Elsevier

Journal of Ethnopharmacology

Volume 198, 23 February 2017, Pages 417-431
Journal of Ethnopharmacology

Reverse ethnopharmacology and drug discovery

https://doi.org/10.1016/j.jep.2016.12.044Get rights and content

Abstract

Ethnopharmacological relevance

Ethnopharmacological investigations of traditional medicines have made significant contributions to plant-derived drugs, as well as the advancement of pharmacology. Drug discovery from medicinal flora is more complex than generally acknowledged because plants are applied for different therapeutic indications within and across cultures. Therefore we propose the concept of “reverse ethnopharmacology” and compare biomedical uses of plant taxa with their ethnomedicinal and popular uses and test the effect of these on the probability of finding biomedical and specifically anticancer drugs.

Materials and methods

For this analysis we use data on taxonomy and medical indications of plant derived biomedical drugs, clinical trial, and preclinical trial drug candidates published by Zhu et al. (2011) and compare their therapeutic indications with their ethnomedicinal and popular uses as reported in the NAPRALERT® database. Specifically, we test for increase or decrease of the probability of finding anticancer drugs based on ethnomedicinal and popular reports with Bayesian logistic regression analyses.

Results

Anticancer therapy resulted as the most frequent biomedicinal indication of the therapeutics derived from the 225 drug producing higher plant taxa and showed an association with ethnomedicinal and popular uses in women's medicine, which was also the most important popular use-category. Popular remedies for dysmenorrhoea, and uses as emmenagogues, abortifacients and contraceptives showed a positive effect on the probability of finding anticancer drugs. Another positive effect on the probability of discovering anticancer therapeutics was estimated for popular herbal drugs associated with the therapy of viral and bacterial infections, while the highest effect was found for popular remedies used to treat cancer symptoms. However, this latter effect seems to be influenced by the feedback loop and divulgence of biomedical knowledge on the popular level.

Conclusion

We introduce the concept of reverse ethnopharmacology and show that it is possible to estimate the probability of finding biomedical drugs based on ethnomedicinal uses. The detected associations confirm the classical ethnopharmacological approach where a popular remedy for disease category X results in a biomedical drug for disease category X but does also point out the existence of cross-over relationships where popular remedies for disease category X result in biomedical therapeutics for disease category Y (Zhu et al., 2011).

Introduction

Human exploitation of plant diversity for the provision of medicine follows two main strategies. Herbal medicine depends on synergistic effects of mostly water soluble complex multi-compound mixtures, while biomedicine generally relies on the application of single compound drugs derived from plants, whether medicinally used or not. The discovery of new drugs from biological diversity and plants in particular has been allegorized with the search for the needle in the haystack (Cordell et al., 1991). In contrast follow-up to discoveries from random screening guided by taxonomy, chemotaxonomy draws on phylogenetic relatedness in the search of identical or similar bioactive compounds (e.g. taxol, Denis et al., 1988) while therapeutic indications of indigenous drugs are used as a lead in the ethnopharmacological approach (e.g. Artemisinin: Klayman, 1985 and Tu, 2011; Cyclotides: Koehbach et al., 2013).

The empirical value and legacy of medicinal plant use is often uncritically seen as proof of effectiveness and safety (c.f. Heinrich et al. 2004). More specifically Fabricant and Farnsworth (2001) claim that from 122 plant-derived clinical drugs 88 have the same or similar indications as the medicinal plants from which they are sourced. This claim is, however, a simplification of a more complex reality and statistically not correct in that it does not consider that medicinal plants often have many different uses within and across cultures, including those listed in Fabricant and Farnsworth (2001). Multipurpose medicinal applications, at times even apparently contradictive, complicate the selection of bioassays and the ethnopharmacological search for new drugs. Although ethnopharmacology uses anthropological concepts and tools such as cross-cultural comparisons and consensus analysis in order to assess the culturally most accepted uses of medicinal plant species (Berlin and Berlin, 2005, Leonti and Weckerle, 2015), the question remains as to how far consensus on ethnomedicinal uses and their scientific interpretations correlate with biomedical disease concepts, phytochemical profiles and meaningful screening results (Gyllenhaal et al., 2012, Leonti et al., 2013a). As a practical example, exemplifying the inherent difficulty in using ethnomedicinal information as a guide for drug discovery, the case of Catharanthus roseus may serve: Madagascar periwinkle is widely used against diabetes in traditional medicines and consequently initially investigated for its alleged hypoglycaemic activity. Instead of finding effects on the blood glucose level, during the biological validation the cytotoxic properties became apparent, giving way to the development of vincristine and vinblastine into anticancer drugs (van der Heijden et al., 2004; Guéritte and Fahy, 2005).

Natural products, including plant metabolites, are known for their importance in the development of anticancer remedies (Pezzuto, 1997, Butler, 2005; Cragg and Newman, 2005). A query of the NAPRALET® database for species ethnomedicinally used against cancer and cancer related symptoms yielded over 500 distinct records for more than 350 plant species (Graham et al., 2000), which corresponds to a relatively low consensus. Symptoms of oncological diseases are, in fact, multifaceted, potentially affecting all kinds of body parts and organs, and therefore, “cancer” is generally poorly recognized in ethnomedicinal systems, which complicates the search for anticancer drugs with ethnopharmacological resources (Cragg and Newman, 2005).

Intriguingly, Spjut and Perdue (1976) found during a retrospective study of the NCI vaults that plants with an ethnomedicinal background, particularly those associated with poisonous uses were more likely to show cytotoxic effects than biomedical collections in general. The highest cytotoxic activity was for plants used as anthelmintics (29.3%), fish- (38.6%) and arrow-, ordeal- and homicidal poisons (45.7%; Cordell et al., 1991; Spjut and Perdue, 1976; Spjut, 2005). Spjut (2005) concluded that poisonous plants, including the ones used in local and traditional medicines, have a higher probability of exhibiting significant cytotoxic activity with respect to plants collected at random.

The observations made by Spjut and Perdue (1976) as well as Spjut (2005) and the case of C. roseus was our inspiration for testing here, in analogy to “reverse pharmacognosy” (see Do and Bernard, 2004; Vaidya, 2006; Patwardhan et al. 2008) the concept of “reverse ethnopharmacology”.

A biomedical perspective during the ethnomedicinal enquiry defines the “reverse” in terms that we look for patterns and associations between therapeutic indications of plant derived biomedical drugs and the ethnomedicinal use of the source plants. We propose “reverse ethnopharmacology” as a drug discovery tool for visualizing hidden associations between ethnomedicinal uses and biomedical indications of plant derived drugs. To this end we compare the therapeutic indications of angiosperm and gymnosperm derived biomedical drugs with the ethnomedicinal uses of the same taxa (mostly species) with a set of statistical tools.

For the biomedical uses of plant derived drugs we rely on the census by Zhu et al. (2011). Inspired by the seminal work of Newman and Cragg (2007), Zhu et al. (2011) compiled taxonomic and therapeutic data on all approved, clinical trial and preclinical natural product drugs. From the 457 existing angiosperm and gymnosperm families (APG IV) 62 families account for all 225 angio- and gymnosperm drugs, clinical trial or preclinical drug producing taxa (see Zhu et al. 2011). These 62 families are largely widespread taxa, embracing, according to the APG system, 152.712 species altogether or more than half of the 286.467 existing angiosperms and gymnosperms. A possible explanation for this over proportional share of species is that plant taxa distributed over wider geographical extensions experience more diverse ecological interactions resulting in the production of secondary metabolites able to interfere with a broad spectrum of biological targets (Leonti et al., 2013a, Leonti et al., 2013b). Corresponding ethnomedicinal uses were quantitatively extracted from the NAPRALERT® database and available for 186 of the 225 taxa. Special attention was given to the relation between the largest therapeutic category of use of the popular/ethnomedicinal domain (pGYN, i.e. women's medicine) and the largest biomedical domain (bCAN, i.e. biomedical cancer therapy).

With a dual statistical approach, one based on citations (i) and a second based on plant taxa (ii) we tested for (i) associations between the biomedical and ethnomedicinal uses by means of usual statistical association tests on joint frequencies of biomedical and ethnomedicinal uses. Moreover, (ii) we estimated the increment/decrement generated by ethnomedicinal categories on the probability of finding biomedical drugs and specifically anticancer drugs and leads (bCAN). Analyses estimating the increment or decrement of probabilities of biomedical use were performed by means of logistic regression. A problem we did not account for in this approach is the chicken and egg situation and the questions, which use established first (the popular or the biomedical?) and which use influenced which (the popular the biomedical or vice versa?).

Section snippets

Research question

Our research questions are: (i) Are there associations between therapeutic indications of plant derived biomedical drugs and the ethnomedicinal uses of the source plants? And if so (ii) what kind of associations can be found?In the ethnopharmacological approach for drug discovery or the validation of traditional medicines a drug used against disease category X is tested in a biological model representing characteristics of disease category X. Here we estimate the probability of finding a

Acknowledgements

This paper is based on an oral presentation held at the 16th ISE Congress in Yulin, China (May 16–18th, 2016) and we would like to thank the organizers for the excellent congress and their hospitality. We are thankful also to Rich Spjut for the valuable comments on a draft version of this paper. GIS was a recipient of The Swiss Government Excellence Postdoctoral Scholarship awarded by the Federal Commission for Scholarships for Foreign Students (FCS). ML and CW acknowledge funding from the

References (84)

  • W.N. Jefferson et al.

    Disruption of the female reproductive system by the phytoestrogen genistein

    Reprod. Toxicol.

    (2007)
  • S. Jin et al.

    Daidzein induces MCF-7 breast cancer cell apoptosis via the mitochondrial pathway

    Ann. Oncol.

    (2010)
  • M. Leonti

    The future is written: impact of scripts on the cognition, selection, knowledge and transmission of medicinal plant use and its implications for ethnobotany and ethnopharmacology

    J. Ethnopharmacol.

    (2011)
  • M. Leonti et al.

    Bioprospecting: evolutionary implications from a post-olmec pharmacopoeia and the relevance of widespread taxa

    J. Ethnopharmacol.

    (2013)
  • M. Leonti et al.

    Bioprospecting: evolutionary implications from a post-olmec pharmacopoeia and the relevance of widespread taxa

    J. Ethnopharmacol.

    (2013)
  • L.F. Liu et al.

    Action of solamargine on human lung cancer cells - enhancement of the susceptibility of cancer cells to TNFs

    FEBS Lett.

    (2004)
  • L. Pan et al.

    The continuing search for antitumor agents from higher plants

    Phytochem. Lett.

    (2010)
  • J.M. Pezzuto

    Plant-derived anticancer agents

    Biochem. Pharm.

    (1997)
  • J. Rouëssé et al.

    Phase II study of elliptinium acetate salvage treatment of advanced breast cancer

    Eur. J. Cancer

    (1993)
  • P.O. Staub et al.

    Classifying diseases and remedies in ethnomedicine and ethnopharmacology

    J. Ethnopharmacol.

    (2015)
  • L. Yeruva et al.

    Perillyl alcohol and perillic acid induced cell cycle arrest and apoptosis in non small cell lung cancer cells

    Cancer Lett.

    (2007)
  • H. Abdelkafi et al.

    Natural products from Cephalotaxus sp.: chemical diversity and synthetic aspects

    Nat. Prod. Rep.

    (2012)
  • Aliotta, G., Piomelli, D., Pollio, A., Touwaide, A., 2003. Le piante medicinali del Corpus Hippocraticum. Guerini e...
  • APG IV, An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG...
  • C. Atkinson et al.

    Gut bacterial metabolism of the soy isoflavone daidzein: exploring the relevance to human health

    Exp. Biol. Med.

    (2005)
  • O.J. Bandele et al.

    Dietary polyphenols as topoisomerase II poisons: B ring and C ring substituents determine the mechanism of enzyme-mediated DNA cleavage enhancement

    Chem. Res. Toxicol.

    (2008)
  • J.T. Belanger

    Perillyl alcohol: applications in oncology

    Alter. Med. Rev.

    (1998)
  • Y. Benjamini et al.

    Controlling the false discovery rate: a practical and powerful approach to multiple testing

    J. R. Stat. Soc. Ser. B (Methodol.)

    (1995)
  • E.A. Berlin et al.

    Some field methods in medical ethnobiology

    Field Methods

    (2005)
  • B. Böck

    Medicinal plants and medicaments used for conception, abortion, and fertility control in ancient Babylon

    J. Asiat.

    (2013)
  • K.B. Brookes et al.

    Uteroactive constituents from Combretum kraussii

    South Afr. J. Chem.

    (1999)
  • M.S. Butler

    Natural products to drugs: natural product derived compounds in clinical trials

    Nat. Prod. Rep.

    (2005)
  • C. Ciganda et al.

    Herbal infusions used for induced abortion

    J. Toxicol. Clin. Toxicol.

    (2003)
  • A. Cirla et al.

    Combretastatins: from natural products to drug discovery

    Nat. Prod. Rep.

    (2003)
  • P. Congdom

    Applied Bayesian Modelling

    (2003)
  • A.I. Constantinou et al.

    Phenoxodiol [2H-1-benzopyran-7-ol, 3-(4-hydroxyphenyl)], a novel isoflavone derivative, inhibits DNA topoisomerase II by stabilizing the cleavable complex

    Anticancer Res.

    (2002)
  • G.M. Cragg et al.

    Impact of natural products on developing new anti-cancer agents

    Chem. Rev.

    (2009)
  • A. Crozier et al.

    Phenols, polyphenols and tannins: an overview

  • J.N. Denis et al.

    Highly efficient, practical approach to natural taxol

    J. Am. Chem. Soc.

    (1988)
  • Q.-T. Do et al.

    Reverse pharmacognosy: a new concept for accelerating natural drug discovery

    Adv. Phytomed.

    (2004)
  • J.N. Eloff et al.

    Isolation of an antibacterial stilbene from Combretum woodii (Combretaceae) leaves

    Afr. J. Biotechnol.

    (2005)
  • D.S. Fabricant et al.

    The value of plants used in traditional medicine for drug discovery

    Environ. Health Perspect.

    (2001)
  • Cited by (24)

    • Natural products against acute respiratory infections: Strategies and lessons learned

      2020, Journal of Ethnopharmacology
      Citation Excerpt :

      In modern pharmacognostic research with its simplified aim to track down the overall effect to one or a few (co)effectors, this habit multiplies researchers’ difficulties to unravel the complexity in terms of bioactive constituents and involved molecular mechanisms as well as additive or even synergistic effects (see Hochu-ekki-to, Sinupret®, Esberitox®). Ethnopharmacology converts traditional cultural and cross-cultural knowledge of medicinal plants and their therapeutic applicability into a helpful tool in drug discovery (Leonti et al., 2017). Going back centuries in the history of traditional medicine, Echinacea purpurea has been one of the most prominent examples for the treatment of ARIs (Barrett, 2003).

    • Extraction, isolation and in vitro evaluation of affinisine from Tabernaemontana catharinensis in human melanoma cells

      2019, Bioorganic Chemistry
      Citation Excerpt :

      Compounds obtained from plants have been recognized as prototypes for the development of new drugs. The adaptation of plants leads to the synthesis of secondary metabolites whose substances also are present pharmacological activities, among which antibacterial, anti-inflammatory, antihypertensive and anticancer activities are the most frequently reported [1,2]. Along this line, indole alkaloids having rich structural diversity have been isolated from natural resources.

    View all citing articles on Scopus
    View full text