Introduction

Herbal medicines have been traditionally used to treat and prevent various human diseases. Interest in and acceptance of naturopaths has surged in developing and developed countries in the past, and it is estimated that approximately 4 billion people in developing countries rely on herbal products as a primary means of health care (Ekor, 2014). The value of the growing global herbal medicine market is predicted to reach $129.6 billion by 2023 (Market Research Future, 2018). However, as the use of herbal medicines increases globally, new products are being released, and the safety of herbal medicines is threatened by insufficient quality control and improper labeling (Raynor et al., 2011). Therefore, it is necessary to identify and clearly indicate the exact ingredients of herbal medicines.

Numerous plants belonging to the family Lamiaceae (also referred to as mint family) are used as medicinal plants and aromatic crops worldwide. The respective medicinal plants have been reported to exert strong analgesic effects, thus, various Lamiaceae crops are promising with respect to novel potential therapeutics with regard to pain treatments (Uritu et al., 2018). Among plants of the Lamiaceae family, the dried above-ground part of the annual Agastache rugosa (Fisch. & C.A.Mey.) Kuntze is used as an herbal medicine referred to as “Kwakhyang” (Korean designation), whereas dried above-ground parts of the perennial Pogostemon cablin Bentham are used as “Gwanggwakhyang” (Korean name; Kim et al., 2003).

Agastache rugosa is known as Korean mint and is cultivated in East Asian countries, including Korea (Shin et al., 2018). This species has been used medicinally to treat common cold, anorexia, cholera, vomiting, poisoning, and other afflictions, and its various biological and pharmacological activities, including antibacterial, antifungal, anti-inflammatory, anticancer, and antioxidant effects, have been documented (Li et al., 2013). Pogostemon cablin exerts similar effects to those of A. rugosa, however, it is an entirely different plant. In the Japanese Pharmacopeia, the generic drug designation of P. cablin is “Kwakhyang”, thus there is some confusion as to the use of this compound (Ministry of Food and Drug Safety, 2022).

A comparative analysis of the components of essential oils of A. rugosa and P. cablin (Wu et al., 2013) and a comparative analysis of physicochemical properties through HPLC (Kim et al., 2003) were published; however, no such study has been performed to elucidate differences in DNA sequences.

Various techniques, including spectroscopy and sensory, chromatographic, and DNA-based analyses, have been used to detect inappropriate ingredients in processed foods (Hong et al., 2017). Multiplex PCR and quantitative real-time PCR (qPCR) have been successfully used to detect food fraud because of their timeliness and cost effectiveness, compared to other methods (An et al., 2018; Oh et al., 2022; Kane and Hellberg, 2016). qPCR, which can be used to detect target DNA fragments with high sensitivity and specificity in complex food products, can generally be designed as probe-based qPCR (e.g., TaqMan analysis) and dye-based qPCR (e.g., SYBR Green analysis) (Arya et al., 2005). Probe-based qPCR is used to detect a target sequence through a fluorescent reporter probe and primers (Navarro et al., 2015). However, this method requires many SNPs and indels when designing probes to differentiate between species, thus making it difficult to optimize qPCR conditions (Safdar and Junejo, 2015; Şakalar and Kaynak, 2016). By contrast, SYBR Green-based qPCR can be a more convenient, flexible, and inexpensive method than probe-based qPCR as it is based on DNA intermediate dye that binds to a fluorescent dye independent of double-stranded DNA; however, non-specific PCR products may occur (Safdar and Junejo, 2015).

In this study, molecular markers were developed to distinguish between A. rugosa and P. cablin through SYBR green-based qPCR. Blind tests and specificity tests showed that this assay is reliable, and it is easily applicable to commercial products.

Materials and methods

Sample preparation

Seeds of A. rugosa were kindly provided by the Gangwon-do Agricultural Research and Extension Services (Cheorwon, Gangwondo, Korea). Dried P. cablin samples were provided by the Herb Resource Bank of Traditional Korea Medicine (Seoul, Korea). Common non-target plant species were included to estimate marker specificity. The samples used for evaluating specificity and applicability of the developed markers were purchased at a local market. DNA was isolated from fresh or dried leaves.

Genomic DNA extraction

DNA was extracted using a DNeasy Plant Pro Kit (QIGEN, Hilden, Germany) according to the manufacturer’s instructions. For large-scale samples, such as in binary mixing tests, DNA was extracted according to a modified large-scale CTAB-based method (Minas et al., 2011). DNA concentration and purity were measured using a SPECTROstar Nano Leader (BMG Labtech, Otenberg, Germany). The DNA purity ratio was 1.7–2.0. At a ratio < 1.7, high-quality DNA was obtained using a Wizard DNA Clean-up System (Promega, Madison, WI, USA).

Binary mixture samples

To test applicability of the developed markers for processed commercial products, a quantitative binary mixture was produced using 2 g of a mixture of A. rugosa powder containing P. cablin [at 0.1% (2 mg), 1% (20 mg), 10% (200 mg), or 100% (2 g)]. The binary mixture samples were then treated as follows: (1) drying at 55 °C for 48 h in a drying oven, (2) boiling at 100 °C for 15 min, and (3) autoclaving at 121 °C for 20 min at 150 kPa in an autoclave.

Blind samples

A total of 25 blind mixtures were provided by the National Institute of Food and Drug Safety (Cheongju, Korea). The samples were mixtures of A. rugosa and P. cablin at random proportions. A total of 400 mg of the final mixture was mixed with A. rugosa powder or P. cablin powder at a concentration of 0–10%, respectively.

Sequence analysis and primer design

Chloroplast DNA sequences of matK, rbcL, and rpoB of the two target species (A. rugosa accession NC_053706.1 and P. cablin accession NC_042796.1) and internal transcribed spacer (ITS) sequences of A. rugosa (accession MH711397.1) and P. cablin (accession KR608755) were downloaded from the National Center for Biotechnology Information (NCBI) database to design specific primers. Nucleotide sequences were aligned using BioEdit 7.2 (Ibis Biosciences, Carlsbad, CA, USA) and ClusteralW2 (EMBL-EBI, Hinston, Cambridgeshire, UK). Species-specific primers were designed using Beacon Designer software (Primer Biosoft, Palo Alto, CA, USA) based on SNP between A. rugosa and P. cablin and were inspected manually. Primers were synthesized by a commercial company (Seoul Macrogen, Seoul, Korea).

Cloning of PCR amplicons and sequencing

Conventional PCR was performed using a VeritiPro Thermal Cycler (Thermo Fisher Scientific, Waltham, MA, USA). A mixture of DNA was prepared using TaKaRa Ex Taq DNA Polymerase (TaKaRa, Kusatsu, Shiga, Japan). The PCR conditions were as follows: pre-denaturation for 5 min at 95 °C, 35 cycles of 95 °C for 10 s, 53–61 °C for 30 s (depending on the primers), and 72 °C for 1 min, and a final step of 72 °C for 5 min. The PCR product was cloned using an RBC T&A cloning vector (Real Biotech Corporation, Taipei, Taiwan). Plasmid DNA extracted using the DokDo-Prep Plasmid Mini-Kit (ELPISB) was sequenced by a commercial service (Seoul Macrogen).

Quantitative real-time PCR assay

qPCR was performed using AccuPower® 2X GreenStar qPCR Master Mix (Bioneer, Daejeon, Korea). The reaction mixture comprised 10 μL 2X GreenStar Master Mix, 0.5 μL of each primer (10 pmol), 1 μL 10 ng μL–1 genomic DNA, 0.25 μL ROX Dye, and distilled water to 20 μL final volume. The qPCR reaction was performed using a Quantstudio 3 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with the following thermocycling protocol: 10 min of 95 °C, followed by 40 cycles of 95 °C for 30 s, 58–60 °C for 20 s (depending on the melting temperature of the respective primers), and 72 °C for 30 s. All qPCRs were performed in triplicates.

qPCR standard curve analysis

A standard curve for evaluating the efficiency and sensitivity of the developed primer set was produced by applying a tenfold diluted genomic DNA mix at five concentrations (10 ng/µL to 1 pg/µL) of target gDNA and non-target gDNA. The extracted DNA was diluted to a final concentration of 10 ng/µL, and the thresholds and baselines were set for further experiments (Yuan et al., 2006). The correlation between the cycle threshold (Ct) standard curve and diluted DNA was estimated using basic parameters. The standard curve was calculated as y =  − ax + b (where a is the standard curve slope and b is the y-intercept). The percentage efficiency was evaluated as (E – 1) × 100%, and the reaction efficiency (E) was calculated as E = (10–1/a) (European Network of GMO Laboratories, ENGL, 2015; Lo and Shaw, 2018).

Interlaboratory validation

Interlaboratory validation was performed using two laboratories to confirm the reproducibility and accuracy of the developed assay. Interlaboratory validation was performed using the Applied Biosystems 7500 Fast Real-Time PCR Instrument System (Applied Biosystems) and the Step 1 Plus Real-Time PCR System (Applied Biosystems) under the same qPCR conditions.

Results and discussion

DNA markers in variable regions

Primer sets for distinguishing A. rugosa and P. cablin were designed in the nuclear ITS region of ribosomal DNA and in chloroplast genes such as matK, rbcL, and rpoB. The ITS region of the rRNA gene, which is a ribosomal gene essential for protein synthesis, is indispensable for organisms and is well conserved in plants. Therefore, because ITS fragments occur in almost all taxa, such primers are well established and can be used for a wide range of organisms (Baldwin et al., 1995). Chloroplast genes such as matK, rbcL, and rpoB have also been used for species identification in other studies (CBOL Plant Working, 2009). Four genes (matK, rbcL, rpoB, and ITS) of both species were downloaded from the NCBI database to design species-specific primer sets. Subsequently, the genes were cloned using conventional PCR-amplicon from both plants and sequenced to verify the downloaded sequences. There was no difference in the sequences between the NCBI data and the cloned genes. Species-specific primer sets were designed based on various SNPs between the aligned sequences of nuclear DNA and chloroplast genomes of A. rugosa and P. cablin (Supplementary Fig. 1). Food processing may affect DNA quality as DNA may be broken down into short pieces during drying, heating, and blending (Lo and Shaw, 2018). Therefore, we developed species-specific primers to detect target species in commercial products by amplifying relatively short amplicons (109–180 bp; Table 1).

Table 1 Developed primer sets and established universal 18S rRNA primers for all plants
Full size table

Amplification efficiency of the designed primer sets

The sensitivity of six primers pairs (AR_matK, AR_rbcL, AR_ITS, PC_matK, PC_rbcL, and PC_rpoB) was evaluated by constructing a standard curve using tenfold serially diluted DNA extracted from leaves (10–0.001 ng/µL) and performing regression analyses. The correlation coefficients of the six primer sets exceeded 0.99, and the slopes were between − 3.36 and − 3.58. Also, the efficiency, according to the slope, was estimated to range from 90.13 to 98.52% (Fig. 1). According to the guidelines of the ENGL, qPCR assays were considered acceptable based on the linear dynamic range and amplification efficiency. The linear operating range should ideally extend to a four log10 concentration with a coefficient of correlation (R2) > 0.98, and the amplification efficiency should be in the range of 110%–90%, corresponding to a slope of − 3.1 to − 3.6 (ENGL, 2015). Our results corresponded to the guidelines with a coefficient of determination of ≥ 0.98 and a slope range of − 3.36 to − 3.58 (Fig. 1). To test assay reproducibility, different qPCR devices were used at different laboratories. The correlation coefficient exceeded 0.99, the slope was − 3.38 to − 3.58, and PCR efficiency was 90.16–97.69% (Supplementary Table 1). Therefore, amplification efficiency and reproducibility of the assay were considered acceptable.

Fig. 1
figure 1

A standard curve was obtained based on the efficiency and correlation of coefficient (R2) from genomic DNA serially diluted tenfold using primers targeting A. rugosa and P. cablin. The x-axis indicates the log-transformed DNA concentration (ng) and the y-axis represents means of Ct values ± SD. A Agastache rugosa targeting primer sets (AR_ITS, AR_,matK, and AR_rbcL). B Pogostemon cablin targeting primer sets (PC_rpoB, PC_matK, and PC_rbcL). Blue and yellow dots indicate A. rugosa and P. cablin, respectively. qPCRs were carried out in triplicates (n = 3)

Full size image

Applicability and specificity of the developed real-time PCR assay

In general, commercial products of medicinal herbs, including A. rugosa and P. cablin, can be processed by heat, dry, and autoclave treatments. Food processing can cause severe DNA degradation in herbs. For example, heat treatment is known to cause severe DNA fragmentation, rendering PCR-based analysis inefficient (Hwang et al., 2015). The slope, correlation coefficient and efficiency of the designed assay were investigated by drying, heating (100 °C), and autoclaving samples in a reference binary mixture to confirm assay applicability for commercial herbal medicine foods. Six standard curves of A. rugosa and P. cablin in non-treated, dried, heated, and autoclaved samples were produced (Table 2). All slopes ranged between − 3.10 and − 3.58, and the correlation coefficient was ≥ 0.99 for the six developed primer sets. The slope efficiencies were estimated to range from 90.13 to 109.63%. These results suggested that the developed markers can detect target species in processed food products. Subsequently, DNA extracted from dried, heated, and autoclaved leaf binary mixtures (0.1–100% w/w) were diluted to 10 ng/μL and were used for qPCR analysis. Ct values ​​were obtained for each sample and processing method (Supplementary Table 2). Binary mixtures using dried leaves showed higher Ct values than those of the untreated samples. However, higher Ct values ​​were observed in heated and autoclaved samples than in the dried samples. These results indicated less DNA degradation in dried than in heated and autoclaved samples, suggesting the use of different Ct values, ​​depending on the food processing method. In general, impurities of < 0.1% are not considered illegal mixtures for economic gain. Therefore, the target species Ct values of 0.1% of all binary mixtures with dried samples were used as the cut-off for detecting target DNA in commercial food products. The obtained cut-off Ct values ranged from 22.99 to 23.75 for the three developed primer sets targeting A. rugosa and 28.00 to 29.17 for the three sets targeting P. cablin (Supplementary Fig. 2).

Table 2 Evaluation of slope, R2, and efficiency obtained with three processing methods (dried, heated, and autoclaved) of a reference binary mixture using the developed primers
Full size table

Fourteen plant species were used to confirm primer specificity. In particular, Lamiaceae herbal medicine crops such as peppermint (Mentha × piperita), apple mint (M. suaveolens), and spearmint (M. spicata) were included because of their intended or non-intended use as spice ingredients in processed foods containing A. rugosa from P. cablin. The 18S rRNA primer set was used as a positive control, and qPCR amplification was observed with all tested species at < 20 cycles. qPCR analysis with the six developed primers was performed on 16 species including A. rugosa and P. cablin (Table 3). Except for the AR_ITS and PC_rbcL primers, no PCR products were amplified at < 40 cycles. The AR_ITS and PC_rbcL primers amplified Mentha species with the high Ct (22.96) and Lavandula stoechas with the high Ct (24.34) respectively, but less than 40 cycles. Taken together, all developed primer sets can be used to detect the target species, without cross-reactivity from other plant species.

Table 3 Specificity test results using other plants and the developed qPCR primer sets
Full size table

Application of the developed qPCR assay to blind samples

A blinded test was performed on 25 samples to estimate qPCR assay reliability. An independent study group randomly mixed 25 blind samples and did not provide any information on how much each A. rugosa and P. cablin powder was mixed. The 18S rRNA primer set used as a positive amplification control showed low Ct values ​​(11.97–13.34 cycles), indicating DNA quality was sufficient for evaluation (Table 4). Whether P. cablin powder was present in the A. rugosa samples was determined based on the cut-off Ct values of the developed primer sets (PC_rpoB, 29.17; PC_matK, 28.89; and PC_rbcL, 28.00). High Ct values of the P. cablin primer set were observed in seven samples, indicating that the P. cablin powder was not mixed with the A. rugosa powder in these cases. The Ct of the remaining 18 samples was lower than the cu-off value of the A. rugosa and P. cablin primer sets, indicating that the samples were mixed with A. rugosa and P. cablin powder. The results from 25 blind samples indicated that the qPCR assay can help confirm authenticity of A. rugosa and P. cablin products.

Table 4 Blind test results for measuring the reliability of the developed primer set
Full size table

Application of the developed assay to commercial herbal medicines

A total of 14 commercial products, including nine samples designated as A. rugosa and five designated as P. cablin, were purchased for authenticity tests (Table 5). DNA quality was evaluated using the 18S rRNA primer set as the positive control. The 18S rRNA primer set showed low Ct values (12.87–15.30), indicating sufficient DNA quality for examining the developed assay. Therefore, the commercial products were examined using the cut-off Ct values of the six primer sets (Table 5). A. rugosa products (samples 1–9) showed lower Ct values (14.29–15.95) than those of the A. rugosa-specific primers. However, they showed higher Ct values (29.63–35.79) than those of the P. cablin-specific primer sets. The nine examined A. rugosa products did not contain P. cablin. P. cablin products (samples 10–14) showed lower Ct values (16.19–18.62) than the cut-off Ct values of the P. cablin-specific primers and were amplified with higher Ct values (28.02–33.35) than the cut-off Ct values of the A. rugosa-specific primer sets. Therefore, the five tested P. cablin products appeared to not contain A. rugosa. These results suggest that the developed qPCR assay can be used to detect both the target species (A. rugosa and P. cablin) in commercial herbal medicine products.

Table 5 Results of qPCR analysis using 14 commercial herbs
Full size table

The qPCR assay developed here is a rapid and sensitive method for detecting target species in processed herbal medicines. To quantify the target species, a standard curve was constructed using binary mixture tests. Additionally, the specificity of the developed primer set was evaluated using 15 other species. Assay reliability was verified by blind testing and application to commercial products. The developed qPCR assay constitutes an effective identification method for distinguishing A. rugosa and P. cablin and may thus contribute to food safety and the protection of consumer rights.