HPLC-UV, Metabarcoding and Genome Skims of Botanical Dietary Supplements: A Case Study in Echinacea

Abstract

The use of DNA-based methods to authenticate botanical dietary supplements has been vigorously debated for a variety of reasons. More comparisons of DNA-based and chemical methods are needed, and concordant evaluation of orthogonal approaches on the same products will provide data to better understand the strengths and weaknesses of both approaches. The overall application of DNA-based methods is already firmly integrated into a wide array of continually modernizing stand alone and complementary authentication protocols. Recently, the use of full-length chloroplast genome sequences provided enhanced discriminatory capacity for closely related species of Echinacea compared to traditional DNA barcoding approaches (matK and rbcL). Here, two next-generation sequencing approaches were used: (1) genome skimming and (2) PCR amplicon (metabarcoding). The two genetic approaches were then combined with HPLC-UV to evaluate 20 commercially available dietary supplements of Echinacea representing “finished” products. The trade-offs involved in different DNA approaches were discussed, with a focus on how DNA methods support existing, accepted chemical methods. In most of the products (19/20), HPLC-UV suggested the presence of Echinacea spp. While metabarcoding was not useful with this genus and instead only resolved 7 products to the family level, genome skimming was able to resolve to species (9) or genus (1) with the 10/20 products where it was successful. Additional ingredients that HPLC-UV was unable to identify were also found in four products along with the relative sequence proportion of the constituents. Additionally, genome skimming was able to identify one product that was a different Echinacea species entirely.

Utilizing Big Data to Identify Tiny Toxic Components: Digitalis

“…We simulated mixed samples based on practical data limits (~2 Gb per sample, equivalent to a MiSeq 600-cycle kit multiplexed eight ways) and at this depth unambiguous genus-level Digitalis detection was possible even in trace amounts when Digitalis data made up only 0.05% of the mixture (~1 Mb of spike-in data) (Figure 2B). Additionally, we see a highly-correlated increase in genus-level SNP recovery with respect to the amount of Digitalis data spiked into the mixture (Figure 2A), suggesting that this method is semiquantitative, supporting previous work [21]. Relative to the genus-informative SNP sets, there were fewer species-informative SNPs, and robust identification of species required around 10 times more data (0.5% of the mixture, or just under 10Mb) (Figure 3).…”

“…Metabarcoding approaches often rely on well-curated reference databases, but due to the limited size of many of these datasets, care must be taken to choose marker regions that both (1) provide appropriate taxonomic discrimination while (2) also accounting for issues such as amplification errors, bias, and artifacts introduced by PCR. Additionally, metabarcoding approaches often preclude quantification beyond relative abundance, and relying on highly targeted techniques can also fail due to fragmented DNA [21], which is often found in highly processed food products and dietary supplements.…”

“…On the other end of the spectrum, whole-genome sequencing (WGS) approaches, including low-coverage genome skimming, circumvent many of the limitations associated with traditional metabarcoding or targeted locus methods. WGS data provides an untargeted snapshot of DNA in a sample; it bypasses the need for a priori marker selection, reduces negative biases associated with PCR amplification, and provides a more holistic, semiquantitative (with caveats), representation of sample composition [21]. Especially in the absence of a reference genome, analysis of WGS data typically relies on kmer binning [45][46][47], BLAST [48], or metagenomic assembly and classification of the resulting contigs using a database [49,50].…”

“…The world of eukaryotic DNA-based species identification in food and dietary supplements products has undergone a revolution in the past 10 years, and while progress has been swift it has not been without challenges [8][9][10]. Contemporary methods of DNAbased identification include DNA fingerprinting techniques like restriction fragment length polymorphisms [11], microarrays [12], a full suite of polymerase chain reaction (PCR) methods, [13][14][15][16][17], in addition to traditional DNA barcoding [9,18] and a variety of next generation sequencing (NGS) approaches [19][20][21].…”

“…NGS methods have recently been investigated to distinguish between closely related eukaryotic species [21,22] and to detect food origins [23], the presence of toxic species [24], allergens [25], as well as other cases of adulteration [20,26,27], both accidental and economically motivated. However, the application of NGS technology is not without limitations.…”