Next-Generation
More Successful than Sanger Sequencing in DNA Identification of Herbal Raw Materials
Reviewed: Ivanova NV,
Kuzmina ML, Braukmann TW, Borisenko AV, Zakharov EV. Authentication of herbal supplements using next-generation sequencing. PLoS One. 2016;11(5):e0156426. doi:
10.1371/journal.pone.0156426.
Keywords: DNA barcoding, Sanger sequencing, next-generation
sequencing, echinacea, Echinacea angustifolia,
ginkgo, Ginkgo biloba, St. John’s wort, Hypericum perforatum, fenugreek, Trigonella
foenum-graecum, valerian, Valeriana officinalis
DNA-based authentication of herbal
ingredients and finished products has become more prominent in recent years.
While genetic methods for quality control have been adapted by a number of
dietary supplement manufacturing companies, questions about the standardization
of DNA extraction, amplification, and sequencing have not been resolved. In addition,
rapidly evolving improvements in instrumentation have led to a situation in
which the selection of the most appropriate genetic method has been a matter of
much scientific disagreement.
In this study, the identification success
of two sequencing methods, Sanger sequencing and next-generation sequencing
(NGS), were compared using 15 dietary supplements made with ingredients from
five medicinal plants—Echinacea purpurea
(Asteraceae), Valeriana officinalis (Caprifoliaceae),
Ginkgo biloba (Ginkgoaceae), Hypericum perforatum (Hypericaceae), and Trigonella foenum-graecum (Fabaceae). Sanger sequencing
generally amplifies DNA from only one species, carrying the risk that
non-target DNA (e.g., DNA from a small amount of foreign organic matter in a
dietary ingredient or supplement) may be preferentially amplified, in
particular when the target DNA is degraded. NGS has enabled the amplification
of DNA from multiple species, e.g., those in a dietary supplement containing an
ingredient mixture, regardless of the relative amounts of DNA present.
For Sanger sequencing, two genomic regions
(rbcL and ITS2) were amplified, while
only ITS2 was used with NGS. After optimization of the DNA extraction protocol,
the 15 samples (three samples of each species including powdered crude raw
material [n=8] and extracts [n=7]) were amplified and then sequenced using the
two methods. A fragment of the caterpillar Crinodes ritsemae
COI gene was used to verify amplification success (to exclude the possibility
of interference, e.g., from plant secondary metabolites), and multiple negative
controls (i.e., samples where the protocol was carried out without adding the
dietary supplement ingredients) were included to check for laboratory
contamination.
Overall, DNA of the labeled species was
detected in seven (46.7%), four (26.7%), and eight (53.3%) samples using Sanger
(rbcL), Sanger (ITS2), and NGS,
respectively. Not surprisingly, powdered crude raw material yielded higher
sequencing success than extracts, with Sanger (rbcL)
being successful in seven (87.5%) samples, and NGS in five (62.5%) samples out
of eight. Sequencing of the ITS2 genomic region in the valerian root samples
was particularly challenging due to the high intraspecific variation (>15%)
and the presence of multiple nucleotide (nucleotides, or “bases,” are the DNA
building blocks) insertions/deletions. The results also showed that Sanger
sequencing led to preferential amplification of excipient DNA in two products,
one made with echinacea and one made from ginkgo. All the samples processed
using NGS contained DNA from fungal species and other non-labeled plant
species. Presence of DNA from non-labeled plant species was lowest in ginkgo
extracts, confirming that the extensive purification during the manufacturing
of ginkgo extracts will lead to DNA degradation or loss.
No ginkgo DNA was detected in the two
supplements that were labeled to contain ginkgo extracts. Both samples were
also evaluated by chemical analysis using high-performance liquid
chromatography–mass spectrometry (HPLC-MS), which confirmed the presence of the
characteristic ginkgo terpene lactones and the flavonoids isoquercitrin,
quercitrin, rutin, quercetin, kaempferol, and isorhamnetin.
Comment: The study provides data that
highlight some of the limitations of DNA-based authentication methods. It is
also one of the few papers where DNA results are at least partially verified by
chemical means. The data confirmed that highly processed ingredients have a low
likelihood to be correctly identified by DNA sequencing methods, and that
successful identification depends on the ingredient in question and selection
of the target region, extraction, amplification, and sequencing methods. Even
with crude raw material, accurate results were obtained in only 50-87.5% of the
samples. The best results were obtained using ILB lysis buffer for the
extraction, and NGS as the sequencing method. Preferential amplification of
non-target DNA, e.g., DNA from excipients or fungal species, is a particular
concern when using Sanger’s method of sequencing DNA. This is highly
problematic for a quality control method since it can lead to false positive
results. Based on these results, authentication of botanical raw materials using
DNA sequencing methods should be carried out only on crude raw material, and
results need to be verified by other appropriate methods, such as taxonomic,
microscopic, and/or chemical means.
Concerns about the interpretation of
results obtained using NGS were voiced by Steven Newmaster, PhD, Professor of
Botany at the University of Guelph, et al., in a paper posted on the website of
the American Herbal Products Association.1 The ubiquitous presence
of DNA from other plant species (foreign organic matter) in a botanical
material, which is generally permissible at levels up to 2% according to the
European and United States pharmacopeias, should always be considered when
using genetic methods, and the amounts of such “incidental DNA” can be
overestimated using NGS. Based on results obtained for experimental mixtures of
botanically authenticated ingredients that were analyzed using NGS in several established
and internationally recognized labs, NGS often missed species and/or detected
species which did not match the experimental composition of the ingredients.
Furthermore, the quantitative estimates from NGS mixtures did not correspond to
any of the experimental mixtures. As such, the authors believe that “there are
considerable problems with NGS that present an immediate impediment to
generating scientifically valid test results….”1
Reference
1. Newmaster S, Ragupathy S, Hanner R. A caution to industry and regulators
– “incidental DNA fragments” may be misinterpreted using next generation
sequencing (NGS). American Herbal Products Association website. Available at: http://ahpa.org/Portals/0/PDFs/Incidental_DNA_fragments_misinterpreted_Next_Generation_Sequencing_NGS.pdf. Published 2016. Accessed December 13, 2016.
