicines as alternatives to synthetic drugs, thereby mitigating the potential risks and exorbitant costs associated with them^1,2. A major challenge in this regard is preventing contamination of herbal medicines with mycotoxins, small metabolites naturally secreted by fungi that are toxic to animals and humans because they induce genetic mutations leading to cancer or birth defects, or they harm the functioning of the liver, kidneys, immune system or brain^3–6. Mycotoxins contaminate approximately 25% of global agricultural products each year^7,8, and such contamination is a growing problem in Chinese herbal medicines^9–11.

Contamination of Chinese herbal medicines with mycotoxins can result in acute or chronic mycotoxicosis, thereby elevating the susceptibility to a range of health complications^12,13. Unfortunately most mycotoxins are quite stable, making it difficult to reduce their levels or eliminate them entirely by heating or other processing^14–16. Therefore, it is important to determine the frequency and levels of mycotoxin contamination in herbal medicines in order to define exposure limits so that consumers can continue to receive beneficial treatment while managing the risk of adverse effects.

The dry root of Radix Dipsaci Wall is a Chinese herbal medicine for which exposure limits have not been defined because no data exist on its contamination with mycotoxins, even though it is highly susceptible to contamination with soil fungi during its growth^17,18 and uterine bleeding Radix Dipsaci is commonly employed in the treatment of hepatic or renal disorders, lumbar muscle or knee weakness, rheumatic arthralgia, fall-related injuries, tendon damage, and uterine hemorrhage due to its potent therapeutic properties^19–21. However, the complex matrix of traditional Chinese medicine increases the difficulty of sample detection, and the pre-treatment method of the sample determines the quality of the detection result. At present, immunoaffinity column is the main method used for pretreatment, but it is expensive and can only detect Radix Dipsaci single toxin, which cannot meet the screening of mycotoxins in A. Therefore, it is of great significance to optimize sample pretreatment methods and develop detection methods for multiple mycotoxins in Radix Dipsaci.

We have developed an analytical technique to simultaneously detect eight mycotoxins in Radix Dipsaci, which we subsequently utilized for the analysis of multiple batches of this herb sourced from various markets across China. Our results indicate a high prevalence of single and multiple contamination, based on which we estimated exposure limits for consumers. Our findings and our analytical method may guide efforts to ensure the safety and quality of Radix Dipsaci preparations.

Materials and methods

Reagents and chemicals

The following mycotoxin standards of purity ≥ 99% were purchased from Pribolab (Qingdao, Shandong, China) and stored at – 20 °C in the dark: aflatoxin B1 (catalog no. MSS1003), aflatoxin B2 (MSS1014), aflatoxin G1 (MSS1005), aflatoxin G2 (MSS1006), ochratoxin A (MSS1020), sterigmatocystin (MSS1022), zearalenone (MSS1024) and T-2 Toxin (MSS1023).

Reagents were chromatography-grade and obtained from various sources as follows: NaCl, catalog no. 20200919, Shangon Biotech (Shanghai, China); MgSO₄, 22030, Tianjin Zhiyuan Chemical Reagent (Tianjin, China); octadecylsilyl, 17606, Tianjin Kemio (Tianjin, China); N-(N-propyl) ethylenediamine, P20-00557, Tianjin Kemio; Graphitized carbon black (GCB), C11495332, Shanghai Maclin Biochemical Technology (Shanghai, China); methanol, 200343, Fisher, USA; acetonitrile, 202888, Fisher; and formic acid, 166511, Fisher.

Preparation of standard solutions

Stock solutions of all mycotoxins were prepared by dissolving the powders in 50% (v/v) acetonitrile in water to final concentrations of 1.25 mg/mL in the case of zearalenone or 1 mg/mL in the case of the others. Aliquots of 0.10–10.0 mL of these stock solutions were used to prepare working solutions of 12.5 µg/mL in the case of zearalenone or 10 µg/mL in the case of the others. Aliquots of 1.0–25 mL from these working solutions were mixed together to prepare a standard cocktail containing zearalenone at 500 ng/mL and all other mycotoxins at 400 ng/mL.

Preparation of Radix Dipsaci samples

We sampled Radix Dipsac being sold at five major herbal medicine markets in five major cities spanning five Chinese provinces or autonomous region. At each market, we randomly selected 7–10 outlets to purchase samples. At least 5 kg was purchased for each batch. Finally, we sampled 45 batches of herbs.

Herb powder (2.0 g) was mixed with 20.0 mL of acetonitrile: water: formic acid (80:19:1 and shaken for 1 h at 180 rpm. MgSO₄ (2 g) and NaCl (1 g) were added to the mixture, which was swirled for 1 min, then centrifuged at 5000 rpm for 5 min. Supernatant (5 mL) was added to a mixture of the following pure powders, which served as “cleaning agents” to remove lipids, pigments, organic acids and sugars (Zhai et al., 2023): Octadecyl trichlorosilane (C18) (0.3 g), primary secondary amine (PSA, 0.1 g), MgSO₄ (0.3 g) and graphitized carbon black (GCB, 0.01 g). The mixture was swirled for 1 min, then an aliquot of supernatant (3 mL) was transferred to a new tube and evaporated nearly to dryness at 40 ℃ under nitrogen. The residue was reconstituted with 1.0 mL 50% acetonitrile in water, filtered through a membrane with pores of diameter 0.22 µm, and used in the analytical procedure described below.

Mycotoxin determination using chromatography-mass spectrometry

We detected all eight mycotoxins using ultra-high performance liquid chromatography followed by tandem mass UPLC-MS/MS. Chromatographic parameters were as follows: column type, Waters HSS T3 (100 × 2.1 mm, 1.8 µm); column temperature, 40 ℃; mobile phases, 0.1% formic acid in water (solvent A) and acetonitrile (solvent B); flow rate, 0.3 mL/min; and sample volume, 5 µL. The following elution gradient was performed: 0–3 min, 85% A; 3–5 min, 65% A; 5–7 min, 55% A; 7–7.5 min, 50% A.

UPLC-MS/MS was performed on a QTRAP 5500 quadrupole system (SCIEX) using electrospray ionization in positive ion mode and multiple reaction monitoring under the following parameters: capillary voltage, 3.5 kV; ionization temperature, 150 ℃; back-blow flow rate in the cone hole, 30 L/h; dissolvent temperature, 500 ℃; and solvent gas flow rate, 600 L/h. The mass spectrometry parameters of each mycotoxin reference solution are shown in Supplementary Tables 1. Elution profiles of mycotoxins from Radix Dipsaci as showed in (Supplementary figure 1).

Potential matrix effects due to ion suppression or enhancement during UPLC-MS/MS were assessed by generating standard curves from the cocktail in section 2.2 after dilution with 0% or 50% acetonitrile in water. The ratio of slopes for the standard curves after dilution with blank or acetonitrile was calculated as the “suppression/enhancement effect”, and a value of 0.8–1.2 was taken to indicate negligible matrix effects²².

Validation of the analytical method for mycotoxin detection

Before sample testing, this study conducted methodological verification in accordance with the requirements of the Chinese Pharmacopoeia 2020 Edition 9101 Guiding Principles for the Validation of Analytical Methods for sample determination.

Specificity examination

Prepare intermittent test solution and mycotoxin control solution of appropriate concentration in accordance with item 2.3. Sample was taken according to item 2.4, and chromatogram was recorded.

Linearity, detection limit and quantification limit

In this study, the intermittent sample extract was chosen to dilute the mixed standard substance into a series of concentrations, and the standard curve was established by testing under specific conditions (2.2). The limit of detection (LOD) was set at S/N ≥ 3, and the limit of quantification (LOQ) at S/N ≥ 10.

Repeatability, stability and precision tests

Weigh 6 blank intermittent samples 2.0g, add 0.4 mL of mixed control solution of 8 mycotoxins, prepare the test solution in parallel according to the method in item 2.3, test according to the condition in item 2.4, calculate the RSD value, and check the repeatability.

A sample of 2.0 blank intermittent samples was weighed, 0.4 mL of mixed control solution of 8 mycotoxins was added, and the test solutions were prepared in parallel according to the method in item 2.3. The samples were injected at 0, 2, 4, 6, 12, 24 h, respectively, and the recovery RSD value of 8 mycotoxins was calculated to check the stability.

The sample solution under item 2.3 was taken for 6 consecutive samples under item 2.4, and the RSD value of mycotoxin peak area was calculated to check the precision of the instrument.

Sample recovery experiment

The recovery rate was investigated using the blank spiked recovery method. The test solution of blank Radix Dipsaci was prepared according to the method described above. The solution was combined with “low” (the concentration of ZEN is 100ng/ml, and the concentration of other mycotoxins is 80ng/ml), “medium” (the concentration of ZEN is 200ng/ml, and the concentration of other mycotoxins is 80ng/ml), and “high” (ZEN concentration of 300 ng/ml, and the rest of the fungus toxin concentration at 240 ng/ml) in a mixture containing mycotoxin standard products at different concentrations. Each concentration was performed in triplicate, resulting in a total of nine samples. Mycotoxin recovery and its relative standard deviation were determined.

Determination of mycotoxin transfer from Radix Dipsaci into water decoction

The decoctions were prepared by addition of 15 g Radix Dipsaci slices, which were the clinically daily dosage of Radix Dipsaci recommended by Chinese Pharmacopoeia Commission (2020). These Radix Dipsaci slices was suspended in 300 mL of water in an automated decoction apparatus and allowed to stand for 30 min, (25 ℃) then subjected to 2 h decoction. The mixture was filtered through multiple layers of gauze and concentrated to 50 mL. The content of mycotoxins in Radix Dipsaci slices and their decoctions were evaluated as previous description²³.

The rate of transfer T of each of the three mycotoxins from raw Radix Dipsaci into water decoction was determined using the formula²⁴ as Eq. (1):

$$T=\frac{\mathit{Md}}{Md+Ms}\times 100,$$ 1

where Md is the ng of the mycotoxin in the decoction and Ms is the ng of mycotoxin in herb.

Risk of consumer exposure to mycotoxins in Radix Dipsaci and acceptable exposure limits

We estimated risk of exposure to aflatoxin A1, aflatoxin G1 and ochratoxin A based on the estimated daily intake (EDI) per kg of body weight²⁵, as Eq. (2):

$$EDI=\frac{C\times IR\times T}{\mathit{BW}},$$ 2

where C refers to the average content of mycotoxins in Radix Dipsaci, BW is the average weight of an adult (63 kg) and IR to daily intake of Radix Dipsaci (9–15 g/day) recommended by Chinese Pharmacopoeia Commission (2020); and based on the estimated lifetime (70 years) per kg of body weight²³.

Based on these estimated daily intakes, we used a “margin of exposure” (MOE) method^24,26 to evaluate the mean, 50th and 95th percentile values of the MOE risk factors for health effects of AFB1, AFG1 or OTA when the consumers received a low or high dose of Radix Dipsaci treatment for 1, 5, 10, 20, 40 years or lifetime. We calculated the MOE as Eq. (3):

$$MOE=\frac{{\mathit{BMDL}}_{10}}{\mathit{EDI}}$$ 3

where BMDL₁₀ refers to the benchmark dose for a 10% response. The BMDL₁₀ of 0.40 μg per kg of body weight for AFB1 and AFG1^24,27, 4.73 μg per kg of body weight for OTA²³, we estimated the MOE risk factors for health effects of the three mycotoxins. Higher MOE values indicate lower potential risk, as exposure levels fall below those judged to be harmful based on animal studies. The MOE’s threshold (TMOE) of 10,000 was applied for health effects of AFB1, AFG1 and OTA²⁸. We calculated risk quotients for the three mycotoxins since consumers may consume them not only in Radix Dipsaci but also in others foods and herbal medicines. The risk threshold was calculated using Eq. (4)

$$TR=\frac{BMD{L}_{10}}{\mathit{TMOE}},$$ 4

The risk quotient was calculated using Eq. (5)

$$RQ=\frac{\mathit{EDI}}{\mathit{TR}}$$ 5

Where TR is risk thresholds for mycotoxins with different effects, TMOE the MOE’s threshold for each effect of mycotoxins; and RQ the risk quotient of mycotoxins in Radix Dipsaci.

The distribution of the three mycotoxins in Radix Dipsaci showed a log-normal distribution. The model was developed in Monte Carlo software (version 4.0.2) using 200 simulations × 10,000 iterations to simulate the risk of AFB1, AFG1 and OTA in Radix Dipsaci. The mean, 50th and 95th percentile values of EDI and MOE were extracted from Monte Carlo simulations, each involving 10⁴ iterations. The 50th and 95th percentile values from 10⁴ simulations defined the corresponding 95% confidence interval (CI) for the mean value of EDI and MOE.

Based on this modeling of exposure risk, we estimated limits of consumption of Radix Dipsaci to ensure that intake of aflaxatoxin B1, aflatoxin G1 and ochratoxin A would not lead to unacceptable risk of neoplastic effects.

Results

Optimization of mycotoxin extraction from Radix Dipsaci

We optimized several aspects of mycotoxin extraction to ensure their sensitive detection, given that they are normally present only in trace amounts. Among the three sample preparation methods, the oscillation method demonstrated superior recovery rates (90–114%) for all eight fungal toxin extractions in comparison to the ultrasound or homogenization methods. (Fig. 1A), while extracting the mycotoxins with acetonitrile: water: formic acid (70:29:1) led to better recoveries (70–120%) than the same co-solvents in a different ratio or with methanol instead of acetonitrile (Fig. 1B). An extraction time of 1 h gave better results than 30 min and similar results as 90 min (Fig. 1C). Tests with different amounts of the same set of four cleaning agents showed the best formulation to be 0.3 g carbon 18 + 0.1 g primary secondary amine + 0.3 g magnesium sulfate + 0.01 g graphitized carbon black (Fig. 1D).

Fig. 1.

Optimization of conditions for extracting mycotoxins from Radix Dipsaci: (A) extraction method, (B) extraction solvent, (C) extraction time and (D) cleaning agent. Recovery (R) between 70% and 125% represent relatively reliable extraction methods. A acetonitrile, AFB1 aflatoxin B1, AFB2 aflatoxin B2, AFG1 aflatoxin G1, AFG2 aflatoxin G2, C18 Octadecyltrichlorosilane, FA formic acid, GCB graphitized carbon black, OTA ochratoxin A, PSA primary secondary amine, R recovery, ST Sterigmatocystin, W Water, ZEN zearalenone.

Validation of the simultaneous detection method of multiple mycotoxins in Radix Dipsaci

In our method, the mycotoxins showed well-separated retention times ranging from 3.9 to 5.1 min (Fig. 2A), and no peaks were observed in the blank sample (data not shown). The linear range of the method was good with R² > 0.9990, the detection limit was 0.064–0.151 μg/kg, and the limit of quantitation was 0.221–0.501 μg/kg (Fig. 2B). During quantitation, we took into account matrix effects on suppression and enhancement of ions during analysis of each of the mycotoxins, based on analysis of the suppression/enhancement value (Fig. 2B). The repeatability RSD was 0.365–5.939%; The RSD value of 24 h stability was 0.936–3.762% and less than 3%. In the precision test, the RSD of precision was 0.739–2.936%, were lower than 15% of the requirements of Chinese Pharmacopoeia (2020 edition), indicating that the method had good repeatability, stability and precision (Fig. 2C). And sample recoveries ranged from 74.765 to 114.543%, with relative standard deviations between 3.283 and 9.495%, which were lower than the 15% requirement stipulated in the Chinese Pharmacopoeia (2020 edition) (Fig. 2D).

Fig. 2.

Validation of the analytical method (A) Specific chromatogram of eight mycotoxins in Radix Dipsaci. cps, counts per second. (B) Calibration and validation of the analytical method for simultaneous detection of eight mycotoxins in Radix Dipsaci. *Calibration curves were adjusted for matrix effects (see Methods). LOD limit of detection, LOQ limit of quantitation, SSE suppression/enhancement values. (C) Repeatability, stability and precision for determination of eight mycotoxins in Radix Dipsaci. (D) Recovery of eight kinds of mycotoxins from Radix Dipsaci. RSD, relative standard deviation. Mycotoxin abbreviations are defined in Fig. 1.

Determination of mycotoxins in market samples of Radix Dipsaci

Of 45 batches of Radix Dipsaci sampled from major herbal medicine markets in five parts of China, 22 samples (48.9%) contained at least one of the eight mycotoxins, 11 samples (24.4%) were found to be contaminated with one mycotoxin, 8 samples (17.8%) with two and 3 samples (6.7%) with three among the eight mycotoxins that we analyzed (Fig. 3A–C). The predominant contaminants identified in the samples were aflatoxin B1 (16 samples), ochratoxin A (10 samples), aflatoxin G1 (7 samples), aflatoxin G2 (1 sample), zearalenone (1 sample), and sterigmatocystin (1 sample). The contamination levels varied from 0.25 to 143.38 µg/kg. We did not detect aflatoxin B2 or T-2 in any batches (Fig. 3D–F). No mycotoxins were detected in the 22 (48.9%) samples (Supplementary figure 2).

Fig. 3.

Determination of eight mycotoxins in 45 batches of Radix Dipsaci from major herbal medicine markets in five parts of China. (A) Heat map of mycotoxin content. Red shading indicates relative levels of contamination, while turquoise indicates no detectable mycotoxin. (B) Numbers of samples contaminated with each mycotoxin. (C) Percentages of samples contaminated with one, two or three mycotoxins. (D) Percentages of samples contaminated with different mycotoxins, alone or in combination. (E) Percentages of samples containing the indicated mycotoxins. (F) Levels of the indicated mycotoxins in contaminated samples. Mycotoxin abbreviations are defined in Fig. 1. ND, not detected.

Transfer of mycotoxins from raw Radix Dipsaci into water decoction

Considering that the patient mainly took the water decoction of Radix Dipsaci, the water decoction was prepared according to the dosage and procedure proposed by the Chinese Pharmacopoeia Commission. We calculated the rates of transfer of the three most frequent mycotoxins contaminents from raw herb into water decoction (Figs. 4, 5, 6, 7). We found that about 20.2% of the aflatoxin B1 was transferred from the herbs to their decoction (Fig. 4A–C). The transfer rate of aflatoxin B1 positively corrected with the content of aflatoxin B1 in raw Radix Dipsaci (r > 0.3, P ˂ 0.05) (Fig. 4D,E). About 29.1% of the aflatoxin G1 and 24.8% of the ochratoxin A was transferred from the herbs to their decoction (Figs. 5A–C and 6A–C). The transfer rate of these two mycotoxins did not correct with the content of aflatoxin G1 in raw Radix Dipsaci (r² ˂ 0.09, P > 0.05) (Figs. 5D and 6D). These results indicated that aflatoxin B1, aflatoxin G1 and could be transferred from the raw herbs into their decoction, which might pose a health risk to consumers.

Fig. 4.

Transfer of aflatoxin B1 (AFB1) from raw Radix Dipsaci into water decoction. (A-B) The changes in the content of AFB1 between the herb slice (A) and their water decoction (B) in AFB1-contaminated Radix Dipsaci. (C) The transfer was measured in terms of change in absolute content and percentage of total content that transferred. (D) Correlation analysis between the transfer rate of AFB1 from herb to water decoction and the content of AFB1 in the raw Radix Dipsaci.

Fig. 5.

Transfer of aflatoxin G1 (AFG1) from raw Radix Dipsaci into water decoction. (A-B) The changes in the content of AFG1 between the herb slice (A) and their water decoction (B) in AFG1-contaminated Radix Dipsaci. (C) The transfer was measured in terms of change in absolute content and percentage of total content that transferred. (D) Correlation analysis between the transfer rate of AFG1 from herb into water decoction and the content of AFG1 in the raw Radix Dipsaci.

Fig. 6.

Transfer of ochratoxin A (OTA) from raw Radix Dipsaci into water decoction. (A-B) The changes in the content of OTA between the herb slice (A) and their water decoction (B) in OTA-contaminated Radix Dipsaci. (C) The transfer was measured in terms of change in absolute content and percentage of total content that transferred. (D) Correlation analysis between the transfer rate of OTA from herb into water decoction and the content of OTA in the raw Radix Dipsaci.

Fig. 7.

Distribution of chronic exposure to AFB1, AFG1 or OTA in Radix Dipsaci (one of 10000 simulations). P50, 50th percentile; P95, 95th percentile; bw, body weight.

Risk of consumer exposure to mycotoxins in Radix Dipsaci and acceptable exposure limits

The chronic exposure to AFB1, AFG1 or OTA in low or high dose of Radix Dipsaci was shown in Fig. 7. The estimated mean daily intake of AFB1 obtained from the exposure assessment was 0.0545 ng/kg bw (50th–95th: 0.0542–0.0677 ng/kg bw) in low dose of Radix Dipsaci (Fig. 7A) and 0.0917 ng/kg bw (50th–95th: 0.0908–0.1243 ng/kg bw) in high dose of Radix Dipsaci (Fig. 7B). The estimated mean daily intake of AFG1 obtained from the exposure assessment was 0.0793 ng/kg bw (50th–95th: 0.0782–0.1039 ng/kg bw) in low dose of Radix Dipsaci (Fig. 7C) and 0.1325 ng/kg bw (50th–95th: 0.1300–0.1836 ng/kg bw) in high dose of Radix Dipsaci (Fig. 7D). The estimated mean daily intake of OTA obtained from the exposure assessment was 0.1964 ng/kg bw (50th–95th: 0.1932–0.2658 ng/kg bw) in low dose of Radix Dipsaci (Fig. 7C) and 0.3283 ng/kg bw (50th–95th: 0.3222–0.4493 ng/kg bw) in high dose of Radix Dipsaci (Fig. 7D). The mean, 50th and 95th percentile values of chronic exposure to AFB1, AFG1 or OTA for 1, 5, 10, 20, 50 years or lifetime was also evaluated in low or high dose of Radix Dipsaci (Table 1).

Table 1.

Estimated daily intake (EDI) in humans due to exposure to aflatoxin B1 from Radix Dipsaci over different periods.

Mycotoxins	Duration of consumption,yr	Estimated mean and 95%Cl of quantiles for AFB1/AFG1/OTA exposure levels in Radix Dipsaci
		Low dose (×10−2)			High dose (×10−2)
		50th percentile	Mean	95th percentile	50th percentile	Mean	95th percentile
AFB1	1	0.078 (0.076–0.078)	0.078 (0.076–0.080)	0.097 (0.094–0.099)	0.128 (0.125–0.132)	0.131 (0. 130–0.141)	0.184 (0. 177–0.191)
	5	0.388 (0.381–0.393)	0.389 (0.382–0.403)	483 (0. 473–0.494)	0. 642 (0.630–0.653)	0.649 (0.635–0.681)	0. 844 (0.817–0.874)
	10	0.775 (0.754–0.788)	0.779 (0.760–0.803)	0 966 (0. 941–0.985)	1.289 (1.257–1 314)	1.310 (1. 296–1.397)	1.783 (1.717–1. 847)
	20	1.550 (1. 530–1.573)	1. 557 (1.533–1 609)	1 932 (1.891–1.978)	2.579 (2.512–2.620)	2.620 (2.593–2.781)	3.565 (3.441–3.664)
	40	3.100（3.054–3.146）	3.115（3.037–3.219）	3.864（3.732–3.954）	5.518 (5.032–5.228)	5.240 (5.180–5.558)	7.130 (6. 871–7.362)
	Lifetime	5.421（5.330–5.504）	5.451（5.316–5.619）	6.772（6.546–6.895）	9.018 (8.763–9.178)	9.171 (9.044–9.771)	12.434 (11. 999–12.976)
AFG1	1	0.011 (0.011–0.011)	0.011 (0.0111–0.012)	0.013 (0.013–0.013)	0.186 (0.182–0.190)	0.189 (0.186–0.202)	0.257 (0.244–0.268)
	5	0.056 (0.056–0.057)	0.056 (0.052–0.057)	0.066 (0.064–0.066)	0.929 (0.909–0.947)	0.945 (0.933–1.006)	1.286 (1. 238–1.330)
	10	0.112 (0.110–0.113)	0.112 (0. 109–0.114)	0.130 (0. 128–0.133)	1. 858 (1.808–1. 889)	1.890 (1.858–2.011)	2.563 (2 458–2.663)
	20	0.225 (0.222–0.227)	0.224 (0.120–0.242)	0.262 (0.257–0.265)	3.710 (3.644–3.786)	3.775 (3.740–4.052)	5.146 (4. 940–5.383)
	40	0.446 (0. 437–0 458)	0.455 (0. 449–0483)	0.619 (0.596–0.637)	7. 422 (7.261–7. 546)	7. 565 (7. 482–8.081)	10.310 (9.907–10.693)
	Lifetime	0.786 (0.777–0.793)	0.785 (0.771–0.798)	0.912 (0.900–0.928)	12.989 (12 685–13.288)	13.226 (13. 123–14.097)	17.955 (17.480–18 605)
OTA	1	0.274 (0.271–0.280)	0.278 (0.274–0.267)	0.363 (0.354–0.377)	0. 461 (0.453–0.475)	0. 469 (0. 467–0.503)	0.639 (0.618–0.662)
	5	1 .388 (1.354–1.405)	1. 400 (1.376–4.469)	1.829 (1.773–1.867)	2.306 (2.262–2.350)	2. 346 (2. 326–2.495)	3.197 (3.099–3.311)
	10	2.755 (2.709– 2807)	2.781 (2.738–2. 925)	3.625 (3. 509–3.738)	4.612 (4.513–4 .694)	4.691 (4. 635–5.024)	6.394 (6.117–6.659)
	20	5.505 (5.429–5.638)	5.570 (5.449–5.859)	7.283 (7. 049–7.492)	9.225 (9.050–9.414)	9.382 (9.297–10.050)	12.789 (12 204–13.159)
	40	11. 010 (10.826–11.234)	11. 152 (10.986–11. 699)	14.575 (14.123–14. 993)	18.449 (18.099–18.807)	18.765 (18. 587–19. 987)	25.577 (24. 623–26.343)
	Lifetime	19.284 (18.954–19.792)	19.504 (19. 228–20.467)	25.401 (24.740–26.034)	32.223 (31.593–32.907)	32.838 (32.502–34 940)	44.937 (42.881–46.072)

Based on these data and a BMDL₁₀ of 0.40 μg per kg of body weight for AFB1 and AFG1, 4.73 μg per kg of body weight for OTA, we estimated the MOE risk factors for health effects of the three mycotoxins. The 95th percentile values of the MOE risk factors for health effects of AFB1 was less than 10000 at low doses of Radix Dipsaci treatment (Fig. 8A) but more than 10000 at high doses of Radix Dipsaci treatment (Fig. 8B). The mean and 50th percentile values of the MOE risk factors for health effects of AFB1 and AFG1 exceed 10000 when consumers received either low or high doses of Radix Dipsaci treatment (Fig. 8A–D), suggesting the health risk of AFG1 and AFB1 in Radix Dipsaci was relatively high, and it should be concerned.

Fig. 8.

Margin of exposure (MOE) value of chronic exposure to AFB1, AFG1 or OTA in Radix Dipsaci (one of 10000 simulations). P50, 50th percentile; P95, 95th percentile.

The mean, 50th and 95th percentile values of the MOE risk factors for health effects of AFB1, AFG1 or OTA was also evaluated when the consumers received a low or high dose of Radix Dipsaci treatment for 1, 5, 10, 20, 40 years or lifetime (Table 2). With the shortening of exposure years, the MOE values of AFB1 and AFG1 also gradually returned to the acceptable range, such as when the exposure years of AFB1 was less than 40 years or the exposure years of AFG1 was less than 20 years, the mean, 50th and 95th percentile values of the MOE risk factors for health effects were more than 10000 either at low or high doses of Radix Dipsaci treatment. The mean, 50th and 95th percentile values of the MOE risk factors for health effects of OTA were more than 10000 either the consumers received a low or high dose of Radix Dipsaci treatment for 1, 5, 10, 20, 40 years or lifetime (Table 2 and Fig. 8E,F), suggesting the health risk of OTA in Radix Dipsaci was relatively low.

Table 2.

Margin of exposure (MOE) values for health effects in humans due to exposure to AFB1, AFG1 and OTA from Radix Dipsaci for different durations and dose of exposure.

Mycotoxins	Duration of consumption,yr	Estimated mean and 95%Cl of quantiles for risk quotients of AFB1/AFG1/OTA
		Low dose (×104)			High dose (×104)
		50th percentile	Mean	95th percentile	50th percentile	Mean	95th percentile
AFB1	1	51.345 (50.390–52.788)	52 400 (52.099–56 541)	71.736 (69.159–74.130)	30.768 (30.197–31.535)	31.750 (31.705–34.569)	45.139 (42.885–46.749)
	5	10.292 (10.047–10.503)	10.502 (10. 407–11 279)	14.393 (13. 884–15.035)	6.245 (6.047–6 397)	6.395 (6.307–6.923)	9.050 (8.562–9.341)
	10	5.122 (5.037–5.249)	5.233 (5.195–5.625)	7.161 (6. 882–7 388)	3.103 (3.029–3.175)	3.189 (3.174–3.466)	4.521 (4.317–4.708)
	20	2.577 (2.519–2 625)	2.628 (2.591–2.801)	3.415 (3.457–3.729)	1. 543 (1. 507–1.580)	1.587 (1.583–1.700)	2.253 (2.157–2.329)
	40	1. 285 (1.254–1.326)	1.311 (1.304–1.403)	1.784 (1.734–1.850)	0.777 (7.545–8.001)	0.797 (7. 950–8.657)	0.113 (0.101–0.117)
	Lifetime	0.732 (0.717–0.750)	0.747 (0.745–0.806)	1.015 (0. 985–1.068)	0. 443 (0.432–0. 454)	0. 455 (0. 455–0495)	0.647 (0.618–0.672)
AFG1	1	35.671 (35.033–36.723)	36.601 (30 680–40.228)	51.795 (50.231–54.195)	21.401 (21.031–22.081)	22.011 (22.001–23.967)	31.152 (30.037–32.450)
	5	7.152 (6.960–7.307)	7.338 (7.317–8.007)	1.038 (0. 955–1.085)	4.296 (4.179–4.395)	4.419 (4. 394–4.824)	6.255 (5.953–6.625)
	10	3.574 (3. 494–3.661)	3.670 (3.665–4.002)	5.225 (5.001–5 468)	2.146 (2.099–2.195)	2.203 (1.876–2.395)	3.106 (3.006–3.244)
	20	1. 800 (1.752–1.830)	1. 841 (1.838–1 995)	2.585 (2497–2 691)	1.074 (1.045–1.097)	1.102 (0.939–1.199)	1.550 (1.479–1.622)
	40	1.306 (0.872–0.916)	0.919 (0.911–0 999)	0.894 (1.258–1.366)	0.536 (0.526–0.547)	0.552 (0.546–0.596)	0.785 (0.749–0.812)
	Lifetime	0.508 (0. 425–0 444)	0.523 (0. 438–0.475)	0.724 (0.577–0.724)	0.306 (0.230–0.316)	0.315 (0.305––0.343)	0.444 (0.431–0.471)
OTA	1	170.177 (166.222–174.259)	174.799 (174.285–189.873)	248.493 (237.805–257.205)	101.800 (100.194–105.40)	105.107 (104.953–114.831)	148.802 (142. 388–155.728)
	5	34.138 (33.312–34.926)	34.992 (34.985–37.922)	49.445 (47.832–51.467)	20.360 (19.944–20. 898)	20.957 (20.931–22.775)	20.976 (20.845–30.924)
	10	8.514 (8.297–8.732)	8.744 (8.738–9.561)	12.397 (11. .869–12.946)	10.180 (10.005–10.441)	10.478 (10.436–11.492)	14.880 (14.301–15.467)
	20	8.552 (8.306–8.718)	8.782 (8.744–9.469)	12.488 (11.872–12.801)	5.090 (4.895–5.228)	5.239 (5.218–5.753)	7.440 (7.114–7.911)
	40	4.247 (4.140–4.334)	4. 381 (4.367–4.734)	6.170 (5.979–6.419)	2.545 (2.497–2.624)	2.631 (2.621–2.879)	3.720 (3.577–3.895)
	Lifetime	2. 449 (2.365–3.392)	2.507 (2.487–2.726)	3.538 (3.392–3.697)	1.460 (1.429–1.492)	1.501 (1.500–1.622)	2.142 (2.040–2.193)

The risk quotient of AFB1, AFG1 and OTA was also evaluated when the consumers received a low or high dose of Radix Dipsaci treatment for 1, 5, 10, 20, 40 years or lifetime (Table 3). Within the time range commonly used for high dose of Radix Dipsaci treatment (5–20 years), the mean risk quotient of AFB1, AFG1 and OTA exposure in Radix Dipsaci ranged from 0.165 to 0.656, 0.431 to 1.727, 0.050 to 0.198, respectively. The 50th percentile risk quotient of AFB1, AFG1 and OTA exposure in Radix Dipsaci ranged from 0.160 to 0.643, 0.422 to1.699, 0.049 to 0.193, respectively. And the 95th percentile risk quotient of AFB1, AFG1 and OTA exposure in Radix Dipsaci ranged from 0.223 to 0.889, 0.589 to 2.350, 0.070 to 0.281, respectively.

Table 3.

Risk quotients for health effects in humans due to exposure to AFB1, AFG1 and OTA from Radix Dipsaci for different durations and dose of exposure.

Mycotoxins	Duration of consumption,yr	Estimated mean and 95%Cl of quantiles for risk quotients of AFB1/AFG1/OTA
		Low dose (×10−2)			High dose (×10−2)
		50th percentile	Mean	95th percentile	50th percentile	Mean	95th percentile
AFB1	1	0.019 (0.019–0.020)	0.020 (0.019–0.021)	0.027 (0.026–0.028)	0.032 (0.032–0.033)	0.033 (0.032–0.035)	0.044 (0.043–0.040)
	5	0.097 (0.094–0.098)	0.098 (0.097–0.105)	0.134 (0.129–0.139)	0.160 (0.157–0.165)	0.165 (0.162–0.220)	0.223 (0.189–0.225)
	10	0.193 (0.189–0.198)	0.197 (0.194–0.210)	0.267 (0.259–0.277)	0.320 (0.315–0.327)	0.327 (0.324–0.350)	0.448 (0.431–0.462)
	20	0.388 (0.378–0.393)	0.394 (0.388–0.417)	0.533 (0.515–0.550)	0.643 (0.630–0.658)	0.655 (0.647–0.705)	0.889 (0 .861–0.925)
	40	0.773 (0.753–0.796)	0.784 (0.782–0.836)	1.064 (1.039–1.102)	0.644 (0.632–0.656)	0.656 (0.648–0.696)	0.893 (0.856–0.919)
	Lifetime	1.349 (1.319–1.380)	1.377 (1.367–1.465)	1.885 (1.822–1.928)	2.209 (2.232–2.656)	2.301 (1.648–3.696)	3.123 (3.856–6.919)
AFG1	1	0.051 (0.050–0.052)	0.052 (0.051–0.055)	0.070 (0.068–0.073)	0.085 (0.082–0.087)	0.086 (0086–0.092)	0.117 (0.114–0.122)
	5	0.254 (0.249–0.201)	0.260 (0.255–0.278)	0.354 (0.340–0.368)	0.422 (0.417–0.433)	0.431 (0.426–0.458)	0.589 (0.559–0.601)
	10	0.509 (0.498–0.520)	0.517 (0.515–0.552)	0.702 (0.685–0.727)	0.853 (0.831–0.867)	0.868 (0.855–0.922)	1.177 (1.134–1.218)
	20	1.020 (0.997–1.038)	1.040 (1.025–1.100)	1.412 (1.358–1.453)	1.699 (1. 659–1.725)	1.727 (1.714–1 .842)	2.350 (2.285–2.434)
	40	2.034 (1.994–2.080)	2.077 (2.054–2.214)	2.844 (2.718–2.930)	3.383 (3.317–3.483)	3. 444 (3.429–3.698)	4.670 (4.553–4.858)
	Lifetime	3.554 (3. 499–3.660)	3.617 (3.577–3.874)	4.916 (4.719–5.112)	5.931 (5.813–6.060)	6.040 (5.990–6.465)	8.215 (7.957–8.461)
OTA	1	0.006 (0.006–0.006)	0.006 (0.006–0.008)	0.008 (0.008–0.008)	0.010 (0.010–0.010)	0.010 (0.010–0.014)	0.014 (0.014–0.014)
	5	0.029 (0.028–0.030)	0.030 (0.029–0.032)	0.042 (0.040–0.043)	0.049 (0.047 –0.050)	0.050 (0.049–0.053)	0.070 (0.067–0.072)
	10	0.058 (0.057–0.060)	0.060 (0.059–0.065)	0.084 (0.080–0.088)	0.097 (0.094–0.099)	0.100 (0.098–0.101)	0.140 (0.134–0.145)
	20	0.116 (0.114–0.119)	0.119 (0.118–0.128)	0.167 (0.161–0.173)	0.193 (0.189–0.199)	0.198 (0.198–0.214)	0.281 (0.270–0.290)
	40	0.232 (0.226–0.239)	0.238 (0.236–0.256)	0.334 (0.323–0.347)	0.387 (0.380–0.397)	0.396 (0.395–0.425)	0.559 (0.537–0.584)
	Lifetime	0.407 (0.395–0.417)	0.416 (0.411–0.450)	0.587 (0.562–0.608)	0.678 (0.658–0.695)	0.694 (0.688–0.750)	0.976 (0.937–1.011)

By setting an MOE value of 10,000 as the threshold for maximum mycotoxin intake, we estimated the maximum acceptable content (µg/kg) of the three mycotoxins in Radix Dipsaci over different periods and dose of consumption. When the consumers received a high dose of Radix Dipsaci treatment for 1, 5, 10, 20, 40 years or lifetime, limit levels ranged from approximately 0.83 to 58.22 µg/kg for AFB1, 0.57 to 40.35 µg/kg for AFG1, and 8.01 to 560.73 µg/kg for OTA (Fig. 9). When the consumers received a low dose of Radix Dipsaci treatment for 1years to lifetime, limit levels ranged from approximately 1.38 to 97.05 µg/kg for AFB1, 0.96 to 67.26 µg/kg for AFG1, and 13.35 to 934.55 µg/kg for OTA (Fig. 9). Based on this exposure and a typical human diet, we estimated respective 20 years exposure limits for the three mycotoxins to be 5.821, 4.035 and 56.073 μg/kg in Radix Dipsaci.

Fig. 9.

Limit levels of AFB1, AFG1 and OTA in Radix Dipsaci for different durations and dose of exposure. Loose and strict limit levels were based on scenarios of intake low and high dose of Radix Dipsaci, respectively.

Discussion

Here we develop and validate an analytical method for simultaneously detecting eight mycotoxins dangerous to humans, and we apply it to the first systematic study of the medicinal herb Radix Dipsaci on the Chinese market. Our results suggest that mycotoxin contamination of this herb is prevalent, especially contamination with aflatoxin B1, aflatoxin G1 and ochratoxin A. Our work indicates that substantial fractions of these mycotoxins in raw herb pass into the water decoctions used in traditional Chinese medicine. We also estimate the risk of mycotoxin exposure for consumers of the herb and we propose intake limits for the mycotoxins to ensure acceptable health risk. Our analysis emphasizes the need for careful monitoring of Dipsaci Radix preparations for mycotoxin contamination in order to ensure consumer safety, and our method may prove useful for detecting contamination of other herbal medicines.

Due to the prevalent fungal contamination in the herbal medicines, the contamination ratio of mycotoxins in Traditional Chinese medicine was extremely high²⁹. The exposure risk exists in the whole supply chain, the primary contamination comes from planting and processing. The toxigenic fungi in root herbal medicines mainly come from the indigenous fungi in the soil where they were grown²⁹. After the introduction of toxigenic fungi from soil to herbal medicines, these fungi may proliferate and produce mycotoxins during processing and storage⁴. The medicinal herb Radix Dipsaci, characterized by its root structure, is susceptible to fungal contamination during its growth cycle^18,30. Furthermore, the humid and warm conditions required for the “sweating” process of Radix Dipsaci favor the proliferation of toxic fungi^31,32. The exposure limits of Radix Dipsaci have not been defined because no data exist on its contamination with mycotoxins. In this study, we were the first to investigate mycotoxin contamination in Radix Dipsaci.

Currently, over 300 different mycotoxins have been identified, exhibiting toxicity to humans and animals¹². Common mycotoxins in medicinal herb include aflatoxins, ochratoxins, zearalenone, sterigmatocystin and so on^33,34. When choosing the eight mycotoxins in our study, we focused on frequent contaminants of herbal medicines. Aflatoxin B1 and ochratoxin A have been classified as human carcinogens with strong toxic effects³⁵, while zearalenone can cause disorders of the human reproductive system^33,36,37. In this study, through the detection of mycotoxins in Radix Dipsaci, we found that the most frequent mycotoxins in our samples of Radix Dipsaci were aflatoxin B1, present in 35.56% of batches (at 0.25–34.84 μg/kg); aflatoxin G1, 15.56% (1.99–44.05 μg/kg); and ochratoxin A, 22.22% (16.11–143.38 μg/kg), which are produced by Aspergillus fungi. Similarly, high prevalence of contamination with ochratoxin A, aflatoxin G1 and other aflatoxins has been reported in ginseng, polygala, liquorice and other herbal medicines^14,18. We also found that the phenomenon of combined contamination by multiple mycotoxins is common in Radix Dipsaci. Of the 45 batches of Radix Dipsaci, 48.89% contained at least one of the eight mycotoxins, 24.44% contained one, 17.78% contained two and 6.67% contained three. This suggests that Radix Dipsaci can be infected by various toxin-producing fungi throughout the processes of cultivation, harvesting, processing, and storage.

Our work supports the usefulness of liquid chromatography and tandem mass spectrometry for specific, sensitive, rapid detection of multiple mycotoxins at the same time^29,38. It also supports the usefulness of methanol and acetonitrile for mycotoxin extraction³⁹: these solvents show low toxicity, they are relatively inexpensive, and they are highly soluble over a broad range of concentrations. We diluted these solvents in water to help them remain soluble when mixed with the sample^39,40. We purified the sample with a combination of C18 column, which can efficiently remove non-polar organic compounds such as fat and lipids in the sample to achieve the separation and purification of various organic compounds^39,41; and primary secondary amine, which are mainly used to remove substances such as organic acids and sugars from samples⁴². We also added GCB to adsorb pigments in the sample that might interfere with mycotoxin detection. Careful pretreatment and extraction are important for minimizing matrix interference, particularly when detecting mycotoxins^39,43.

Studies have shown that toxic fungi can be transferred from Chinese herbal medicine to water decoction, and then harm the health of users. Previous studies have reported aflatoxins were detected to be transferred from the herbal medicines to decoctions with significantly different transfer rates in the different types of herbal medicines. Quietly high transfer rates of 7.26–115.36% for AFG2, 4.37–26.37% for AFB1 and 9.64–47.68% for AFB2 were reported⁹. Yin Zhu et al. assessed the transfer of OTA from Astragali Radix to their decoction, and the transfer rate was calculated to be 56.1 %²³. In this study, mycotoxins were found to migrate into the water decoction from Radix Dipsaci, with migration rates of 16.94%, 29.14%, and 16.78% for AFB1, AFG1, and OTA, respectively¹⁸. AFB1 as the most toxic aflatoxin expressed the lowest transfer rate, which may be related to its structural characteristics such as high thermal stability and low water solubility⁹.

Mycotoxins possess carcinogenic, teratogenic, mutagenic, hepatorenal toxicities, as well as immunotoxicity and neurotoxicity. The thermal stability of these mycotoxins makes them impossible to reduce or eliminate by simple processing or roasting techniques⁴⁴. Consumption of Chinese herbal medicine contaminated with mycotoxins can lead to acute or chronic mycotoxicosis, resulting in various severe health hazards^44,45. Unlike other agricultural products and foods, herbal medicines consumers mainly take the water decoction, so in this study, we introduced the transfer rate of mycotoxins from raw herb into their decoction to assess the health risk of Radix Dipsaci consumers on the basis of the traditional risk assessment method²³. We found the mean the estimated daily intake of AFB1 was estimated to be 0.054 and 0.091 ng/kg body weight (bw)/day for the lower bound and upper bound scenarios, respectively, which is comparable to the mean estimated daily intake of AFG1. The mean and 50th percentile values of the MOE risk factors for health effects of AFB1 and AFG1 exceed 10000 when consumers received either low or high doses of Radix Dipsaci treatment, suggesting the health risk of AFG1 and AFB1 in Radix Dipsaci was relatively high. If the scenario of combined contamination by these two mycotoxins is considered, the risk will be further exaggerated and it should be concerned. As a medicinal herb used in a specific scenario, the dose and taking duration by Radix Dipsaci consumers varies according to the different diseases it treats. With the shortening of exposure years, we found the MOE values of AFB1, AFG1 and OTA also gradually returned to the acceptable range. We therefore conducted separate risk assessments and limit levels of mycotoxins for different doses and durations to make our findings generalizable to multiple use scenarios. For example, Radix Dipsaci is used in the treatment of osteoporosis at a dose of 15g per day for about 5–20 years, the corresponding exposure limits for AFB1, AFG1 and OTA to be 5.821, 4.035 and 56.073 μg/kg in Radix Dipsaci.

Our analysis highlights the need to identify factors influencing risk of Radix Dipsaci contamination with mycotoxins during cultivation, processing, transportation, and storage in order to develop appropriate prevention measures. Future work should also apply mixed exposure models given the substantial prevalence of simultaneous contamination with multiple mycotoxins and the potential for synergistic toxicological effects. These considerations are likely to apply to many other traditional herbal medicines.

Conclusion

Many commercial batches of Radix Dipsaci are contaminated with one or more mycotoxins, which transfer from the raw herb into the water decoctions used in traditional medicine. Samples should be systematically screened simultaneously for the most common or dangerous mycotoxins, and exposure limits for consumers of herbal medicine should be defined. The present study provides an analytical method and exposure modeling that may prove useful in such work.

Abbreviations

AFs

Aflatoxins

LOD

Limit of detection

LOQ

Limit of quantification

MRM

Multiple reaction monitoring

MOE

Margin of exposure

OTA

Ochratoxin A

RSD

Relative standard deviation

ST

Sterigmatocystin

T-2

T-2 toxin

ZEN

Zearalenone

Author contributions

Lulu Wang: Investigation, Methodology, Software, Writing – original draft, Writing – review & editing, Validation. Dapeng Su: Investigation, Methodology, Validation. Qingsong Yuan: Visualization. Chenghong Xiao: Data curation. Min Hu: Investigation. Lanping Guo: Conceptualization, Supervision. Chuangzhi Kang: Software. Jinqiang Zhang: Conceptualization, Methodology, Supervision, Writing – review & editing, Project administration. Tao Zhou: Resources, Project administration.

Funding

This work was supported by Key project at central government level for the ability establishment of sustainable use for valuable Chinese medicine resources (2060302), the University Science and Technology Innovation Team of the Guizhou Provincial Department of Education [Qian-jiao-ji (2023)071], the Scientific and Technological Innovation Project of the China Academy of Chinese Medical Science (CI2021B0 13), and the National and Provincial Scientific and Technological Innovation Talent Team of Guizhou University of Traditional Chinese Medicine [Gui-zhong-yi TD he-zi (2022)003]; the High-level Innovative Talents Program of the Guizhou Province of China [Qian Ke He Platform and Talent (2018)5638-2], the Research Platform Team Project of the Education Department of Guizhou Province [Qian-Jiao-Ji(2022)021], the National Technical System of Traditional Chinese Medicine Industry [grant number CARS-21].

Data availability

No datasets were generated or analysed during the current study.

Competing Interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-73597-0.