Isolation and Characterization of Yuremamine, a New Phytoindole

• Abstract
• Introduction
• Materials and Methods
• Isolation
• High-pressure liquid chromatography
• Liquid chromatography-mass spectrometry
• Nuclear magnetic resonance spectroscopy
• Results and Discussion
• Acknowledgements
• References

Abstract

Yuremamine was isolated and characterized from the stem bark of Mimosa tenuiflora. This plant is still used by indigenous peoples in North-eastern Brazil to make yurema, a psychoactive beverage that is used for medico-religious purpose (jurema preta or vinho da jurema, in Portuguese). The characterization of this novel compound by NMR and mass spectrometry introduces a new class of phytoindoles.

Introduction

Yurema is the name of a plant, a deity, certain geographical areas and a sacramental beverage that is made from the root bark of Mimosa tenuiflora (Willd.) Poir. This species may be identical to Mimosa hostilis (Mart.) Benth. [1]. The beverage, as jurema preta, vinho da jurema and other names, is still used by people of North-eastern Brazil for medico-religious purposes, and may contain other plant additives [2]. Yurema also means ‘thorny’ in the local Tupi dialect of Pernambuco, Brazil, and this botanical feature is also indicated by the Latin class name of hostilis, i. e., a mimosa with thorns. The bark of M. tenuiflora is also used in Mexico to prepare tepescohuit, a traditional antimicrobial agent that is used to treat burns and other skin problems [3]. An extract of M. tenuiflora inhibited the peristaltic reflex of the isolated guinea-pig ileum, which is densely populated with serotonin receptors [4]. Research efforts had already identified both M. hostilis and M. tenuiflora as a source of N,N-dimethyltryptamine (DMT), which is a potent, psychoactive serotonin agonist [5], [6]. Other phytochemical analyses have also identified the presence of triterpenoid glycosides (saponins) in the root bark of M. tenuiflora, which may help promote damaged skin growth as a topical ointment in folk medicine [7].

The purpose of the present study was to re-examine the phytochemistry of this species with advanced instrumentation and methodologies, particularly ¹H-¹³C nuclear magnetic resonance (NMR) and liquid chromatography-mass spectrometry (LC-MS) under mildly acidic pH.

Materials and Methods

Isolation

Stem bark from Mimosa teniuflora was identified and collected in Oaxaca, Mexico, by ethnobotanists Rob Montgomery and Jonathan Ott. A voucher specimen (2005 - 7RM) of this sample has been deposited in the University of Kuopio Research Garden herbarium. The dried bark was milled, and 2.265 g were soaked in 250 mL methanol at + 5 °C for 2 weeks in a refrigerator with occasional stirring. Subsequent aliquots were removed and evaporated to dryness under nitrogen at 35 °C as needed. The resulting residue was re-dissolved in an equivalent amount of 15 % water and 85 % acetonitrile to avoid precipitation. All samples were passed through a 0.2 μm membrane filter before analyses.

High-pressure liquid chromatography

The HPLC system consisted of a Shimadzu pump LC-8A Preparative Liquid Chromatograph coupled to the SCL-10A VP System Controller, an SPD-10A UV-Vis diode array detector and the FRC-10A Shimadzu Fraction Collector. The analytical column was a Kromasil 100 C-8, (5 microns 150 × 4.6 mm analytical, or 5 microns 150 × 20 mm for preparative experiments) with the corresponding Kromasil C-8 guard column, using an isocratic mobile phase consisting of 80 % water with 0.1 % trifluoroacetic acid and 20 % acetonitrile. The flow rate was 1 mL/min for analytical experiments, with 20 μL injections, and a 20 mL/min flow rate with 3 - 4 mL injection volumes for preparative experiments. The analytes were detected at a wavelength of 254 nm during 20-minute runs. For the preparative runs, the retention time for the compound of interest (1) was 12.9 minutes (k′ = 10.1), and 16.3 minutes (k′ = 13.8) for analytical runs. This peak was collected from 28 preparative chromatographic runs, combined, evaporated, re-dissolved in 3 mL and passed once again as a single solution through the preparative column for the final purification of compound 1. Eventually, about 12 mg (0.11 % w/w) of a dark, red-purple amorphous solid was recovered in pure form, according to HPLC, LC-MS and NMR.

Liquid chromatography-mass spectrometry

HPLC-MS data were recorded on an LCQ quadrupole ion trap mass spectrometer (Finnigan, San Jose, CA, USA) connected to a Rheos 4000 HPLC system (Rheos, Danderyd, Sweden). A Supersphere 60 RP-select B column was used to separate the compounds. The eluent consisted of 0.1 % aqueous trifluoroacetic acid in a gradient of 7.5 % - 35 % acetonitrile over 20 minutes. The spray needle was set to 4.5 kV and the spray was stabilized by a nitrogen sheath flow. The inlet capillary temperature was 225 °C. The MS/MS and MS³ fragmentations were further studied by a LTQ quadrupole ion trap mass spectrometer (Finnigan, San Jose, CA, USA). In this study HPLC-purified fractions were injected with a manual Rheodyne injector (5 μL loop model 7125, Cotati, CA, USA) using flow injection analysis (without a column). The mobile phase was 20 % ACN with aqueous 0.1 % trifluoroacetic acid, at a flow rate of 10 μL/min. The full scan mass spectra were recorded in the positive mode over a range of m/z = 150 - 1000. The MS/MS traces were measured using helium as the collision gas, and the collision energy was optimized individually for each parent ion. The accurate mass of the purified phytoindole was measured using an Ultima Autospec high-resolution mass spectrometer (Micromass, UK) in the electrospray ionization mode.

Nuclear magnetic resonance spectroscopy

NMR experiments were performed at 300 K on a Bruker Avance DRX 500 spectrometer equipped with a normal 5 mm QNP probe. The entire sample (ca. 12 mg) was dissolved in 130 μL of CD₃OD and measured in a 2.5 mm microtube, using TMS as an internal standard. Structure assignment was based on normal ¹H proton decoupled ¹³C and DEPT-135 (where CH and CH₃ are up and CH₂ signals down) NMR spectra, two-dimensional (2D) homonuclear ¹H-¹H correlated COSY (cosygpqf) and long range (LR) ¹H-¹H COSY (cosylrgf), heteronuclear ¹ J _CH correlated ¹H-¹³C HSQC (hsqcetgpsi) and long range ¹H-¹³C HMBC (hmbcgpndqf) NMR spectra, measured by standard Bruker pulse programs [8]. Exact chemical shifts and coupling constants for all protons were determined by PERCH software [9]. Assigned chemical shifts and coupling constants for protons and carbons are listed in Table [1] as long-range ^{2 - 4} J _CH (HMBC) correlations. Strong correlation signals, such as ³ J _CH couplings in aromatic systems, are in bold type.

Results and Discussion

Our initial investigations into M. tenuiflora bark samples began in late 1994 with analytical HPLC, using diode array detection, which revealed the presence of DMT at a retention time of 7.7 minutes (k′ = 5.2) plus another strong signal of near equivalent response with a retention time of 16.3 minutes (k′ = 13.8) (see Supporting Information Fig. [1]). Subsequently, an authentic sample of vinho de jurema [sic] was provided by Prof. Emeritus Bo Holmstedt of the Karolinska in the Spring of 1995. This sample was reportedly prepared from M. hostilis in Paraiba, Brazil by traditional methods, and given to Prof. Holmstedt by a Yurema priestess in 1983. Analysis of this sample in his laboratory by GC-MS had detected only DMT and no other major alkaloids. Surprisingly similar results were noticed between our earlier HPLC analysis of the M. tenuiflora bark sample from Mexico and the authentic sample of yurema beverage from Prof. Holmstedt.

The LC-MS chromatogram obtained from the M. tenuiflora extract also showed two major compounds; DMT with a retention time of 8.50 (k′ = 5.0) min and compound 1 with a retention time of 12.9 min (k′ = 7.6) (Fig. [1] A). The mass spectrum of DMT gave a protonated molecular ion at m/z = 189 and a fragment ion at m/z = 144, caused by the immediate loss of HN(CH₃)₂ (Fig. [1] B). The mass spectrum of compound 1 also shows a protonated molecular ion at m/z = 477 with no major fragment ions (Fig. [1] C). These two main components from the plant extract were cleanly separated by reversed phase HPLC prior to subsequent off-line electrospray ionization LC-MS. A measured accurate mass of 477.2041 corresponded well with an elemental composition of C₂₇H₂₉N₂O₆ (calculated 477.2026). The MS/MS trace of compound 1 gave an m/z = 432 ion as a base peak, indicating a loss of HN(CH₃)₂ from the ethylamine side chain, with very few other ions (see Supporting Information Fig. [2] A). The MS³ trace (Fig. [2]) was used to further verify the molecular structure of compound 1, showing ions at m/z = 264 and 414 (see Supporting Information Fig. [2] B). A characteristic parent ion at m/z = 189.1380 and fragmentation was consistent with DMT, the other major alkaloid in this plant sample. High resolution LC-MS and NMR experiments eventually led to the structure of compound 1, which is proposed in Fig. [3].

The ethylamine substituent was easily identified by ¹H- and ¹³C-NMR experiments for compound 1, which was confirmed by results from HMBC correlations (Table [1]). The remaining difficulty was to assign carbon C9a, as C9 had a small ⁴ J _CH correlation signal with proton B8, while C9a had only a weak correlation to proton D1 and a strong correlation to D2, with no detectable couplings to the CH₂-protons of the indole ethylamine side chain. Following this assignment strategy, chemical shifts for the indole system were found to be in good agreement with literature values [10].

Proton assignments for the phenolic E ring of compound 1 were characteristic and straight-forward from ¹H chemical shifts. Based on the ¹³C-NMR shifts, the hydroxy groups on the E ring had to be meta to each other, due to large downfield shifts. The pyrogallol ring A was more difficult to recognize, as equivalent carbon signals A3′′/5′′ and A2′′/6′′ were only ca. 30 % higher, and not twice as high, when compared to related neighbouring peaks.

Aliphatic methyl and methylene signals were typical and easily assigned from the ¹³C-NMR DEPT spectrum [10]. The only difficulty remained in the assignment of non-equivalent chemical shifts for the two methyl substituents on the aliphatic nitrogen and NCH₂-protons, which is typically difficult for chiral systems. There is an obvious hydrogen bond in compound 1 from the E2′-OH proton with the lone pair electrons on the aliphatic nitrogen (Fig. [3]), which could protect this aliphatic amine from enzymatic metabolism by monoamine oxidase (MAO).

Based on the 2D ¹H-¹H- and ¹H-¹³C-NMR correlated experiments, the remaining three aliphatic protons and carbons were observed to form a -CH-CH-CH- chain with the required substituents. Based on long-range 2D ¹H-¹³C correlated results, one end of this chain was found to be adjacent to indole C and phenolic E rings, while the other end was close to the pyrogallol A ring. There were also strong connections between protons D1-E6′ and D3-A3′′/5′′ in the ¹H-¹H LR COSY. The ³ J _HH coupling constants for protons D1-D2 and D2-D3 were typical for trans-trans-substituted protons in five-membered rings [11].

Initially we applied standard acid-base extraction methods to the stem bark sample, which resulted in a wide variety of complex, unidentified degradation products. Compound 1 is clearly unstable, which may account for the identification of only DMT in previous phytochemical investigations of both M. hostilis and M. tenuiflora when using basic extraction conditions during sample preparation. The earliest phytochemical analysis of this Mimosa species (as Mimosa hostilis) identified only one alkaloid, which may or may not have been DMT [2]. The proposed structure in Fig. [3] also explains the chemical instability of this molecule, especially under basic conditions.

Compound 1 is not only a new compound, but also a new class of phytoindoles. According to IUPAC guidelines, this new structure can be named as 1′′-[1-(2′,4′-dihydroxyphenyl)-9-(2-dimethylaminoethyl)-2-hydroxy-2,3-dihydro-1H-3a-azacyclopenta[a]inden-1-yl]benzene-3′′,4′′,5′′-triol, with a suggested common name of yuremamine. The structural assignment was based on mass spectral data and both one and two dimensional ¹H- and ¹³C-NMR experiments.

Yurema is still used for its psychoactive properties in traditional religions of Brazil [1]. However, a pharmacological explanation for the visionary effects from this sacred beverage is lacking, as DMT is not orally active because of its rapid enzymatic metabolism by MAO. Ayahuasca, another sacred visionary beverage from Brazil, typically contains DMT (from the leaves of Psychotria viridis) but another plant is required to render DMT orally active; Banisteriopsis caapi, which contains the potent MAO inhibitor harmine [12], [13]. It is suggested herein that intramolecular hydrogen bonding of the tertiary aliphatic nitrogen of yuremamine protects it from metabolism and could allow it to act as an inhibitor of MAO, thus facilitating the oral activity of DMT in this single-plant formulation. Presently, the putative pharmacology of purified yuremamine is unknown.

Acknowledgements

Our thanks to Anita Hemmilä for linguistic contributions, Prof. Ulysses Paulino de Albuquerque for cultural information on yurema/jurema and Jukka Leppänen, Ph. D. for technical assistance with HPLC. This work is dedicated to the memory of Prof. Emeritus Bo Holmstedt. This work was partly supported by grant number 204 755 from the Academy of Finland.

References

• 1 Paulino de Albuquerque U. The use of medicinal plants by the cultural descendents of African people in Brazil.  Acta Farm Bonaerense. 2001;  20 139-44
  PubMedSearch in Google Scholar
• 2 Gonçalves de Lima O. Archivos do Instituto de Pesquisas.  Agronomicas. 1946;  4 45
  PubMedSearch in Google Scholar
• 3 Lozoya X, Navarro V, Arnason J T, Kourany E. Experimental evaluation of Mimosa tenuiflora (Willd.) Poir. (Tepescohuite) I. Screening of the antimicrobial properties of bark extracts.  Arch Invest Med (Méx). 1989;  20 87-93
  PubMedSearch in Google Scholar
• 4 Meckes-Lozoya M, Lozoya X, Gonzales J L, Martinez M. Effect produced by the alkaloid fraction of Mimosa tenuiflora (tepescohuite) on the peristaltic reflex of the guinea pig ileum.  Arch Invest Med (Méx). 1990;  21 171-4
  PubMedSearch in Google Scholar
• 5 Pachter I L, Zacharias D E, Ribeiro O. Indole alkaloids of Acer saccharinum (the silver maple), Dictyoloma incanescens, Piptadenia colubrine, and Mimosa hostilis .  J Org Chem. 1959;  24 1285-7
  CrossrefPubMedSearch in Google Scholar
• 6 Meckes-Lozoya M, Lozoya X, Marles R J, Souchy-Breau C, Varasen A, Arnason J T. N,N-Dimethyltryptamine alkaloid in Mimosa tenuiflora bark (tepescohuite).  Arch Invest Med (Méx). 1990;  21 175-7
  PubMedSearch in Google Scholar
• 7 Jiang Y, Massiot G, Lavaud C, Teulon J -M, Guéchot C, Haag-Berrurier M. et al . Triterpenoid glycosides from the bark of Mimosa tenuiflora .  Phytochemistry. 1991;  30 2357-60
  CrossrefPubMedSearch in Google Scholar
• 8 Parella T. In: Pulse program catalogue for Topspin v1.2. NMR Quide v 4.0 Barcelona; Bruker BioSpin GmbH 2004
  Search in Google Scholar
• 9 Laatikainen R, Niemitz M, Weber U, Sundelin J, Hassinen T, Vepsäläinen J. General strategies for total-lineshape-type spectral analysis of NMR spectra using integral-transform iterator (NMR).  J Magn Reson A. 1996;  120 1-10
  PubMedSearch in Google Scholar
• 10 Pretsch E, Bühlmann P, Affolter C. In: Structure Determination of Organic Compounds. Heidelberg; Springer Verlag 2000
  Search in Google Scholar
• 11 Padwa A, Fryxell G E, Gasdaska J R, Venkatramanan M K, Wong G SK. A dipolar cycloaddition approach to pyrrolo[1,2-a]indoles using N-[(trimethylsily1)methyl]-substituted indoles.  J Org Chem. 1989;  54 644-53
  CrossrefPubMedSearch in Google Scholar
• 12 Callaway J C, McKenna D J, Grob C S, Brito G S, Raymon L P. et al . Pharmacokinetics of Hoasca alkaloids in healthy humans.  J Ethnopharmacol. 1999;  65 243-56
  CrossrefPubMedSearch in Google Scholar
• 13 Ott J. Ayahuasca analogues - Pangæan Entheogens. Kennewick, WA, USA; Natural Products 2004
  Search in Google Scholar

J. C. Callaway, Ph. D.

Department of Pharmaceutical Chemistry

University of Kuopio

PL 1627

FIN-70211 Kuopio

Finland

Phone: +358-40-725-2534

Fax: +358-17-162-456

Email: callaway@uku.fi

References

• 1 Paulino de Albuquerque U. The use of medicinal plants by the cultural descendents of African people in Brazil.  Acta Farm Bonaerense. 2001;  20 139-44
  PubMedSearch in Google Scholar
• 2 Gonçalves de Lima O. Archivos do Instituto de Pesquisas.  Agronomicas. 1946;  4 45
  PubMedSearch in Google Scholar
• 3 Lozoya X, Navarro V, Arnason J T, Kourany E. Experimental evaluation of Mimosa tenuiflora (Willd.) Poir. (Tepescohuite) I. Screening of the antimicrobial properties of bark extracts.  Arch Invest Med (Méx). 1989;  20 87-93
  PubMedSearch in Google Scholar
• 4 Meckes-Lozoya M, Lozoya X, Gonzales J L, Martinez M. Effect produced by the alkaloid fraction of Mimosa tenuiflora (tepescohuite) on the peristaltic reflex of the guinea pig ileum.  Arch Invest Med (Méx). 1990;  21 171-4
  PubMedSearch in Google Scholar
• 5 Pachter I L, Zacharias D E, Ribeiro O. Indole alkaloids of Acer saccharinum (the silver maple), Dictyoloma incanescens, Piptadenia colubrine, and Mimosa hostilis .  J Org Chem. 1959;  24 1285-7
  CrossrefPubMedSearch in Google Scholar
• 6 Meckes-Lozoya M, Lozoya X, Marles R J, Souchy-Breau C, Varasen A, Arnason J T. N,N-Dimethyltryptamine alkaloid in Mimosa tenuiflora bark (tepescohuite).  Arch Invest Med (Méx). 1990;  21 175-7
  PubMedSearch in Google Scholar
• 7 Jiang Y, Massiot G, Lavaud C, Teulon J -M, Guéchot C, Haag-Berrurier M. et al . Triterpenoid glycosides from the bark of Mimosa tenuiflora .  Phytochemistry. 1991;  30 2357-60
  CrossrefPubMedSearch in Google Scholar
• 8 Parella T. In: Pulse program catalogue for Topspin v1.2. NMR Quide v 4.0 Barcelona; Bruker BioSpin GmbH 2004
  Search in Google Scholar
• 9 Laatikainen R, Niemitz M, Weber U, Sundelin J, Hassinen T, Vepsäläinen J. General strategies for total-lineshape-type spectral analysis of NMR spectra using integral-transform iterator (NMR).  J Magn Reson A. 1996;  120 1-10
  PubMedSearch in Google Scholar
• 10 Pretsch E, Bühlmann P, Affolter C. In: Structure Determination of Organic Compounds. Heidelberg; Springer Verlag 2000
  Search in Google Scholar
• 11 Padwa A, Fryxell G E, Gasdaska J R, Venkatramanan M K, Wong G SK. A dipolar cycloaddition approach to pyrrolo[1,2-a]indoles using N-[(trimethylsily1)methyl]-substituted indoles.  J Org Chem. 1989;  54 644-53
  CrossrefPubMedSearch in Google Scholar
• 12 Callaway J C, McKenna D J, Grob C S, Brito G S, Raymon L P. et al . Pharmacokinetics of Hoasca alkaloids in healthy humans.  J Ethnopharmacol. 1999;  65 243-56
  CrossrefPubMedSearch in Google Scholar
• 13 Ott J. Ayahuasca analogues - Pangæan Entheogens. Kennewick, WA, USA; Natural Products 2004
  Search in Google Scholar

J. C. Callaway, Ph. D.

Department of Pharmaceutical Chemistry

University of Kuopio

PL 1627

FIN-70211 Kuopio

Finland

Phone: +358-40-725-2534

Fax: +358-17-162-456

Email: callaway@uku.fi

• Supplementary Material