In total, 35 specimens, including the original plants of genuine Angelicae Sinensis Radix, Astragali Radix, and their common adulterants from different production areas in China, were collected (sampling information is listed in Table 1). The specimens were identified by Associate Professor Enwei Tian and Professor Zhi Chao from the School of Traditional Chinese Medicine, Southern Medical University (SMU), and were deposited in the herbarium of the School of TCM, SMU. The fresh leaves were preserved in gel-dried silica for genomic DNA extraction and further used to evaluate the specificity of the designed primers. To optimize the reaction conditions of the multiplex PCR method, the Danggui Buxue pill was handcrafted according to the guidelines of the Drug standard of the Ministry of Health of the People’s Republic of China—traditional Chinese medicine formulation (Pharmacopoeia Committee of the People’s Republic of China, 1995) and named DGBXP. The detailed preparation procedure is as follows: two different doses (160 g of Angelicae Sinensis Radix and 400 g of Astragali Radix) were crushed into fine powder, which was further sieved and mixed. Then, 30 g–40 g of refined honey was added, an appropriate amount of water was added to the powder, and the pills were panned. Afterward, the pills were dried. Four batches of adulterated Danggui Buxue pills were handcrafted referring to the aforementioned preparation (Batch 1: genuine Angelicae Sinensis Radix and Astragali Radix plus all adulterants; Batch 2: genuine Angelicae Sinensis Radix plus adulterants of Astragali Radix and Angelicae Sinensis Radix; Batch 3: genuine Astragali Radix plus adulterants of Angelicae Sinensis Radix and Astragali Radix; Batch 4: all adulterants), which were used to further test the specificity of the designed primes.

To determine the stability and practicability of the multiplex PCR assay, 14 batches of commercial Danggui Buxue pills were purchased from two manufacturers in China and numbered BX1–BX8 (Jilin Zixin Pharmaceutical Industrial Co., Ltd., China; batch nos: 190204, 190302, 190306, 190308, 190311, 190809, and 190506) and BX9–BX14 (Shangqiu Jinma Pharmaceutical Co., Ltd., China; batch nos: 20092021, 21062221, 19123021, 21032722, 21041122, and 20112321).

The modified CTAB method (4% CTAB for cell lysis and polysaccharide removal and 2 volumes of ice-cold 70% methanol for DNA precipitation) was used to extract the total genomic DNA of the dried leaves from the 35 samples (Murray and Thompson, 1980; He et al., 2011). For the 14 batch samples of commercial Danggui Buxue pills, handcrafted samples (DGBXP), and four batches of handcrafted and adulterated Danggui Buxue pills (B1∼ B4), the aforementioned modified CTAB method combined with the silica-based spin column DNA purification kit (Cat.#DP204-03 and Lot#X0902; Tiangen, Bejing, China) was used for genomic DNA extraction. Considering the potential issues of the DNA degradation of commercial and handcrafted samples, we increased the dosage of samples (2 g powder) for DNA extraction, which was much more than that (50 mg–100 mg) of dried leaves using the classic CTAB method (Murray and Thompson, 1980). The concentration and integrity of genomic DNA were assessed using a NanoDrop 1000 UV/Vis spectrophotometer (Thermo Scientific, Wilmington, DE, USA) and 1.5% agarose gel electrophoresis, respectively.

Fourteen and sixteen nrITS sequences for the original plants of Angelicae Sinensis Radix, Astragali Radix, and their adulterants were downloaded from GenBank (NCBI: https://www.ncbi.nlm.nih.gov/) (Supplementary Table S1, Supplementary Table S2). Quality control for these sequences complied with the five simple quality control guidelines for establishing basic authenticity and reliability (Nilsson et al., 2012). Afterward, the sequences were aligned using MEGA v7.0.14 software (Kumar et al., 2016) and used to search for highly variable SNP sites. Primer Premier 5 software (https://www.premierbiosoft.com/primerdesign/index.html) was used to design primers for the specific ITS regions of A. sinensis, A. membranaceus var. mongholicus, and A. membranaceus. The designed primers were evaluated according to the standards of Li et al. (2015) and Sathirapongsasuti et al. (2020). The primers were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). After the pre-experiment, two species-specific primer pairs (DG and HQ) were tentatively selected for identifying Angelicae Sinensis Radix and Astragali Radix (Table 2). The conservation of the specific SNPs for Angelicae Sinensis Radix and Astragali Radix had been determined by phylogenetic trees (neighbor-joining tree), which were generated by the downloaded ITS sequences, respectively (Supplementary Figure S1, Supplementary Figure S2).

The two selected primers were used in conventional PCR amplification of genuine Angelicae Sinensis Radix (six samples: MX1∼MX3 and GS1∼GS3) and its adulterants (nine samples: ODG1, ODG2, ODG3, ZN1, ZN2, ZN3, AG1, AG2, and AG3) and Astragali Radix (six samples: GM1, GM2, DT1, DT2, ZY1, and ZY2) and its adulterants (JJE1, JJE2, ZMX1, ZMX2, ZMX3, SWZ1, SWZ2, SK1, SK2, and SK3), respectively (Table 1). Four batches (B1∼B4) of adulterated Danggui Buxue pills were used to further test the specificity of the two selected primers using multiplex PCR. The 20 uL PCR reaction system contained 2 uL of the DNA template, 1.6 μL of 10 × buffer (Mg²⁺ free), 2.5 uM of MgCl₂, 300 uM of dNTP mixture, 2U of rTaq DNA polymerase (Takara, Dalian, China) and 0.5 uM of HQ primers, or/and 0.25 uM of DG primers. PCR amplification was performed with the following procedure: one pre-denaturation cycle of 5 min at 94°C, 35 cycles of 94°C for 50 s, 65°C for 30 s, 72°C for 90 s, and followed a final extension at 72°C for 7 min. PCR products were analyzed on 1.2% agarose gel and visualized by the electrophoresis gel imaging system (Peiqing, Shanghai, China).

The genomic DNA extracted from plant materials performed well on a single DNA band in agarose gel, but not for handcrafted and commercial Danggui Buxue pill samples, which suggested that serious DNA degradation occurred within these samples due to the processing of raw materials and the addition of various excipients. The concentration detection results showed that they varied from 60 to 200 ng/μL and from 10 to 400 ng/μL for plant materials and Danggui Buxue pill samples (including handcrafted samples), respectively. The A260/A280 information and the DNA electrophoretogram of handcrafted DGBXP, of the 14 samples of commercial Danggui Buxue pills, and of the four samples of handcrafted Danggui Buxue pills (adulterated) are shown in Supplementary Table S3 and Supplementary Figure S3.

Conventional PCR amplification with the single pair of primers (DG and HQ) generated DNA bands of 349 bp and 381 bp for Angelicae Sinensis Radix and Astragali Radix, respectively, and no amplification signals were observed for the adulterant samples (Figures 1A,B; Supplementary Figure S4). It demonstrated that DG and HQ primers had strong specificity and could be used for identifying genuine Angelicae Sinensis Radix and Astragali Radix. The consistent results were also presented on the specific amplification of four batches (B1∼B4) of adulterated Danggui Buxue pills (Supplementary Figure S5).

The results of the multiplex PCR amplification showed that the two biological ingredients within the handcrafted sample (DGBXP) could be successfully amplified at the temperature range of 50°C–67.6°C (Figure 2). Based on the amplification signals, the optimal temperature range of the multiplex PCR was 54.6°C–65.1°C for both ingredients. To ensure the specificity of the primers, we chose 65°C as the optimal temperature to detect the sensitivity of the multiplex PCR, which was also further applied for the identification of the biological ingredients within commercial Danggui Buxue pills.

To assess the sensitivity of the multiplex PCR assay, six DNA concentration gradients of DGBXP samples were set, which started at a concentration of 40 ng/μL (initial concentration) and were successively diluted 10 times until the terminal concentration of 4.0 × 10⁻⁴ ng/μL. The multiplex PCR assay was capable of amplifying both targeted fragments (349 bp and 381 bp) when the genomic DNA of the DGBXP sample was diluted from 40.0 to 4.0 × 10 ^–2 ng/μL. Then, the genomic DNA was further diluted at a concentration of 4.0 × 10⁻³ ng/μL, and the targeted gene fragment (HQ, 381 bp) for Astragali Radix was still successfully amplified, but with a weak signal. However, no amplification signal was present at the 349 bp position (DG) for Angelicae Sinensis Radix (Figure 3). The results suggested the lowest detection concentration was 4.0 × 10 ^–2 ng/μL and 4.0 × 10 ^–3 ng/μL for Angelicae Sinensis Radix and Astmgali Radix, respectively.

Fourteen genomic DNA samples of commercial Danggui Buxue pills were respectively amplified with both species-specific primers (DG and HQ) in one reaction according to the optimal reaction conditions of the multiplex PCR assay. The results showed that all the commercial samples could be simultaneously amplified and showed two bands (349 bp and 381 bp) (Figure 4), which suggested that the multiplex PCR assay was stable, accurate, and suitable for detecting Angelicae Sinensis Radix and Astragali Radix within commercial Danggui Buxue pills.

The quality of genomic DNA is the key to the molecular identification of the biological ingredients within CPM (Jing et al., 2021). However, during the preparation, the raw medical materials are always processed by all kinds of processing methods, such as decocting, calcining, stir-frying, and fermenting (Chen, 2018), which may result in severe DNA degradation and affect DNA quality and quantity (Karni et al., 2013). Meanwhile, the addition of a variety of pharmaceutical excipients, such as honey, starch, and sugar, also causes significant difficulties in DNA extraction (Sajali et al., 2018). Therefore, the methods formulated for genomic DNA extraction depend on the different dosages and specific preparation processes of CPM. In this study, we used the modified CTAB method, which has been widely proven to be suitable for DNA extraction of plant-derived materials (He et al., 2011). Approximately 4% CTAB (the concentration of NaCl was increased to 2.5 M) was used for complete cell lysis and the removal of polysaccharides (Cui et al., 2006), and 2 volumes of ice-cold 70% methanol were added for DNA precipitation overnight, which allowed obtaining high quality of genomic DNA. In order to extract sufficient genomic DNA for further study, we increased the dosage of commercial Danggui Buxue pill samples. After comparing different dosages of samples (0.1 g, 0.5 g, 1 g, and 2 g), we determined the sample dosage of 2 g for DNA extraction, which has been proved to be sufficient for subsequent PCR amplification. Many previous studies also suggested that more than 100 mg should be used for DNA extraction (Gou et al., 2018). DNA purity is another very important factor that affects PCR amplification. Some impurities, such as polysaccharides, polyphenols, and proteins mixed within genomic DNA, will also inhibit PCR amplification (Hedman et al., 2013; Sajali et al., 2018). In order to obtain high-purity genomic DNA, we carried out the CTAB method combined with the silica-based spin column DNA purification kit to extract genomic DNA from commercial samples, and our validation results of PCR amplification with species-specific primers suggested that the method was effective.

To select an appropriate DNA barcode for designing species-specific primers, we aligned the sequences of four universal barcodes (ITS, matK, rbcL, and trnH-psbA) among the original plant species of Angelicae Sinensis Radix, Astragali Radix, and adulterants and further checked their variation information. Consistent with previous studies (Parvathy et al., 2015), ITS featured much more variables and informative sites, which had already been recognized as a perfect molecular marker in the identification of angiosperms. On the other hand, SNP genotyping with allele-specific PCR is a powerful tool for the authentication of botanical drugs (Wang et al., 2012). In this study, we focused on the variable targeted gene ITS to design allele-specific primers. The reasonable product size of the SNP-based allele-PCR amplification to balance the amplification success rate and resolution of the targeted gene fragments is worth mentioning. Generally, the smaller targeted gene fragment usually shows a higher amplification success rate for the highly processed raw medicinal materials, which makes it much more difficult for designing appropriate primers because of insufficient variation sites harbored in limited gene fragments (Yang J. F. et al., 2018). In this study, the product sizes of amplified gene fragments were less than 400 bp (349 bp and 381 bp, respectively) in length, and these results suggested that both pairs of primers had excellent amplification success rates. Chen Y. et al. (2005) found that although mechanical grinding and high-heat processes would cause DNA degradation, they still amplify approximately 400 bp of DNA fragments, which is consistent with the results of PCR amplification in this study. However, for some other CPMs requiring stricter processing, it is recommended that the size of the amplified product should be around 100–250 bp in length (Yang F. et al., 2018; Gao et al., 2019). Moreover, the difference between product sizes of multiplex PCR amplification with multiple primers is also a major concern. Liu et al. (2018) recommended that the differences in multiple PCR product sizes visualized by gel electrophoresis should not be less than 50 bp, and Zhou et al. (2005) used capillary electrophoresis to distinguish the difference (17 bp) of multiple PCR products. In fact, in this study, the product size difference of approximately 30 bp could still be distinguished by gel electrophoresis. To ensure the specific binding of primers with targeted DNA fragments, we suggest keeping the difference sizes of different targeted gene amplification products as large as possible, which will be conducive to the separation of different targeted fragments in gel electrophoresis. In summary, the product size differences of multiplex PCR for different biological ingredients depend on the number of primers, the difficulties of primer designing, the degree of degradation of the DNA template, and the detecting method.

Detection of Angelicae Sinensis Radix and Astragali Radix within 14 batches of commercial Danggui Buxue pills using multiplex PCR assay yielded positive results for both ingredients. In order to further verify the accuracy of the multiplex PCR assay, the PCR products were sequenced and BLASTed in NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The similarity compared with ITS sequences of Angelica sinensis and Astragalus mongholicus in GenBank was 100% (sequence coverage: 100%). Our results indicated that the multiplex PCR assay was highly specific and could simultaneously authenticate Angelicae Sinensis Radix and Astragali Radix with the two pairs of primers. Although all the genuine crude drugs were added in the 14 batches of commercial samples, the sampling errors should not be ignored, and the supervision department should still strengthen the monitoring, as many adulterants of Angelicae Sinensis Radix and Astragali Radix have been reported (Chen, 2010; Wang et al., 2013; Zhang et al., 2018). Accordingly, there are 277 and 255 compound preparations containing Angelicae Sinensis Radix or Astragali Radix, respectively, which have been mentioned by the National Medical Products Administration of China (https://www.nmpa.gov.cn/datasearch/search-result.html). The number of compound preparations that were simultaneously added with Angelicae Sinensis Radix and Astragali Radix was up to 39 (Supplementary Table S4). Thus, the developed specific SNP primers and the multiplex PCR assay will have a wide application in the authentication of Angelicae Sinensis Radix or/and Astragali Radix of the compound preparations.

Regarding the advantages of the multiplex PCR assay, such as being simple, effective, economical, time-saving, and performed without any complicated bioinformatics analysis like DNA metabarcoding and shotgun genomics, it also has some limitations: 1) this method could not be used for authenticating adulterants mixed in authentic crude drugs. It is necessary to further design species-specific primers for all common adulterants, which allow identifying both genuine crude drugs and adulterants within CPM simultaneously. 2) The success of the multiplex PCR assay is closely related to the degree of DNA degradation and the specificity of primers (Wang et al., 2017). We have to fully understand the processing process for producing various kinds of CPM, which will be helpful in obtaining high-quality genomic DNA as much as possible. Likewise, we should also estimate the genetic relationship among the original species of the biological ingredients within CPMs so as to screen high-resolution targeted gene fragments and design highly species-specific primers and ensure the success rate of identification. 3) The quality control of CPM not only requires quantitative evaluation but also qualitative evaluation. In order to determine whether the manufacturer really provides the biological ingredients according to the dosage of the CPM formula, the molecular quantitative method needs to be further developed in the future. Previous studies have reported that the amount of some remnant DNA fragments (DNA copy numbers) could reflect the number of crude drugs required for animal-derived CPM (Ye et al., 2018). Therefore, to simultaneously realize qualitative and quantitative monitoring of biological components within CPM, the fluorescence quantitative PCR and ddPCR assay based on the changes in DNA copy number might be an alternative choice.