HerbalEGram: Volume 10, Number 11, November 2013

By Stefan Gafner, PhD,^a Mark Blumenthal,^a Danica Harbaugh Reynaud, PhD,^b Steven Foster,^c Natascha Techen, PhD^d

^a  American Botanical Council, Austin, TX

^b  Authentechnologies LLC, Richmond, CA

^c  Steven Foster Group, Eureka Springs, AR

^d  National Center for Natural Products Research, University of Mississippi, Oxford, MS

Editor’s note: The American Botanical Council (ABC) previously has published several peer-reviewed articles in HerbalGram describing the valuable role that DNA-based analytical techniques can provide to quality control efforts in the botanical products industry. Consequently, we believed it was important to provide ABC members and others in the botanical research and products communities with this critique and commentary of a recently published paper that used DNA-based methods to analyze commercial herbal products.^1,2

The open-access journal BMC Medicine published an article online titled “DNA barcoding detects contamination and substitution in North American herbal products” on October 11, 2013.³ Steven Newmaster, PhD, associate professor in the Department of Integrative Biology, and his colleagues from the University of Guelph in Ontario, Canada, conducted research on deoxyribonucleic acid (DNA) barcoding of North American herbal products using two DNA regions sanctioned by the Consortium for the Barcode of Life* as core or supplemental barcodes to differentiate all plant species.

The study purports that “most of the herbal products tested were of poor quality, including considerable product substitution, contamination and use of fillers,” and that “some of the contaminants we found pose serious health risks to consumers.”³ The article eventually was picked up by many media outlets with headlines such as “Herbal product contamination 'considerable', DNA tests find,” “That herbal supplement may not be what you think it is, scientists find,” “Herbal supplements often contain unlisted ingredients,” and “Herbal supplements may contain toxic contaminants,” respectively.^4-7 Another article, “Herbal supplements are often not what they seem,” was published online by the New York Times on November 3.⁸

The study’s results are based on a method that uses plant DNA barcodes (barcodes are short genetic markers in an organism's DNA to identify it as belonging to a particular species) as an unbiased reproducible method of species identification. Since DNA collects mutations that are subsequently inherited throughout the process of biological evolution, these genetic markers can be used to determine the exact plant species and the relationship among species. In the present study, the authors used the rbcL and ITS2 regions to obtain DNA barcodes, which were then compared to the barcodes of what the authors refer to as their own “standard reference material (SRM)” herbal barcode library to identify the plant species. (This is one of numerous anomalies in the paper. Standard Reference Material^® is a trademark registered with the United States Patent and Trademark Office by the US National Institute of Standards and Technology (NIST). It is a certified reference material issued by NIST that also meets additional NIST-specific certification criteria and is issued with a certificate or certificate of analysis that reports the results of its characterizations and provides information regarding the appropriate use(s) of the material. A knowledgeable peer reviewer should have detected this error in the Newmaster et al. paper.)

To be clear, this review and critique is not a criticism of the applicability of DNA as an appropriate form of botanical identification, which the authors of this critique consider reliable when properly executed.

The paper by Newmaster et al. is problematic on a number of levels. The approach taken by the authors was to create a number of homemade definitions and then evaluate materials against those definitions using DNA fingerprinting. This would perhaps be acceptable if a review of the list of cited references had not indicated that the authors are apparently deeply unaware of herbal quality-assurance procedures and programs and the nature of commercial botanical products. There are internationally recognized definitions for identity, authenticity, contamination, and substitution. Invention of new definitions for these terms by the authors in order to demonstrate the novelty of their approach and their technical virtuosity is self-referential and, unfortunately, very possibly self-serving. Their apparent lack of adequate knowledge in this field has allowed them to create a virtual problem and then, figuratively, ride to the rescue and solve it.

While there are several clinical trial review articles and many DNA papers cited in the publication, there are no publications at all from recent and/or the past century’s phytochemistry or pharmacognosy literature, and there seems to be no awareness at all that legally binding compendial standards (including specifications and tests for identity and quality, as noted in various officially recognized pharmacopeias) have existed for the better part of two centuries. Included in these monographs are chemical and physical tests for identification of crude botanicals, powdered botanicals, and botanical extracts. In addition to the individual plant monographs found in the official Pharmacopoeia of the People’s Republic of China, the official Japanese Pharmacopoeia, the official Ayurvedic Pharmacopoeia, the official United States Pharmacopeia (USP), the official European Pharmacopoeia (EP), and the unofficial but highly regarded USP Dietary Supplements Compendium (USP-DSC) and American Herbal Pharmacopoeia (AHP), each compendium provides several chapters that describe the nomenclature of different forms of herbs in commerce (extracts, native extracts, standardized extracts, etc.).

Many herbal materials in commerce are extracts, or extracts that are spray-dried onto carriers (such as rice powder) rather than simply dried powdered plant material. These dried extracts are frequently sold in the form of capsules and/or tablets. Since the authors did not provide label information for any of the commercial herbal products they tested, it is not possible to adequately evaluate their claims that the fillers they found were ‘unlabeled.’ In addition, heat or other preservation technologies applied to herbal materials can denature DNA while maintaining the natural phytochemical profile of an herb. Despite the authors' contention to the contrary, there is no evidence that DNA can be obtained from botanical extracts, and DNA from relatively highly processed materials such as finished supplements in tablets and capsules is of poor quality. As a result, the DNA that is detected in these cases is typically either accidental environmental contamination, cross-contamination among samples in the lab, and/or the DNA from the product’s carrier or filler (soy, potato, or rice). Only if the proper controls are used — including negative controls for the extraction and polymerase chain reaction (PCR) — can accurate data be generated.

While some of the authors’ findings on adulterated botanical products may indeed seem very troublesome, like the substitutions of cinnamon (Cinnamomum spp., Lauraceae) with plant material from a member of the Amaryllis family (Amaryllidaceae), burdock (Arctium lappa, Asteraceae) with a member of the buttercup family (Ranunculaceae), or ginkgo (Ginkgo biloba, Ginkgoaceae) with black walnut (Juglans nigra, Juglandaceae), the absence of DNA from the plant species indicated on the product label may be explained due to extensive processing of said material before it was packaged for consumer use. Furthermore, the presence of the adulterants can, in some cases, be explained by cross-contamination from other samples or the lack of specificity in the regions analyzed. The authors’ discussion failed entirely to address these potential confounders.

The authors provide no quantitative data for their results. We, the authors of this critique, understand that accurate quantification would be challenging, if not impossible, since the ITS2 region, which was used as a barcode by Newmaster et al., is present in up to several thousands of identical copies in plant cells — some plants having more copies, some having less. The more copies that are present, the higher the chance they are amplified by PCR. Therefore, it would be misleading to conclude that if more ITS2 PCR product of one species is detected, then more of this plant material is present. Also, the length of the PCR product from authentic plant material versus an adulterant is important. In the same PCR reaction mix, a shorter fragment (e.g., from the adulterant) is preferably amplified over a larger fragment, which also could be misinterpreted to mean that the sample has more adulterant material than authentic material.

We thus are at a loss to know whether or not non-target DNA found in a product is present at levels that would constitute a significant amount of extraneous material or perhaps a few dandelion (Taraxacum officinale, Asteraceae) leaves commingled with a hectare’s worth of harvested crop. Similarly, PCR amplification could well be capable of picking up residual rice DNA left on a piece of processing equipment from the previous production run. Other principles, such as a systematic approach to documenting botanical collections and evaluating the sensitivity, accuracy, and precision of the identification methods the study’s authors used, also are lacking, although such approaches are readily available in the scientific literature.^9-11

The following are various points or actual statements made by the study’s authors, and the American Botanical Council’s clarification and/or criticism of them:

1. In the background information, the authors write:
 

[T]here are currently no best practices in place for identifying the species of the various ingredients used in herbal products. This is because the diagnostic morphological features of the plants on which the current Linnaean taxonomic system is based cannot typically be assessed from powdered or otherwise processed biomaterials. As a result, the marketplace is prone to contamination and possible product substitution, which dilute the effectiveness of otherwise useful remedies, lowering the perceived value of all related products because of a lack of consumer confidence in them.³

The authors continue that “currently, there are no standards for authentication of herbal products." 

5. Similarly, according to Table 1 in the article, from the “DNA barcode results listed for individual samples from blind testing of the 44 herbal products and 50 herbal leaf samples representing 42 medicinal species of plants,” one of the St. John’s wort (Hypericum perforatum, Hypericaceae) products was allegedly substituted with a member of the bean family (Fabaceae). Apparently, the species could not be identified since “some of the DNA sequences recovered were only identified to family as our SRM [sic] herbal DNA barcode library is not yet complete.” However, in the discussion, the authors state that “one product labeled as St. John’s wort was substituted with Senna alexandrina (Fabaceae): it contained only Senna barcodes and no St. John’s wort barcodes. This is a serious health risk as Senna is a Food and Drug Administration (FDA) approved non-prescription herbal laxative, which is not for prolonged use.” So which is true? Why was S. alexandrina not listed as adulterant in Table 1 if the authors were able to identify the material to the species level? Or are the statements in the discussion based on speculation?

6. Furthermore, in Table 1, the authors used the scientific name "Plantago ovate." Presumably, they are referring to P. ovata (Plantaginaceae), and their spellchecker automatically tried to correct “ovata” to “ovate,” an error that the authors, and, presumably, peer reviewers, should have detected. Even though this might seem trivial, and, in some cases, it might be, it implies that either the authors and/or the peer reviewers are not adequately familiar with the scientific botanical nomenclature, and presumably, other aspects of botany and medicinal plants that could have informed a more cohesive and reliable publication.

It is not clear from the article which of the scientific names listed in Table 1 were actually listed on the labels of the products they tested, and which might be assumptions of species present if the label provided only a common name. For example, Sonchus arvensis (Asteraceae) is listed associated with the common name “Thistle.” This is a sow thistle species so it could be present in a field-harvested herbal product as a weedy contaminant but the table indicates that two samples (HP67C and HP87C) were commercially available products. However, a search of the online National Institutes of Health Dietary Supplement Label Database for U.S. products and the Licensed Natural Health Products Database for Canadian products did not find a single product with any species of Sonchus as an ingredient. Searching these databases with the term “thistle” returned results for Blessed Thistle (Cnicus benedictus, Asteraceae) and Milk Thistle (Silybum marianum, Asteraceae), and a Google search turned only one seller of S. arvensis raw material up located in Indonesia, but no dietary supplement containing this material. Table 1 would be much more useful if it were revised to avoid these ambiguities.