Authentication of Brazilian Ginseng using Bar-HRM analysis

Abstract Hebanthe eriantha (Martius) Kuntze and Pfaffia glomerata (Spreng) Pedersen are medicinal plants popularly known as “Brazilian Ginseng” due to their similarity to Panax ginseng. In Brazil, they are sold as the same herb, despite their different pharmacological and toxicological properties. The morphological identification is difficult, which facilitates their adulteration. We report the application of the Barcode DNA High-Resolution Melting (Bar-HRM) using matK gene to differentiate both species in samples sold in the Brazilian market. Using the proposed method, we could discriminate and identify both species. Bar-HRM analysis allowed discriminating and identifying both species. It allowed the identification of H. eriantha and P. glomerata in 43.6% and 56.4% of the amplified samples, respectively. Of these, only seven samples were authenticated and, in 71.4% of the cases, adulterated. We concluded that Bar-HRM has proven to be a fast alternative method to authenticate plants under the common name “Brazilian Ginseng”.

Some studies on the medicinal use of H. eriantha and P. glomerata do not distinguish these plants, despite the presence of some chemicals that lead to differentiated pharmacological and toxicological properties.Their morphological identification is difficult due to their similarity, which can lead to the adulteration of products (Oliveira, 1986;Vigo et al., 2004).Besides, "Brazilian Ginseng" products are usually purchased, processed and unlabeled or unpackaged, which makes their identification even more challenging.
Several DNA regions from the chloroplast genome have enough variation to be used as DNA barcodes to identify plant species (Hollingsworth, 2011).The maturase K (matK) gene is one of the high-level discriminatory regions (Yu et al., 2011).Analysis of this region is usually performed by sequencing method, which is expensive to be largely used.To overcome this problem, the High-Resolution Melting (HRM) has been recently proposed to exploit nucleotide polymorphisms as simple, cost-effective, and fast alternative.This analysis has already proven to be a reliable molecular approach for species identification and authentication (Osathanunkul et al., 2015;Costa et al., 2016;Osathanunkul, Madesis, 2019).Given this context, we aimed to evaluate if the Bar-HRM is useful for discriminating H. eriantha and P. glomerata.

Plant materials
Six voucher specimes (three specimens of P. glomerata and three of H. eriantha) were morphologically identified by a specialist and used as reference species (Table I).In total, sixty commercial samples, sold as "Brazilian Ginseng," were bought in physical or online markets from several Brazilian cities (Table IS).DNA extraction was performed with using NucleoSpin® Plant II kit (Machereley-Nagel, Germany) following the manufacturer's protocol, using 30 mg of roots.DNA was amplified using matK-1KIM-F 5'-ACCCAGTCCATCTGGAAATCTTGGTTC-3' and matK-3KIM-R 5'-CGTACAGTACTTTTGTG TTTACGAG-3' primers.PCR reactions was carried out with 1x PCR Buffer (Invitrogen), 1.5 nM of MgCl 2 (Invitrogen), 1,25pM of each primer, 0.25 mM of dNTP (Promega) and 1U of Platinum™ Taq DNA Polymerase (Invitrogen), 30 ng of DNA and ultrapure lab grade water up to final 12.25 µL.PCR was achieved under the following conditions: 94 ºC for 1 min, then 35 cycles of 95 ºC for 30 s, 52 ºC for 20 s, 72 ºC for 1min, with a final extension at 72 ºC for 5 min and 4 ºC.Cycle sequencing of both strands was conducted using BigDye ® Terminator v3.1 (Thermo Fischer Scientific) in an ABI PRISM ® 3500 Genetic Analyzer.Electropherograms were checked using BioEdit software v.7.0 (Hall, 1999).DNA sequencing results were compared with HRM analysis.

Bar-HRM design, analysis, and validation
Chloroplast matK gene sequence was generated from reference sequences of both species and aligned using BioEdit Sequence Alignment Editor version 7.2 (Hall, 1999).Primers were made manually by flanking polymorphic regions, and the annealing temperature and possible primer-dimer were checked using Perlprimer version 1.21 (Marshall, 2004).The uMelt-DNA Melting  Bar-HRM with pre-amplification was performed with a LightCycler® 96 (ROCHE) real-time PCR system.PCR amplifications were performed using EvaGreen® supermix SsoFast™ PCR, in a total volume of 10 µL, containing 5 µL of the supermix; 500nM of each primer, 5'-YTTCTTGAACGAATMYATTTCTAY-3'F and 5'-ACCTAACATAATGCRKGAAG-3'R (Table II); and 1 µL of ultrapure water to complete the final volume.Three concentrations of DNA amount (0.14, 1.4, and 14 ng) were tested for each one of the reference species to check the better condition of amplification, following the manufacturer's instructions.The concentration of 14 ng DNA was used for all run analysis.
The PCR temperature program included a denaturation step at 95°C for 2 min; followed by 40 cycles at 95°C for 30 s; 56°C for 30 s; and 72°C for 30 s. Melting analysis was performed at 95°C for 1 min, followed by 40°C for 1 min, then 65°C until 97°C, increasing at 0,07 °C/sec with 15 acquisitions of dye per grade.Reactions were performed in triplicate for each sample, including a negative control.Data were analyzed using LightCycler®96 SW v.1.1 (Roche Diagnostics, Risch-Rotkreuz, Switzerland).Genotypes were identified by examining normalized melting curves, difference, and derivative plots of the melting data.Melting temperature (Tm) data were statistically analyzed using Microsoft Excel 2010 to the calculation of standard deviation and confidence interval for each sample run in triplicate (Table IIS).

RESULTS AND DISCUSSION
Bar-HRM performed with the specific matK primer successfully differentiated and identified both species commonly known as "Brazilian Ginseng".The melting profile allowed differentiation between P. glomerata and H. eriantha based on three SNPs, two transversions, T to G and G to C, and one transitions, T to C (Figure1).Our results from the Bar-HRM analysis were consistent when compared with DNA sequencing (Table IIS), as previously reported by Jilberto et al. (2017).The sensitivity of our reactions ranged between 92-93%, indicating a high rate of individuals identified by the proposed method.This means that of all samples analyzed, 92% of those identified by barcode sequencing as P. glomerata also amplified and belonged to the correct cluster when analyzed by the HRM method.And the same was true for 93% of the samples identified as H. eriantha by barcode sequencing.
The chosen primer also had 100% specificity for both species, showing that the test has a maximum performance to exclude the individuals of a particular species correctly.Due to the high specificity value found for the species, the positive likelihood ratio (LR+) was undefined, indicating a high probability that the individual belong to the species identified when the HRM analysis positively identifies a specific species.The values of the negative likelihood ratio (LR-) varied between 7-8%, indicating a high probability that the individual did not belong to the species when the HRM resulted in a negative value for a specific species.The accuracy was 100% for the HRM-matK primer, showing that the analysis did not have false positives.
In total, 43.6% of amplified samples were identified as H. eriantha and 56.4% as P. glomerata using Bar-HRM.Of these, only seven samples had the species listed on the label and, in 71.4% of the cases, were adulterated (they contained the wrong species).
Bar-HRM stands out as a fast process.After only 2 h of amplification and subsequent dissociation (about 2 minutes), the results are available.By promoting a DNA amplification of short duration, HRM analyzes small DNA fragments, which is beneficial for degraded material that is usually present in powder and dry plants (Kool et al., 2012;Särkinen et al., 2012).This method has also been applied with success to differentiate other plant species (Ganopoulos et al., 2012;Singtonat, Osathanunkul, 2015;Song et al., 2016).
We showed that the Bar-HRM was useful to identify both species, H. eriantha and P. glomerata, popularly grouped under the name "Brazilian Ginseng".Since this study is faster and cheaper than DNA sequencing, its use becomes easier in a wide variety of samples.HRM may be included as a quality control protocol for raw materials of plants and promote consumer confidence.However, its use still requires regulation in each country.Based on the conventional melting analysis, two groups of melt peaks could be observed: 75.0ºC and 75.5ºC for P. glomerata and H. eriantha, respectively (Figure 2A).The normalized melting curves from HRM analysis allowed the classification of the samples into two clusters, differentiating the species (Figure 2B).

FIGURE 1 -
FIGURE 1 -Sequence alignment of the matK gene region showing nucleotide differences between species (red box) and primer design regions (red arrow).

FIGURE 2 -
FIGURE 2 -A)Melting peaks of standard curves and B)normalized melting curves of Hebanthe eriantha and Pfaffia glomerata using Bar-HRM.

TABLE I -
Reference samples of Hebanthe eriantha and Pfaffia glomerata used in this study

TABLE I -
Reference samples of Hebanthe eriantha and Pfaffia glomerata used in this study

TABLE II -
matK partial gene sequences of Pfaffia glomerata and Hebanthe eriantha amplified by HRM primers in this study

TABLE IS -
Commercial samples sold as "Brazilian Ginseng" used in this study

TABLE IIS -
Application of the developed method on commercial samples

TABLE IIS -
Application of the developed method on commercial samples

TABLE IIS -
Application of the developed method on commercial samples Note: *Sample with mixture of other unrelated species.NRC (No Reference Cluster).