513
NOTE Authentication of Ginkgo biloba herbal dietary supplements using DNA barcoding Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
Damon P. Little
Abstract: Ginkgo biloba L. (known as ginkgo or maidenhair tree) is a phylogenetically isolated, charismatic, gymnosperm tree. Herbal dietary supplements, prepared from G. biloba leaves, are consumed to boost cognitive capacity via improved blood perfusion and mitochondrial function. A novel DNA mini-barcode assay was designed and validated for the authentication of G. biloba in herbal dietary supplements (n = 22; sensitivity = 1.00, 95% CI = 0.59–1.00; specificity = 1.00, 95% CI = 0.64–1.00). This assay was further used to estimate the frequency of mislabeled ginkgo herbal dietary supplements on the market in the United States of America: DNA amenable to PCR could not be extracted from three (7.5%) of the 40 supplements sampled, 31 of 37 (83.8%) assayable supplements contained identifiable G. biloba DNA, and six supplements (16.2%) contained fillers without any detectable G. biloba DNA. It is hoped that this assay will be used by supplement manufacturers to ensure that their supplements contain G. biloba. Key words: digital PCR, matK, mini-barcode. Résumé : Le Ginkgo biloba L. (connu sous le nom ginkgo’ ou arbre aux quarante écus’) est une gymnosperme fascinante et isolée sur le plan phylogénétique. Des suppléments alimentaires naturels, préparés a` partir des feuilles du G. biloba, sont consommés pour accroître les capacités cognitives via une meilleure perfusion sanguine et une fonction mitochondriale améliorée. Un nouveau test d’ADN a été conçu et validé pour l’authentification du G. biloba dans les suppléments alimentaires (n = 22; sensibilité = 1,00, IC 95 % = 0,59–1,00; spécificité = 1,00, IC 95 % = 0,64–1,00). Ce test a ensuite été employé pour mesurer la fréquence de suppléments alimentaires faussement étiquetés comme étant a` base de ginkgo aux États-Unis. Un ADN propice a` l’amplification PCR n’a pu être obtenu pour trois (7,5 %) des 40 suppléments examinés. Pour 31 des 37 suppléments testés (83,8 %), de l’ADN du G. biloba a été détecté, tandis que les six autres suppléments (16,2 %) contenaient des agents de remplissage sans aucun ADN détectable de G. biloba. Les auteurs espèrent que les manufacturiers de suppléments utiliseront ce test pour assurer que leurs suppléments contiennent du G. biloba. [Traduit par la Rédaction] Mots-clés : PCR digital, matK, mini-code-barre.
Introduction Ginkgo biloba L.—referred to as ginkgo or maidenhair tree in English vernacular—is a phylogenetically isolated (Lee et al. 2011; Wu et al. 2013), charismatic, gymnosperm tree. Although it was once widespread throughout the northern hemisphere (Traula 1967; Zhou 2009), G. biloba was, until recently, presumed extinct in the wild (reviewed in Del Tredici et al. 1992). Incontrovertible evidence of relic wild populations in southwest China has recently been amassed (Tang et al. 2012). Over the last two millennia, a subset of the genetic diversity found in wild G. biloba populations has been cultivated in China, Korea, and Japan (Zhao et al. 2010). Cultivated G. biloba in Europe and North America were propagated from Korean and Japanese trees in the early 18th century (Zhao et al. 2010). Humans ingest G. biloba herbal dietary supplements to boost cognitive capacity via improved blood perfusion and mitochondrial function (Diamond and Bailey 2013). Standardized leaf extract of G. biloba contains two major pharmacologically active fractions: flavonoids (mainly glycosides of isorhamnetin, kaempferol, and quercetin) and terpene trilactones (bilobalide and ginkgolide compounds; Abdel-Kader et al. 2007). Working in concert, these compounds protect and enhance mitochondrial function via direct stabilizing effects (Eckert et al. 2003; Abdel-Kader et al.
2007) and by scavenging reactive molecules such as hydroxyl radicals (Pincemail and Deby 1986), nitric oxide (Marcocci et al. 1994), peroxyl radicals (Maitra et al. 1995), and superoxide anions (Pincemail et al. 1989). Impaired mitochondrial function is thought to be a leading cause of cognitive decline (reviewed in Hauptmann et al. 2006). Hence G. biloba is marketed as a cognitive enhancer. Meta-analyses of clinical trial data show that G. biloba decreases blood viscosity and increases blood perfusion (Kellermann and Kloft 2011); that G. biloba is an effective adjunctive therapy for chronic schizophrenia patients (Singh et al. 2010); that cognitive function significantly improves in dementia patients after G. biloba treatment (Wang et al. 2010; Weinmann et al. 2010); and that G. biloba does not significantly improve concentration, memory, or executive function in healthy individuals (Laws et al. 2012). Few minor adverse events are reported from clinical trials that use standardized G. biloba leaf extract (e.g., gastro-intestinal complaints, headache, nausea; Singh et al. 2010; Weinmann et al. 2010; Kellermann and Kloft 2011; Diamond and Bailey 2013; Nicolaï et al. 2013). Sales of G. biloba herbal supplements, in the United States of America, totaled more than US$25.8 million during 2012 (the most
Received 28 August 2014. Accepted 3 November 2014. Corresponding Editor: P. Gulick. D.P. Little. Lewis B. and Dorothy Cullman Program for Molecular Systematics, The New York Botanical Garden, Bronx, NY 10458-5126, USA. E-mail for correspondence: [email protected]. Genome 57: 513–516 (2014) dx.doi.org/10.1139/gen-2014-0130
Published at www.nrcresearchpress.com/gen on 11 December 2014.
514
recent data available; Lindstrom et al. 2012)—making G. biloba the fifth best-selling herbal supplement. The purpose of this study is to validate a novel DNA barcode assay for the authentication of G. biloba in herbal dietary supplements and to use this novel assay to estimate the frequency of mislabeled ginkgo supplements on the market in the United States of America.
Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
Materials and methods A barcode reference database was constructed from all gymnosperm matK sequences deposited in GenBank (8 August 2013). Sequences were trimmed to the barcoding region—defined by priming sites Gym_F1A and Gym_R1A (Li et al. 2011). Sequences that did not span the entire barcoding region, those that contained ambiguous nucleotides, and those with inferred stop codons were discarded (supplementary data, Table S11). Nucleotide sequences were aligned as amino acids using translatorX 1.1 (Abascal et al. 2010) and KALIGN 2.04 (Lassmann and Sonnhammer 2005). To ensure more reliable PCR amplification from degraded DNA samples, a 166-bp matK mini-barcode was designed using PRIMER3 1.1.4 (Koressaar and Remm 2007; Fig. 1). To validate the assay, morphologically identifiable G. biloba (n = 10) and related gymnosperms (n = 12) were collected. Leaf tissue was preserved in silica gel. Voucher specimens were deposited in the Steere Herbarium (NY; supplementary data, Table S2). After validation, the assay was applied to herbal supplements purchased in New York area retail stores and on-line. The supplements were either dried and powdered leaves of G. biloba (n = 8) or G. biloba leaf extract dried on an “inert” filler (e.g., rice, soy, etc.; n = 32). Samples (10 mg) of herbal supplements or silica dried leaf tissue were disrupted in a 1.6 mL tube using a Qiagen TissueLyser (30 Hz for 2 × 1.5 min) and two stainless steel ball bearings (3 mm). Samples were incubated with 50 rpm horizontal shaking at 42 °C in 600 L extraction buffer (8 mmol/L NaCl, 16 mmol/L sucrose, 5.8 mmol/L EDTA, 0.5% (w/v) sodium dodecyl sulphate, 12.4 mmol/L tris pH 9.1, 200 g/mL proteinase K; Alexander et al. 2007). Following 18 h of incubation, 200 L of 3 mol/L potassium acetate (pH 4.7) was added to each sample and the samples were cooled to 0 °C. After 10 min, samples were centrifuged at 14 000g for 5 min. A 900 L sample of 2 mol/L guanidine hydrochloride in 95% (v/v) ethanol was mixed with 600 L of each sample’s aqueous phase. The mixtures were applied, 500 L at a time, to silica spin columns (Epoch Life Science). After each application, columns were centrifuged at 7000g for 1 min. Each sample was washed twice with 500 L of buffer (50% (v/v) ethanol, 10 mmol/L tris pH 7.4, 0.5 mmol/L EDTA, 50 mmol/L NaCl; Elphinstone et al. 2003). Ethanol residue was removed by centrifugation at 7000g for 2 min. Total DNA was eluted, by centrifugation (7000g for 1 min), in 2 × 100 L 10 mmol/L tris (pH 8.0). The matK mini-barcode was PCR amplified in a 15 L volume containing the following: 20 mmol/L tris pH 8.8, 10 mmol/L KCl, 10 mmol/L (NH4)2SO4, 2 mmol/L MgSO4, 0.1% (v/v) Triton X-100, 5% (w/v) sucrose, 0.025% (w/v) cresol red, 0.025 g/L BSA, 0.2 mmol/L dNTPs, 0.5 mol/L of each primer (5=-TTTCAGGGCGTCCAATTAGT-3= and 5=-TTAGCACATGGAAGTCGGAGT-3=), 0.5 U Taq polymerase, and 0.5 L DNA. Reactions were incubated for 150 s at 95 °C, cycled 35 times (30 s at 95 °C, 120 s at 58 °C), and then incubated 10 min at 60 °C. Residual primers and dNTPs were neutralized with ExoSAP-IT (Affymetrix) and bidirectionally sequenced using the amplification primers and BigDye v3.1 (1 L neutralized PCR product in each 5 L reaction; Life Technologies) on a 3730 automated se-
1
Genome Vol. 57, 2014
Fig. 1. Nucleotides diagnostic of Ginkgo biloba within the matK mini-barcode. Positions are numbered vertically in reference to the aligned gymnosperm mini-barcode reference database (supplementary data, Table S1). Bases identical to the first sequence are indicated with “.”. Variable bases are indicated with standard International Union of Pure and Applied Chemistry (IUPAC) codes. Ordinal and familial circumscriptions follow Christenhusz et al. (2011). The number of sequences summarized (n) for each family is indicated.
quencer (Life Technologies; High-Throughput Genomics Unit, University of Washington). Herbal supplements that only produced sequences from filler materials or that produced sequencing chromatograms with multiple signals were reexamined using digital PCR (Vogelstein and Kinzler 1999). Samples were diluted (1:5–1:50 000) in 10 mmol/L tris (pH 8.0) to approximately 1 template molecule per L. A minimum of 16 independent PCR reactions (as described above) were conducted on each diluted sample. PCR products were sequenced as described above. Base calls and quality values (QV) were generated by KB 1.4 (Life Technologies). Contigs were assembled, trimmed, and edited using Sequencer 5.2.3 (Gene Codes). Sequence quality was assessed using the barcode quality index (B; Little 2010; version 1.2) with quality threshold (q) set to 30, expected coverage (x) set to the number of reads, and contig size (c) set to the observed size. BRONX 2.0 (Little 2011) was used to compare newly generated matK mini-barcode sequences to reference sequences (the gymnosperm reference library plus species commonly used as fillers; supplementary data, Tables S1, S3). Descriptive statistics were computed with R 3.0.3 (http://www.R-project.org).
Results and discussion Ginkgo biloba mini-barcode sequences are unambiguously differentiated from all other gymnosperms by the presence of a C at nucleotide position 107 (all other gymnosperms have a T; Fig. 1). Within the mini-barcode, G. biloba sequences are variable at four nucleotide positions (Fig. 1, positions 50, 59, 87, and 95)—each of the variants is restricted to a unique accession. These variable positions do not distinguish among the haplotype groups described by Gong et al. (2008). PCR amplification was successful for all of the morphologically identifiable validation samples (n = 22). Mini-barcode sequence quality, B30, was imperfect (median = 0.596, IQR = 0.476–0.648), but comparable to that obtained by a similar study (median = 0.530–0.633; Little and Jeanson 2013; supplementary data, Table S2). Using mini-barcode sequences, BRONX was able to correctly identify all G. biloba (n = 10) validation samples and exclude G. biloba as a possible identification for all of the other gymnosperm samples (n = 12; sensitivity = 1.00, 95% CI = 0.59–1.00; specificity = 1.00, 95% CI = 0.64–1.00). The non-ginkgo validation samples were
Supplementary data are available with the article through the journal Web site at http://nrcresearchpress.com/doi/suppl/10.1139/gen-2014-0130. Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
Little
always correctly identified to taxonomic order, but not necessarily to family, genus, or species. Thus, in terms of PCR universality, sequence quality, and discriminatory power this assay is suitable for the identification of G. biloba samples. DNA amenable to PCR could be extracted from 37 (92.5%) of the 40 herbal dietary supplements. Usable DNA was extracted from seven of the eight (87.5%) samples labeled as containing ginkgo leaves and 30 (93.8%) of the 32 samples labeled as containing leaf extract. The DNA extraction success rate is similar to that reported for dried herbal supplements (77.3%–92.0%; Baker et al. 2012; Little and Jeanson 2013; Newmaster et al. 2013). Thus, the process used to produce G. biloba leaf extract is not a notable source of DNA degradation. Nine (24.3%) of the 37 herbal supplements required digital PCR to separate filler DNA from possible G. biloba DNA. All nine samples were labeled as G. biloba leaf extract. Digital PCR produced amplicons of G. biloba, Oryza sativa L. (Poaceae), and an unidentifiable species (similar to Bromus commutatus Schrad. (Poaceae)): one (11.1%) of the nine samples produced O. sativa amplicons and amplicons belonging to the unidentifiable species, three (33.3%) samples produced both G. biloba and O. sativa amplicons, and the remaining five (55.6%) samples produced only O. sativa amplicons. Mini-barcode sequence quality for herbal supplement samples was lower than that of the validations samples (median = 0.400, IQR = 0.292–0.524; supplementary data, Table S4). In total, 31 of 37 (83.8%) assayable herbal dietary supplements contained identifiable G. biloba DNA and six supplements (16.2%) contained fillers without any detectable G. biloba DNA. Similar percentages were reported from a test of seven ginkgo supplements: G. biloba DNA was found in five (71.4%) samples; the remaining two samples (28.6%) contained Juglans nigra L. (Juglandaceae) and O. sativa (Newmaster et al. 2013). Investigations of other herbal supplements using DNA barcoding have shown similar patterns: 75% of black cohosh (Actaea racemosa L.) and 85% of saw palmetto (Serenoa repens (W.Bartram) Small) supplements contained the correct species (Baker et al. 2012; Little and Jeanson 2013). The matK mini-barcode assay cannot distinguish between samples that do not contain any G. biloba and samples that contain G. biloba processed in a way that rendered the G. biloba DNA incompatible with PCR (e.g., prolonged high temperature treatment). Thus, the assay cannot determine if the six samples (16.2%) without trace of G. biloba DNA do not in fact contain any G. biloba. It is hoped that the matK mini-barcode assay described here will be used by supplement manufacturers to ensure that their supplements not only contain the classes of compounds used for standardization but also contain G. biloba.
Acknowledgements I thank Joan Deutsch and Rolando Rojas for providing excellent technical assistance. Funding from the Alfred P. Sloan Foundation (2010-6-02) is gratefully acknowledged.
References Abascal, F., Zardoya, R., and Telford, M.J. 2010. TranslatorX: multiple alignment of nucleotide sequences guided by amino acid translations. Nucleic Acids Res. 38: W7–W13. doi:10.1093/nar/gkq291. PMID:20435676. Abdel-Kader, R., Hauptmann, S., Keil, U., Scherping, I., Leuner, K., Eckert, A., and Müller, W.E. 2007. Stabilization of mitochondrial function by Ginkgo biloba extract (EGb 761). Pharmacol. Res. 56: 493–502. doi:10.1016/j.phrs.2007.09.011. PMID:17977008. Alexander, P.J., Rajanikanth, G., Bacon, C.D., and Bailey, C.D. 2007. Recovery of plant DNA using a reciprocating saw and silica-based columns. Mol. Ecol. Notes, 7: 5–9. doi:10.1111/j.1471-8286.2006.01549.x. Baker, D.A., Stevenson, D.W., and Little, D.P. 2012. DNA barcode identification of black cohosh herbal dietary supplements. J. AOAC Int. 95: 1023–1034. doi:10. 5740/jaoacint.11-261. PMID:22970567. Christenhusz, M.J.M., Reveal, J.L., Farjon, A., Gardner, M.F., Mill, R.R., and Chase, M.W. 2011. A new classification and linear sequence of extant gymnosperms. Phytotaxa, 19: 55–70. Del Tredici, P., Ling, H., and Yang, G. 1992. The ginkgos of Tian Mu Shan. Conserv. Biol. 6: 202–209. doi:10.1046/j.1523-1739.1992.620202.x.
515
Diamond, B.J., and Bailey, M.R. 2013. Ginkgo biloba: indications, mechanisms, and safety. The Psychiatric clinics of North America, 36: 73–83. doi:10.1016/j.psc. 2012.12.006. PMID:23538078. Eckert, A., Keil, U., Kressmann, S., Schindowski, K., Leutner, S., Leutz, S., and Müller, W.E. 2003. Effects of EGb 761 Ginkgo biloba extract on mitochondrial function and oxidative stress. Pharmacopsychiatry, 36: S15–S23. PMID:13130384. Elphinstone, M.S., Hinten, G.N., Anderson, M.J., and Nock, C.J. 2003. An inexpensive and high-throughput procedure to extract and purify total genomic DNA for population studies. Mol. Ecol. Notes, 3: 317–320. doi:10.1046/j.14718286.2003.00397.x. Gong, W., Chen, C., Dobes, C., Fu, C.-X., and Koch, M.A. 2008. Phylogeography of a living fossil: Pleistocene glaciations forced Ginkgo biloba L. (Ginkgoaceae) into two refuge areas in China with limited subsequent postglacial expansion. Mol. Phylogenet. Evol. 48: 1094–1105. doi:10.1016/j.ympev.2008.05.003. PMID:18554931. Hauptmann, S., Keil, U., Scherping, I., Bonert, A., Eckert, A., and Müller, W.E. 2006. Mitochondrial dysfunction in sporadic and genetic Alzheimer’s disease. Exp. Gerontol. 41: 668–673. doi:10.1016/j.exger.2006.03.012. PMID: 16677790. Kellermann, A.J., and Kloft, C. 2011. Is there a risk of bleeding associated with standardized Ginkgo biloba extract therapy? A systematic review and metaanalysis. Pharmacotherapy, 31: 490–502. doi:10.1592/phco.31.5.490. PMID: 21923430. Koressaar, T., and Remm, M. 2007. Enhancements and modifications of primer design program Primer3. Bioinformatics, 23: 1289–1291. doi:10. 1093/bioinformatics/btm091. PMID:17379693. Lassmann, T., and Sonnhammer, E.L. 2005. Kalign—an accurate and fast multiple sequence alignment algorithm. BMC Bioinform. 6: 298. doi:10.1186/14712105-6-298. PMID:16343337. Laws, K.R., Sweetnam, H., and Kondel, T.K. 2012. Is Ginkgo biloba a cognitive enhancer in healthy individuals? A meta-analysis. Hum. Psychopharmacol. 27: 527–533. PMID:23001963. Lee, E.K., Cibrian-Jaramillo, A., Kolokotronis, S.O., Katari, M.S., Stamatakis, A., Ott, M., et al. 2011. A functional phylogenomic view of the seed plants. PLoS Genet. 7: e1002411. doi:10.1371/journal.pgen.1002411. PMID:22194700. Li, Y., Gao, L.-M., Poudel, R.C., Li, D.-Z., and Forrest, A. 2011. High universality of matK primers for barcoding gymnosperms. Journal of Systematics and Evolution, 49: 169–175. doi:10.1111/j.1759-6831.2011.00128.x. Lindstrom, A., Ooyen, C., Lynch, M.E., and Blumenthal, M. 2012. Herb supplement sales increase 5.5% in 2012: herbal supplement sales rise for 9th consecutive year; turmeric sales jump 40% in natural channel. HerbalGram, 99: 60–65. Little, D.P. 2010. A unified index of sequence quality and contig overlap for DNA barcoding. Bioinformatics, 26: 2780–2781. doi:10.1093/bioinformatics/ btq507. PMID:20837570. Little, D.P. 2011. DNA barcode sequence identification incorporating taxonomic hierarchy and within taxon variability. PLoS ONE, 6: e20552. doi:10.1371/ journal.pone.0020552. PMID:21857897. Little, D.P., and Jeanson, M.L. 2013. DNA barcode authentication of saw palmetto herbal dietary supplements. Sci. Rep. 3: 3518. PMID:24343362. Maitra, I., Marcocci, L., Droy-Lefaix, M.T., and Packer, L. 1995. Peroxyl radical scavenging activity of Ginkgo biloba extract EGb 761. Biochem. Pharmacol. 49: 1649–1655. doi:10.1016/0006-2952(95)00089-I. PMID:7786306. Marcocci, L., Maguire, J.J., Droy-Lefaix, M.T., and Packer, L. 1994. The nitric oxide-scavenging properties of Ginkgo biloba extract EGb 761. Biochem. Biophys. Res. Comm. 201: 748–755. doi:10.1006/bbrc.1994.1764. PMID:8003011. Newmaster, S., Grguric, M., Shanmughanandhan, D., Ramalingam, S., and Ragupathy, S. 2013. DNA barcoding detects contamination and substitution in North American herbal products. BMC Med. 11: 222. doi:10.1186/1741-701511-222. PMID:24120035. Nicolaï, S.P.A., Kruidenier, L.M., Bendermacher, B.L.W., Prins, M.H., Stokmans, R.A., Broos, P.P.H.L., and Teijink, J.A.W. 2013. Ginkgo biloba for intermittent claudication. Cochrane Database of Systematic Reviews, 6: CD006888. PMID:23744597. Pincemail, J., and Deby, C. 1986. Propriétés antiradicalaires de l’extrait de Ginkgo biloba. La Presse Médicale, 15: 1475–1479. PMID:2947084. Pincemail, J., Dupuis, M., Nasr, C., Hans, P., Haag-Berrurier, M., Anton, R., and Deby, C. 1989. Superoxide anion scavenging effect and superoxide dismutase activity of Ginkgo biloba extract. Experientia, 45: 708–712. doi:10.1007/ BF01974564. PMID:2547645. Singh, V., Singh, S.P., and Chan, K. 2010. Review and meta-analysis of usage of ginkgo as an adjunct therapy in chronic schizophrenia. Int. J. Neuropsychopharmacol. 13: 257–271. doi:10.1017/S1461145709990654. PMID:19775502. Tang, C.Q., Yang, Y., Ohsawa, M., Yi, S.-R., Momohara, A., Su, W.-H., et al. 2012. Evidence for the persistence of wild Ginkgo biloba (Ginkgoaceae) populations in the Dalou Mountains, southwestern China. Am. J. Bot. 99: 1408–1414. doi:10.3732/ajb.1200168. PMID:22847538. Traula, H. 1967. The phytogeographic evolution of the genus Ginkgo L. Botaniska Notiser, 120: 409–422. Vogelstein, B., and Kinzler, K.W. 1999. Digital PCR. Proc. Natl. Acad. Sci. U.S.A. 96: 9236–9241. doi:10.1073/pnas.96.16.9236. PMID:10430926. Wang, B.-S., Wang, H., Song, Y.-Y., Qi, H., Rong, Z.-X., Zhang, L., and Chen, H.-Z.
®
Published by NRC Research Press
516
Wu, C.-S., Chaw, S.-M., and Huang, Y.-Y. 2013. Chloroplast phylogenomics indicates that Ginkgo biloba is sister to cycads. Genome Biol. Evol. 5: 243–254. doi:10.1093/gbe/evt001. PMID:23315384. Zhao, Y., Paule, J., Fu, C., and Koch, M.A. 2010. Out of China: distribution history of Ginkgo biloba L. Taxon, 59: 495–504. Zhou, Z.-Y. 2009. An overview of fossil Ginkgoales. Palaeoworld, 18: 1–22. doi:10. 1016/j.palwor.2009.01.001.
Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
2010. Effectiveness of standardized Ginkgo biloba extract on cognitive symptoms of dementia with a six–month treatment: a bivariate random effect meta–analysis. Pharmacopsychiatry, 43: 86–91. doi:10.1055/s-0029-1242817. PMID:20104449. Weinmann, S., Roll, S., Schwarzbach, C., Vauth, C., and Willich, S. 2010. Effects of Ginkgo biloba in dementia: systematic review and meta-analysis. BMC Geriatrics, 10: 14. doi:10.1186/1471-2318-10-14. PMID:20236541.
Genome Vol. 57, 2014
Published by NRC Research Press
NOTE Authentication of Ginkgo biloba herbal dietary supplements using DNA barcoding Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
Damon P. Little
Abstract: Ginkgo biloba L. (known as ginkgo or maidenhair tree) is a phylogenetically isolated, charismatic, gymnosperm tree. Herbal dietary supplements, prepared from G. biloba leaves, are consumed to boost cognitive capacity via improved blood perfusion and mitochondrial function. A novel DNA mini-barcode assay was designed and validated for the authentication of G. biloba in herbal dietary supplements (n = 22; sensitivity = 1.00, 95% CI = 0.59–1.00; specificity = 1.00, 95% CI = 0.64–1.00). This assay was further used to estimate the frequency of mislabeled ginkgo herbal dietary supplements on the market in the United States of America: DNA amenable to PCR could not be extracted from three (7.5%) of the 40 supplements sampled, 31 of 37 (83.8%) assayable supplements contained identifiable G. biloba DNA, and six supplements (16.2%) contained fillers without any detectable G. biloba DNA. It is hoped that this assay will be used by supplement manufacturers to ensure that their supplements contain G. biloba. Key words: digital PCR, matK, mini-barcode. Résumé : Le Ginkgo biloba L. (connu sous le nom ginkgo’ ou arbre aux quarante écus’) est une gymnosperme fascinante et isolée sur le plan phylogénétique. Des suppléments alimentaires naturels, préparés a` partir des feuilles du G. biloba, sont consommés pour accroître les capacités cognitives via une meilleure perfusion sanguine et une fonction mitochondriale améliorée. Un nouveau test d’ADN a été conçu et validé pour l’authentification du G. biloba dans les suppléments alimentaires (n = 22; sensibilité = 1,00, IC 95 % = 0,59–1,00; spécificité = 1,00, IC 95 % = 0,64–1,00). Ce test a ensuite été employé pour mesurer la fréquence de suppléments alimentaires faussement étiquetés comme étant a` base de ginkgo aux États-Unis. Un ADN propice a` l’amplification PCR n’a pu être obtenu pour trois (7,5 %) des 40 suppléments examinés. Pour 31 des 37 suppléments testés (83,8 %), de l’ADN du G. biloba a été détecté, tandis que les six autres suppléments (16,2 %) contenaient des agents de remplissage sans aucun ADN détectable de G. biloba. Les auteurs espèrent que les manufacturiers de suppléments utiliseront ce test pour assurer que leurs suppléments contiennent du G. biloba. [Traduit par la Rédaction] Mots-clés : PCR digital, matK, mini-code-barre.
Introduction Ginkgo biloba L.—referred to as ginkgo or maidenhair tree in English vernacular—is a phylogenetically isolated (Lee et al. 2011; Wu et al. 2013), charismatic, gymnosperm tree. Although it was once widespread throughout the northern hemisphere (Traula 1967; Zhou 2009), G. biloba was, until recently, presumed extinct in the wild (reviewed in Del Tredici et al. 1992). Incontrovertible evidence of relic wild populations in southwest China has recently been amassed (Tang et al. 2012). Over the last two millennia, a subset of the genetic diversity found in wild G. biloba populations has been cultivated in China, Korea, and Japan (Zhao et al. 2010). Cultivated G. biloba in Europe and North America were propagated from Korean and Japanese trees in the early 18th century (Zhao et al. 2010). Humans ingest G. biloba herbal dietary supplements to boost cognitive capacity via improved blood perfusion and mitochondrial function (Diamond and Bailey 2013). Standardized leaf extract of G. biloba contains two major pharmacologically active fractions: flavonoids (mainly glycosides of isorhamnetin, kaempferol, and quercetin) and terpene trilactones (bilobalide and ginkgolide compounds; Abdel-Kader et al. 2007). Working in concert, these compounds protect and enhance mitochondrial function via direct stabilizing effects (Eckert et al. 2003; Abdel-Kader et al.
2007) and by scavenging reactive molecules such as hydroxyl radicals (Pincemail and Deby 1986), nitric oxide (Marcocci et al. 1994), peroxyl radicals (Maitra et al. 1995), and superoxide anions (Pincemail et al. 1989). Impaired mitochondrial function is thought to be a leading cause of cognitive decline (reviewed in Hauptmann et al. 2006). Hence G. biloba is marketed as a cognitive enhancer. Meta-analyses of clinical trial data show that G. biloba decreases blood viscosity and increases blood perfusion (Kellermann and Kloft 2011); that G. biloba is an effective adjunctive therapy for chronic schizophrenia patients (Singh et al. 2010); that cognitive function significantly improves in dementia patients after G. biloba treatment (Wang et al. 2010; Weinmann et al. 2010); and that G. biloba does not significantly improve concentration, memory, or executive function in healthy individuals (Laws et al. 2012). Few minor adverse events are reported from clinical trials that use standardized G. biloba leaf extract (e.g., gastro-intestinal complaints, headache, nausea; Singh et al. 2010; Weinmann et al. 2010; Kellermann and Kloft 2011; Diamond and Bailey 2013; Nicolaï et al. 2013). Sales of G. biloba herbal supplements, in the United States of America, totaled more than US$25.8 million during 2012 (the most
Received 28 August 2014. Accepted 3 November 2014. Corresponding Editor: P. Gulick. D.P. Little. Lewis B. and Dorothy Cullman Program for Molecular Systematics, The New York Botanical Garden, Bronx, NY 10458-5126, USA. E-mail for correspondence: [email protected]. Genome 57: 513–516 (2014) dx.doi.org/10.1139/gen-2014-0130
Published at www.nrcresearchpress.com/gen on 11 December 2014.
514
recent data available; Lindstrom et al. 2012)—making G. biloba the fifth best-selling herbal supplement. The purpose of this study is to validate a novel DNA barcode assay for the authentication of G. biloba in herbal dietary supplements and to use this novel assay to estimate the frequency of mislabeled ginkgo supplements on the market in the United States of America.
Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
Materials and methods A barcode reference database was constructed from all gymnosperm matK sequences deposited in GenBank (8 August 2013). Sequences were trimmed to the barcoding region—defined by priming sites Gym_F1A and Gym_R1A (Li et al. 2011). Sequences that did not span the entire barcoding region, those that contained ambiguous nucleotides, and those with inferred stop codons were discarded (supplementary data, Table S11). Nucleotide sequences were aligned as amino acids using translatorX 1.1 (Abascal et al. 2010) and KALIGN 2.04 (Lassmann and Sonnhammer 2005). To ensure more reliable PCR amplification from degraded DNA samples, a 166-bp matK mini-barcode was designed using PRIMER3 1.1.4 (Koressaar and Remm 2007; Fig. 1). To validate the assay, morphologically identifiable G. biloba (n = 10) and related gymnosperms (n = 12) were collected. Leaf tissue was preserved in silica gel. Voucher specimens were deposited in the Steere Herbarium (NY; supplementary data, Table S2). After validation, the assay was applied to herbal supplements purchased in New York area retail stores and on-line. The supplements were either dried and powdered leaves of G. biloba (n = 8) or G. biloba leaf extract dried on an “inert” filler (e.g., rice, soy, etc.; n = 32). Samples (10 mg) of herbal supplements or silica dried leaf tissue were disrupted in a 1.6 mL tube using a Qiagen TissueLyser (30 Hz for 2 × 1.5 min) and two stainless steel ball bearings (3 mm). Samples were incubated with 50 rpm horizontal shaking at 42 °C in 600 L extraction buffer (8 mmol/L NaCl, 16 mmol/L sucrose, 5.8 mmol/L EDTA, 0.5% (w/v) sodium dodecyl sulphate, 12.4 mmol/L tris pH 9.1, 200 g/mL proteinase K; Alexander et al. 2007). Following 18 h of incubation, 200 L of 3 mol/L potassium acetate (pH 4.7) was added to each sample and the samples were cooled to 0 °C. After 10 min, samples were centrifuged at 14 000g for 5 min. A 900 L sample of 2 mol/L guanidine hydrochloride in 95% (v/v) ethanol was mixed with 600 L of each sample’s aqueous phase. The mixtures were applied, 500 L at a time, to silica spin columns (Epoch Life Science). After each application, columns were centrifuged at 7000g for 1 min. Each sample was washed twice with 500 L of buffer (50% (v/v) ethanol, 10 mmol/L tris pH 7.4, 0.5 mmol/L EDTA, 50 mmol/L NaCl; Elphinstone et al. 2003). Ethanol residue was removed by centrifugation at 7000g for 2 min. Total DNA was eluted, by centrifugation (7000g for 1 min), in 2 × 100 L 10 mmol/L tris (pH 8.0). The matK mini-barcode was PCR amplified in a 15 L volume containing the following: 20 mmol/L tris pH 8.8, 10 mmol/L KCl, 10 mmol/L (NH4)2SO4, 2 mmol/L MgSO4, 0.1% (v/v) Triton X-100, 5% (w/v) sucrose, 0.025% (w/v) cresol red, 0.025 g/L BSA, 0.2 mmol/L dNTPs, 0.5 mol/L of each primer (5=-TTTCAGGGCGTCCAATTAGT-3= and 5=-TTAGCACATGGAAGTCGGAGT-3=), 0.5 U Taq polymerase, and 0.5 L DNA. Reactions were incubated for 150 s at 95 °C, cycled 35 times (30 s at 95 °C, 120 s at 58 °C), and then incubated 10 min at 60 °C. Residual primers and dNTPs were neutralized with ExoSAP-IT (Affymetrix) and bidirectionally sequenced using the amplification primers and BigDye v3.1 (1 L neutralized PCR product in each 5 L reaction; Life Technologies) on a 3730 automated se-
1
Genome Vol. 57, 2014
Fig. 1. Nucleotides diagnostic of Ginkgo biloba within the matK mini-barcode. Positions are numbered vertically in reference to the aligned gymnosperm mini-barcode reference database (supplementary data, Table S1). Bases identical to the first sequence are indicated with “.”. Variable bases are indicated with standard International Union of Pure and Applied Chemistry (IUPAC) codes. Ordinal and familial circumscriptions follow Christenhusz et al. (2011). The number of sequences summarized (n) for each family is indicated.
quencer (Life Technologies; High-Throughput Genomics Unit, University of Washington). Herbal supplements that only produced sequences from filler materials or that produced sequencing chromatograms with multiple signals were reexamined using digital PCR (Vogelstein and Kinzler 1999). Samples were diluted (1:5–1:50 000) in 10 mmol/L tris (pH 8.0) to approximately 1 template molecule per L. A minimum of 16 independent PCR reactions (as described above) were conducted on each diluted sample. PCR products were sequenced as described above. Base calls and quality values (QV) were generated by KB 1.4 (Life Technologies). Contigs were assembled, trimmed, and edited using Sequencer 5.2.3 (Gene Codes). Sequence quality was assessed using the barcode quality index (B; Little 2010; version 1.2) with quality threshold (q) set to 30, expected coverage (x) set to the number of reads, and contig size (c) set to the observed size. BRONX 2.0 (Little 2011) was used to compare newly generated matK mini-barcode sequences to reference sequences (the gymnosperm reference library plus species commonly used as fillers; supplementary data, Tables S1, S3). Descriptive statistics were computed with R 3.0.3 (http://www.R-project.org).
Results and discussion Ginkgo biloba mini-barcode sequences are unambiguously differentiated from all other gymnosperms by the presence of a C at nucleotide position 107 (all other gymnosperms have a T; Fig. 1). Within the mini-barcode, G. biloba sequences are variable at four nucleotide positions (Fig. 1, positions 50, 59, 87, and 95)—each of the variants is restricted to a unique accession. These variable positions do not distinguish among the haplotype groups described by Gong et al. (2008). PCR amplification was successful for all of the morphologically identifiable validation samples (n = 22). Mini-barcode sequence quality, B30, was imperfect (median = 0.596, IQR = 0.476–0.648), but comparable to that obtained by a similar study (median = 0.530–0.633; Little and Jeanson 2013; supplementary data, Table S2). Using mini-barcode sequences, BRONX was able to correctly identify all G. biloba (n = 10) validation samples and exclude G. biloba as a possible identification for all of the other gymnosperm samples (n = 12; sensitivity = 1.00, 95% CI = 0.59–1.00; specificity = 1.00, 95% CI = 0.64–1.00). The non-ginkgo validation samples were
Supplementary data are available with the article through the journal Web site at http://nrcresearchpress.com/doi/suppl/10.1139/gen-2014-0130. Published by NRC Research Press
Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
Little
always correctly identified to taxonomic order, but not necessarily to family, genus, or species. Thus, in terms of PCR universality, sequence quality, and discriminatory power this assay is suitable for the identification of G. biloba samples. DNA amenable to PCR could be extracted from 37 (92.5%) of the 40 herbal dietary supplements. Usable DNA was extracted from seven of the eight (87.5%) samples labeled as containing ginkgo leaves and 30 (93.8%) of the 32 samples labeled as containing leaf extract. The DNA extraction success rate is similar to that reported for dried herbal supplements (77.3%–92.0%; Baker et al. 2012; Little and Jeanson 2013; Newmaster et al. 2013). Thus, the process used to produce G. biloba leaf extract is not a notable source of DNA degradation. Nine (24.3%) of the 37 herbal supplements required digital PCR to separate filler DNA from possible G. biloba DNA. All nine samples were labeled as G. biloba leaf extract. Digital PCR produced amplicons of G. biloba, Oryza sativa L. (Poaceae), and an unidentifiable species (similar to Bromus commutatus Schrad. (Poaceae)): one (11.1%) of the nine samples produced O. sativa amplicons and amplicons belonging to the unidentifiable species, three (33.3%) samples produced both G. biloba and O. sativa amplicons, and the remaining five (55.6%) samples produced only O. sativa amplicons. Mini-barcode sequence quality for herbal supplement samples was lower than that of the validations samples (median = 0.400, IQR = 0.292–0.524; supplementary data, Table S4). In total, 31 of 37 (83.8%) assayable herbal dietary supplements contained identifiable G. biloba DNA and six supplements (16.2%) contained fillers without any detectable G. biloba DNA. Similar percentages were reported from a test of seven ginkgo supplements: G. biloba DNA was found in five (71.4%) samples; the remaining two samples (28.6%) contained Juglans nigra L. (Juglandaceae) and O. sativa (Newmaster et al. 2013). Investigations of other herbal supplements using DNA barcoding have shown similar patterns: 75% of black cohosh (Actaea racemosa L.) and 85% of saw palmetto (Serenoa repens (W.Bartram) Small) supplements contained the correct species (Baker et al. 2012; Little and Jeanson 2013). The matK mini-barcode assay cannot distinguish between samples that do not contain any G. biloba and samples that contain G. biloba processed in a way that rendered the G. biloba DNA incompatible with PCR (e.g., prolonged high temperature treatment). Thus, the assay cannot determine if the six samples (16.2%) without trace of G. biloba DNA do not in fact contain any G. biloba. It is hoped that the matK mini-barcode assay described here will be used by supplement manufacturers to ensure that their supplements not only contain the classes of compounds used for standardization but also contain G. biloba.
Acknowledgements I thank Joan Deutsch and Rolando Rojas for providing excellent technical assistance. Funding from the Alfred P. Sloan Foundation (2010-6-02) is gratefully acknowledged.
References Abascal, F., Zardoya, R., and Telford, M.J. 2010. TranslatorX: multiple alignment of nucleotide sequences guided by amino acid translations. Nucleic Acids Res. 38: W7–W13. doi:10.1093/nar/gkq291. PMID:20435676. Abdel-Kader, R., Hauptmann, S., Keil, U., Scherping, I., Leuner, K., Eckert, A., and Müller, W.E. 2007. Stabilization of mitochondrial function by Ginkgo biloba extract (EGb 761). Pharmacol. Res. 56: 493–502. doi:10.1016/j.phrs.2007.09.011. PMID:17977008. Alexander, P.J., Rajanikanth, G., Bacon, C.D., and Bailey, C.D. 2007. Recovery of plant DNA using a reciprocating saw and silica-based columns. Mol. Ecol. Notes, 7: 5–9. doi:10.1111/j.1471-8286.2006.01549.x. Baker, D.A., Stevenson, D.W., and Little, D.P. 2012. DNA barcode identification of black cohosh herbal dietary supplements. J. AOAC Int. 95: 1023–1034. doi:10. 5740/jaoacint.11-261. PMID:22970567. Christenhusz, M.J.M., Reveal, J.L., Farjon, A., Gardner, M.F., Mill, R.R., and Chase, M.W. 2011. A new classification and linear sequence of extant gymnosperms. Phytotaxa, 19: 55–70. Del Tredici, P., Ling, H., and Yang, G. 1992. The ginkgos of Tian Mu Shan. Conserv. Biol. 6: 202–209. doi:10.1046/j.1523-1739.1992.620202.x.
515
Diamond, B.J., and Bailey, M.R. 2013. Ginkgo biloba: indications, mechanisms, and safety. The Psychiatric clinics of North America, 36: 73–83. doi:10.1016/j.psc. 2012.12.006. PMID:23538078. Eckert, A., Keil, U., Kressmann, S., Schindowski, K., Leutner, S., Leutz, S., and Müller, W.E. 2003. Effects of EGb 761 Ginkgo biloba extract on mitochondrial function and oxidative stress. Pharmacopsychiatry, 36: S15–S23. PMID:13130384. Elphinstone, M.S., Hinten, G.N., Anderson, M.J., and Nock, C.J. 2003. An inexpensive and high-throughput procedure to extract and purify total genomic DNA for population studies. Mol. Ecol. Notes, 3: 317–320. doi:10.1046/j.14718286.2003.00397.x. Gong, W., Chen, C., Dobes, C., Fu, C.-X., and Koch, M.A. 2008. Phylogeography of a living fossil: Pleistocene glaciations forced Ginkgo biloba L. (Ginkgoaceae) into two refuge areas in China with limited subsequent postglacial expansion. Mol. Phylogenet. Evol. 48: 1094–1105. doi:10.1016/j.ympev.2008.05.003. PMID:18554931. Hauptmann, S., Keil, U., Scherping, I., Bonert, A., Eckert, A., and Müller, W.E. 2006. Mitochondrial dysfunction in sporadic and genetic Alzheimer’s disease. Exp. Gerontol. 41: 668–673. doi:10.1016/j.exger.2006.03.012. PMID: 16677790. Kellermann, A.J., and Kloft, C. 2011. Is there a risk of bleeding associated with standardized Ginkgo biloba extract therapy? A systematic review and metaanalysis. Pharmacotherapy, 31: 490–502. doi:10.1592/phco.31.5.490. PMID: 21923430. Koressaar, T., and Remm, M. 2007. Enhancements and modifications of primer design program Primer3. Bioinformatics, 23: 1289–1291. doi:10. 1093/bioinformatics/btm091. PMID:17379693. Lassmann, T., and Sonnhammer, E.L. 2005. Kalign—an accurate and fast multiple sequence alignment algorithm. BMC Bioinform. 6: 298. doi:10.1186/14712105-6-298. PMID:16343337. Laws, K.R., Sweetnam, H., and Kondel, T.K. 2012. Is Ginkgo biloba a cognitive enhancer in healthy individuals? A meta-analysis. Hum. Psychopharmacol. 27: 527–533. PMID:23001963. Lee, E.K., Cibrian-Jaramillo, A., Kolokotronis, S.O., Katari, M.S., Stamatakis, A., Ott, M., et al. 2011. A functional phylogenomic view of the seed plants. PLoS Genet. 7: e1002411. doi:10.1371/journal.pgen.1002411. PMID:22194700. Li, Y., Gao, L.-M., Poudel, R.C., Li, D.-Z., and Forrest, A. 2011. High universality of matK primers for barcoding gymnosperms. Journal of Systematics and Evolution, 49: 169–175. doi:10.1111/j.1759-6831.2011.00128.x. Lindstrom, A., Ooyen, C., Lynch, M.E., and Blumenthal, M. 2012. Herb supplement sales increase 5.5% in 2012: herbal supplement sales rise for 9th consecutive year; turmeric sales jump 40% in natural channel. HerbalGram, 99: 60–65. Little, D.P. 2010. A unified index of sequence quality and contig overlap for DNA barcoding. Bioinformatics, 26: 2780–2781. doi:10.1093/bioinformatics/ btq507. PMID:20837570. Little, D.P. 2011. DNA barcode sequence identification incorporating taxonomic hierarchy and within taxon variability. PLoS ONE, 6: e20552. doi:10.1371/ journal.pone.0020552. PMID:21857897. Little, D.P., and Jeanson, M.L. 2013. DNA barcode authentication of saw palmetto herbal dietary supplements. Sci. Rep. 3: 3518. PMID:24343362. Maitra, I., Marcocci, L., Droy-Lefaix, M.T., and Packer, L. 1995. Peroxyl radical scavenging activity of Ginkgo biloba extract EGb 761. Biochem. Pharmacol. 49: 1649–1655. doi:10.1016/0006-2952(95)00089-I. PMID:7786306. Marcocci, L., Maguire, J.J., Droy-Lefaix, M.T., and Packer, L. 1994. The nitric oxide-scavenging properties of Ginkgo biloba extract EGb 761. Biochem. Biophys. Res. Comm. 201: 748–755. doi:10.1006/bbrc.1994.1764. PMID:8003011. Newmaster, S., Grguric, M., Shanmughanandhan, D., Ramalingam, S., and Ragupathy, S. 2013. DNA barcoding detects contamination and substitution in North American herbal products. BMC Med. 11: 222. doi:10.1186/1741-701511-222. PMID:24120035. Nicolaï, S.P.A., Kruidenier, L.M., Bendermacher, B.L.W., Prins, M.H., Stokmans, R.A., Broos, P.P.H.L., and Teijink, J.A.W. 2013. Ginkgo biloba for intermittent claudication. Cochrane Database of Systematic Reviews, 6: CD006888. PMID:23744597. Pincemail, J., and Deby, C. 1986. Propriétés antiradicalaires de l’extrait de Ginkgo biloba. La Presse Médicale, 15: 1475–1479. PMID:2947084. Pincemail, J., Dupuis, M., Nasr, C., Hans, P., Haag-Berrurier, M., Anton, R., and Deby, C. 1989. Superoxide anion scavenging effect and superoxide dismutase activity of Ginkgo biloba extract. Experientia, 45: 708–712. doi:10.1007/ BF01974564. PMID:2547645. Singh, V., Singh, S.P., and Chan, K. 2010. Review and meta-analysis of usage of ginkgo as an adjunct therapy in chronic schizophrenia. Int. J. Neuropsychopharmacol. 13: 257–271. doi:10.1017/S1461145709990654. PMID:19775502. Tang, C.Q., Yang, Y., Ohsawa, M., Yi, S.-R., Momohara, A., Su, W.-H., et al. 2012. Evidence for the persistence of wild Ginkgo biloba (Ginkgoaceae) populations in the Dalou Mountains, southwestern China. Am. J. Bot. 99: 1408–1414. doi:10.3732/ajb.1200168. PMID:22847538. Traula, H. 1967. The phytogeographic evolution of the genus Ginkgo L. Botaniska Notiser, 120: 409–422. Vogelstein, B., and Kinzler, K.W. 1999. Digital PCR. Proc. Natl. Acad. Sci. U.S.A. 96: 9236–9241. doi:10.1073/pnas.96.16.9236. PMID:10430926. Wang, B.-S., Wang, H., Song, Y.-Y., Qi, H., Rong, Z.-X., Zhang, L., and Chen, H.-Z.
®
Published by NRC Research Press
516
Wu, C.-S., Chaw, S.-M., and Huang, Y.-Y. 2013. Chloroplast phylogenomics indicates that Ginkgo biloba is sister to cycads. Genome Biol. Evol. 5: 243–254. doi:10.1093/gbe/evt001. PMID:23315384. Zhao, Y., Paule, J., Fu, C., and Koch, M.A. 2010. Out of China: distribution history of Ginkgo biloba L. Taxon, 59: 495–504. Zhou, Z.-Y. 2009. An overview of fossil Ginkgoales. Palaeoworld, 18: 1–22. doi:10. 1016/j.palwor.2009.01.001.
Genome Downloaded from www.nrcresearchpress.com by UNIVERSITY OF PITTSBURGH on 03/08/15 For personal use only.
2010. Effectiveness of standardized Ginkgo biloba extract on cognitive symptoms of dementia with a six–month treatment: a bivariate random effect meta–analysis. Pharmacopsychiatry, 43: 86–91. doi:10.1055/s-0029-1242817. PMID:20104449. Weinmann, S., Roll, S., Schwarzbach, C., Vauth, C., and Willich, S. 2010. Effects of Ginkgo biloba in dementia: systematic review and meta-analysis. BMC Geriatrics, 10: 14. doi:10.1186/1471-2318-10-14. PMID:20236541.
Genome Vol. 57, 2014
Published by NRC Research Press
