Roots of Atractylodes macrocephala, Coptis chinensis, Radix Isatidis and Flos Genkwa were purchased from Caizhiling, a reputable Chinese medicinal herb store in Guangzhou, China. Their authenticity was confirmed by Professor Ma Zhiguo of the Pharmacy College of Jinan University. The aqueous extract was prepared by a general method. After cutting the herbs into small pieces, 100 g dried plant material was boiled in 1,000 ml distilled water for 1 h. The decoction was collected and the residue was boiled another two times. The decoction obtained from the three separated extractions was mixed, filtered and lyophilized by freeze drying. The powdered forms of the extracts were stored at −20°C.

The dried extracts were dissolved in double-distilled water until the initial concentration was 1 g/ml and centrifuged at 14,000 g for 5 min before filtration sterilization to obtain a clear, sterile supernatant for testing. The chemicals were dissolved in appropriate solvents. The chemical 5-fluorouracil (5-FU; Sigma-Aldrich, St. Louis, MO, USA) was dissolved in dimethylsulphoxide. Phenytoinum natricum (DPH; Sigma-Aldrich) and saccharin (SAC; Sigma-Aldrich) were dissolved in phosphate-buffered saline (PBS). The final solvent concentrations applied in the differentiation and cytotoxicity assays demonstrated no undesired background effects.

Undifferentiated mouse ESCs of the OG₂ cell line were purchased from the Chinese Academy of Sciences. Continuous cultures of the cell line were grown on mitomycin C inactivated mouse embryonic fibroblast (MEF) feeders in a standard culture medium consisting of 80% high-glucose (4.5 g glucose/l) Dulbecco’s modified Eagle’s medium (DMEM; Gibco Life Technologies, Karlsruhe, Germany), 15% fetal calf serum (Hyclone, Erembodegem-Aalst, Belgium), as well as 2 mM glutamine, 2 mM sodium pyruvate, antibiotics (50 U/ml penicillin and 50 µg/ml streptomycin), 1% non-essential amino acids, 0.1 mM β-mercaptoethanol and 1,000 U/ml murine leukemia inhibitory factor, which were all purchased from Gibco Life Technologies. BALB/c 3T3 fibroblasts, purchased from the cell bank of Sun Yat-sen University of Medical Sciences, were cultured in DMEM supplemented with 10% fetal calf serum, 50 U/ml penicillin and 50 µg/ml streptomycin. The cells were maintained under 5% CO₂ and 95% humidity at 37°C.

The ESCs were subjected to alkaline phosphatase (AKP) staining on day 5 of passage. The original medium was discarded from the culture plates, and the cells were washed with PBS and fixed in 4% paraformaldehyde for 10 min. They were then washed again with PBS and stained with 75 mg/ml nitrotetrazolium blue chloride (Inogent, Hyderabad, India) and 50 mg/ml 5-bromo-4-chloro-3-indolylphosphate (Innogent) for 1 h at room temperature.

The cytotoxic effects on the mESCs and 3T3 cells were detected by performing a CCK8 (Mbchem, Mumbai, India) cell proliferation assay on day 10. In brief, on day 1, 500 cells were seeded into each well of a 96-well tissue plate in a routine cell culture medium without leukemia inhibitory factor and incubated for 2 h. Following incubation, 200 µl culture medium containing the appropriate dilution of extracts from the four medicines were added to each well. On days 3 and 5, the medium was renewed. On day 10, CCK8 assay was carried out. Cytotoxicity was expressed as the concentration of the compound that reduced the viability of cells to 50% of the control level (IC₅₀3T3 and IC₅₀ES), determined from a concentration-response curve.

Differentiation assays were performed to detect compound-induced changes in the differentiation of the mESCs into contracting cardiomyocytes. Briefly, on day 0, 20 µl stem cell suspension containing 750 cells was placed as hanging drops on the inner side of the lid of a petri dish filled with 6 ml PBS, then incubated for three days at 37°C, with 5% CO₂ and 95% humidity, in the presence of the test chemicals at various concentrations. During this period, the cells formed aggregates referred to as embryonic bodies (EBs). After three days, the aggregates that formed were transferred to bacterial petri dishes and exposed to the appropriate concentration of the test chemical for another two days. On day 5, the EBs were placed individually into six wells of a Falcon tissue culture plate. On day 10, the EBs were collected for quantitative polymerase chain reaction (qPCR) detection to observe the expression of related genes.

The cell samples for analysis were collected on day 10 of the differentiation assay, and total RNA was extracted using an EZNA™ Total RNA kit II (Omega Bio-Tek, Norcross, GA, USA) according to the instructions of the manufacturer. cDNA was synthesized using 1 µg RNA and PrimeScript™ RT Master mix (Takara, Otsu, Japan) as per the instructions of the manufacturer. The reaction mixture was incubated at 37°C for 30 min, followed by 5 sec at 85°C. To verify the undifferentiated marker genes Sox2 and Oct4 and the myocardial-specific marker β-myosin heavy chain (β-MHC), qPCR was performed using SsoAdvanced™ SYBR-Green (Bio-Rad Laboratories, Hercules, CA, USA). qPCR reactions were conducted in a 20-µl mixture that included 10 µl 2X qPCR Master mix, 8 µl nuclease-free water with mouse-specific primers for glyceral-dehyde-3-phosphate dehydrogenase (GAPDH; 5′-GCCTTC TCCATGGTGGTGAA-3′ and 5′-GCACAGTCAAGGCCGA GAAT-3′) and β-MHC (5′-TCCGCAACCGAGAGAATCAG-3′ and 5′-TGTCGCCAGAAATTGTGCCTT-3′), and 2 µl cDNA template. The thermal cycle profile consisted of an initial 30-min step at 95°C, followed by 40 cycles of 95°C for 5 sec, and 60°C for 20 min. The CFX Connect Real-Time system and CFX Manager Software (version 2.0; Bio-Rad) were used to collect the PCR data. The RNA levels of each gene were normalized to GAPDH.

Two permanent mouse cell lines of ESCs and 3T3 cells were used to predict the cytotoxicity of the test compounds using the EST. The concentrations of the compounds that inhibited the proliferation of 50% of the ESCs and 50% of the 3T3 cells (IC₅₀ES and IC₅₀3T3) were determined with CCK8 assay. In addition, the concentration that inhibited the differentiation of 50% of the ESCs (ID₅₀ES) was obtained from the results of qPCR in a differentiation assay. The three endpoints obtained in each experiment were used to calculate linear discriminant functions (I, II, III) for each extract.

Depending on the variables of the three functions I, II and III, the embryotoxicity of a test compound can be classed as non-embryotoxic (Class 1), weakly embryotoxic (Class 2) or strongly embryotoxic (Class 3).

The classification criteria were as follows: Class 1: if I>II and I>III; Class 2: if II>I and II>III; and Class 3: if III>I and III>II.

We first used a modified EST to determine the effect of 5-FU, DPH and SAC on differentiation and to classify the embryotoxic potential of each compound, which was revealed to be strong, weak and non-existent, respectively, to verify the predictive validity of the EST we established. Subsequently, the embryotoxic classification of the four herbal extracts was determined with the modified EST.

The statistical analysis was performed using GraphPad Prism 5 (GraphPad Prism Software, Inc., San Diego, CA, USA). Each data point represented three independent experiments. Data are given as the mean ± SEM. A one-way ANOVA was used to assess the statistical significance. P<0.05 was considered to indicate a statistically significant difference.

We verified the undifferentiated cells through morphology, AKP staining and qPCR to determine the expression levels of undifferentiated markers in undifferentiated cells. Observed under the microscope, the ESCs were nest-like, compact clones, with clear boundaries (Fig. 1A). They were AKP-positive and deep brown in color (Fig. 1B). The expression levels of the undifferentiated markers Sox2, Oct4 and Nanog were much higher than in the negative control, MEF (Fig. 1C). These results indicated that the ESCs were undifferentiated, and they were used in subsequent experiments.

The modified EST model was verified by determining the embryotoxicity of 5-FU (Fig. 2), DPH (Fig. 3) and SAC (Fig. 4). Three endpoint values were obtained for each compound. Detailed data on the endpoint values are shown in Table I. We confirmed that 5-FU had strong embryotoxicity, DPH was weakly embryotoxic and SAC was non-embryotoxic.

The four TCMs (Rhizoma Atractylodes macrocephala, Radix Isatidis, Coptis chinensis and Flos Genkwa) were decocted with water, and the EST was performed to determine the embryotoxic potential of the extracts. Three endpoints were obtained for each extract from three independent experiments. The cytotoxicity of the extracts on the OG₂ cells and the 3T3 cells (IC₅₀3T3 and IC₅₀ES) was determined with a CCK8 assay, and the inhibition of cardiomyocyte differentiation of the ESCs (ID₅₀ES; based on the expression levels of the β-MHC gene) was determined by qPCR. The three parameters of each extract were substituted into three linear discriminant functions (I, II and III). Detailed data on the classification of the embryotoxic potential of each extract using the criteria for embryotoxicity are provided in Table I.

In the cytotoxicity assay, the IC₅₀3T3 value was 4.82 mg/ml, which was ~1/10 that of the IC₅₀ES value (42.42 mg/ml; Fig. 5A and B). The ID₅₀ value for Rhizoma Atractylodes macrocephala was 37.67 mg/ml (Fig. 5C). It promoted the survival of ESCs and led to cardiac differentiation of ESCs at a low concentration. Thus, it was defined as a non-embryotoxic compound.

ESCs increased at a low Radix Isatidis concentration. As the concentration increased, the survival rate of the ESCs declined. The proliferation of the 3T3 cells was suppressed at a low concentration, whereas the proliferation of the ESCs was promoted (Fig. 6A and B). The IC₅₀ values were 1.76 and 27.68 mg/ml for the 3T3 fibroblasts and the ESCs, respectively. The expression levels of β-MHC decreased in a concentration-dependent manner following exposure to Radix Isatidis extract (Fig. 6C). The ID₅₀ values were 15 mg/ml. Radix Isatidis extract was classified as non-embryotoxic.

In both cell lines, the Coptis chinensis extract inhibited the survival of cells in a dose-dependent manner (Fig. 7A and B). The cytotoxic sensitivity of the 3T3 fibroblasts was much higher than that of the ESCs exposed to Coptis chinensis extract. The IC₅₀ values were 0.398 and 2.73 mg/ml for the 3T3 fibroblasts and the ESCs, respectively. The ID₅₀ value was 1.06 mg/ml. The Coptis chinensis extract was defined as weakly embryotoxic.

The 3T3 cells and ESCs were both sensitive to Flos Genkwa. The IC₅₀ values for the 3T3 cells (0.51 mg/ml) and those for the ESCs (0.75 mg/ml) were below 1 mg/ml (Fig. 7A and B). The ESCs were slightly more sensitive than the 3T3 cells when the concentration was under 0.25 mg/ml. However, as the concentration was increased, the opposite trend was observed. The value of ID₅₀ was 1.15 mg/ml. Flos Genkwa was classified as strongly embryotoxic.