PeroxiBase: A class III plant peroxidase database

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/7330310

PeroxiBase: A Class III plant peroxidase database ARTICLE in PHYTOCHEMISTRY · APRIL 2006 Impact Factor: 2.55 · DOI: 10.1016/j.phytochem.2005.12.020 · Source: PubMed

CITATIONS

READS

61

296

7 AUTHORS, INCLUDING: Vassilios Ioannidis

Claudia Cosio

Swiss Institute of Bioinformatics

University of Geneva

19 PUBLICATIONS 1,315 CITATIONS

25 PUBLICATIONS 1,251 CITATIONS

SEE PROFILE

SEE PROFILE

Laurent Falquet

Christophe Dunand

Université de Fribourg

Laboratoire De Recherche En Sciences Vég…

58 PUBLICATIONS 6,148 CITATIONS

59 PUBLICATIONS 3,001 CITATIONS

SEE PROFILE

SEE PROFILE

Available from: Laurent Falquet Retrieved on: 05 February 2016

PHYTOCHEMISTRY

www.elsevier.com/locate/phytochem

Phytochemistry Vol. 67, No. 6, 2006

Contents MOLECULES OF INTEREST Polygalacturonase-inhibiting protein (PGIP) in plant defence: a structural view

pp 528–533

Adele Di Matteo, Daniele Bonivento, Demetrius Tsernoglou, Luca Federici, Felice Cervone * Polygalacturonase-inhibiting proteins are produced by plants to counteract the activity of pathogen-derived polygalacturonases and limit the aggressive potential of phytopathogenic microorganisms. Here, we describe the structure of the inhibitor and focus on our current understanding of its interaction with polygalacturonases.

UPDATE IN BIOINFORMATICS PeroxiBase: A class III plant peroxidase database

pp 534–539

Nenad Bakalovic, Filippo Passardi, Vassilios Ioannidis, Claudia Cosio, Claude Penel, Laurent Falquet, Christophe Dunand * PeroxiBase is a database entirely devoted to the class III plant peroxidase and accessible through a web server http://peroxidase.isb-sib.ch. An exhaustive data mining performed from various EST projects has allowed drawing up the evolution of the class III across land plants evolution.

PROTEIN BIOCHEMISTRY Substrate specificity of acyl-D6-desaturases from Continental versus Macaronesian Echium species Federico Garcı´a-Maroto, Aurora Man˜as-Ferna´ndez, Jose´ A. Garrido-Ca´rdenas, Diego Lo´pez Alonso * Biased substrate specificity of the acyl-D6-desaturases was found to be one of the determinants of the different fatty acid profiles exhibited by Macaronesian versus Continental Echium (Boraginaceae) species.

pp 540–544

524

Contents / Phytochemistry 67 (2006) 523–527

Purification and primary structure determination of two Bowman–Birk type trypsin isoinhibitors from Cratylia mollis seeds

pp 545–552

P.M.G. Paiva, M.L.V. Oliva *, H. Fritz, L.C.B.B. Coelho, C.A.M. Sampaio

CmTI1: C-T-Y-S-M-P-G-M-C The separation of isoinhibitors from Cratylia mollis seeds (CmTI1 and CmTI2) by reverse phase chromatography allowed sequence determination and characterization of structural heterogeneity.

CmTI2: C-T-L-S-F-P-A-Q-C

MOLECULAR GENETICS AND GENOMICS Overexpression of the Saussurea medusa chalcone isomerase gene in S. involucrata hairy root cultures enhances their biosynthesis of apigenin

pp 553–560

Feng-Xia Li, Zhi-Ping Jin, De-Xiu Zhao *, Li-Qin Cheng, Chun-Xiang Fu, Fengshan Ma

Saussurea involucrata Chalcone isomerase

Chalcone isomerase (CHI) is a key enzyme of the flavonoid biosynthesis pathway. The chi gene from Saussurea medusa was introduced into hairy roots of Saussurea involucrata using Agrobacterium rhizogenes. Transgenic hairy root lines produced significantly higher levels of total flavonoids as well as apigenin.

Agrobacterium rhizogenes

Transformed root cultures

Enhanced production of apigenin

METABOLISM Natural and directed biosynthesis of communesin alkaloids

pp 561–569

Lucy J. Wigley, Peter G. Mantle *, David A. Perry Roles for acetate, mevalonate, methionine, tryptophan and tryptamine have been established for natural communesins, containing two indole moieties, but addition of D L -6-fluorotryptophan directed biosynthesis only to monofluorinated analogues.

In vitro shoot and root organogenesis, plant regeneration and production of tropane alkaloids in some species of Schizanthus

pp 570–578

Miguel Jordan *, Munir Humam, Stefan Bieri, Philippe Christen, Estrella Poblete, Orlando Mun˜oz

N

1

7

Ten alkaloids ranging from simple pyrrolidine derivatives to tropane esters derived from angelic, tiglic, senecioic or methylmesaconic acids were obtained from in vitro regenerated plantlets Schizanthus hookeri Gill. (Solanaceae), an endemic Chilean plant. One of them, 3a-methylmesaconyloxytropane, is hitherto unknown.

6

4

O

3

O

O O

Contents / Phytochemistry 67 (2006) 523–527

525

C-prolinylquercetins from the yellow cocoon shell of the silkworm, Bombyx mori

pp 579–583

ECOLOGICAL BIOCHEMISTRY

Chikara Hirayama *, Hiroshi Ono, Yasumori Tamura, Masatoshi Nakamura Two C-prolinylquercetins, prolinalin A and B, were isolated from the cocoon shell of the silkworm, Bombyx mori. Their structures were elucidated by spectroscopic methods and chemical evidence.

A hemiterpene glucoside as a probing deterrent of the bean aphid, Megoura crassicauda, from a non-host vetch, Vicia hirsuta

pp 584–588

Naohiro Ohta, Naoki Mori, Yasumasa Kuwahara, Ritsuo Nishida

*

A specific probing deterrent for the bean aphid, Megoura crassicauda, was isolated from a non-host tiny vetch, Vicia hirsuta, and identified as (E)-2-methyl-2-butene-1,4-diol 4-O-b-D glucopyranoside.

HO HO HO

O O OH OH

Volatile oil from Guarea macrophylla ssp. tuberculata: Seasonal variation and electroantennographic detection by Hypsipyla grandella

pp 589–594

Joa˜o Henrique G. Lago *, Marisi G. Soares, Luciane G. Batista-Pereira, M. Fa´tima G.F. Silva, Arlene G. Correˆa, Joa˜o B. Fernandes, Paulo C. Vieira, Nı´dia F. Roque Leaf oil of Guarea macrophylla (Meliaceae) is composed mainly of sesquiterpenes, whose degree of oxygenation is seasonally dependent. Analysis of the oil by GC-electroantennographic detection employing antennae of Hypsipyla grandella females revealed five components that could be responsible for the attraction of this insect pest to the plant.

BIOACTIVE PRODUCTS Gluconic acid: An antifungal agent produced by Pseudomonas species in biological control of take-all Rajvinder Kaur, John Macleod, William Foley, Murali Nayudu

pp 595–604 COOH

*

Pseudomonas strain AN5 produces D -gluconic acid which acts as an antifungal agent in suppressing the take-all fungal pathogen, both on potato dextrose media and in the wheat rhizosphere. Transposon mutants which have lost biocontrol do not produce gluconic acid, but restoration of gluconic acid production by complementation restores take-all biocontrol.

H

C

OH

HO

C

H

H

C

OH

H

C

OH

CH2 OH D-Gluconic acid

526

Contents / Phytochemistry 67 (2006) 523–527

Newbouldiaquinone A: A naphthoquinone–anthraquinone ether coupled pigment, as a potential antimicrobial and antimalarial agent from Newbouldia laevis Kenneth Oben Eyong, Gabriel Ngosong Folefoc *, Victor Kuete, Veronique Penlap Beng, Karsten Krohn *, Hidayat Hussain, Augustin Ephram Nkengfack, Michael Saeftel, Salem Ramadan Sarite, Achim Hoerauf

pp 605–609

O 1

8 7

9

6

10 5

2 4

3

O O

3'

O

A naphthoquinone–anthraquinone ether-coupled pigment, Newbouldiaquinone A, a potential antimicrobial and antimalarial agent, was isolated from Newbouldia laevis together with 14 known compounds.

RO

1 : R=H

5' 4'

6'

1'

7'

2'

8'

O

Vasodilatory and hypoglycaemic effects of two pyrano-isoflavone extractives from Eriosema kraussianum N. E. Br. [Fabaceae] rootstock in experimental rat models

pp 610–617

John A.O. Ojewole *, Siegfried E. Drewes, Fatima Khan Two pyrano-isoflavones (e.g., 1) produced dose-related initial transient contractions followed by secondary longer-lasting significant relaxation of rat isolated portal veins. The two compounds also caused dose-dependent hypoglycaemia in rats.

O

O

OH

HO

O

(1)

An angiotensin-I converting enzyme inhibitor from buckwheat (Fagopyrum esculentum Moench) flour Yasuo Aoyagi

pp 618–621

*

A potent angiotensin-I converting enzyme innibitor 200 -hydroxynicotianamine was isolated from buckwheat (Fagopyrum esculentum Moench) flour.

CHEMISTRY Structure of anthocyanin from the blue petals of Phacelia campanularia and its blue flower color development Mihoko Mori, Tadao Kondo, Kenjiro Toki, Kumi Yoshida

*

The dicaffeoyl anthocyanin, phacelianin, was isolated from the blue petals of Phacelia campanularia and its structure was determined. Phacelianin may take both an inter- and intramolecular stacking form and shows the blue petal color by molecular association and the co-existence of a small amount of metal ions.

pp 622–629

Contents / Phytochemistry 67 (2006) 523–527

527

OTHER CONTENTS Book review Announcement: Phytochemical Society of North America Author Index Guide for Authors

p 630 pI p II pp III–IV

* Corresponding author

The Editors encourage the submission of articles online, thus reducing publication times. For further information and to submit your manuscript, please visit the journal homepage at http://www.elsevier.com/locate/phytochem

IN D E X E D /A B S T R A C T E D I N : Current Awareness in Biological Sciences (CABS), Curr Cont ASCA. Chem. Abstr. BIOSIS Data, PASCALCNRS Data, CAB Inter, Cam Sci Abstr, Curr Cont/Agri Bio Env Sci, Curr Cont/Life Sci, Curr Cont Sci Cit Ind, Curr Cont SCISEARCH Data, Bio Agri Ind ISSN 0031-9422

PHYTOCHEMISTRY Phytochemistry 67 (2006) 528–533 www.elsevier.com/locate/phytochem

Molecules of Interest

Polygalacturonase-inhibiting protein (PGIP) in plant defence: a structural view Adele Di Matteo a,b, Daniele Bonivento b, Demetrius Tsernoglou b, Luca Federici c, Felice Cervone a,* a

Dipartimento di Biologia Vegetale, Universita’ di Roma ‘‘La Sapienza’’, Piazzale Aldo Moro 5, 00185 Roma, Italy Dipartimento di Scienze Biochimiche, Universita’ di Roma ‘‘La Sapienza’’, Piazzale Aldo Moro 5, 00185 Roma, Italy Ce.S.I. Centro Studi sull’Invecchiamento and Dipartimento di Scienze Biomediche, Universita’ di Chieti ‘‘G. D’Annunzio’’, Via dei Vestini 31, 66013 Chieti, Italy b

c

Received 14 December 2005; accepted 16 December 2005 Available online 3 February 2006

Abstract Polygalacturonase-inhibiting proteins are plant extracellular leucine-rich repeat proteins that specifically bind and inhibit fungal polygalacturonases. The interaction with PGIP limits the destructive potential of polygalacturonases and might trigger the plant defence responses induced by oligogalacturonides. A high degree of polymorphism is found both in PGs and PGIPs, accounting for the specificity of different plant inhibitors for PGs from different fungi. Here, we review the structural features and our current understanding of the PG–PGIP interaction. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Plant innate immunity; Polygalacturonase; Polygalacturonase inhibiting protein; Crystal structure; Protein–protein interaction

1. Introduction Plant innate immunity is based on an ancient system of molecules that defend the host against infection. Defence relies on the capability of each cell to recognize the presence of pathogens and activate downstream responses. The first defence line is the plant cell wall, a physiological barrier that has a critical role in controlling pathogen invasion. Microorganisms have a number of enzymes that degrade the polysaccharides of the cell wall. Endo-polygalacturonase (PG), one of the enzymes secreted at the early stages of infection, depolymerizes the homogalacturonan, the main component of pectin, by cleaving the a-1,4 glycosidic bonds between the galacturonic acid units (De Lorenzo et al., 2001).

*

Corresponding author. E-mail address: [email protected] (F. Cervone).

0031-9422/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.025

PGs are produced by bacteria, fungi, nematodes and insects (De Lorenzo and Ferrari, 2002; Jaubert et al., 2002; Girard and Jouanin, 1999) and their involvement in pathogenesis has been demonstrated for several fungi such as Botrytis cinerea (ten Have et al., 1998; Kars et al., 2005), Aspergillus flavus (Shieh et al., 1997), Alternaria citri (Isshiki et al., 2001), Claviceps purpurea (Oeser et al., 2002) and Sclerotinia sclerotiorum (Li et al., 2004) and bacteria such as Ralstonia solanacearum (Tans-Kersten et al., 2001) and Agrobacterium tumefaciens (Rodriguez-Palenzuela et al., 1991). PGs from salivary glands of phytophagous insects are considered a main cause of plant damage (Girard and Jouanin, 1999; Boyd et al., 2002; Frati et al., in press). Many plants produce extracellular polygalacturonase-inhibiting proteins (PGIPs) that specifically recognize and inhibit fungal and insect PGs (De Lorenzo and Ferrari, 2002; D’Ovidio et al., 2004). The PG–PGIP interaction limits the destructive potential of polygalacturonases and leads to the accumulation of elicitor active oligogalacturo-

A. Di Matteo et al. / Phytochemistry 67 (2006) 528–533

nides as demonstrated in vitro. These oligosaccharides may activate plant defence responses such as synthesis of phytoalexins, lignin and ethylene, expression of proteinase inhibitor I and b-1,3-glucanase and production of reactive oxygen species (Ridley et al., 2001). The importance of PGIPs in plant defence has been corroborated by in vivo studies. The overexpression of the genes Atpgip1 and Atpgip2 in Arabidopsis limits the colonization by B. cinerea and reduces disease symptoms (Ferrari et al., 2003). A significant increase of PG-inhibitory activity and a decrease in susceptibility to B. cinerea has been found in transgenic tomato and grapevine plants overexpressing a pear pgip (Powell et al., 2000; Agu¨ero et al., 2005), and in tobacco and Arabidopsis plants overexpressing a bean pgip (Pvpgip2) (Manfredini et al., in press). Arabidopsis plants with antisense expression of pgip genes have a reduced inhibitory activity in response to biotic and abiotic stress and are more susceptible to B. cinerea (our unpublished data). These findings suggest that PGIPs are important players in plant innate immunity. To accommodate pathogenesis to different environmental conditions and on various hosts, fungi produce PG isoenzymes variable in terms of sequence, specific activity, pH optimum and substrate preference (De Lorenzo et al., 2001; Poinssot et al., 2003). Conversely, plants have evolved PGIPs with different recognition specificities encoded by differentially regulated pgip genes (De Lorenzo and Ferrari, 2002; Ferrari et al., 2003). Also plants produce PGs that play a role in the cell wall development (Torki et al., 2000) but these PGs do not interact with PGIPs, suggesting that the inhibitors are specialized for plant defence (Federici et al., 2001). The finding that transgenic tobacco and Arabidopsis plants overexpressing PGIPs do not show morphological alterations is consistent with this vision (Ferrari et al., 2003; Capodicasa et al., 2004; Manfredini et al., in press). PGIPs belong to the leucine-rich repeat (LRR) superfamily of proteins (Kobe and Kajava, 2001) and are characterized by the tandem repetition of a consensus motif rich of leucines (Mattei et al., 2001). LRR proteins are ubiquitous in the life kingdoms and provide a versatile recognition surface for protein–protein interactions. The LRR domain of plant proteins is often fused with other domains and acts in recognition of hormonal signals and pathogenderived elicitors. Plant LRR proteins involved in plant defence include: (i) those containing an extracellular LRR domain and a single membrane-spanning helix, such as the Cf proteins of tomato (van der Hoorn et al., 2005); (ii) proteins containing an extracellular LRR domain and a cytoplasmic serine-threonine kinase, connected by a transmembrane helix, such as the rice resistance gene products Xa21 (Song et al., 1995) and Xa26 (Sun et al., 2004); (iii) cytoplasmic proteins containing a LRR domain, a putative nucleotide binding sites (NBS) and a N-terminal putative leucine-zippers (LZ) or Toll interleukine receptor (TIR) domain (Staskawicz et al., 2001). LRRs of the extracellular type homologous to PGIP are found in pattern-recognition receptors involved in non-host

529

specific defence such as the Arabidopsis flagellin receptor FLS2 (Gomez-Gomez et al., 2001), in carrot proteins with antifreeze activity (Worrall et al., 1998), in the Arabidopsis protein SHY with a role in pollen tube growth (Guyon et al., 2004), in several receptors involved in development, perception of hormones (Becraft, 2002) and in bacterial and fungal symbiosis (Kistner and Parniske, 2002). Plant proteins with extracytoplasmic LRRs (eLRRs) are specifically characterized by the tandem repetition of the same 24-residues motif: xxLxLxxNxLt/sGxIPxxLxxLxxL. This repeat differs from those found in plant cytoplasmic LRRs as well as other LRR subfamilies thus suggesting that eLRR proteins might share the same three-dimensional arrangement (Kobe and Kajava, 2001; Di Matteo et al., 2003).

2. The structure of PGIP is the prototype of the plant eLRR proteins The crystal structure of the isoform 2 of PGIP from Phaseolus vulgaris (PvPGIP2) is the only available structure of plant LRR proteins (Di Matteo et al., 2003) (Fig. 1). PvPGIP2 displays the right-handed superhelical

Fig. 1. Ribbon representation of the crystal structure of PGIP2 from P. vulgaris.

530

A. Di Matteo et al. / Phytochemistry 67 (2006) 528–533

fold typical of LRR proteins such as the porcine and human ribonuclease inhibitors (Kobe and Deisenhofer, 1993; Papageorgiou et al., 1997), U2A 0 (Price et al., 1998), RanGAP rna1p (Hillig et al., 1999), Internalin B (Marino et al., 1999), YopM (Evdokimov et al., 2001), Skp2 (Schulman et al., 2000) and decorin (Scott et al., 2004). The fold of PvPGIP2 consists of a central LRR domain (residues 53–289) flanked by the N- and C-terminal cysteine-rich regions (residues 1–52 and residues 290–313, respectively). The LRR domain is characterized by the tandem repetition of 10 coils matching the consensus sequence xxLxLxxNxLt/sGxIPxxLxxLxxL. An extended parallel bsheet (B1), conserved in all known LRR protein structures (Kobe and Kajava, 2001), occupies the concave inner side of the protein solenoid. B1 is the b-sheet where the residues determining the affinity and the specificity of PGIP2 reside (Leckie et al., 1999; Sicilia et al., 2005). Nine 310-helices are located on the convex side of the protein. While the majority of LRR proteins have only one b-sheet connected with the helices on the convex side by loops or b-turns, PGIP2, instead, has an additional parallel b-sheet (B2). B2 is distorted because of the twisted shape of the molecule and the variable length of the b-strands. Specific positions (3, 5, 10, 18, 21 and 24) of the LRR repeats are occupied by hydrophobic amino acids, mostly leucines, that point into the interior of the protein scaffold and stabilize the overall fold topology thorough van der Waals interactions (Fig. 2). Position 8 is usually occupied by asparagine residues that are oriented towards the core of the protein and form hydrogen bonds with the mainchain carbonyls or amide groups originating the typical ‘‘asparagine ladder’’. An additional cooperative stabilization pattern across the LRR domain is generated by the conserved serine or threonine residues at position 17. The glycine at position 12, the isoleucine at position 14 and the proline at position 15 are conserved in plant-derived eLRR proteins. In PGIP2 the characteristic sequence Lt/ sGxIP is partially involved in the formation of the sheet B2. Interestingly, glycines show a stereochemistry that is forbidden to any other residue and determine the peculiar bending of sheet B2. Sequence alignments suggest that, like PGIP, other plant eLRR proteins display a second b-sheet that may form an additional surface for interaction (Di Matteo et al., 2003). The convex face of the LRR region of PGIP2 is mostly occupied by unstructured segments that are stabilized through water molecules; these are organized in spines along the structure and form H-bond interactions with the protein backbone (Marino et al., 1999; Evdokimov et al., 2001). While contributing to stabilize the protein scaffold, the water network provides a structural flexibility to the molecule and might facilitate the adaptation of the PGIP scaffold to the surface of its interacting partners (Di Matteo et al., 2003). The N-terminal region contains two disulphide bridges (Cys3–Cys33, Cys34–Cys43) and consists of a 15 residuelong a-helix and a short b-strand that forms H-bonds with residues of the sheet B1. The C-terminal region contains

Fig. 2. A typical LRR of PvPGIP2 matching the consensus motif: 1 xxLxLxxNxLt/sGxIPxxLxxLxxL24.

two disulphide bonds (Cys281–Cys303, Cys305–Cys312) and consists of the last two 310 helices, the last strand of the sheet B2 and a short loop. The LRR-flanking regions play a structural role in capping the hydrophobic core of the protein (see Fig. 1). PGIP is ionically anchored to the cell wall and interacts in vitro with unesterified homogalacturonan and with blockwise de-esterified homogalacturonan by means of positive residues, located between the sheets B1 and B2 and protruding into the solvent to create a regular distribution of charges (Fig. 3). The interaction of PGIP with pectin is competed in vitro by PGs, suggesting that PGIP use overlapping although not identical regions to interact with pectin and PGs. The interaction of PGIP with the cell wall may be instrumental to ensure the presence of the inhibitor where infection initiates and microbial PGs degrade homogalacturonan (our unpublished data).

3. The pgip gene families Gene families code for PGIP isoforms with homologous structures but different specificities (De Lorenzo et al., 2001; Li et al., 2003; Ferrari et al., 2003; D’Ovidio et al., 2004). The arrangement and the similarity among the genes of Phaseolus vulgaris suggest that they derive from a com-

A. Di Matteo et al. / Phytochemistry 67 (2006) 528–533

531

Pvpgip3 responds to OGs but not to fungal glucan, salicylic acid or wounding. Pvpgip1 responds only to wounding while Pvpgip2, that encodes the most efficient inhibitor of fungal PGs, is upregulated by all these stress stimuli (D’Ovidio et al., 2004). Both Arabidopsis inhibitors are induced in response to wounding and B. cinerea infection. Atpgip1 expression responds to oligogalacturonides and is independent of salicylic acid, ethylene or jasmonic acid; Atpgip2 is instead induced by jasmonic acid (Ferrari et al., 2003). Moreover Atpgip2 is induced by the fungus Alternaria brassicicola but not by insect attack (Reymond et al., 2000). In Brassica napus PGIP-encoding genes respond differentially to biotic and abiotic stimuli (Li et al., 2003). The ability of pgip genes to switch on in response to different stress-related signals ensures the expression of at least one inhibitor if a pathogen evolves the capacity of blocking or avoiding the activation of a particular transduction pathway.

4. The PG–PGIP interaction

Fig. 3. Electrostatic potential surface of PvPGIP2. Residues are coloured according to their electrostatic potential: blue indicates positive charge while red indicate negative charge. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

mon ancestor as a result of a sequence of duplication– divergence–duplication events. The four paralogues Pvpgip1–4 form two pairs Pvpgip1/Pvpgip2 and Pvpgip3/Pvpgip4 that code for functionally distinct classes of PGIPs, each of which differentially devoted to the recognition of PGs from fungi or insects (D’Ovidio et al., 2004). In Arabidopsis two closely related genes code for two distinct products (AtPGIP1 and AtPGIP2) (Ferrari et al., 2003). In B. napus the pgip family is composed of four different members; two products (BnPGIP1 and BnPGIP2) have been characterized (Li et al., 2003). Pgip genes are different not only in terms of recognition specificity of their products but also because they are differentially regulated (Ferrari et al., 2003). For instance

PGIPs specifically interact with PGs by forming a bimolecular complex. The PG–PGIP interaction varies in terms of inhibition kinetics and strength, and reflects the counteradaptation occurring in both enzymes and inhibitors (Federici et al., in press). Pathogens have evolved different PGs to maximize their offensive potential and, conversely, plants have evolved various PGIPs with different specificities to counteract the many forms of PG existing in nature. For instance PGs from B. cinerea (BcPGs) and Colletotrichum acutatum are inhibited with different efficiencies by all members of the bean PGIP family, while PG from Aspergillus niger (AnPGII) is inhibited by PvPGIP1 and PvPGIP2 but not by PvPGIP3 and PvPGIP4 (D’Ovidio et al., 2004). PG from Fusarium moniliforme (FmPG) is inhibited only by PvPGIP2 (Leckie et al., 1999). PGs from the phytophagous insects Lygus rugulipennis and Adelphocoris lineolatus are inhibited by PvPGIP3 and PvPGIP4 (D’Ovidio et al., 2004). PGIPs from Arabidopsis inhibit PGs from Colletotrichum gloesporioides, Stenocarpella maydis and B. cinerea (Ferrari et al., 2003; Manfredini et al., in press) but are uneffective against FmPG and AnPGII (D’Ovidio et al., 2004). Computational analysis predicts that the hypervariable LxxLxLxx region spanning the sheet B1 in the concave surface of PvPGIP2 has a strong propensity to be engaged in protein–protein interactions (Sicilia et al., 2005; Federici et al., in press). This is the same area where the determinants of PGIP’s affinity and specificity are located (Leckie et al., 1999; Sicilia et al., 2005). Homologous regions of other LRR proteins such as the ribonuclease inhibitor (Papageorgiou et al., 1997), internalin A (Schubert et al., 2002) and RanGAP (Seewald et al., 2002) are also involved in recognition. Phylogenetic codon-substitution analysis performed on PvPGIPs showed that only nine out of 313 residues are positively selected and seven of them are

532

A. Di Matteo et al. / Phytochemistry 67 (2006) 528–533

located within the LRR domain (Bishop, 2005). The small number of positively selected residues suggests that PGIPs require the maintenance of features necessary for the basic recognition of PGs through a network of multiple and weak contacts. A limited number of ‘‘hot spots’’ may be responsible for specificity and their mutations lead to the recognition of new molecules. Variability of recognition and function of PGIPs is not only reflected by their specificity but also by the variable inhibition kinetics played against different fungal PGs: tomato PGIP inhibits AnPGII in a non-competitive manner (Stotz et al., 2000); PGIPs from bean (Lafitte et al., 1984) and raspberry (Johnston et al., 1993) are noncompetitive inhibitors of PGs from Colletotrichum lindemuthianum and B. cinerea, respectively; PGIPs from pear (AbuGoukh and Labavitch, 1983) and orange fruits (Barmore and Nguyen, 1985) are competitive inhibitors of PG from Diplodia natalensis. PvPGIP2 acts competitively against FmPG (Federici et al., 2001), non-competitively against AnPGII (King et al., 2002) and has a mixed mode of inhibition against BcPG1 (Sicilia et al., 2005). This suggests that the orientation of PGIP in the PG–PGIP complexes differs depending on different PG partners. Consistently, molecular docking analysis of the interaction of PvPGIP2 with FmPG, AnPG and BcPG1 predicts that different, although partially overlapping, surfaces of these enzymes are recognized by PGIPs (Sicilia et al., 2005; Federici et al., in press).

Acknowledgements Research is supported by the Armenise-Harvard Foundation, the Institute Pasteur – Fondazione Cenci-Bolognetti and the Ministero per l’Istruzione, l’Universita´ e la Ricerca Scientifica (PRIN 2005).

References Abu-Goukh, A.A., Labavitch, J.M., 1983. The in vivo role of ‘‘Bartlett’’ pear fruit polygalacturonase inhibitors. Physiol. Plant Pathol. 23, 123– 135. Agu¨ero, C.B., Uratsu, S.L., Greve, C., Powell, A.L.T., Labavitch, J.M., Meredith, C.P., Dandekar, A.M., 2005. Evaluation of tolerance to Pierce’s disease and Botrytis in transgenic plants of Vitis vinifera L. expressing the pear PGIP gene. Mol. Plant. Pathol. 6, 43–51. Barmore, C.R., Nguyen, T.K., 1985. Polygalacturonase inhibition in rind of Valencia orange infected with Diplodia natalensis. Phytopathology 75, 446–449. Becraft, P.W., 2002. Receptor kinase signaling in plant development. Annu Rev. Cell. Dev. Biol. 18, 163–192. Bishop, JG., 2005. Directed mutagenesis confirms the functional importance of positively selected sites in polygalacturonase inhibitor protein. Mol. Biol. Evol. 22, 1531–1534. Boyd Jr., D.W., Cohen, A.C., Alverson, D.R., 2002. Digestive enzymes and stylet morphology of Deraeocoris nebulosus (hemiptera: Miridae), a predacious plant bug. Ann. Entomol. Soc. Am. 95, 395–401. Capodicasa, C., Vairo, D., Zabotina, O., McCartney, L., Caprari, C., Mattei, B., Manfredini, C., Aracri, B., Benen, J., Knox, J.P., De

Lorenzo, Cervone, F., 2004. Targeted modification of homogalacturonan by transgenic expression of a fungal polygalacturonase alters plant growth. Plant Physiol. 135, 1294–1304. De Lorenzo, G., Ferrari, S., 2002. Polygalacturonase-inhibiting proteins in defense against phytopathogenic fungi. Curr. Opin. Plant Biol. 5, 295–299. De Lorenzo, G., D’Ovidio, R., Cervone, F., 2001. The role of polygalacturonase-inhibiting proteins (PGIPs) in defense against pathogenic fungi. Annu. Rev. Phytopathol. 39, 313–335. Di Matteo, A., Federici, L., Mattei, B., Salvi, G., Johnson, K.A., Savino, C., De Lorenzo, G., Tsernoglou, D., Cervone, F., 2003. The crystal structure of polygalacturonase-inhibiting protein (PGIP), a leucinerich repeat protein involved in plant defense. Proc. Natl. Acad. Sci. U.S.A. 100, 10124–10128. D’Ovidio, R., Raiola, A., Capodicasa, C., Devoto, A., Pontiggia, D., Roberti, S., Galletti, R., Conti, E., O’Sullivan, D., De Lorenzo, G., 2004. Characterization of the complex locus of bean encoding polygalacturonase-inhibiting proteins reveals subfunctionalization for defense against fungi and insects. Plant Physiol. 135, 2424–2435. Evdokimov, A.G., Anderson, D.E., Routzahn, K.M., Waugh, D.S., 2001. Unusual molecular architecture of the Yersinia pestis cytotoxin YopM: a leucine-rich repeat protein with the shortest repeating unit. J. Mol. Biol. 312, 807–821. Federici, L., Caprari, C., Mattei, B., Savino, C., Di Matteo, A., De Lorenzo, G., Cervone, F., Tsernoglou, D., 2001. Structural requirements of endopolygalacturonase for the interaction with PGIP (polygalacturonase-inhibiting protein). Proc. Natl. Acad. Sci. U.S.A. 98, 13425–13430. Federici, L., Di Matteo, A., Fernandez-Recio, J., Tsernoglou, D., Cervone, F., in press. Polygalacturonase inhibiting proteins: players in plant innate immunity? Trends Plant Sci. Ferrari, S., Vairo, D., Ausubel, F.M., Cervone, F., De Lorenzo, G., 2003. Tandemly duplicated Arabidopsis genes that encode polygalacturonase-inhibiting proteins are regulated coordinately by different signal transduction pathways in response to fungal infection. Plant Cell 15, 93–106. Frati, F., Gallerri, R., De Lorenzo, G., Salerno, G., Conti, E., in press. Endopolygalacturonase activity in mired bugs and their inhibition by plant cell wall proteins (PGIPs). Eur. J. Entomol. Girard, C., Jouanin, L., 1999. Molecular cloning of cDNAs encoding a range of digestive enzymes from aphytophagous beetle, Phaedon cochleariae. Insect Biochem. Mol. Biol. 29, 1129–1142. Gomez-Gomez, L., Bauer, Z., Boller, T., 2001. Both the extracellular leucine-rich repeat domain and the kinase activity of FSL2 are required for flagellin binding and signaling in Arabidopsis. Plant Cell 13, 1155–1163. Guyon, V., Tang, W.H., Monti, M.M., Raiola, A., De Lorenzo, G., McCormick, S., Taylor, L.P., 2004. Antisense phenotypes reveal a role for SHY, a pollen-specific leucine-rich repeat protein, in pollen tube growth. Plant J. 39, 643–654. Hillig, R.C., Renault, L., Vetter, I.R., Drell IV, T., Wittinghofer, A., Becker, J., 1999. The crystal structure of rna1p: a new fold for a GTPase-activating protein. Mol. Cell. 3, 781–791. Isshiki, A., Akimitsu, K., Yamamoto, M., Yamamoto, H., 2001. Endopolygalacturonase is essential for citrus black rot caused by Alternaria citri but not brown spot caused by Alternaria alternata. Mol. Plant-Microbe Interact. 14, 749–757. Jaubert, S., Laffaire, J.B., Abad, P., Rosso, M.N., 2002. A polygalacturonase of animal origin isolated from the root-knot nematode Meloidogyne incognita. FEBS Lett. 522, 109–112. Johnston, D.J., Ramanathan, V., Williamson, B., 1993. A protein from immature raspberry fruits which inhibits endopolygalacturonases from Botrytis cinerea and other micro-organisms. J. Exp. Bot. 44, 971–976. Kars, I., Krooshof, G.H., Wagemakers, L., Joosten, R., Benen, J.A., van Kan, J.A., 2005. Necrotizing activity of five Botrytis cinerea endopolygalacturonases produced in Pichia pastoris. Plant J. 43, 213–225. King, D., Bergmann, C., Orlando, R., Benen, J.A., Kester, H.C., Visser, J., 2002. Use of amide exchange mass spectrometry to study

A. Di Matteo et al. / Phytochemistry 67 (2006) 528–533 conformational changes within the endopolygalacturonase II-homogalacturonan–polygalacturonase inhibiting protein system. Biochemistry 41, 10225–10233. Kistner, C., Parniske, M., 2002. Evolution of signal transduction in intracellular symbiosis. Trends Plant Sci. 7, 511–518. Kobe, B., Deisenhofer, J., 1993. Crystal structure of porcine ribonuclease inhibitor, a protein with leucine-rich repeats. Nature 366, 751–756. Kobe, B., Kajava, A.V., 2001. The leucine-rich repeat as a protein recognition motif. Curr. Opin. Struct. Biol. 11, 725–732. Lafitte, C., Barthe, J.P., Montillet, J.L., Touze, A., 1984. Glycoprotein inhibitors of Colletotrichum lindemuthi-anum endopolygalacturonase in near isogenic lines of Phaseolus vulgaris resistant and susceptible to anthrac-nose. Physiol. Plant Pathol. 25, 39–53. Leckie, F., Mattei, B., Capodicasa, C., Hemmings, A., Nuss, L., Aracri, B., De Lorenzo, G., Cervone, F., 1999. The specificity of polygalacturonase-inhibiting protein (PGIP): a single amino acid substitution in the solvent-exposed beta-strand/beta-turn region of the leucine-rich repeats (LRRs) confers a new recognition capability. EMBO J. 18, 2352–2363. Li, R., Rimmer, R., Yu, M., Sharpe, A.G., Seguin-Swartz, G., Lydiate, D., Hegedus, D.D., 2003. Two Brassica napus polygalacturonase inhibitory protein genes are expressed at different levels in response to biotic and abiotic stresses. Planta 217, 299–308. Li, R., Rimmer, R., Buchwaldt, L., Sharpe, A.G., Seguin-Swartz, G., Hegedus, D.D., 2004. Interaction of Sclerotinia sclerotiorum with Brassica napus: cloning and characterization of endo- and exopolygalacturonases expressed during saprophytic and parasitic modes. Fungal Genet. Biol. 41, 754–765. Manfredini, C., Sicilia, F., Ferrari, S., Pontiggia, D., Salvi, G., Caprari, C., Lorito, M., De Lorenzo, D., in press. Polygalacturonase-inhibiting protein 2 of Phaseolus vulgaris inhibits BcPG1, a polygalacturonase of Botrytis cinerea important for pathogenicity, and protects transgenic plants from infection. Physiol. Mol. Plant Pathol. Marino, M., Braun, L., Cossart, P., Ghosh, P., 1999. Structure of the lnlB leucine-rich repeats, a domain that triggers host cell invasion by the bacterial pathogen L. monocytogenes. Mol. Cell 4, 1063–1072. Mattei, B., Bernalda, M.S., Federici, L., Roepstorff, P., Cervone, F., Boffi, A., 2001. Secondary structure and post-translational modifications of the leucine-rich repeat protein PGIP (polygalacturonase-inhibiting protein) from Phaseolus vulgaris. Biochemistry 40, 569–576. Oeser, B., Heidrich, P.M., Muller, U., Tudzynski, P., Tenberge, K.B., 2002. Polygalacturonase is a pathogenicity factor in the Claviceps purpurea/rye interaction. Fungal Genet. Biol. 36, 176–186. Papageorgiou, A.C., Shapiro, R., Acharya, K.R., 1997. Molecular recognition of human angiogenin by placental ribonuclease inhibitor – an X-ray crystallographic study at 2.0 A resolution. EMBO J. 16, 5162–5177. Poinssot, B., Vandelle, E., Bentejac, M., Adrian, M., Levis, C., Brygoo, Y., Garin, J., Sicilia, F., Coutos-Thevenot, P., Pugin, A., 2003. The endopolygalacturonase 1 from Botrytis cinerea activates grapevine defense reactions unrelated to its enzymatic activity. Mol. PlantMicrobe Interact. 16, 553–564. Powell, A.L., van Kan, J., ten Have, A., Visser, J., Greve, L.C., Bennett, A.B., Labavitch, J.M., 2000. Transgenic expression of pear PGIP in tomato limits fungal colonization. Mol. Plant-Microbe Interact. 13, 942–950. Price, S.R., Evans, P.R., Nagai, K., 1998. Crystal structure of the spliceosomal U2B00 –U2A 0 protein complex bound to a fragment of U2 small nuclear RNA. Nature 394, 645–650. Reymond, P., Weber, H., Damond, M., Farmer, E.E., 2000. Differential gene expression in response to mechanical wounding and insect feeding in Arabidopsis. Plant Cell. 12, 707–720.

533

Ridley, B.L., O’Neill, M.A., Mohnen, D., 2001. Pectins: structure, biosynthesis, and oligogalacturonide-related signaling. Phytochemistry 57, 929–967. Rodriguez-Palenzuela, P., Burr, T.J., Collmer, A., 1991. Polygalacturonase is a virulence factor in Agrobacterium tumefaciens biovar 3. J. Bacteriol. 173, 6547–6552. Schubert, W.D., Urbanke, C., Ziehm, T., Beier, V., Machner, M.P., Domann, E., Wehland, J., Chakraborty, T., Heinz, D.W., 2002. Structure of internalin, a major invasion protein of Listeria monocytogenes, in complex with its human receptor E-cadherin. Cell 111, 825– 836. Schulman, B.A., Carrano, A.C., Jeffrey, P.D., Bowen, Z., Kinnucan, E.R., Finnin, M.S., Elledge, S.J., Harper, J.W., Pagano, M., Pavletich, N.P., 2000. Insights into SCF ubiquitin ligases from the structure of the Skp1–Skp2 complex. Nature 408, 381–386. Scott, P.G., McEwan, P.A., Dodd, C.M., Bergmann, E.M., Bishop, P.N., Bella, J., 2004. Crystal structure of the dimeric protein core of decorin, the archetypal small leucine-rich repeat proteoglycan. Proc. Natl. Acad. Sci. U.S.A. 101, 15633–15638. Seewald, M.J., Korner, C., Wittinghofer, A., Vetter, I.R., 2002. RanGAP mediates GTP hydrolysis without an arginine finger. Nature 415, 662– 666. Shieh, M.T., Brown, R.L., Whitehead, M.P., Cary, J.W., Cotty, P.J., Cleveland, T.E., Dean, R.A., 1997. Molecular genetic evidence for the involvement of a specific polygalacturonase, P2c, in the invasion and spread of Aspergillus flavus in cotton bolls. Appl. Environ. Microbiol. 63, 3548–3552. Sicilia, F., Fernandez-Recio, J., Caprari, C., De Lorenzo, G., Tsernoglou, D., Cervone, F., Federici, L., 2005. The polygalacturonase-inhibiting protein PGIP2 of Phaseolus vulgaris has evolved a mixed mode of inhibition of endopolygalacturonase PG1 of Botrytis cinerea. Plant Physiol. 139, 1380–1388. Song, W.Y., Wang, G.L., Chen, L.L., Kim, H.S., Pi, L.Y., Holsten, T., Gardner, J., Wang, B., Zhai, W.X., Zhu, L.H., Fauquet, C., Ronald, P., 1995. A receptor kinase-like protein encoded by the rice disease resistance gene Xa21. Science 270, 1804–1806. Staskawicz, B.J., Mudgett, M.B., Dangl, J.L., Galan, J.E., 2001. Common and contrasting themes of plant and animal diseases. Science 292, 2285–2289. Stotz, H.U., Bishop, J.G., Brgman, C.W., Koch, M., Albersheim, P., Darwill, A.G., Labavitch, J.M., 2000. Identification of target amino acids that affect interactions of fungal polygalacturonases and their plant inhibitors. Phys. Mol. Plant Pathol. 56, 117–130. Sun, X., Cao, Y., Yang, Z., Xu, C., Li, X., Wang, S., Zhang, Q., 2004. Xa26, a gene conferring resistance to Xanthomonas oryzae pv. oryzae in rice, encodes an LRR receptor kinase-like protein. Plant J. 37, 517–527. Tans-Kersten, J., Huang, H., Allen, C., 2001. Ralstonia solanacearum needs motility for invasive virulence on tomato. J. Bacteriol. 183, 3597–3605. ten Have, A., Mulder, W., Visser, J., van Kan, J.A., 1998. The endopolygalacturonase gene Bcpg1 is required for full virulence of Botrytis cinerea. Mol. Plant-Microbe Interact. 11, 1009–1016. Torki, M., Mandaron, P., Mache, R., Falconet, D., 2000. Characterization of a ubiquitous expressed gene family encoding polygalacturonase in Arabidopsis thaliana. Gene 242, 427–436. van der Hoorn, R.A., Wulff, B.B., Rivas, S., Durrant, M.C., van der Ploeg, A., de Wit, P.J., Jones, J.D., 2005. Structure–function analysis of cf-9, a receptor-like protein with extracytoplasmic leucine-rich repeats. Plant Cell 17, 1000–1015. Worrall, D., Elias, L., Ashford, D., Smallwood, M., Sidebottom, C., Lillford, P., Telford, J., Holt, C., Bowles, D., 1998. A carrot leucinerich-repeat protein that inhibits ice recrystallization. Science 282, 115– 117.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 534–539 www.elsevier.com/locate/phytochem

Update in Bioinformatics

PeroxiBase: A class III plant peroxidase database Nenad Bakalovic a,1, Filippo Passardi a,1, Vassilios Ioannidis b, Claudia Cosio a, Claude Penel a, Laurent Falquet b, Christophe Dunand a,* a

Laboratory of Plant Physiology, University of Geneva, Quai Ernest-Ansermet 30, CH-1211 Geneva 4, Switzerland b Swiss Institute of Bioinformatics, CH-1066 Epalinges/Lausanne, Switzerland Received 13 September 2005; received in revised form 9 December 2005 Available online 26 January 2006

Abstract Class III plant peroxidases (EC 1.11.1.7), which are encoded by multigenic families in land plants, are involved in several important physiological and developmental processes. Their varied functions are not yet clearly determined, but their characterization will certainly lead to a better understanding of plant growth, differentiation and interaction with the environment, and hence to many exciting applications. Since there is currently no central database for plant peroxidase sequences and many plant sequences are not deposited in the EMBL/GenBank/DDBJ repository or the UniProt KnowledgeBase, this prevents researchers from easily accessing all peroxidase sequences. Furthermore, gene expression data are poorly covered and annotations are inconsistent. In this rapidly moving field, there is a need for continual updating and correction of the peroxidase superfamily in plants. Moreover, consolidating information about peroxidases will allow for comparison of peroxidases between species and thus significantly help making correlations of function, structure or phylogeny. We report a new database (PeroxiBase) accessible through a web server (http://peroxidase.isb-sib.ch) with specific tools dedicated to facilitate query, classification and submission of peroxidase sequences. Recent developments in the field of plant peroxidase are also mentioned. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Database; Multigenic family; Evolution; Phylogeny; Peroxidases

1. Introduction Class III plant peroxidases (EC 1.11.1.7, donor:hydrogen-peroxide oxidoreductase) are present in all land plants (Table 1). Genes encoding this enzyme family are particularly numerous in Angiosperms. The high number of isoenzymes and their remarkable catalytic versatility allow them to be involved in a broad range of physiological and developmental processes all along the plant life cycle (Passardi et al., 2005). Plant peroxidases have been shown to be involved in the cross-linking of cell wall constituents, lignin polymerization, the catabolism of auxin – a hormone having a critical role in plant growth and development – and

*

1

Corresponding author. Tel.: +41 223793012; fax: +41 223793017. E-mail address: [email protected] (C. Dunand). The two first authors have contributed equally to this work.

0031-9422/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.020

the formation of reactive oxygen species (superoxide, hydroxyl radical). They also play a prominent role in defence reactions against many pathogenic organisms. However, until now the in vivo functions of a particular peroxidase have not been reported. This knowledge is however crucial to understand the evolution, the roles and the regulations of this key multifunctional enzyme. Plant peroxidases are an example of a multigenic family whose number of members increased since the conquering of land by plants due to constant evolution. The Arabidopsis genome contains 73 genes encoding a peroxidase (Tognolli et al., 2002) and rice contains 138 (Passardi et al., 2004). The homology between paralogs in a plant ranges from 30% to 100%, but very close orthologs exist, even between evolutionarily distant plants. All plant peroxidases contain invariant amino acids essential for their catalytic properties and for their proper folding (Welinder et al., 2002). They are structurally related to other heme-containing

N. Bakalovic et al. / Phytochemistry 67 (2006) 534–539

535

Table 1 Representation of the major plant lineages found in the PeroxiBase Order (number of species/number of sequences): Genus Angiosperms Dicotyledons Rosids Cucurbitales (3/7): Cucumis (2x), Cucurbita Fabales (14/314): Arachis, Cicer, Glycine, Lotus, Lupinus (2x), Medicago (2x), Phaseolus (2x), Pisum, Stylosanthes, Trifolium, Vigna Rosales (3/8): Ficus, Malus, Urtica Fagales (1/2): Quercus Malpighiales (10/92): Euphorbia, Linum, Manihot, Mercurialis, Populus (6x) Malvales (3/58): Gossypium, Theobroma Sapindales (3/20): Citrus (2x), Poncirus Brassicales (7/103): Arabidopsis, Brassica (2x), Armoracia, Raphanus, Thellungiella (2x) Vitaceae (1/21): Vitis Asterids Gentianales (3/10): Coffea, Hedyotis (2x) Lamiales (6/10): Avicennia, Eucommia, Orobanche (2x), Scutellaria, Striga Solanales (11/199): Capsicum (2x), Ipomoea (2x), Lycopersicon (2x), Nicotiana (3x), Petunia, Solanum Asterales (8/73): Artemisia, Cichorium, Helianthus (2x), Gerbera, Lactuca, Stevia, Zinnia Apiales (1/1): Petroselinum Ericales (1/1): Vaccinium Saxifragales (1/6): Ribes Caryophyllales (4/56): Beta, Mesembryanthemum, Mirabilis, Spinacia Ranunculales (2/37): Eschscholzia, Aquilegia Monocotyledons Magnolids Laurales (1/10): Persea Magnoliales (1/11): Liriodendron Piperales (1/0): Saruma Liliopsida Poales (15/655): Aegilops, Ananas, Avena, Cenchrus, Hordeum, Lolium, Oryza, Saccharum (2x), Secale, Setoria, Sorghum, Triticum (2x), Zea Zingiberales (0/0) EST project in run Liliales (2/2): Lilium, Alstroemeria Acorales (1/5): Acorus Alismatales (1/1): Spirodela Arecales (1/1): Elaeis Asparagales (3/36): Allium, Asparagus, Hyacinthus Basal magnoliophyta Austrobaileyales (1/0): Illicium Nymphaeales (1/3): Nuphar Amborellaceae (1/13): Amborella Gymnosperms Gnetopsida Gnetales Welwitschiales (1/1): Welwitschia Coniferales (6/24): Cryptomeria, Pinus (3x), Picea (2x) Ginkgoales (1/8): Ginkgo Cycadales (3/12): Cycas, Zamia (2x) Cryptogames (4/21) Marchantia, Physcomitrella, Selaginella, Ceratopteris Algae Charales (0/0) no sequence available/activity found Chlamydomonadales (0/0) no sequence available/no activity found Number of species and of peroxidases found in each lineage are represented in brackets.

proteins, like peroxidases from prokaryotes, fungi. The key amino acids that interact with heme are also found in hemoglobins and cytochromes. The broad molecular diversification of plant peroxidases mainly results from

gene duplication events. Newly duplicated genes were likely conserved because they acquired new modes of expression, regulation (subfunctionalization) or novel functions (neofunctionalization).

536

N. Bakalovic et al. / Phytochemistry 67 (2006) 534–539

The automated annotation of the whole genomes of Arabidopsis (Arabidopsis Genome Initiative, 2000) and Oryza sativa (Goff et al., 2002; Yu et al., 2002), the automated clustering and assembling of EST sequences, and numerous EST projects led to the identification of a large number of sequences coding for class III plant peroxidases. We decided to construct a database devoted to this large, multigenic family because in our experience automated processing sometimes yields sequences of poor quality. Specificity is compromised and BLAST searching often requires manual sorting. Using the highly conserved motifs of the class III peroxidases (Welinder, 1992), manual annotation and editing can retrieve whole peroxidase sequences, that are unrecognized in automation due to poor quality sequences. Arabidopsis and rice are completely sequenced and are considered to be plant models, but they are not representative of plant diversity. The large number of EST projects developed with more diverse plants will provide a better overview of peroxidase evolution throughout green plants. The first goal of the PeroxiBase is to centralize most of the annotated and non-annotated class III peroxidase-encoding sequences and to make them publicly available, so that the research community has a unique tool for discovery, comparison, and exchange of peroxidase sequences. The second goal is to compile information concerning putative function and transcription regulation in order to facilitate cross-checking between close paralogs and orthologs. The final goal is to confirm the hypothesis that the number of class III isoforms increased after the emergence of the land plants.

2. Construction of the database The database was constructed following two parallel procedures: one exhaustive and another more specific (Fig. 1). Firstly, the plant/fungal/bacterial heme peroxidase proteins are characterized by the motif PEROXIDASE_4 (http://ca.expasy.org/cgi-bin/nicesite.pl? PS50873). Using this signature, systematic data mining with MyHits (Pagni et al., 2004) from different predicted protein databases (TrEST, TrGEN) (Pagni et al., 2001) provides a global view of the peroxidase encoding sequences. The resulting hits are already treated data (assembled and translated sequences) with the risk of automatic compilation and translation. We have then used AtPrx42 and OsPrx73, two sequences potentially related to an ancestral sequence (Passardi et al., 2004), for a second, more specific approach by scanning numerous public sources of plant ESTs and genome sequences in order to obtain a large collection of peroxidase-encoding sequences. For rare species, a tBLASTn was first performed against the NCBI whole database using limited queries for the date and for the organism. The Plant Genome Project (http://www.pgn.cornell.edu, 2004), Plant GDB (Dong et al., 2004) and Sputnik (Rudd, 2005) were used to complete the short sequences obtained from TIGR and to find new ones. Each sequence obtained (assembly or singleton sequences) was individually translated and the presence of characteristic peroxidase motifs was verified using FingerPRINTScan and InterPro Scan softwares. Low quality

Other nucleotide databases

EMBL/GenBank/ DDBJ

tBLASTn with AtPrx42/OsPrx73

SIB assembly procedure

Complete cDNA sequences

EST collections

CAP3 assembly UniProtKB

TrEST, TrGEN, trome

Genomic sequences Splicing

Translated sequences

Presence of the specific peroxidase motifs + BLASTp against Peroxi Base

PeroxiBase

Cross-checking data

Appearance and evolution

Homology between species

Fig. 1. Procedure of data analysis for generation of the PeroxiBase. Various EST and genomic databases have been used as sources of peroxidase encoding sequence.

N. Bakalovic et al. / Phytochemistry 67 (2006) 534–539

3. Web interface The PeroxiBase web interface includes four main modules. (i) Search: this module enables a text query from the entire dataset with keywords such as tissue type, accession number, inducer/repressor, and name of sequences and organisms. (ii) Organism: each organism possesses a link

Dicotyledons

with its taxonomic identity. Information for the peroxidases present in each organism can be viewed independently. Each file contains a direct link to the corresponding database (NCBI, TIGR, PGN, Sputnik) and to Swiss-Prot and DNA sequences when these entries exist. In addition, numbers of ESTs, cellular localization and tissue type are all included. The three closest homologous sequences, with their corresponding score and E-value are also described in the files. (iii) BLAST: two BLAST searches can be performed against the whole peroxidase database (Altschul et al., 1997), BLASTp for protein sequence and BLASTx for nucleotide sequences. The alignments are visualized and linked to the entry of each peroxidase. (iv) FingerPrintscan: this tool helps to find out which peroxidase family the sequence belongs to. Three minor modules Tissue type, Inducers/repressors and Cellular localization are used as alternative ways of viewing sequences.

4. Modelization of the number of class III isoforms evolution Class III peroxidase encoding sequences and peroxidase activity are both absent from the green alga Chlamydomonas reinhardtii (Passardi et al., 2004). On the contrary, in various Chara species, guaiacol oxidation in the presence of H2O2, specific to class III peroxidase activity (Greppin et al., 1986), can be detected (data not shown). The exhaustive data mining performed for the setup of the PeroxiBase confirms that the class III peroxidases are present in all land plants. For some species the total EST count is low (less than 1000), yet several independent isoforms were identified. In the case of large EST libraries (over 10,000), multigenic families with numerous putative class III peroxidases sequences can be found confirming previous results obtained with Arabidopsis and Oryza (Tognolli et al., 2002; Duroux and Welinder, 2003). In addition, the case of Physcomitrella patens seems to further validate the Monocotyledons Gymnosperms

3 2.5

Rosids

Asterids

Poales

2

Ly cop hytes

sequences are not included in TIGR consensus sequences due to the high sequence stringency TIGR uses. Using peroxidase motifs as a guide, manual inspection of these poor quality sequences allows for increasing their length and assembling new sequences. All distinct sequences, even short ones, are kept in the database. We have also compared the UTR regions to confirm that sequences with high homology are truly distinct. In addition, as in numerous other genera, ESTs from Gossypium (cotton), Picea (spruce), Populus (poplar) have been assembled by TIGR. However, the sequences used for the construction of the contigs are treated as if derived from a single species although they originate from different species. For example, cotton is derived from Gossypium arboreum and hirsutum, spruce from a collection of Picea abies, Picea glauca and Picea sitchensis and poplar from a mix of Populus alba, Populus balsamifera, Populus euphratica, Populus kitakamiensis, Populus nigra, Populus tremula, Populus tremuloides, and Populus trichocarpa. The Sputnik database (http://sputnik.btk.fi/) helped us unscramble this mixture of sequences for Gossypium and Populus and Picea species. The other species have been assembled directly from the NCBI entries. After analysis of sequence alignment within each species (ClustalW and BioEdit), the protein sequences were individually entered in the database with their corresponding accession numbers as well as various information concerning the putative functions and transcription regulation (localization, induction and repression).

537

1.5 1 0.5 Ar ab M idop ed s ic is ag th Po o t alia nc run na iru ca tu s Ri V trifo la be iti lia s s t He am vin a lia er ife n Ly ic ra co Ste thu an pe via s a um Ni rsic re nnu co on ba us tia e ud na sc ian be ule a nt ntu h m Ca Be am S a p t a ia n cc sic vu a ha um lg a ru m an ris of nuu f A c O ic in m or ry aru Am us za m a s b W orel mer ativ el la ic a wi tr an ts ich us ch o ia po m d Pi ira a b n Cy us ilis c a ta Ph s e ys Za ru da co m mp m ia f hi it r is i el ch la e pa ri te ns

0

Fig. 2. Evolution of the peroxidase encoding genes versus the total number of genes in various key organisms. The y-axis represents a ratio obtained as following: (number of peroxidase encoding EST/number of independent peroxidase encoding genes)/(number of total EST/number of unigenes).

538

N. Bakalovic et al. / Phytochemistry 67 (2006) 534–539

hypothesis that the number of peroxidase encoding genes increases over evolution. With more than 100,000 EST and only 12 peroxidase encoding sequences, the number of peroxidases sequences for this moss is far below those from other organisms such as Arabidopsis. A potential evolution of peroxidase gene numbers can be drawn up based on information concerning each species such as total EST count and number of unigenes with the following formula: (number of peroxidase encoding EST/ number of independent peroxidase encoding genes)/(number of total EST/number of unigenes). The value is proportional to the number of peroxidase genes in each species and gives information regarding the putative evolution of the peroxidase isoforms. Other completed genome sequencing and increasing EST sequences should confirm this hypothesis. Independently of the EST number, the size of the family seems also to follow a gradual increase from Charales (few isoforms) to the higher plants (numerous isoforms) confirming the previous hypothesis. The species can be classified in two major groups following the value of this ratio (Fig. 2). Rosids, Asterids and Poales considered to be higher plants, show a high diversification rate value over 1. On the other hand, species issued from basal Gymnosperms, and from small Mono- and Dicotyledons orders such as Vitaceae, Saxifragales, Caryophyllales, and Ranunculales have values around 0.5 or smaller.

5. Current status and future developments The first goal of the PeroxiBase was to develop an efficient tool for the study of the evolution of a plant multigenic family. We tried not to be exclusive and to include as many sequences as possible from different organisms. The base currently consists of a core dataset containing over 2000 complete or partial peroxidase-encoding sequences from 125 organisms (Table 1), and it is still in constant evolution. New peroxidase encoding sequences can be easily and directly added to the database by external people with individual user name and password. Continuous data mining will be performed until a putative complete analysis of the available sequences is achieved (EST and genomic sequences). At this point, a semi-automatic update will be set up to collect the peroxidase encoding sequences newly submitted to general databases (NCBI, Swiss-Prot). Information concerning the expression profile will also be updated and new features such as results of knock-out, knock-down or overexpression studies will be added when available. The superfamily of plant, fungal and bacterial heme peroxidases contains class I (Cytochrome C peroxidase (EC 1.11.1.5), catalase peroxidase (EC 1.11.1.6) and ascorbate peroxidase (EC 1.11.1.11)), class II (lignin peroxidases (EC 1.11.1.14) and manganese peroxidases (EC 1.11.1.13)) and class III. The next update will include class I and class II peroxidases in the database. To our knowl-

edge, the PeroxiBase will become the first repository devoted exclusively to a superfamily composed of multigenic families. The database could help to confirm the hypothesis that the three classes evolved from a single ancestral sequence (Zamocky, 2004). Another major addition will be to relate the major lineages containing peroxidase-encoding sequences to a schematic evolutionary tree. Peroxidases may become key markers for the evolution of plants, from as early as the first moments of land colonization to the human impact on genetics of cultivated plants today. The varied functions of peroxidases will be characterized and lead to a better understanding of plant growth, differentiation and interaction with the environment, and eventually to many exciting applications.

6. Useful web links BioEdit: http://www.mbio.ncsu.edu/BioEdit/bioedit. html ClustalW: http://www.ebi.ac.uk/clustalw/ Expasy translate: http://us.expasy.org/tools/dna.html FingerPRINTScan: http://www.bioinf.man.ac.uk/finger PRINTScan/ InterPro Scan: http://www.ebi.ac.uk/InterProScan/ MyHits: http://myhits.isb-sib.ch/cgi-bin/motif_query NCBI: http://www.ncbi.nlm.nih.gov/ PlantGDB: http://zmdb.iastate.edu/PlantGDB/ Plant Genome Network: http://pgn.cornell.edu/ Reverse-Complement: http://bioinformatics.org/sms/ rev_comp.html SoftBerry- FGENESH: http://www.softberry.com/berry. phtml?topic=fgenesh&group=programs&subgroup= gfind Sputnik: http://sputnik.btk.fi/ests TIGR: http://www.tigr.org/tdb/tgi/plant.shtml

Acknowledgments We thank Sonia Guimil for her critical reading. The financial support of the Swiss National Science Foundation (Grant 31-068003.02) to C.P. and C.D. is gratefully acknowledged, N.B. is paid by the Office Cantonal de l’Emploi. References Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. Arabidopsis Genome Initiative, 2000. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796–815. Dong, Q., Schlueter, S.D., Brendel, V., 2004. PlantGDB, plant genome database and analysis tools. Nucleic Acids Res. 32, D354–D359.

N. Bakalovic et al. / Phytochemistry 67 (2006) 534–539 Duroux, L., Welinder, K.G., 2003. The peroxidase gene family in plants: a phylogenetic overview. J. Mol. Evol. 57, 397–407. Goff, S.A., Ricke, D., Lan, T.H., Presting, G., Wang, R., Dunn, M., Glazebrook, J., Sessions, A., Oeller, P., Varma, H., Hadley, D., Hutchison, D., Martin, C., Katagiri, F., Lange, B.M., Moughamer, T., Xia, Y., Budworth, P., Zhong, J., Miguel, T., Paszkowski, U., Zhang, S., Colbert, M., Sun, W.L., Chen, L., Cooper, B., Park, S., Wood, T.C., Mao, L., Quail, P., Wing, R., Dean, R., Yu, Y., Zharkikh, A., Shen, R., Sahasrabudhe, S., Thomas, A., Cannings, R., Gutin, A., Pruss, D., Reid, J., Tavtigian, S., Mitchell, J., Eldredge, G., Scholl, T., Miller, R.M., Bhatnagar, S., Adey, N., Rubano, T., Tusneem, N., Robinson, R., Feldhaus, J., Macalma, T., Oliphant, A., Briggs, S., 2002. A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296, 92–100. Greppin, H., Penel, C., Gaspar, T., 1986. Molecular and Physiological Aspects of Plant Peroxidases. University of Geneva, Switzerland. Pagni, M., Iseli, C., Junier, T., Falquet, L., Jongeneel, V., Bucher, P., 2001. trEST, trGEN and Hits: access to databases of predicted protein sequences. Nucleic Acids Res. 29, 148–151. Pagni, M., Ioannidis, V., Cerutti, L., Zahn-Zabal, M., Jongeneel, C.V., Falquet, L., 2004. MyHits: a new interactive resource for protein annotation and domain identification. Nucleic Acids Res. 32, W332– W335. Passardi, F., Longet, D., Penel, C., Dunand, C., 2004. The class III peroxidase multigenic family in rice and its evolution in land plants. Phytochemistry 65, 1879–1893. Passardi, F., Cosio, C., Penel, C., Dunand, C., 2005. Peroxidases have more functions than a Swiss army knife. Plant Cell Rep. 24, 255–265.

539

Rudd, S., 2005. openSputnik: a database to establish comparative plant genomics using unsaturated sequence collections. Nucleic Acids Res. 33, D622–D627. Tognolli, M., Penel, C., Greppin, H., Simon, P., 2002. Analysis and expression of the class III peroxidase large gene family in Arabidopsis thaliana. Gene 288, 129–138. Welinder, K.G., 1992. Plant peroxidases: structure–function relationships. In: Penel, C., Gaspar, T., Greppin, H. (Eds.), Plant Peroxidases. University of Geneva, Switzerland. Welinder, K.G., Justesen, A.F., Kjaersgard, I.V., Jensen, R.B., Rasmussen, S.K., Jespersen, H.M., Duroux, L., 2002. Structural diversity and transcription of class III peroxidases from Arabidopsis thaliana. Eur. J. Biochem. 269, 6063–6081. Yu, J., Hu, S., Wang, J., Wong, G.K., Li, S., Liu, B., Deng, Y., Dai, L., Zhou, Y., Zhang, X., Cao, M., Liu, J., Sun, J., Tang, J., Chen, Y., Huang, X., Lin, W., Ye, C., Tong, W., Cong, L., Geng, J., Han, Y., Li, L., Li, W., Hu, G., Li, J., Liu, Z., Qi, Q., Li, T., Wang, X., Lu, H., Wu, T., Zhu, M., Ni, P., Han, H., Dong, W., Ren, X., Feng, X., Cui, P., Li, X., Wang, H., Xu, X., Zhai, W., Xu, Z., Zhang, J., He, S., Xu, J., Zhang, K., Zheng, X., Dong, J., Zeng, W., Tao, L., Ye, J., Tan, J., Chen, X., He, J., Liu, D., Tian, W., Tian, C., Xia, H., Bao, Q., Li, G., Gao, H., Cao, T., Zhao, W., Li, P., Chen, W., Zhang, Y., Hu, J., Liu, S., Yang, J., Zhang, G., Xiong, Y., Li, Z., Mao, L., Zhou, C., Zhu, Z., Chen, R., Hao, B., Zheng, W., Chen, S., Guo, W., Tao, M., Zhu, L., Yuan, L., Yang, H., 2002. A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science 296, 79–92. Zamocky, M., 2004. Phylogenetic relationships in class I of the superfamily of bacterial, fungal, and plant peroxidases. Eur. J. Biochem. 271, 3297–3309.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 540–544 www.elsevier.com/locate/phytochem

Substrate specificity of acyl-D6-desaturases from Continental versus Macaronesian Echium species Federico Garcı´a-Maroto a, Aurora Man˜as-Ferna´ndez a, Jose´ A. Garrido-Ca´rdenas a, Diego Lo´pez Alonso b,* b

a Area de Bioquı´mica, Facultad de Ciencias Experimentales, Universidad de Almerı´a, 04120 Almerı´a, Spain Departamento de Biologı´a Aplicada, Facultad de Ciencias Experimentales, Universidad de Almerı´a, 04120 Almerı´a, Spain

Received 19 October 2005; received in revised form 25 November 2005 Available online 7 February 2006

Abstract Echium (Boraginaceae) species from the Macaronesian islands exhibit an unusually high level of c-linolenic acid (18:3n-6; GLA) and relatively low content of octadecatetraenoic acid (18:4n-3; OTA) in the seed, while the amounts of both fatty acids in their Continental (European) relatives are rather similar. We have tested the hypothesis of whether a different specificity of the acyl-D6-desaturases (D6DES) towards their respective usual substrates, linoleic acid (18:2n-6; LA) for GLA and a-linolenic acid (18:3n-3; ALA) for OTA, was partly responsible for this composition pattern. To this aim we have expressed in yeast the coding sequences of the D6DES genes for the Continental species Echium sabulicola, and the Macaronesian Echium gentianoides. When the yeast cultures are supplemented with the two fatty acid substrates (LA and ALA), a similar utilization of both compounds was found for the D6DES of E. sabulicola, while a preference for LA over ALA was observed for the enzyme of E. gentianoides. This substrate preference must contribute to the increased accumulation of GLA in the seeds of the Macaronesian Echium species. Comparison among the amino acid sequences of these desaturases and other related enzymes, allowed us the discussion about the possible involvement of some specific positions in the determination of substrate specificity.
2005 Elsevier Ltd. All rights reserved. Keywords: Echium gentianoides; Echium sabulicola; Boraginaceae; Acyl specificity; Gene cloning; D6-Desaturase

1. Introduction Polyunsaturated fatty acids are receiving considerable attention due their involvement in human health (Gunstone, 1998). Consequently, efforts have been addressed to the obtaining of improved new sources of those compounds. Genetic engineering of oilseed crops using appropriate heterologous genes from the lipid biosynthetic pathway has become a strategy that is giving promising results (Thelen and Ohlrogge, 2002; Sayanova and Napier, 2004; Surinder et al., 2005). One of the key genes encodes the acyl-D6-desaturase (D6DES) an enzyme that catalyses *

Corresponding author. Tel.: +34 950015033; fax: +34 950015476. E-mail address: [email protected] (D.L. Alonso).

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.005

the first committed step in the biosynthesis of essential fatty acids acting as eicosanoid precursors (Napier et al., 1997; Lo´pez Alonso and Garcı´a-Maroto, 2000). This enzyme has two preferred substrates in plants, the linoleic acid (LA, 18:2n-6) and the a-linolenic acid (ALA, 18:3n-3), giving rise to either c-linolenic acid (GLA, 18:3n-6) or octadecatetraenoic acid (OTA, 18:4n-3), respectively. GLA and OTA are unusual in most plant species but they accumulated at a high level by the seeds of some families such as the Boraginaceae. Among them, Borago officinalis and several Echium species have been used as a source for cloning of D6DES genes, and the enzymatic activity of their products has been assayed by using heterologous expression systems (Sayanova et al., 1997; Garcı´a-Maroto et al., 2002). Though substrate availability (e.g., LA/ALA content)

F. Garcı´a-Maroto et al. / Phytochemistry 67 (2006) 540–544

541

represented by five common European taxa, contain similar amounts of GLA (10% average) and OTA (13% average), with GLA/OTA ratios below the unit and ranging from 0.38 to 0.95. Conversely, Macaronesian taxa represented by five endemics from the Canary islands contain a higher level of GLA (21% average) relative to the OTA content (5% average), with GLA/OTA ratios ranging from 3.3 to 5.8. Similar results have been reported for other Echium species (Guil-Guerrero et al., 2001a, 2003). Based on the available biochemical data at least two explanations may account for this composition pattern. One of them is a different substrate (LA/ALA) availability that would favor the synthesis of the respective product. This is likely since opposite n-6 to n-3 ratios are found in both groups of plants, probably due to a different activity of the x3-desaturase that regulates the flux towards n-3 or n-6 fatty acids during seed development. Thus, the elevated n-6:n-3 ratio indicates that LA availability is higher in Macaronesian taxa, therefore resulting in an increased GLA content, and the contrary applies to the Continental Echium. Another but not excluding explanation for the observed GLA/OTA ratios is a different specificity of the D6DES enzymes that determines a different utilization rate of the two substrates. This mechanism was previously illustrated in Primula species (Sayanova et al., 2003). To explore this possibility we have investigated the specificity of the D6DES enzymes of the Continental representative E. sabulicola and the Macaronesian endemic E. gentianoides, using the coding sequences from their respective genes in a yeast expression system. The D6DES gene from E. gentianoides (EGD6DES) was available from a previous work (Garcı´a-Maroto et al., 2002). Here we have cloned the gene for the D6DES of E. sabulicola (ESD6DES; see Section 4). As described for other D6DES genes (Lo´pez Alonso et al., 2003; Sayanova et al., 2003) its genomic sequence is intron-less and encodes a 448 amino acids product bearing typical features of plant D6-desaturases such as a cytochrome-b5 domain an the three histidine boxes (Fig. 1).

seems to be an important determinant in the relative synthesis of GLA/OTA in the seeds of Boraginaceae species (Guil-Guerrero et al., 2001a), there are evidences that other factors such as substrate (i.e., acyl) specificity could also play a significant role. Thus, a distinct substrate preference has been demonstrated for the D6DES enzymes of two Primulaceae species (Sayanova et al., 2003). This is an interesting characteristic that could be usefully exploited to channel the synthesis to a particular product. However, finding of the molecular determinants of this specificity, by comparison of the amino acid sequences, was not possible due to the considerable divergence among them. A survey of the fatty acid composition in seeds of Echium species revealed that taxa from the Macaronesian archipelago (Canary, Cape Verde and Madeira islands) show an unusually high level of GLA and low OTA content, as compared to Continental (European and North African) species where similar amounts of both fatty acids are accumulated (Guil-Guerrero et al., 2000, 2001a,b, 2003). We decided to investigate if different properties of the D6DES enzymes leading to a different utilization of LA/ALA as substrates could be responsible, at least to some extent, for the observed bias in the GLA/OTA ratio. To this aim we have cloned a gene encoding the D6DES from a Continental representative, E. sabulicola Pomel and tested its substrate specificity in a heterologous yeast expression system, as compared to the D6DES from E. gentianoides, an endemic from the Canary islands for which the gene was available (Garcı´a-Maroto et al., 2002). Comparison among the amino acid sequences of these desaturases, allowed us a discussion about the possible involvement of specific positions in the determination of substrate specificity.

2. Results and discussion Fatty acid compositions of seed oils from different Echium species are shown in Table 1. Continental Echium, as

Table 1 Fatty acid composition of seed oils from Macaronesian versus Continental Echium species Echium species

Fatty acid 16:0

18:0

18:1

LA

GLA

ALA

OTA

GLA/OTA ratio

7.6 6.7 5.5 6.5 6.4

3.1 2.5 2.3 3.7 2.8

8.3 16.4 14.7 9.4 13.0

19.1 16.3 8.6 16.9 13.8

11.5 10.9 5.5 10.9 9.2

38.2 33.3 47.1 39.3 36.6

12.1 12.3 14.3 13.3 13.0

0.95 0.89 0.38 0.82 0.71

Macaronesian (Canary Islands) E. gentianoidesa 7.4 6.8 E. pitardiic E. auberianumd 7.0 E. strictumd 6.9 E. giganteumd 7.6

4.5 3.8 2.5 4.2 3.5

10.6 13.0 19.2 11.6 12.3

22.2 24.0 29.1 28.8 23.3

26.7 18.9 17.5 18.8 21.7

21.9 25.9 18.9 23.0 23.1

6.7 5.7 3.0 4.4 5.7

3.98 3.31 5.83 4.27 3.81

Continental (Europe) E. sabulicolaa E. lusitanicumb E. boissieric E. vulgareb E. plantagineumd

The 16:1 fatty acid was under 0.3% for all Echium species; The data were obtained from: (a) this work; (b) Guil-Guerrero et al. (2003); (c) Guil-Guerrero et al. (2001a); (d) Guil-Guerrero et al. (2000).

542

F. Garcı´a-Maroto et al. / Phytochemistry 67 (2006) 540–544

Fig. 1. Amino acid sequences of D6DES enzymes from Boraginaceae species. The amino acid sequences of the D6-desaturases of E. sabulicola (ESD6DES; acc. no. DQ067612) E. gentianoides (EGD6DES, AY055117), Echium pitardii (EPD6DES; AAL23581), and Borago officinalis (BOD6DES; AAC49700), were aligned with ClustalX (v1.7) (Thompson et al., 1997). The ‘‘Boxshade’’ v3.21 software is used to highlight the homology between protein sequences. Shading is applied when there is agreement for a fraction of sequences above 0.5. Amino acids identical to ESD6DES are enclosed in black boxes, and similar residues in gray. Conserved histidine boxes and the cytochrome-b5 domain are indicated. Amino acid differences among ESD6DES and EGD6DES proteins are marked by asterisks.

The whole coding sequence of ESD6DES and EGD6DES were cloned in the yeast expression vector pYES-2 under the control of a galactose inducible promoter, and further introduced in the yeast strain INVSc1. The yeast transformed with the empty vector was used in negative controls. Both fatty acid substrates of the D6DES enzymes (LA and ALA) were exogenously supplemented in the culture at equimolar amounts, though ALA was incorporated more efficiently under our experimental conditions (Table 2). Total fatty acids were extracted and analysed from the biomass of the induced cultures, and the amounts of the individual fatty acids were determined by comparison to an internal standard. The conversion

rates of LA into GLA (CRLA) and ALA into OTA (CRALA) were calculated (Table 1). For the D6DES of E. sabulicola the average conversion rates into GLA and OTA are similar (18.5 vs. 15.4, respectively) with a slight preference over the LA substrate. However, the conversion rates for the D6DES of E. gentianoides (CRLA = 28.9 and CRALA = 19.2) indicate a clear preference in the utilization of LA over ALA. These results agree with fatty acid compositions observed in the seeds, and indicate that the higher GLA content in E. gentianoides is the result of both a preference in the utilization of LA by the D6DES enzyme and a higher availability of the n-6 substrate as compared to E. sabulicola.

F. Garcı´a-Maroto et al. / Phytochemistry 67 (2006) 540–544 Table 2 Fatty acid composition (%) and conversion rates of yeast expressing either the empty vector (pYES-2), the D6DES of E. sabulicola (ESD6DES) or the D6DES of E. gentianoides (EGD6DES) Fatty acid

16:0 16:1n-7 18:0 18:1n-9 LA GLA ALA OTA CRLA CRALA

Expression constructa pYES-2

ESD6DES

EGD6DES

20.2 20.2 6.2 11.9 15.2 – 25.2 – – –

21.3 21.0 6.7 12.0 10.9 2.5 19.9 3.7 18.5 15.4

21.0 19.6 6.7 11.9 10.3 4.2 19.9 4.7 28.9 19.2

(1.4) (1.4) (0.6) (0.6) (1.1) (1.6)

(0.5) (1.4) (0.2) (0.4) (0.8) (0.2) (1.2) (0.2) (1.5) (1.3)

(0.3) (0.4) (0.2) (0.1) (0.4) (0.1) (0.2) (0.3) (1.3) (1.0)

a

Yeast cultures were supplemented with equimolar amounts of both LA and ALA to determine the conversion rates (CR) from each substrate. The conversion rate (CR) of the desaturase substrates is defined as: CR(%) = [molar % of product/molar % of non-transformed substrate + molar % of product] · 100, under our experimental conditions. Total fatty acids were extracted and analysed as indicated in Section 4.3 after washing of cells. Results are the average of three independent experiments with standard error indicated within the brackets.

Previous data obtained for the Borago desaturases indicated that the determinants of the substrate specificity were scattered along the whole molecule, so that they remain elusive (Libisch et al., 2000; Sayanova et al., 2003). When the amino acid sequences of EGD6DES and ESD6DES are compared (Fig. 1) nine amino acid changes (four of them being conservative) are recorded, mainly distributed along the first 180 amino acids. Among the non-conservative changes there are three potentially informative sites corresponding to positions 74, 143 and 334. Interestingly, those amino acid replacements in ESD6DES are coincident in the protein from B. officinalis, a desaturase that neither shows substrate preference (Libisch et al., 2000), while those positions in EGD6DES are occupied by the same amino acid in the desaturase from Echium pitardii, another ‘‘GLA-rich’’ Macaronesian taxa. Site-directed mutagenesis of these amino acids could be used to confirm the involvement of those sites in the determination of the desaturase specificity, and eventually to increase or modify substrate discrimination.

3. Concluding remarks Our experimental data indicate that both substrate availability and specificity of the D6-desaturase enzyme are responsible for the observed bias in fatty acid compositions of Echium species with different geographic distributions. Moreover, comparison of the desaturase sequences suggested possible amino acid positions that could be further investigated to check their involvement on substrate selectivity. This is interesting regarding the possibility of using genetic engineering to obtain more suitable D6-desaturase enzymes.

543

4. Experimental 4.1. Biological material Seeds of Echium sabulicola Pomel were collected from plants located in their natural habitat at El Alquian (Almerı´a, Spain). The yeast strain INVSc1 (Invitrogen) was used to assay D6-desaturase activities by heterologous expression. 4.2. Cloning of the D6DES gene of Echium sabulicola Cloning of the D6DES gene from E. sabulicola (ESD6DES) was achieved by PCR amplification using the following flanking primers with sequences derived from D6DES genes of other Echium species: ESDES-UP (5 0 -CTACATATGGCTAATGCAATCAAGAAGTACATTAC-3 0 ) and ESDES-DOWN (5 0 -CATAGGATCCAACAAGTAGAACCAATGCAAGC-3 0 ). The PCR program consisted of a denaturation step of 3 min at 94 C, followed by 38 cycles of 15 s at 94 C, 45 s at 55 C and 1 min 45 s at 72 C, ending with a 5 min step at 72 C. The product was initially cloned into the pGEM-T-Easy vector (Promega), and fully sequenced on both strands using a Perkin–Elmer ABI-310 DNA automated sequencer and the Big Dye v3.1 chemistry. The sequence of ESD6DES was deposited in the GenBank under the accession number DQ067612. 4.3. Functional assays in yeast The whole D6DES coding sequences from E. gentianoides or E. sabulicola were transcriptionally fused to the GAL1 inducible promoter of the pYES2 expression vector (Stratagene), and the resulting plasmid used to transform S. cerevisiae by the LiAcO method (Elble, 1992). Cultures were grown at 28 C in standard minimal medium supplemented with the auxotrophic requirement of the strain plus 1% (w/v) raffinose, and expression was further induced on a 0.4 OD600 culture by the addition of galactose 2% (w/v). To assay D6-desaturase activity, induction was maintained for 48 h at 22 C in the presence of the two substrates, linoleic acid and a-linolenic acid at 0.5 mM, and 1% Tween-40. Yeast cells were collected by centrifugation, further washed with 1.3% NaCl, and the resulting biomass subjected to lyophilization. Simultaneous lipid extraction and generation of fatty acid methyl esters were performed as described elsewhere (Rodrı´guez-Ruiz et al., 1998). Fatty acid composition was determined by GC as in Garcı´a-Maroto et al. (2002).

Acknowledgements This research was supported by a Spanish Grant AGL2002-00390 from the DGICYT, Ministerio de Ciencia y Tecnologı´a (MCYT). A. Man˜as-Ferna´ndez and J.A. Garrido-Ca´rdenas were supported by fellowships from

544

F. Garcı´a-Maroto et al. / Phytochemistry 67 (2006) 540–544

the Junta de Andalucı´a and MCYT (Programa Nacional de Formacio´n de Profesorado Universitario), respectively.

References Elble, R., 1992. A simple and efficient procedure for transformation of yeasts. BioTechniques 13, 18–20. Garcı´a-Maroto, F., Garrido-Ca´rdenas, J.A., Rodrı´guez-Ruiz, J., VilchesFerro´n, M., Adam, A.C., Polaina, J., Lo´pez Alonso, D., 2002. Cloning and molecular characterization of the D6-desaturase from two Echium plant species: production of c-linolenic acid by heterologous expression in yeast and tobacco. Lipids 37, 417–426. Guil-Guerrero, J.L., Go´mez-Mercado, F., Garcı´a-Maroto, F., CampraMadrid, P., 2000. Occurrence and characterization of oils rich in clinolenic acid part I: Echium seeds from Macaronesia. Phytochemistry 53, 451–456. Guil-Guerrero, J.L., Go´mez-Mercado, F., Rodrı´guez-Garcı´a, I., CampraMadrid, P., Garcı´a-Maroto, F., 2001a. Occurrence and characterization of oils rich in c-linolenic acid (III): the taxonomical value of the fatty acids in Echium (Boraginaceae). Phytochemistry 58, 117–120. Guil-Guerrero, J.L., Garcı´a-Maroto, F., Jime´nez Jime´nez, A., 2001b. Fatty acid profiles from forty-nine plant species that are potential new sources of c -linolenic acid. J. Am. Oil Chem. Soc. 78, 677–684. Guil-Guerrero, J.L., Garcı´a-Maroto, F., Vilches-Ferro´n, M.A., Lo´pez Alonso, D., 2003. Gamma-linolenic acid from fourteen Boraginaceae species. Ind. Crop Prod. 18, 85–89. Gunstone, F.D., 1998. Movements towards tailor-made fats. Prog. Lipid Res. 37, 277–305. Libisch, B., Michaelson, L.V., Lewis, M.J., Shewry, P.R., Napier, J.A., 2000. Chimeras of D6-fatty acid and D8-sphingolipid desaturases. Biochem. Bioph. Res. Co. 279, 779–785.

Lo´pez Alonso, D., Garcı´a-Maroto, F., 2000. Plants as ’chemical factories’ for the production of polyunsaturated fatty acids. Biotechnol. Adv. 18, 481–497. Lo´pez Alonso, D., Garcı´a-Maroto, F., Rodriguez-Ruiz, J., Garrido, J.A., Vilches, M.A., 2003. Evolution of the membrane-bound fatty acid desaturases. Biochem. Syst. Ecol. 31, 1111–1124. Napier, J.A., Sayanova, O., Stobart, A.K., Shewry, P.R., 1997. A new class of cytochrome b5 fusion proteins. Biochem. J. 328, 717– 720. Rodrı´guez-Ruiz, J., Belarbi, E.-H., Garcı´a Sa´nchez, J.L., Lo´pez Alonso, D., 1998. Rapid simultaneous lipid extraction and transesterification for fatty acid analyses. Biotechnol. Techniq. 12, 689–691. Sayanova, O.V., Napier, J.A., 2004. Eicosapentaenoic acid: biosynthetic routes and the potential for synthesis in transgenic plants. Phytochemistry 65, 147–158. Sayanova, O., Smith, M.A., Lapinskas, P., Stobart, A.K., Dobson, G., Christie, W.W., Shewry, P.R., Napier, J.A., 1997. Expression of a borage desaturase cDNA containing an N-terminal cytochrome b5 domain results in the accumulation of high levels of D6-desaturated fatty acids in transgenic tobacco. Proc. Nat. Acad. Sci. USA 94, 4211– 4216. Sayanova, O., Beaudoin, F., Michaelson, L.V., Shewry, P.R., Napier, J.A., 2003. Identification of Primula fatty acid D6-desaturases with n-3 substrate preferences. FEBS Lett. 542, 100–104. Surinder, P.S., Xue-Rong, Z., Qing, L., Stymne, S., Green, A.G., 2005. Metabolic engineering of new fatty acids in plants. Curr. Opin. Plant Biol. 8, 197–203. Thelen, J.J., Ohlrogge, J.B., 2002. Metabolic engineering of fatty acid biosynthesis in plants. Metabol. Eng. 4, 12–21. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The clustal X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucl. Acids Res. 24, 4876–4882.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 545–552 www.elsevier.com/locate/phytochem

Purification and primary structure determination of two Bowman–Birk type trypsin isoinhibitors from Cratylia mollis seeds P.M.G. Paiva a, M.L.V. Oliva

b,*

, H. Fritz c, L.C.B.B. Coelho a, C.A.M. Sampaio

b

a

c

Departamento de Bioquı´mica, CBB/UFPE, Av. Moraes Rego, S/N, Cidade Universita´ria, Recife-PE, CEP 50670-420, Brazil b Universidade Federal de Sa˜o Paulo-Escola Paulista de Medicina, Departamento de Bioquı´mica, Rua Treˆs de Maio, 100, 04044-020 Sa˜o Paulo, SP, Brazil Abteilung fu¨r Klinische Chemie und Klinische Biochemie, Chirurgische Klinik und Poliklinik, LMU Mu¨nchen, Mu¨nchen, Germany Received 11 August 2004; accepted 4 November 2005 Available online 26 January 2006

Abstract Two Bowman-Birk type trypsin inhibitors (CmTI1 and CmTI2) were purified from Cratylia mollis seeds by acetone precipitation, ion exchange, gel filtration and reverse-phase chromatography. CmTI1 and CmTI2, with 77 and 78 amino acid residues, respectively, were sequenced in their entirety and show a high structural similarity to Bowman-Birk inhibitors from other Leguminosae. The putative reactive sites of CmTI1 are a lysine residue at position 22 and a tyrosine residue at position 49. Different reactive sites, as identified by their alignment with related inhibitors, were found for CmTI2: lysine at position 22 and leucine at position 49. The dissociation constant Ki of the complex with trypsin is 1.4 nM. The apparent molecular mass is 17 kDa without DDT and 11 kDa with reducing agent and heating.
2005 Elsevier Ltd. All rights reserved. Keywords: Cratylia mollis; Leguminosae; Bowman-Birk inhibitor purification; Primary sequence; Trypsin inhibitor

1. Introduction Plant serine proteinase inhibitors are divided into certain protein groups according to primary structure homology, position of the reactive sites and number or location of disulphide bonds (Park et al., 2000; Silva et al., 2001). Bowman-Birk inhibitors (BBIs) from dicotyledonous seeds are comprised of a cysteine-rich polypeptide chain which is bridged by seven conserved disulphide bonds, a fact that probably explains the high degree of stability that BBIs maintain in the presence of denaturing agents and heat (Singh and Appu, 2002).

*

Corresponding author. Tel.: +55 11 576 44 44; fax: +55 11 55723006. E-mail address: [email protected] (M.L.V. Oliva). 0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.017

Dicotyledonous BBIs contain two functional inhibitory domains within disulphide-linked loops each composed of 7 amino acid residues (Mello et al., 2003). Characteristic amino acid residues of the inhibitory regions determine their inhibition specificity, e.g., for trypsin or chymotrypsin (Singh and Appu, 2002). Comparisons among BBIs from different species showed that the first N-terminally located reactive site is more conserved than the second C-terminally located one: Lysine is normally the P1 residue of the first site, whereas at the second P1 site there may be present amino acids of different structures (Tanaka et al., 1997). Alignment analysis showed that the N-terminal amino acid of these inhibitors is usually serine, and that the N- and C-terminal regions are highly hydrophilic (Wu and Whitaker, 1991; Prakash et al., 1996). C- and N-terminal extensions determine the number of amino acid residues of Vigna unguiculata and Pisum sativum BBIs (Mello et al., 2003).

546

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

The interaction between BBIs and their target enzymes follows a tight-binding inhibitory mechanism. Synthetic peptides derived from the BBI reactive site loop retain this inhibitory ability and were used to study interaction between BBIs and enzymes (Scarpi et al., 2002), stability towards proteolytic hydrolysis (Gariani and Leatherbarrow, 1997) as well as tools for pharmacological studies (Dittmann et al., 2001). Analysis of the BBIs chymase interaction showed that the complex did not dissociate in SDS– PAGE under non-reducing conditions (Ware et al., 1997). Leguminosae seeds contain multiple BBIs differing in the number of amino acid residues (Park et al., 2000) and hydrophobicity (Wu and Whitaker, 1991). The isoinhibitors of these BBIs may vary due to differences in genetic polymorphism or to post-translational protein modification (Quillien et al., 1997). In soybean, the soybean cultivar determines the number of BBI isoforms present with trypsin inhibitory activity (Gladysheva et al., 2000). Trypsin inhibitors have been isolated by affinity chromatography through the interaction inhibitor reactive site and matrix (Silva et al., 2001) or by their glycosylated moiety using lectins (Paiva et al., 2003). However, for inhibitor structure analysis ion exchange and reverse-phase chromatography were used aiming to avoid undesirable inhibitor modification by enzyme matrix (Shibata et al., 1986). The biotechnological potential of BBIs has been intensely investigated. These inhibitors are thought to play important roles in plant defense mechanisms against insects (Birk, 1996) and can be used for biological control. The potential of serine protease activity during tumour development has stimulated various lines of research. For example, Chu et al. (1997) investigated the proteolytic activities of transformed cells inhibited by BBIs. Garcia-Gasca et al. (2002) analysed the action of Phaseolus acutifolius BBI on in vitro cell proliferation and cell adhesion of transformed cells, and Kennedy and Wan (2002) evaluated the anticarcinogenic activity of soybean BBIs on prostate cancer cells. BBI detection in Leguminosae seeds is important for the evaluation of the nutritive value of unheated seeds (Hossain and Becker, 2002). Cratylia mollis, or camaratu bean, is a legume that is native to the semi-arid Northeast region of Brazil and is highly resistant to desiccation. The seeds have been an alternative source of nutrition in livestock feed, contributing to regional development and a better quality of life for the local inhabitants. In this study, we describe the purification process of two Bowman-Birk type trypsin isoinhibitors derived from C. mollis seeds, CmTI1 and CmTI2 as well as their properties and amino acid sequences.

2. Experimental 2.1. Proteases, markers and columns Bovine trypsin (EC 3.4.21.4), chymotrypsin (EC 3.4.21.1), DL-Bz-Arg-pNan and Suc-Phe-pNan were purchased from Sigma Chemical Company (St. Louis, MO, USA), porcine pancreatic elastase (EC 3.4.21.35) and MeO-Suc-Ala-Ala-Pro-Val-pNan from Calbiochem (San Diego, CA, USA). Molecular mass markers (carbonic anhydrase, b-lactoglobulin, lysozyme and ovalbumin) were also from Sigma. DEAE-Sephacel and Superdex 75 (FPLC) columns were from Pharmacia Fine Chemicals (Uppsala, Sweden). Ultrasphere and Vydac C4 columns (15.0 cm · 0.46 mm) as well as Aquapore RP300 C (7– 8 lm) and LiChrospher 100 RP (8.5 lm) columns were from Applied Biosystems (Foster City, CA, USA). 2.2. Crude extract Mature seeds, manually harvested from wild plants in the Northeast region of Brazil (near the city of Ibimirim, in the state of Pernambuco), were milled to a fine powder. The meal (60 g) was then homogenised in a blender with 0.15 M NaCl (600 ml), centrifuged at 5000g, and the supernatant was maintained at 60 C for 15 min. This crude extract was taken as starting material. The proteins in the crude extract were precipitated using acetone–H2O (4:1) at 4 C. The sediment separated by centrifugation was vacuum dried and dissolved in 0.05 M Tris–HCl buffer, pH 8 (acetone fraction). Protein concentration was according to Lowry et al. (1951). The initial specific activity of crude extract may not represent a real value due to eventual loss of protein but was considered as the initial activity of the isolation procedure. 2.3. Ion exchange chromatography The acetone fraction (165 mg proteins) was applied to a DEAE-Sephacel column (1.5 · 9.0 cm) using an FPLC system equilibrated with 0.05 M Tris–HCl buffer, pH 8 (1.5 ml/min flow rate). After extensive washing with equilibrium buffer, the inhibitor was eluted with a linear gradient of NaCl (0–0.6 M in 88 min). Spectrophotometry at 280 nm was used to follow protein elution, and inhibitory activity was determined by trypsin inhibition using DL-Bz-Arg-pNan as substrate (see below). Fractions with trypsin inhibitory activity were pooled, dialysed and concentrated under nitrogen atmosphere using a membrane to exclude the Mr 3000 range. 2.4. Gel filtration chromatography Any protein with inhibitory activity (2.2 mg) was submitted to gel filtration on a Superdex 75 column (1.0 · 25 cm) using an FPLC system equilibrated with

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

547

0.05 M Tris–HCl buffer, pH 8, containing 0.5 M NaCl (0.5 ml/min flow rate). Protein content and inhibitory activity were measured as described above.

concentration was determined by titration with trypsin assuming an equimolecular binding between enzyme and inhibitor.

2.5. Reverse-phase chromatography

2.11. Trypsin inhibition and Ki determination

The gel-filtered CmTI inhibitor (20 lg) was applied onto a C4 reversed-phase column using an HPLC system equilibrated with 0.1% (v/v) CF3CO2H acid (TFA) in water. Separation was achieved using an CH3CN gradient (0–80%, 120 min) in 0.1% (v/v) TFA. The eluted protein was used for amino acid sequence determination.

The equilibrium dissociation constant (Ki) and the inhibitor concentration were determined for trypsin through pre-incubation of the enzyme with increasing concentrations of the inhibitor at 37 C in 0.1 M Tris– HCl, pH 8.0, in a final volume of 200 ll. Residual activity was subsequently measured using 4 mM Bz-Arg-pNan as substrate.

2.6. SDS electrophoresis 2.12. Inhibition of other serine proteinases The CmTI molecular mass was assessed by electrophoresis on 15% SDS–polyacrylamide gels as described by Laemmli (1970) at reducing or non-reducing conditions. CmTI was reduced with 1.3 or 2.6 M DTT and heated to 100 C for 15 or 45 min. Protein was detected by silver staining (Heuskoven and Dernick, 1985). 2.7. Structure determination For CmTI sequence determination, peptides were obtained from the C4-column protein peak. After reduction and carboxymethylation, proteins were subjected to Aquapore RP300 C (7–8 lm column) HPLC system and from enzymatic digestion (chymotrypsin and endoproteases Lys-C, Glu-C, Asp-N) as well as CNBr cleavage (Friedman et al., 1970). The peptides in the mixtures were separated using the same HPLC system described above and their sequences determined by automated Edman degradation on a 477A Applied Biosystems protein sequencer. 2.8. Mass spectroscopy Purified inhibitor was infused into an atmosphericpressure ionization source fitted to a tandem quadruple instrument AP III (Sciex) as previously described (Covey et al., 1988; Mann, 1990).

After pre-incubation with CmTI, inhibition of chymotrypsin and porcine pancreatic elastase was determined by measuring the remaining hydrolytic activity towards 10 mM Suc-Phe-pNan or 5 mM MeO-SucAla-Ala-Pro-Val-pNan, respectively. Enzymes or enzyme-inhibitor mixtures (20 ll) were added to a solution containing substrate in 0.05 M Tris–HCl buffer, pH 8.0, in a final volume of 200 ll at 37 C. The reaction was monitored for 15–30 min and stopped by addition of 50 ll of 30% acetic acid and substrate hydrolysis was followed by absorbance at 405 nm in an ELISA reader. 2.13. Temperature and pH stability CmTI was maintained in 0.15 M NaCl at 20, 37, 60 and 100 C. After 30 min, the inhibitor solutions were incubated for 15 min with trypsin at pH 8.0. Subsequently, the residual trypsin activity was measured with 4 mM DL-Bz-Arg-pNan as substrate. The pH effect on CmTI activity was determined through pre-incubation of CmTI for 30 min at 37 C in buffers with pH values ranging from 2.0 to 10.0. The pH was subsequently adjusted to 8.0, and trypsin inhibition was assayed as described above.

2.9. Sequence comparison

3. Results

The database from the Max-Planck-Institute for Biochemistry at Martinsried, Germany, was accessed using the Lipmann and Pearson fast protein-searching algorithm, FASTP (Lipmann and Pearson, 1985). Alignments were optimised using CLUSTAL (Higgins and Sharp, 1988).

3.1. Inhibitor purification

2.10. Active site and inhibitor titration Trypsin active site concentration was determined by NPGB titration (Sampaio et al., 1984). Inhibitor

The results of the isolation procedure are summarised in Table 1. Trypsin inhibition was only detected after the C. mollis seed extract had been heated. This treatment abolished the endogenous DL-Bz-Arg-pNan hydrolyzing activity. Acetone fractionation was efficient in concentrating the trypsin inhibitor but did not increase biological activity (Table 1). Samples with trypsin inhibitory activity were subjected to DEAE-Sephacel chromatographic column

548

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

Table 1 Purification of CmTI from Cratylia mollis seeds Preparation

Total protein (mg)

Total activity (U)

Specific activity (U/mg)

Purification (x fold)

Yield (%)

Crude extract Acetone fractionation DEAE-Sephacel fraction Superdex 75 fraction

1200 976 280 19

6.6 5.6 2.8 2.0

0.0055 0.0057 0.01 0.10

1 – 1.7 18

100 85 42 30

The protein was measured according to Lowry et al. (1951); inhibitory activity was determined by trypsin inhibition assay whereby 1 U = 1 mg of inhibitor; specific activity means mg of inhibitor/mg of protein; the purification factor was determined via the specific activity values; yield was expressed as percentage of isolated inhibitor (Tanaka et al., 1997). The initial specific activity of crude extract was considered as the initial activity of the isolation procedure.

where the inhibitor was 1.7-fold purified with a yield of 42% (Table 1). The pooled inhibitor material was subsequently loaded onto a Superdex-75 column and activity was recovered in only one peak (Fig. 1).

was not affected within the pH 2–10 range and temperatures of 20, 37, 60 and 100 C did not decrease its effect on trypsin. 3.4. Stoichiometry of trypsin inhibition

3.2. Electrophoretic analysis The purity of CmTI was analysed using polyacrylamide gel electrophoresis. A single band (Mr 17 kDa) was obtained, indicating that CmTI was homogeneous (inset 1A). After 1.3 M dithiothreitol treatment and heating for 15 min, the inhibitor showed the same protein migration behavior; however, a faint band of 11 kDa was also detected at 2.6 M DTT. The increased heating time (45 min) resulted in 11 kDa as main peptide. 3.3. Temperature and pH stability Analysis of the temperature and pH effect revealed that CmTI is remarkably stable. Inhibitory activity

Bovine trypsin was inhibited by CmTI and the dissociation constant of the complex was measured by determination of the enzyme’s residual activity after incubation with the inhibitor. A Ki value of 1.4 · 10 9 M (Fig. 2) was calculated using the equation described by Morrison for slow tight binding. We assayed also the inhibitory effect of CmTI on chymotrypsin and porcine pancreatic elastase but no inhibition was detected. 3.5. Primary structure determination CmTI was loaded onto a C4 reversed-phase column and the eluted protein peak was used to determine the primary structure. Peptides were obtained from the C4 column through reduction followed by enzymatic cleav-

Fig. 1. Gel filtration on Superdex 75 column (FPLC). Protein (2.2 mg) was subjected to Superdex 75 CC (1.0 · 25 cm) equilibrated (0.5 mL/min) with 0.05 M Tris–HCl buffer, pH 8.0. The inset A shows SDS–PAGE of the native (not reduced) inhibitor purified by gel filtration on Superdex 75.

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

1

Residual Activity (%)

Ki = 1.4 nM 0.8

0.6

0.4

0.2

0

0

10

20

30

40

50

60

Inhibitor (nM)

Fig. 2. Trypsin inhibition by CmTI. Bovine trypsin (0.42 lM) preincubated (10 min, 37 C, 0.1 M Tris–HCl buffer, pH 8.0) with increasing amounts of CmTI eluted from Superdex 75 column. The residual trypsin activity was assayed with DL-Bz-Arg-pNan (4 mM) as substrate as described in Section 2.

549

age, CNBr cleavage, and modification with 4-vinyl pyridine. The resulting proteins were identified as CmTI1 and CmTI2 (Fig. 3). The amino acid sequences of CmTI1 and CmTI2 were determined by automated sequence analysis resulting in polypeptides with 77 and 78 amino acid residues, respectively (Fig. 4). Both isoinhibitors showed highest structural similarity to the Bowman-Birk inhibitors from other Leguminosae (Fig. 4) and high homology with each other, but differed in the reactive site of the C-terminally located inhibitor domain. The putative reactive site of CmTI1 is a lysine residue at position 22 and a tyrosine at position 49, whereas CmTI2 shows besides lysine at position 22 a leucine at position 49 (identified by alignment to related inhibitors). The molecular mass of CmTI1, as determined by mass spectrometry, was 8555.5 (elementary composition: C335 H520N101O128S17; pI = 4.6). The molecular mass of CmTI2 was 8625.5 (elementary composition: C341H530N99O133S15; pI = 4.2).

Fig. 3. Amino acid sequences of CmTI1 and CmTI2. The peptides obtained after fragmentation by enzyme digestion with chymotrypsin and endoprotease Lys-C, Glu-C, Asp-N or by cyanogen bromide are indicated. MS, mass spectroscopy.

550

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

CmTI1 CmTI2 DgTI CLTI SBTI TcTI CmTI1 CmTI2 DgTI CLTI SBTI TcTI

01

10

S D H S S D E S E S D E S G D E S E S S S S D D E S S

S S G K S K

21

30

T T T T T T

K K K K K K

S S S S S S

E E E E N I

P P P P P P

P P P P P P

Q Q Q Q Q Q

C C C C C C

Q Q Q Q R H

C C C C C C

E K E K K E

H F A L S S

C C C C C C

M I V L I A

C C C C C C

T T S A A T

Y L H L L H

H D H D D D

F F F F F F

C C C C C C

Y Y H Y Y Y

E K E K E K

P P P P P P

C C C C C C

K K K K K -

41

CmTI1 CmTI2 DgTI CLTI SBTI TcTI

A S A A A A N T T T T T

K P C C K A C C - - C C C C K P C C E A C C

D D D D D D

K D R E Q R

C C C C C C

Q R R K A A

K V Q V S A

D D D D D D

V M V T M I

R R R R R R

L L L L L L

N E N E N N

S S S S S S

C C C C C C

H H H H H H

S S S S S S

M F M F Y I

P P P P P P

G A G A A A

M Q L K Q Q

C C C C C C

R R S R F R

C C C C C C

L V L V V F

D D D D D D

S S S G S S

G G G G E G

D D D G D -

D D D D D -

E D E E K -

E G D D H D D E N -

C C C C C C 40

50

61

CmTI1 CmTI2 DgTI CLTI SBTI TcTI

C C C C C C

20

S S P P S W

S S S S S S 60

T T I T I I

70

S S S S P -

Fig. 4. Amino acid sequences of CmTI1 and CmTI2 and other structurally homolog Bowman-Birk type inhibitors. DgTI (Bueno et al., 1999); CLTICanavalia lineata (Odani and Ikenaka, 1972); SBTI-Glycine max (Lourenc¸o et al., 1989) and TcTI (Tanaka et al., 1997). Amino acids at the P1 positions of the reactive sites are in white lettering on a gray background and cysteine residues are shaded in gray.

4. Discussion The inhibitor (CmTI) material isolated by gel filtration chromatography contains two isoforms with high affinity for bovine trypsin (1.4 · 10 9 M). In contrast to other structurally related inhibitors, CmTI neither inhibited chymotrypsin nor porcine pancreatic elastase (Tanaka et al., 1996; Deveraj and Manjunatha, 1999). Its high specificity for trypsin is comparable to that of Dioclea glabra trypsin inhibitor, DgTI (Bueno et al., 1999). CmTI is remarkably stable over a wide pH range and unusually heat-resistant (data not shown). The large number of cysteine residues that form disulfide bonds render them structurally stable (Singh and Appu, 2002). Comparison of the molecular mass of CmTI (Mr 17,000 Da determined by SDS–PAGE) with the molecular masses determined by sequence of CmTI1 (8554.5 Da) and CmTI2 (8624.4 Da) indicates a molecular association of inhibitor isoforms, a common characteristic of a Bowman Birk inhibitor (Tanaka et al., 1997; de la Sierra et al., 1999). The different migration obtained (11 kDa polypeptide) under SDS–PAGE and 2.6 M DTT could be explained by disulfide bond of different stabilities in the inhibitor structure. In fact, a model has been proposed to localize labile and more stable disulfide bonds (Biewenga and van Run, 1992). The molecular masses (determined by mass spectrometry) of

CmTI1 (8555.5) and CmTI2 (8625.5) are close to the calculated masses from amino acid sequences but posttranslational modifications to the isolated proteins need to be evaluated. The protein sequences established showed clearly that CmTI1 and CmTI2 are structurally related to the Bowman Birk inhibitor family. In particular, the 14 cysteine residues and the lysine residue in P1 position of the Nterminally located reactive site are in conserved positions previously described for other Bowman-Birk inhibitors. The second putative reactive site has a tyrosine in CmTI1 at P1 position and a leucine in CmTI2; however, CmTI preparation containing both inhibitors was able to inhibit only trypsin. A remarkable difference between CmTI inhibitors and known Bowman-Birk inhibitors is that CmTI inhibitors are rich in acidic amino acids, which suggests that these inhibitors are more closely related to DgTI (Bueno et al., 1999). Also lack of chymotrypsin inhibition by CmTI makes it functionally more similar to DgTI and less similar to Torresea cearensis, TcTI, which inhibits chymotrypsin with a Ki of 5.0 · 10 8 M (Tanaka et al., 1997) probably due to a strong negatively charged carboxyl portion that interferes with the chymotrypsin interaction. Bowman-Birk inhibitors are classified according to their structural features and inhibitory characteristics. Isolated from dicotyledons, they are doubled-headed

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

proteins. Inhibitors isolated from monocotyledons are classified as Bowman-Birk type I (Mr 16,000 proteins and two reactive sites) or Bowman-Birk type II inhibitors (Mr 8,000 proteins with only one reactive site). The reactive sites are formed by a consensus sequence C-T-P1-S-X-P-P-Q-C (X being any amino acid residue) with lysine, arginine or serine at P1 position of the first reactive site being effective for trypsin inhibition. At the second reactive site, lysine, arginine, phenylalanine, tyrosine or leucine can occupy the P1 position (Prakash et al., 1996). The lack of a second reactive site in monocotyledon Bowman-Birk type II inhibitors may be due to the loss of the Cys10-Cys11 disulphide bridge that connects the second inhibitory loop to the first inhibitor domain or to the loss of some conserved residues thus resulting in specific trypsin inhibitors (Lin et al., 1993). This explanation serves as well for CmTI1 and CmTI2 as it did for DgTI isolated from dicotyledons. The sequence C-T-K-S-E-P-P-Q-C (8 conserved residues and appropriate P1 residue) comprises the first reactive sites of both CmTI1 and CmTI2. However, the second reactive site sequences C-T-Y-S-M-P-G-M-C (of CmTI1) and C-T-L-S-F-A-Q-M-C (of CmTI2) present only 5 and 4 conserved residues, respectively, and have different residues in P1 position (Y or L). This variability may be responsible for the poor activity of the second reactive site of these inhibitors, allowing specificities similar to those of Bowman-Birk type II inhibitors. But they have not lost the disulphide bridge, as occurs in some monocotyledonous inhibitors that lack cysteine residues in conservative positions (Prakash et al., 1996). Trypsin activity was already detected in C. mollis seeds (data not shown); the presence of CmTI in the same vegetal tissue may minimize digestion of peptides and proteins in quiescent seeds.

Acknowledgements The authors thank Reinhardt Mentele from the Abteilung fu¨r Klinische Chemie und Klinische Biochemie, Chirurgische Klinik und Poliklinik of LMU, Munich, Germany, for performing the structure determination. This study was partially supported by CAPES, CNPq, and SPDM (Brazil) as well as SFB 469 and VW Foundation, Germany.

References Biewenga, J., van Run, P.E.M., 1992. Effects of limited reduction on disulfide bonds in human IgA1 and IgA1 fragments. Molecular Immunology 29, 327–334. Birk, Y., 1996. Protein proteinase inhibitors in legume seeds – overview. Archivos Latinoamericanos de Nutricion 44, 26S–30S. Bueno, N.R., Fritz, H., Auerswald, E.A., Mentele, R., Sampaio, M., Sampaio, C.A., Oliva, M.L., 1999. Primary structure of Dioclea

551

glabra trypsin inhibitor, DgTI, a Bowman-Birk inhibitor. Biochemical and Biophysical Research Communications 261, 838–843. Chu, S.C., Chou, F.P., Liu, J.Y., Lin, L.J., Hsieh, Y.S., 1997. Proteolytic activities of mouse sarcoma 180 cells that are inhibited by Bowman-Birk and Kunitz protease inhibitors. Biochemistry and Molecular Biology International 42, 965–975. Covey, T.R., Bronner, R.F., Shushan, B.I., Henion, J., 1988. The determination of protein, oligonucleotide and peptide molecular weights by ion-spray mass spectrometry. Rapid Communications in Mass Spectrometry 2, 249–256. de la Sierra, L., Quillien, L., Flecker, P., Gueguen, J., Brunie, S., 1999. Dimeric crystal structure of a Bowman-Birk protease inhibitor from pea seeds. Journal of Molecular Biology 285, 1195–1207. Deveraj, V.R., Manjunatha, N.H., 1999. Purification and characterization of a proteinase inhibitor from field bean, Dolichos lablab purpureus L. Journal of Protein Chemistry 18, 47–54. Dittmann, K.H., Gueven, N., Mayer, C., Rodemann, H.P., 2001. Characterization of the amino acids essential for the photo-and radioprotective effects of a Bowman-Birk protease inhibitorderived nonapeptide. Protein Engineering 14, 157–160. Friedman, M., Krull, L.H., Cavins, J.F., 1970. The chromatographic determination of cystine and cysteine residues in proteins as S-beta(4-pyridylethyl)cysteine. Journal of Biological Chemistry 245, 3868–3871. Garcia-Gasca, T., Salazar-Olivo, L.A., Mendiola-Olaya, E., BlancoLabra, A., 2002. The effects of a protease inhibitor fraction from tepary bean (Phaseolus acutifolius) on in vitro cell proliferation and cell adhesion of transformed cells. Toxicology in vitro 16, 229– 233. Gariani, T., Leatherbarrow, R.J., 1997. Stability of protease inhibitors based on the Bowman-Birk reactive site loop to hydrolysis by proteases. Journal of Peptide Research 49, 467–475. Gladysheva, I.P., Balabushevich, N.G., Moroz, N.A., Larionova, N.I., 2000. Isolation and characterization of soybean Bowman-Birk inhibitor from different sources. Biochemistry 65, 198–203. Heuskoven, J., Dernick, R., 1985. Simplified method for silver staining of proteins in polyacrylamide gels and the mechanism of silver staining. Electrophoresis 6, 103–112. Higgins, D.G., Sharp, P.M., 1988. CLUSTAL: a package for performing multiple sequence alignment on a microcomputer. Gene 73, 237–244. Hossain, M.A., Becker, K., 2002. In vitro rumen degradability of crude protein seeds from four Sesbania spp. and the effects of treatments designed to reduce the levels of antinutrients in seeds. Animal Feed Science and Technology 95, 49–62. Kennedy, A.R., Wan, X.S., 2002. Effects of Bowman-Birk inhibitor on growth, invasion, and clonogenic survival of human prostate epithelial cells and prostate cancer cells. Prostate 50, 125–133. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. Lin, G., Bode, W., Huber, R., Chi, C., Engh, R.A., 1993. The 0.25-nm X-ray structure of the Bowman-Birk-type inhibitor from mung bean in ternary complex with porcine trypsin. European Journal of Biochemistry 212, 549–555. Lipmann, D.J., Pearson, W.R. 1985. Improved tools for biological sequence comparison. In: Proceeding of National Academy of Science, vol. 85, pp. 2444–2448. Lourenc¸o, D.M., Sampaio, M.U., Kerbauy, J., Sampaio, C.A.M., 1989. Estimation of plasma kallikrein in Sickle-cell anemia, and its relation to the coagulation and fibrinolytic systems. Advances in Experimental Medicine and Biology 247B, 553–557. Lowry, O.H., Rosebrought, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the folin phenol reagent. Journal of Biological Chemistry 193, 265–275. Mann, M., 1990. Electrospray: its potential and limitations as an ionization method for biomolecules. Organic Mass Spectrometry 25, 575–587.

552

P.M.G. Paiva et al. / Phytochemistry 67 (2006) 545–552

Mello, M.O., Tanaka, A.S., Silva-Filho, M.C., 2003. Molecular evolution of Bowman-Birk type proteinase inhibitors in flowering plants. Molecular Phylogenetics and Evolution 27, 103–112. Odani, S., Ikenaka, T., 1972. Studies on soybean trypsin inhibitors. IV. Complete amino acid sequence and the anti-proteinase sites of Bowman-Birk soybean proteinase inhibitor. Journal of Biochemistry 71, 839–848. Paiva, P.M.G., Souza, A.F., Oliva, M.L.V., Kennedy, J.F., Cavalcanti, M.S.M., Coelho, L.C.B.B., Sampaio, C.A.M., 2003. Isolation of a trypsin inhibitor from Echinodorus paniculatus seeds by affinity chromatography on immobilized Cratylia mollis isolectins. Bioresource Technology 88, 75–79. Park, S.S., Sumi, T., Ohba, H., Nakamura, O., Kimura, M., 2000. Complete amino acid sequences of three proteinase inhibitors from white sword bean (Canavalia gladiata). Bioscience Biotechnology and Biochemistry 64, 2272–2275. Prakash, B., Selvaraj, S., Murthy, M.R.N., Sreerama, Y.N., Rao, D.R., Gowda, L.R., 1996. Analysis of the amino acid sequences of plant Bowman-Birk inhibitors. Journal of Molecular Evolution 42, 560–569. Quillien, L., Ferrasson, E., Molle, D., Gueguen, J., 1997. Trypsin inhibitor polymorphism: multigene family expresion and posttranslational modification. Journal of Protein Chemistry 16, 195– 203. Sampaio, C.A.M., Sampaio, M.U., Prado, E.S., 1984. Active-site titration of horse urinary kallikrein. Hoppe Seyler’s Z Physiology Chemistry 365, 297–302. Scarpi, D., McBride, J.D., Leatherbarrow, R.J., 2002. Inhibition of human beta-tryptase by Bowman-Birk inhibitor derived peptides. Journal of Peptide Research 59, 90–93.

Shibata, H., Hara, S., Ikenaka, T., Abe, J., 1986. Purification and characterization of proteinase inhibitors from winged bean (Psophocarpus tetragonolobus (L.) DC.) seeds. Journal of Biochemistry 99, 1147–1155. Silva, J.A., Macedo, M.L., Novello, J.C., Marangoni, S., 2001. Biochemical characterization and N-terminal sequences of two new trypsin inibitors from Copaifera langsdorffii seeds. Journal of Protein Chemistry 20, 1–7. Singh, R.R., Appu, R.A.G., 2002. Reductive unfolding and oxidative refolding of a Bowman-Birk inhibitor from horsegram seeds (Dolichos biflorus): evidence for ‘‘hyperreactive’’ disulfide bonds and rate-limiting nature of disulfide isomerization in folding. Biochimica et Biophysica Acta 1597, 280–291. Tanaka, A.S., Sampaio, M.U., Marangoni, S., Oliveira, B., Novello, J.C., Oliva, M.L.V., Fink, E., Sampaio, C.A.M., 1997. Purification and primary structure determination of a Bowman-Birk trypsin inhibitor from Torresea cearensis seeds. Biological Chemistry 378, 273–281. Tanaka, A.S., Sampaio, M.U., Mentele, R., Auerswald, E.A., Sampaio, C.A.M., 1996. Sequence of a new Bowman-Birk inhibitor from Torresea acreana seeds and comparison with Torresea cearensis trypsin inhibitor (TcTI2). Journal of Protein Chemistry 15, 553–560. Ware, J.H., Wan, X.S., Rubin, H., Schechter, N.M., Kennedy, A.R., 1997. Soybean Bowman-Birk protease inhibitor is a highly effective inhibitor of human mast cell chymase. Archives of Biochemistry and Biophysics 344, 133–138. Wu, C., Whitaker, J.R., 1991. Homology among trypsin/chymotrypsin inhibitors form red kidney bean, Brazilian pink bean, lima bean, and soybean. Agricultural and Biological Chemistry 39, 1583–1589.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 553–560 www.elsevier.com/locate/phytochem

Overexpression of the Saussurea medusa chalcone isomerase gene in S. involucrata hairy root cultures enhances their biosynthesis of apigenin Feng-Xia Li a

a,b

, Zhi-Ping Jin a,b, De-Xiu Zhao a,*, Li-Qin Cheng a, Chun-Xiang Fu a,b, Fengshan Ma c

Key Laboratory of Photosynthesis and Environmental Molecular Physiology, Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China b Graduate School of the Chinese Academy of Sciences, Beijing 100039, China c Department of Biology, University of Waterloo, Waterloo, Ont., Canada N2L 3G1 Received 31 May 2005; received in revised form 23 November 2005 Available online 20 January 2006

Abstract Saussurea involucrata is a medicinal plant well known for its flavonoids, including apigenin, which has been shown to significantly inhibit tumorigenesis. Since naturally occurring apigenin is in very low abundance, we took a transgenic approach to increase apigenin production by engineering the flavonoid pathway. A construct was made to contain the complete cDNA sequence of the Saussurea medusa chalcone isomerase (CHI) gene under the control of the cauliflower mosaic virus (CaMV) 35S promoter. Using an Agrobacterium rhizogenes-mediated transformation system, the chi overexpression cassette was incorporated into the genome of S. involucrata, and transgenic hairy root lines were established. CHI converts naringenin chalcone into naringenin that is the precursor of apigenin. We observed that transgenic hairy root lines grew faster and produced higher levels of apigenin and total flavonoids than wild-type hairy roots did. Over a culture period of 5 weeks, the best-performing line (C46) accumulated 32.1 mg L
1 apigenin and 647.8 mg L
1 total flavonoids, or 12 and 4 times, respectively, higher than wild-type hairy roots did. The enhanced productivity corresponded to elevated CHI activity, confirming the key role that CHI played for total flavonoids and apigenin synthesis and the efficiency of the current metabolic engineering strategy.
2005 Elsevier Ltd. All rights reserved. Keywords: Saussurea involucrata; Compositae; Hairy root; Genetic transformation; Chalcone isomerase; Flavonoids; Apigenin

1. Introduction Saussurea involucrata Kar. et Kir. ex Maxim. has been widely used in Chinese medicine because of its effect in promoting blood circulation and anti-inflammation, and its analgesic functions (Li et al., 1980). The main bioactive chemical constituents of this plant are syringin, S. involucrata polysaccharides, umbelliferone, and flavonoids (Wang et al., 1986; Zheng et al., 1993; Han, 1995; Zhao and Wang, 2003). The content of total flavonoids has been established as the criterion for quality control (Wang et al., 1996). Among the flavonoids investigated (such as apigenin (5), *

Corresponding author. Tel.: +86 10 62836201; fax: +86 10 62590833. E-mail address: [email protected] (D.-X. Zhao).

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.004

jaceoside, hispidulin, quercetin and rutin), apigenin (5) (5,7,4 0 -trihydroxyflavone) is of particular interest (Fig. 1). Studies on apigenin (5) have demonstrated that it can inhibit the development of several types of cancers (Yin et al., 1999; McVean et al., 2000; Wang et al., 2000) as well as its anti-inflammatory effects (Liang et al., 1999; Gupta et al., 2001). These properties are responsible for a growing demand for S. involucrata and its flavonoid extracts. Supply of this medicinal plant is presently only from the wild populations, since field cultivation of this plant remains unsuccessful. In the wild, S. involucrata grows only on snowy mountains at altitudes of 4000–5000 m, and now this valuable plant species is under enormous pressure for survival. Production of the desired flavonoids by means of chemical synthesis could be an alternative source, only

554

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560 4-Coumaroyl CoA 1

Malonyl CoA 2 (×3) Chalcone Synthase

OH

HO

OH

OH

O

Naringenin Chalcone 3 Chalcone Isomerase

O

HO

OH

OH

Flavone Synthase HO

O

OH

O

OH

O

Apigenin 5 (Flavone)

Naringenin 4 Flavanone 3Hydroxylase

O

HO

OH

DihydroFlavonol Reductase

OH Anthocyanins

OH O

Dihydrokaempferol 6 Flavonol Synthase Flavonols

Fig. 1. The flavonoid pathway.

it will be too costly for commercial application. Still another option is to produce the bioactive compounds by in vitro culture systems (Kieran et al., 1997), but cell lines tend to accumulate only low levels of secondary metabolites and are genetically unstable (Rhodes et al., 1990). Our goal has been to establish an in vitro system for S. involucrata with satisfactory efficiency of flavonoids accumulation in general and apigenin biosynthesis in particular. Hairy roots systems have attracted much attention by their great potential for producing important pharmaceuticals (Hamill et al., 1987; Yoshikawa and Furuya, 1987). Hairy roots are normally formed when the T-DNA of the root-inducing plasmid (pRi) resided in Agrobacterium rhizogenes is integrated into the genome of a plant (White and Nester, 1980). Hairy root cultures are characterized by rapid growth, frequent branching and, more importantly, the ability to synthesize the same compounds as the intact plants. However, in most cases the yields of specific metabolites are so low that commercial exploitation is not possible. Metabolic engineering could be a useful approach to boost accumulation of targeted metabolites. We have chosen chalcone isomerase (CHI; EC 5.5.1.6) to start our investigation in this direction. CHI is one of the key enzymes of the flavonoid pathway (Fig. 1). Specifically, CHI naringenin converts chalcone (3) into (2S)-naringenin (4). Although this conversion can occur spontaneously, CHI catalyzes this conversion by 107-fold faster (Bednar and Hadcock, 1988). In addition, CHI controls the preferred formation of biologically active (2S)-

flavanones as well (Jez et al., 2000; Jez and Noel, 2002). In support of the notion that CHI plays critical roles in flavonoid metabolism, Reuber et al. (1997) reported that a barley mutant of CHI defection showed a dramatic reduction of flavonoid levels in the primary leaf. In a more recent study, Kim et al. (2004) described that inactivation of CHI in onion resulted in a significantly reduced amount of quercetin. Alternatively, Muir et al. (2001) documented that overexpression of petunia chi in tomato, was responsible for up to 78-fold increases in fruit peel flavonols, mainly in the form of quercetinglycosides. These examples suggest that CHI is a very good candidate for genetic manipulation. In the present work, the chi gene from S. medusa was overexpressed in hairy roots of S. involucrata under the control of the cauliflower mosaic virus (CaMV) 35S promoter. The growth rates, as well as apigenin (5) and total flavonoid production capacity, of the engineered hairy root lines were compared with those of wild-type hairy roots.

2. Results and discussion 2.1. Root transformation and establishment of hairy root lines A. rhizogenes strain R1601 carrying pRiA4 and the binary vector pCHI (Fig. 2(a)) was able to transform S. involucrata roots. Hairy roots were formed 2–3 weeks of

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560

555

Fig. 2. The pCHI plasmid used in transformation and PCR analysis of transgenic hairy root lines: (a) schematic representation of the pCHI construct. P35S, CaMV 35S promoter; TEV, tobacco etch virus translational enhancer sequence; chi, S. medusa chalcone isomerase gene; T35S, CaMV 35S terminator; hpt, hygromycin phosphotransferase gene; (b) representative PCR analyses for the chi, rolB and virG genes in transgenic hairy root lines. W, wild-type hairy root induced by A. rhizogenes strain R1601; Cn, transgenic hairy root lines induced by A. rhizogenes strain R1601 with the binary vector pCHI (‘‘n’’ indicates the line number); A, pRi from A. rhizogenes strain R1601; M, DNA marker (DGL-2000).

inoculation on over 80% of the root explants to constitutively express the Saussurea medusa chi gene. Such a high transformation efficiency with A. rhizogenes strain R1601 was unusual, as in our previous experiments on S. medusa with strains R1000, A4 and LBA9402 transformation rates never exceeded 5% (Zhao et al., 2004). About 10–15 d after hairy roots’ emergence, individual root clones were pre-selected in 1/2 MS medium supplemented with 20 mg L
1 hygromycin for at least five generations. Cefotaxime was added to the medium at 300 mg L
1 until the bacteria were eliminated. The resultant hairy root lines were then subcultured for 5–6 weeks in hormone-free, 1/2 MS medium. In all cases, some transgenic roots turned brown and aged considerably faster than wild-type root cultures (transformed by wild-type A. rhizogenes strain R1601 carrying only pRiA4). These brown roots were discarded and only those that showed a good growth capacity were maintained for further characterization. Altogether, 15 chi-transformed lines were maintained; they were designated C1, C7, C12, C17, C18, C20, C27, C29, C31, C35, C41, C46, C52, C57, and C60. The presence of chi, rolB and virG genes in the hairy roots’ genomes was confirmed by PCR (Fig. 2(b)). All selected transgenic lines gave two bands, one at 994 bp and another at 862 bp. The former corresponded to the S. medusa chi fragment and the latter to the rolB fragment. The coexistence of the chi and rolB genes confirmed that our root transformation was achieved as desired, by both the pCHI and pRiA4 plasmids. Wild-type hairy roots gave only one band corresponding to the rolB gene, which is a proof of transformation by the pRiA4 only. Hairy root lines obtained through transformation with A. rhizogenes strain R1601 carrying an empty pCAMBIA1301 vector (without chi), which were used as a second control, also

gave a band corresponding to rolB. Neither chi-transformed root lines nor control hairy roots carried the virG gene, which is present outside the T-DNA region of the Ri-plasmid; thus the possibility of A. rhizogenes R1601 contamination of hairy root cultures was ruled out. Considerable variations in growth capacity among individual lines were observed after the 5-week cultivation (Fig. 3). For example, line C46 achieved twofold higher fresh weight than line C52. Such variations were also observed in both groups of control hairy root lines. As previously reported for a few other hairy root systems (Jouhikainen et al., 1999; Moyano et al., 2003; Palazo´n et al., 2003; Zhang et al., 2004), differences in growth capacity were quite common, since each clone arised from an independent transformation event, depending on the presence of certain T-DNA genes of the Ri-plasmid, mainly the aux1 gene, in the genome of the root line obtained, growth could be very different. There was some indication that chi-transformed hairy roots senesced faster than control hairy roots. Maximum biomasses (dry weight) were achieved at 24–28 days and 30–35 days in chi-transformed and control hairy root lines (data not shown), respectively. 2.2. Apigenin (5) production The chi-transformed hairy roots of S. involucrata displayed a flavonoid spectrum similar to that of control hairy roots, as judged from HPLC analyses (Fig. 4). As mentioned in earlier reports, the profiles of secondary products are often conserved while, sometimes, concentrations of some compounds may alter (Park and Facchini, 2000; Palazo´n et al., 2003). In our system, there was a wide range of apigenin (5) content among the chi-transformed root lines,

556

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560

Control

Control

C1

C1

C7

C7

C12

C12

C17

C17

C18

C18

C20

C20

C27

C27

C29

C29

C31

C31

C35

C35

C41

C41

C46

C46

C52

C52

C57

C57

C60

C60 0

50

100

150

Fresh weight (g L-1)

0

5

10

15

Dry weight (g L-1)

Fig. 3. Fresh and dry weights (g L
1) of chi-transformed hairy root lines at the end of the culture period (35 d). The value for each line is the mean of three to five determinations ±SD, and the value for control represents the mean of five individual wild-type hairy root lines ±SD.

from 3.2 mg L
1(line C31) to 32.1 mg L
1 (line C46). These lines were divided into three groups according to their performance in apigenin production: best- (lines C17 and C46), moderate- (lines C1, C12, C18, C20, C27, C29, C35, C41, C52, C57 and C60), and low-performing (lines C7 and C31). It was exciting that the majority of these lines accumulated higher amounts of apigenin (5) than even the best apigenin-producing wild-type lines did (p < 0.05), except for the low-performance lines (lines C7 and C31) (Fig. 5(a)). Furthermore, a comparison on total flavonoids production indicated that, in spite of the considerable variations observed, the majority of the chi-transformed lines had significantly higher flavonoids than the control lines (p < 0.05) (Fig. 5(b)). Essentially same results were obtained from empty vector lines as wild-type lines (not shown). These observations strongly suggest that CHI is a key enzyme in the flavonoid pathway, and up-regulating CHI leads to enhanced biosynthesis of flavonoids. This conclusion is in agreement with an earlier report for tomato overexpressing petunia chi (Muir et al., 2001). 2.3. Expression of chi and enzyme activity To establish a possible relationship between the constitutive chi overexpression and the capacity of the hairy root lines to synthesize apigenin, chi gene transcript levels and CHI activity were studied in both control and chi-transformed hairy roots. Two lines from each of the three cate-

gories, best-, moderate-, and low-performing, were chosen from chi-transformed hairy roots for RT-PCR. Wild-type hairy roots and empty vector hairy roots were also examined. It was shown that the S. medusa chi gene was transcribed in all transformed hairy root lines of S. involucrata tested, but the transcripts were detected as different richness (Fig. 6 (a)). No detectable difference was found between wild-type hairy roots and empty vector hairy roots and, thus, only results for wild-type hairy roots were presented. In addition, the CHI enzymatic activity in these root lines was measured simultaneously, and a good correlation was observed between chi transcript levels and enzymatic activities. The strongest chi expression and high CHI activity were observed in the best-performing lines (C17 and C46), intermediate expression and activity in moderate-performing lines (C20 and C41), and low expression and activity in the low-producing lines (C7 and C31) (Fig. 6). Combining results on apigenin production (above) and CHI activity, it becomes clear that the two are positively correlated, confirming the pivotal role of chi in the regulation of apigenin synthesis. In summary, the present study provides an effective approach to efficiently increase the end products of secondary metabolic pathways by appropriate genetic engineering strategies. With the S. medusa chi in particular, which is an important gene in the flavonoid pathway, its overexpression in S. involucrata hairy roots did lead to enhanced apigenin biosynthesis.

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560

(a)

140 120 Apigenin

100 80

hygromycin and 50 mg L
1 rifampin. Positive clones as determined by PCR and enzymatic digestion for the presence of the chi gene were used to transform S. involucrata roots. 3.2. Root transformation and hairy root culture

60 40

Absorbance at 337 nm (mAU)

557

20 0 30

32.5

35

37.5

40

42.5

50

(b)

40 30 20 Apigenin

10 0 30

32.5

35

37.5

40

42.5

50

(c)

40 30 Standard apigenin

20 10 0 30

32.5

35

37.5

40

42.5

Retention time (min) Fig. 4. Chromatograms recorded at 337 nm: (a) chi-transformed root line C46; (b) a wild-type root line; (c) standard sample of apigenin (5).

3. Experimental 3.1. Construction of the chi binary expression vector The chi gene from S. medusa was amplified from the plasmid pTripIEx2. The latter contained the complete chi cDNA that was cloned from the Red callus line of S. medusa in our laboratory (Jin et al., 2004). The chi gene coding sequence thus obtained was digested with Sac I and BamH I, and then subcloned into pRTL2 between the CaMV 35S promoter linked to translational enhancement TEV leader sequence and the CaMV 35S terminator. The resultant construct, designated pRTL2-chi, was digested by Pst I and a 2236-bp fragment containing the complete chi expression cassette was obtained. This latter cassette was cloned into the pCAMBIA1301 vector (Cambia, Australia) carrying the hygromycin phosphotransferase gene (hpt) driven by the CaMV 35S promoter. The new binary vector was referred to as pCHI (Fig. 2(a)). The binary vector pCHI was mobilized by electroporation into the armed A. rhizogenes strain R1601 containing genes conferring resistance to kanamycin and rifampin (Mozo and Hooykaas, 1991). Recombinant clones of A. rhizogenes were selected with 100 mg L
1 kanamycin, 50 mg L
1

Genetic transformation of root explants from S. involucrata was carried out following the method used to transform S. medusa (Zhao et al., 2004). Briefly, micropropagated shoots of S. involucrata were rooted on 1/2 MS solid medium supplemented with 2.7 lM NAA. Root sections of about 10 mm in length were excised from these plants and were co-cultivated with the recombinant A. rhizogenes carrying pCHI at 25 C for 2 d in the dark. Next, the roots were incubated on 1/2 MS solid medium supplemented with 20 mg L
1 hygromycin and 500 mg L
1 cefotaxime for 4 weeks. Hairy roots developed at cut ends, which normally took place 2–3 weeks in culture, were excised and selected on 1/2 MS solid or liquid medium supplemented with 20 mg L
1 hygromycin. Cefotaxime was added to the medium at 300 mg L
1 until the bacteria were eliminated. Root cultures were grown at 25 C on a rotary shaker (90 rpm) under a 12/12-h (light/dark) photoperiod (45 mmol photons m
2 s
1, with Osram 40-W fluorescent tubes) and routinely subcultured every 35–42 d. After a minimum of five generations in liquid culture, rapidly growing roots were used to establish hairy root lines. To generate a hairy root line, about 80 mg (fr. wt) root sections of 20 mm were transferred to 30 ml 1/2 MS liquid medium in a 100-ml Erlenmeyer flask. As a control, hairy root lines were established by transforming roots with wild-type A. rhizogenes strain R1601 (wild-type control lines). A second control was also included by transforming roots with A. rhizogenes strain R1601 carrying a pCAMBIA1301 vector (empty vector control lines). Culture conditions were the same as for chi-engineered hairy root lines described above. 3.3. Transgene confirmation by PCR The S. medusa chi gene and the A. rhizogenes rolB gene in chi-transformed hairy roots were examined by PCR. Genomic DNA samples were obtained from both chi-engineered hairy root lines and wild-type hairy roots using the cetyl trimethyl ammonium bromide (CTAB) method (Doyle and Doyle, 1990). A primer pair of 5 0 -CCCACTATCCTTCG-3 0 and 5 0 -GCATGAGAGCTCTGGGCACACAGATGATTT-3 0 was designed to amplify a 994-bp fragment for the S. medusa chi gene (GenBank accession no. AF509335). A second pair of primers, 5 0 -TACTGCAGCAGGCTTCATGCA-3 0 and 5 0 -GCTTTCCCGACCAGAGACTG-3 0 , would amplify an 862-bp fragment of the rolB gene (Slightom et al., 1986). The rolB gene is located in the T-DNA region of pRiA4 plasmid, and thus can be used as a marker for transformation with this plasmid. In addition, primers (5 0 -GGCTTCGCCAACCAATTTGGAGAT-3 0 and 5 0 -TTTTGCTCCTTCAAGGGAG-

558

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560

(a)

(b)

Control

Control C1

C1 C7

C7

C12

C12

C17

C17

C18

C18

C20

C20

C27

C27

C29

C29

C31

C31

C35

C35

C41

C41

C46

C46

C52

C52

C57

C57

C60

C60 0.0

10.0

20.0

30.0

40.0

-1

Apigenin (mg L )

0

200

400

600

800

-1

Flavonoids (mg L )


1

Fig. 5. Apigenin (5) and total flavonoids (mg L ) production by chi-transformed hairy root lines at the end of the culture period (35 d). The value for each line is the mean of three to five determinations ±SD, and the value for control represents the mean of five individual wild-type hairy root lines ±SD.

and 2.5 ll 10· buffer in a total volume of 25 ll. Conditions for the chi amplification were as follows: initial denaturation at 94 C for 4 min, followed by 30 cycles of amplification (94 C 1 min, 56 C 1 min and 72 C 45 s) and 10 min at 72C. For rolB and virG, PCR was carried out in the following conditions: initial denaturation at 94 C for 4 min, 30 cycles of amplification (94 C 1 min, 53 C 1 min, and 72 C for 1 min), and 72 C for 10 min. PCR products were analyzed on 1.0% agarose gels. 3.4. Analyses of apigenin (5) and total flavonoids

Fig. 6. Analysis of chi gene expression and CHI activity: (a) RT-PCR analysis of the chi gene. Data are shown for control, lines C7, C31, C20, C41, C17 and C46. The actin gene was used as a loading control; (b) analysis of enzyme activity of hairy root lines. The value for each line is the mean of five determinations ±SD, and the value for control represents the mean of five wild-type hairy root lines ±SD.

GTGCC-3 0 ) for detecting the virG gene outside the TDNA of pRiA4 were also tested to eliminate the possibility of A. rhizogenes contamination of the hairy root cultures. Each PCR mixture contained 200 ng plant genomic DNA (or 100 pg Ri plasmid DNA), 400 lM each primer, 200 lM dNTPs, 1.0 unit Taq DNA polymerase (TaKaRa)

For all established root lines, roots were collected at 3, 7, 14, 21, 28, 32, 35 and 42 d after inoculation. Following filtration, roots were washed with 10 ml of distilled water and then dried at 60 C to a constant weight. Root samples (1.5 g dry wt) were ground and extracted with EtOH–H2O (30 ml, 7:3, v/v) at 25 C for 24 h. The extracts were cleared by centrifugation at 13,000g for 20 min. Analysis of total flavonoids content in the chi-transformed roots of S. involucrata were performed following the protocol of Guan et al. (1995). Wild-type hairy roots were examined as a control. Apigenin (5) levels were determined using an Agilent 1100 system (Agilent, USA) with a Zorbax 300 SB-C18 reversed-phase analytical column (250 · 4.5 mm, particle size 4 lm). A photodiode array detector (Agilent) was used to record online spectra (from 190 to 400 nm) of compounds eluted from the column. In each case, 5 ll sample was separated at constant ambient temperature using a gradient of MeOH in 0.1% orthophosphoric acid, at a flow rate of 1.0 ml min
1: 5–20% linear in 5 min, 20–30% in 7 min,

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560

30–45% in 20 min and 45–70% in 10 min, followed by a 10 min washing with MeOH–H2O (7:3) in 0.1% orthophosphoric acid. Peak purity, identification, and integration were carried out on Agilent Chemstations software A.10.02. Peak was monitored at 337 nm. The detection limit of this system was 0.05 lg ml
1 extract, corresponding to 5 mg kg
1 dry wt S. involucrata hairy roots. 3.5. Gene expression analysis by RT-PCR Total RNA was isolated using Trizol (Invitrogen, USA) from the hairy root lines after 3 weeks in culture. Reverse transcription was performed with 1 lg total RNA using a Promega reverse transcription kit. An aliquot from the reaction mixture was subjected to PCR with two pairs of primers, one pair was used to amplify the internal standard actin gene for a loading control, the other pair (5 0 -GATAAAGCCATTCCGTCAC-3 0 and 5 0 -ACTCGATTACTGTTTGTGCC-3 0 ) was used to amplify the S. medusa chi gene. PCR was carried out in the following conditions: initial denaturation at 94 C for 4 min, 30 cycles of 94 C 45 s, 50 C 45 s and 72 C 1 min, and terminated after extension at 72 C for 10 min. All products were examined on 1.5% agarose gels. 3.6. CHI enzyme activity assay CHI activity in chi-transgenic and control hairy root lines was measured with the method employed by Fouche´ and Dubery (1994) and Jez and Noel (2002). Hairy roots cultured 3 weeks were used. First, fresh hairy roots (2 g) were collected from the culture and washed with distilled water and homogenized in 6 ml 120 mM potassium phosphate buffer (pH 8.0), containing 3 mM EDTA, 20 mM b-mercaptoethanol, 2% hydrated polyclar AT, and 100 lM jodoacetic acid, followed by centrifugation at 12,000g for 30 min. The protein content of samples was determined by the Bradford method (Bradford, 1976). The standard CHI assay was performed at 25 C by monitoring the isomerization of naringenin chalcone (3) at 390 nm using a Beckman DU-640 spectrophotometer. The reaction mixture contained the following in a final volume of 1 ml: 50 mM Tris–HCl buffer (pH 7.6) containing 1% EtOH, 100 lM narigenin chalcone (3) (4,2 0 ,4 0 ,6 0 -tetrahydroxychalcone), and an appropriate amount of the enzyme. The rate of spontaneous cyclization of the substrate was determined beforehand and was subtracted from the rate in the presence of enzyme. Naringenin chalcone (3) was synthesized from naringenin (4) according to the procedure of Moustafa and Wong (1967) and naringenin (4) was purchased from Herbfine (Nanchang, China).

Acknowledgements We thank Dr. Xiaofeng Xue, Institute of Apiculture Research, Chinese Academy of Agriculture Science, Beijing,

559

for the generous gift of standard sample of apigenin. This work was supported by the National Science Foundation of China (No. 30472158).

References Bednar, R.A., Hadcock, J.R., 1988. Purification and characterization of chalcone isomerase from soybeans. J. Biol. Chem. 263, 9582–9588. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal. Biochem. 72, 248–254. Doyle, J.J., Doyle, J.L., 1990. Isolation of plant DNA from fresh tissue. Focus 12, 13–15. Fouche´, S.D., Dubery, I.A., 1994. Chalcone isomerase from Citrus sinensis: purification and characterization. Phytochemistry 37, 127–132. Guan, J.Y., Wang, W.W., Ma, M.T., 1995. Investigation of technological preparation on extracts of Saussurea involucrata. J. Shenyang Pharma. Univ. 12, 209–211. Gupta, S., Afaq, F., Mukhtar, H., 2001. Selective growth-inhibitory, cellcycle deregulatory and apoptotic response of apigenin in normal versus human prostate carcinoma cells. Biochem. Biophys. Res. Commun. 287, 914–920. Hamill, J.D., Parr, A.J., Rhodes, M.J.C., Robins, R.J., Walton, N.J., 1987. New routes to plant secondary products. Biotechnology 5, 800–804. Han, S.L., 1995. Study on the effect of four chemical constituents against cancer of Sausurea involucrata. Aibian Jibian Tubian 7, 80–83. Jez, J.M., Noel, J.P., 2002. Reaction mechanism of chalcone isomerase. J. Biol. Chem. 277, 1361–1369. Jez, J.M., Bowman, M.E., Dixon, R.A., Noel, J.P., 2000. Structure and mechanism of the evolutionarily unique plant enzyme chalcone isomerase. Nat. Struct. Biol. 7, 786–791. Jin, Z.P., Zhao, D.X., Qiao, C.L., Fu, C.X., 2004. Construction of cDNA library from the callus of Saussurea medusa. Chin. Bull. Bot. 21, 61–65. Jouhikainen, K., Lindgren, L., Jokelainen, T., Hiltunen, R., Teeri, T.H., Oksman-Caldentey, K.-M., 1999. Enhancement of scopolamine production in Hyoscyamus muticus L. hairy root cultures by genetic engineering. Planta 208, 545–551. Kieran, P.M., MacLoughlin, P.F., Malone, D.M., 1997. Plant cell suspension cultures: some engineering considerations. J. Biotechnol. 59, 39–52. Kim, S., Jones, R., Yoo, K.S., Pike, L.M., 2004. Gold color in onions (Allium cepa): a natural mutation of the chalcone isomerase gene resulting in a premature stop codon. Mol. Genet. Genom. 272, 411–419. Li, G.H., Liu, F., Zhao, R.C., 1980. Studies on pharmacological actions of Saussurea involucrata. Kar. et Kir. ex Maxim. Acta Pharmacol. Sin. 15, 368–369. Liang, Y.C., Huang, Y.T., Tsai, S.H., Lin-Shiau, S.Y., Chen, C.F., Lin, J.K., 1999. Suppression of inducible cyclooxygenase and inducible nitric oxide synthase by apigenin and related flavonoids in mouse macrophages. Carcinogenesis 20, 1945–1952. McVean, M., Xiao, H.Y., Isobe, K.I., Pelling, J.C., 2000. Increase in wildtype p53 stability and transactivational activity by the chemopreventive agent apigenin in keratinocytes. Carcinogenesis 21, 633–639. Moustafa, E., Wong, E., 1967. Purification and properties of chalconeflavanone isomerase from soya bean seed. Phytochemistry 6, 625–632. Moyano, E., Jouhikainen, K., Tammela, P., Palazo´n, J., Cusido´, R.M., Pinˇol, M.T., Teeri, T.H., Oksman-Caldentey, K.-M., 2003. Effect of pmt gene overexpression on tropane alkaloid production in transformed root cultures of Datura metel and Hyoscyamus muticus. J. Exp. Bot. 54, 203–211. Mozo, T., Hooykaas, P.J.J., 1991. Electroporation of megaplasmids into Agrobacterium. Plant Mol. Biol. 16, 917–918. Muir, S.R., Collins, G.J., Robinson, S., Hughes, S., Bovy, A., Ric de Vos, C.H., van Tunen, A.J., Verhoeyen, M.E., 2001. Overexpression of

560

F.-X. Li et al. / Phytochemistry 67 (2006) 553–560

petunia chalcone isomerase in tomato results in fruits containing increased levels of flavonols. Nat. Biotechnol. 19, 470–474. Park, S.U., Facchini, P.J., 2000. Agrobacterium rhizogenes-mediated transformation of opium poppy, Papaver somniferum L., and California poppy, Eschscholzia californica Cham., root cultures. J. Exp. Bot. 51, 1005–1016. Palazo´n, J., Moyano, E., Cusido´, R.M., Bonfill, M., Oksman-Caldentey, K.-M., Pin˜ol, M.T., 2003. Alkaloid production in Duboisia hybrid hairy roots and plants overexpressing the h6h gene. Plant Sci. 165, 1289–1295. Reuber, S., Jende-Strid, B., Wray, V., Weissenbo¨ck, G., 1997. Accumulation of the chalcone isosalipurposide in primary leaves of barley flavonoid mutants indicates a defective chalcone isomerase. Physiol. Plant. 101, 827–832. Rhodes, M.J.C., Robins, R.J., Hamill, J.D., Parr, A.J., Hilton, M.G., Walton, N.J., 1990. Properties of transformed root cultures. In: Charlwood, B.V., Rhodes, M.J.C. (Eds.), Secondary Products from Plant Tissue Culture: Proceedings of the Phytochemical Society of Europe, vol. 30. Clarendon, Oxford, pp. 201–225. Slightom, J.L., Durand-Tardif, M., Jouanin, L., Tepfer, D., 1986. Nucleotide sequence analysis of TL-DNA of Agrobacterium rhizogenes agropine type plasmid. Identification of open reading frames. J. Biol. Chem. 261, 108–121. Wang, H.K., Lin, Z.D., He, K., Wan, S.W., 1986. Studies on the chemical constituents of Saussurea involucrata. Yao Xue Xue Bao 21, 680–682. Wang, B.F., Lu, J., Guan, J.Y., 1996. The extraction of Saussurea involucrata injection. Chin. Pharm. J. 31, 299–301.

Wang, W.Q., Heideman, L., Chung, C.S., Pelling, J.C., Koehler, K.J., Birt, D.F., 2000. Cell-cycle arrest at G2/M and growth inhibition by apigenin in human colon carcinoma cell lines. Mol. Carcinogen. 28, 102–110. White, F.F., Nester, E.W., 1980. Hairy root: plasmid encodes virulence in Agrobacterium rhizogenis. J. Bacteriol. 141, 1134–1141. Yin, F., Giuliano, A.E., Van Herle, A.J., 1999. Signal pathways involved in apigenin inhibition of growth and induction of apoptosis of human anaplastic thyroid cancer cells. Anticancer Res. 19, 4297–4303. Yoshikawa, T., Furuya, T., 1987. Saponin production by cultures of Panax ginseng transformed with Agrobacterium rhizogenes. Plant Cell Rep. 6, 449–453. Zhang, L., Ding, R.X., Chai, Y.R., Bonfill, M., Moyano, E., OksmanCaldentey, K.-M., Xu, T.F., Pi, Y., Wang, Z.N., Zhang, H.M., Kai, G.Y., Liao, Z.H., Sun, X.F., Tang, K.X., 2004. Engineering tropane biosynthetic pathway in Hyoscyamus niger hairy root cultures. Proc. Natl. Acad. Sci. USA 101, 6786–6791. Zhao, L., Wang, X.L., 2003. Research on chemical composition, pharmacology and its clinic application of Saussurea involucrata. J. Southwest National Coll. 29, 424–428. Zhao, D.X., Fu, C.X., Chen, Y.Q., Ma, F.S., 2004. Transformation of Saussurea medusa for hairy roots and jaceosidin production. Plant Cell Rep. 23, 468–474. Zheng, R.L., Liu, G.S., Xing, G.X., Jia, Z.J., Du, M., Tan, L.Q., 1993. Free radical scavenging and antifatigue activities of Saussurea involucrata polysaccharides. Acta Pharmacol. Sin. 14 (Suppl.), 47–49.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 561–569 www.elsevier.com/locate/phytochem

Natural and directed biosynthesis of communesin alkaloids Lucy J. Wigley a, Peter G. Mantle a

a,*

, David A. Perry

b

Department of Biochemistry, Imperial College of Science, Technology and Medicine, South Kensington, London SW7 2AZ, UK b Pfizer Global Research and Development, Groton Laboratories, Eastern Point Road, Groton, CT 06340, USA Received 10 August 2005; received in revised form 6 October 2005 Available online 1 December 2005

Abstract A role for tryptophan, acetate, mevalonate and methionine in the biosynthesis of communesins A and B, novel structurally-related and biologically-active Penicillium metabolites, has been established by isotopic labelling techniques. The incorporation of 14C-tryptamine has also been demonstrated. DL-2-13C-tryptophan specifically enriched two carbon atoms in the 13C NMR spectrum, thereby defining the intra-molecular arrangement of the two tryptophan-derived moieties. Feeding differentially labelled precursors during communesin production showed that tryptophan and methionine are involved early in the biosynthesis and that mevalonate provides an isoprene which is added later. A biosynthetic pathway involving an early precursor based on tryptophan is proposed. Indole-N-(13C-methyl) tryptophan was not incorporated into communesins implying that N-methylation of tryptophan is not the first step of the communesin biosynthetic pathway. During deamination of indole-N-(13C-methyl) tryptophan to 1-13C-methylindole-3-carboxylic acid communesin biosynthesis was inhibited. Of several halogenated indoles tested for directed biosynthesis, only DL-6-fluoro-tryptophan and 6-fluorotryptamine caused accumulation of the corresponding monofluoro-analogues of communesins A and B. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Communesin; Alkaloids; Biosynthesis; Radiolabel precursor feeding; Directed biosynthesis; Fluoro-communesin

1. Introduction The structures of two fungal alkaloids communesin A and B (1 and 2), isolated from a Penicillium. sp possessing activity against cultured P-388 lymphocytic leukemia cells, have been described (Numata et al., 1993). Previously in the laboratories of Pfizer UK, structurally related compounds were discovered in an antihelminthic screen exhibiting activity against the free-living nematode Caenorhabditis elegans. The name ‘‘commindolines’’ was used for these metabolites partly on the basis of initial characterization of the producing fungus as Penicillium commune and partly due to the dominant indolic moiety. The structures of the principal compounds, ‘‘commindolines’’ B and A, equated to that of communesins A and B, respectively. Communesins are of mixed biosynthetic origin, predictably derived from tryptophan, mevalonate, acetate and a methyl group from methionine. A biosynthetic relationship *

Corresponding author. Tel.: +44 207 594 5245; fax: +44 207 225 0960. E-mail address: [email protected] (P.G. Mantle).

0031-9422/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.10.011

between communesins and the plant-derived calycanthaceous alkaloids is implied by the presence of a 1,2-diphenylethane group and two aminals. Studies on the biogenesis of calycanthaceous alkaloids (Robinson and Teuber, 1954; Henrickson et al., 1964; Kirby et al., 1969) may have relevance to that of the communesins. The present biosynthetic study was undertaken to seek any biosynthetic features in common with the plant alkaloids and to provide evidence for generating communesin analogues with potentially increased or altered biological activity by directed biosynthesis in microbial fermentation. 2. Results and discussion 2.1. Typical fungal growth and communesin production in submerged fermentation Communesin accumulation commenced during biomass accumulation (Fig. 1). However, decrease in biomass after 70 h is probably through respiration of assimilated reserve

562

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

carbohydrate. Replicatory growth alone is probably complete by 24 h, which is therefore the time point after which it is advantageous to add radiolabelled putative precursors of communesins that are also key intermediates in primary metabolism. The phase during which communesin concentration in broth continued to increase apparently lasted until around 200 h for both 1 and 2. HPLC analysis of cell and broth extracts suggested that 2 is more cell-associated than 1, possibly attributable to the latter
s lipophilic sorbyl side chain constraining its cellular release.

2.2. Communesin biosynthesis: radiochemical experiments

1.4 1 1.2 0.8

1

0.6

0.8 0.6

0.4

0.4 0.2 0.2 0

0 0

20

40

60

80 100 120 140 160 180 200 220 Time (h).

Biomass

1

-1

1.2

Communesin concentration (mg 100ml of starting medium).

1.6

-1

Dry cell weight (g 100ml of starting medium).

Addition of 14C-tryptophan, 14C-acetate, 3H-mevalonate, (methyl-14C)-methionine and 14C-tryptamine to separate submerged fermentations consistently gave rise to

2

Fig. 1. Changes in total communesin yields and biomass over the period of feeding differentially labelled precursors.

radiolabelled 1 and 2, as determined by scintillation counting of HPLC-resolved 1 and 2, indicating that these compounds are precursors of communesins (Table 1). The similar incorporation values recorded for tryptophan, methionine, mevalonate and tryptamine into 1 and 2 (mean of 8.25%, 1.0%, 0.59% and 1.75%, respectively), is consistent with a common structure in these communesins and also provides confidence for the experimental procedure employed. The incorporation of acetate into 2, 2.3%, was approximately double that into 1 and equates closely to the requirement for three acetates in the sorbyl side chain of 2 as compared to the single acetate of 1, particularly when the isoprene, common to both, is taken into account. The high efficiency of the incorporation of radiolabel from tryptophan contrasted with that of methionine. 2-14CTryptophan theoretically only contributes twice as much radiolabel to 1 and 2 as methyl-14C-methionine and therefore the sevenfold greater incorporation of tryptophan inversely reflects the relative metabolic demands for tryptophan and methionine in primary metabolism. The results suggest that the N-methyl group of methylated communesins might arise via donation from S-adenosyl methionine. Reduced competition from other pathways predictably makes tryptamine an even more efficient experimental biosynthetic precursor than tryptophan. However, the percentage incorporation of 14C-tryptamine into 1 and 2 was much lower than that of 14C-tryptophan (3.5% and 16.5%, respectively), possibly reflecting vacuole sequestration (Songstad et al., 1990) or poor uptake of exogenous tryptamine. Alternatively, if communesin biosynthesis is catalysed by a multienzyme complex, the low incorporation could be associated with problems of exogenous tryptamine gaining access to the site of catalysis, especially if biosynthetic intermediates are at some stage enzymebound. 2.3. 13C NMR experiment: incorporation of tryptophan into 1 and 2

DL[2-

13

C]-

Natural abundance 13C NMR spectra of 1 and 2 in acetone-d6showed well-separated signals so that the 13C NMR spectrum of each communesins isolated from a fermentation treated with [2-13C]-tryptophan-showed specific enrichment in the signals at 36.9 and 44.6 ppm, corresponding to C-23 and C-17, respectively. This is consistent with biosynthesis of communesins from two tryptophanderived moieties, the precise orientation of which was thus confirmed. The intensity ratios of signals corresponding to

Table 1 Comparison of percentage incorporation of 14C-tryptophan, [1-14C]acetic acid, RS-[2-3H]mevalonic acid, L-(methyl-14C)methionine and L-14C-tryptamine into 1 and 2 14

14

8.10 8.40 16.50

1.20 2.30 3.50

C-trp (1 lCi)

1 2 Total

C-acetate(1 lCi)

3

14

1.10 0.90 2.00

0.58 0.60 1.18

H-mev (8.5 lCi)

C-met (10 lCi)

14

C-tryptamine (0.14 lCi)

2.00 1.50 3.50

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

C-23 and C-17 in the natural abundance spectra of 1 (1.2:1) and 2 (1.1:1) persist in the enriched spectra, suggesting that the enrichment of each carbon atom occurred in an evenhanded manner. Using background signals C-3 and C-29 as a reference, with intensity similar to that of C-17 and C-23 in the natural abundance 13C NMR spectrum, it was estimated that enrichment in excess of 30-fold for 1 and approximately 18-fold for 2 had been achieved. Such disparate enrichments could have resulted from different rates of accumulation of 1 and 2 during the period of isotope addition (up to ca. 100 h, Fig. 1). 2.4. Dual label experiments, exploring biosynthetic sequence via the mixtures 14C-tryptophan: 3H-methionine or 14Ctryptophan: 3H-mevalonate If two precursors are involved concurrently in biosynthesis of any molecule throughout product accumulation, a mixture of the two precursors, given at any stage, will become incorporated into end product in an approximately constant ratio (Mantle and Shipston, 1987). Therefore, to gain information on the sequence of biosynthetic events, the incorporation of two precursor permutations 14C-tryptophan: 3H-methionine or 14C-tryptophan: 3H-mevalonate into 1 and 2 at various stages in the fermentation was investigated. The approximately constant ratio, 1:1, of 14C-tryptophan- and 3H-methionine-derived label incorporation in 1 and 2 throughout the fermentation indicated that these precursors were involved in the biosynthesis at about the same time (Fig. 2), and demonstrated validity in this case of the

563

principle upon which experiments on temporal involvement of precursors rely. Similar values for 1 and 2 follows from their identical structure in the regions derived from tryptophan and methionine. However, the skewed ratio of label specific activities in communesins following 14Ctryptophan and 3H-mevalonate administration at various stages of the fermentation implies the existence of temporal separation between the biogenic events involving these precursors (Fig. 2). Until 32 h into the fermentation there was an overwhelming preference for tryptophan incorporation into communesin biosynthesis, but later there was a swing in favour of mevalonate, coincident with the first appearance of end product (Fig. 1); the ratio of 14C: 3H decreased between 0 and 48 h from 0.83 to 0.33 in 1 and from 0.75 to 0.33 in 2 (Fig. 2). Isolation of 14C-labelled 1 and 2 from cultures fed with differentially labeled precursors before communesins were detectable, i.e., before 32 h, signifies the involvement of an early, tryptophan-based, biosynthetic intermediate. These results suggest that during the early part of the fermentation there is a build up of early tryptophan-based intermediates
which later give rise to recognisable communesins via completion steps, e.g., prenylation. Synthesis of enzymes catalysing later steps in the biogenesis of these secondary metabolites might be in response to accumulation of critical substrate levels (i.e., accumulation of specific biosynthetic intermediates) and may account for the sudden appearance of communesins, coinciding with the preferential incorporation of mevalonate. Results of administration of differentially-labelled precursors to Penicillium sp. isolate N934-53 therefore indicates that tryptophan and methionine are involved early in communesin biosynthesis and that mevalonate provides the isoprene which is added later. 2.5. Biotransformation of indole-N-(methyl-13C) tryptophan Indole-N-(methyl-13C) tryptophan was not incorporated into 1 or 2, which implied at first sight that methylation of tryptophan was not the first biosynthetic step, consistent with the dual-label experiments. However, exogenous indole-N-(methyl-13C) tryptophan was transformed to a compound identified as 1-methylindole-3-carboxylic acid (Fig. 3), as determined by MS and NMR spectroscopy (Wigley, 1995). Also communesins were not biosynthesised in the presence of indole-N-(methyl-13C) tryptophan as suggested by the absence of radiolabeled communesins in

Fig. 2. Changes in the 14C: 3H ratios in 1 and 2, attributed to radiolabelled communesin precursors from standard 14C-tryptophan: 3H-methionine or 14 C-tryptophan: 3H-mevalonate mixtures, measured at 186 h (solid lines) and 192 h (dotted lines), respectively, following addition to individual cultures at various stages through the fermentation.

Fig. 3. Biotransformation of indole-N-(13C-methyl)-tryptophan to indoleN-(13C-methyl)-3-carboxylic acid.

564

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

cultures treated with indole-N-(methyl-13C) tryptophan spiked with 14C-tryptophan. Plant hormones, indolacetic acid (IAA), 1-naphthaleneacetic acid (NAA) and 2,4-dichlorophenoxyacetic acid (2,4-D) down-regulate transcription of the tryptophan decarboxylase (TDC) gene in cultured Catharanthus roseus cells. Thus negative regulation by IAA has been suggested as a mechanism controlling terpenoid indole alkaloid biosynthesis in plants (Goddijn et al., 1992). It is therefore possible that the structurally related 1-methylindole-3-carboxylic acid could have a similar affect in fungi and may explain why communesins were not produced in the culture fed with indole-N-(methyl-13C) tryptophan. The cellular uptake, fungal transformation capacity and effect upon communesin production of this foreign compound was however, demonstrated. The accumulated evidence concerning communesin biosynthesis suggests that the decarboxylation of tryptophan to yield tryptamine is the first biogenic step (rather than methylation of tryptophan), complementary to reports by various authors that tryptophan decarboxylase forms a regulatory link between primary and secondary metabolism (Aerts et al., 1992 and Noe et al., 1984). The subsequently formed tryptamine is then methylated either before or after bb 0 -oxidative dimerisation with a second tryptamine moiety. Prenylation and acetylation of the putative di-tryptamine intermediate takes place at later stages (Fig. 4). Formation of a di-tryptamine moiety (by bb 0 -oxidative dimerisation of N-methyl tryptamine, hydrolysis of which gives rise to tetra-aminodialdehyde) prior to

methylation, analogous to the biosynthesis of the related calycanthaceous alkaloids (Robinson and Teuber, 1954), is more plausible because methylation does not appear to be a crucial step for the biosynthesis of all communesins. In the biosynthesis of the catharanthaceous alkaloids two asymmetric centres (at the b-indolic positions) are formed upon dimerisation and thus two diastereomers of tetra-aminodialdehyde are possible; one is meso and the other racemic or optically active, each giving rise to the five isomers (a-e) with corresponding optical activity (Robinson and Teuber, 1954). In the meso series however, only the disymmetric e isomer has potential optical activity (Henrickson et al., 1964). a, b and d are the only isomers that have previously been isolated from natural sources, and identified as (+)-calycanthine, (rac)-chimonanthine and iso-calycanthine, respectively (Adjibade et al., 1992). Communesins possibly represent the e isomer isolated, for the first time to our knowledge in this form, from a natural source. 2.6. Suppression by ethionine of the methylation step in communesin biosynthesis to reveal an unmethylated communesin Ethionine, as an exogenous competitive inhibitor of methylation, was used to alter flow of the communesin pathway to 1 and 2 and thus potentially to select for a demethyl-communesin. Preliminary experiments showed incorporation of 14C from 14C-methionine into 1 and 2, and other methionine-derived metabolites (e.g., ergosterol) in

Fig. 4. Partial biosynthetic pathway for communesin alkaloids proposed on the basis of the present experimental findings and known biosynthesis of Catharanthaceous alkaloids.

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

Fig. 5. Autoradiograph showing effect of ethionine (right), added to a fermentation at 48 h, on incorporation of 14C-tryptophan into metabolites at 72 h, when labelled tryptophan was added 30 min after the ethionine and to a control fermentation (left).

cell extract, was markedly reduced after treating fermentations with ethionine (50 mg per 100 ml
1) in early idiophase and adding 14C-methionine 30 min after the ethionine (Wigley, 1995). A similar experiment, substituting 14C-tryptophan for 14C-methionine so as to detect previously unrecognised indole alkaloids according to their biosynthesis from a tryptophan precursor, clearly showed reduced yields of 1 and 2 in broth extract (Fig. 5). Similar reduction in the trace amount of an unidentified compound (b) would be consistent with it being a closely related communesin. However, another minor compound (a), that was not evident on TLC chromatograms under UV254nm light and had only been recognised on account of the radioactivity in the control extract, was much increased in response to ethionine, implying that it may be a de-methyl communesin. 2.7. Directed biosynthesis of communesin analogues Ability of Penicillium sp. N934-53 to accept biosynthetic precursor analogues that could direct biosynthesis towards communesin analogues was explored first using 14C labelled tryptophan analogues as qualitative probes. Since such were not available commercially, sufficient of several probes (14C-bromotryptophan substituted in the 4 or 5 position, 14C-fluorotryptophan substituted in the 4 or 6 position, 14C-5-nitrotryptophan) was prepared by an E. coli expressing high tryptophan synthetase activity (see Section 3.2.8). Radiolabelled 4-fluorotryptophan, 4-bromotryptophan and 5-nitrotryptophan were not incorporated into 1 and 2, showing that the biosynthetic enzymes were too specific for these substrates. However, for the halogenated analogues, 8% and 11%, respectively, of the 14C given was found in fungal protein. Additional amounts were in non-communesin metabolites, indicating uptake and acceptance of these tryptophan analogues into other metabolic pathways. In contrast, 6-fluorotryptophan and 5-bromotryptophan were incorporated into analogues of 1 and 2 and this finding prompted experiments on a larger

565

scale with these halogenated indoles unlabelled. Also tested similarly were 6-fluoro-tryptamine hydrochloride, 5-fluorotryptamine and 5-methyl tryptophan. The mass spectra of 1 and 2 that had been generated in the presence of 6-fluoro-tryptamine hydrochloride contained, in addition to ions characteristic of native 1 and 2, ions corresponding to the communesin molecular ion plus 18 amu (m/z 474 and 526 for 1 and 2, respectively). Similar data was obtained for 2 from culture fed DL-6-fluorotryptophan. The elemental composition of these new ions, deduced from accurate mass measurement, was C28H32N4O2F (m/z 474) and C32H36N4O2F (m/z 526), consistent with the incorporation of only one fluorinated-indole moiety, thus demonstrating the feasibility of generating an analogue of the various communesins by substrate feeding. Absence of ions 36 amu greater than the native communesin molecular ion, implies that di-fluorinated communesins were not formed. Presence of fragment ions in the EI mass spectra corresponding to fluorination of specific native communesin fragment ions, but not of others, is in accordance with the fluoro-group being located exclusively on the ‘‘non-prenylated’’ side of the molecule (Fig. 6). It would seem that the 6-fluorotryptophan and -tryptamine derivatives are unacceptable substrates for insertion into the ultimately-prenylated moiety of communesins. However, insertion specificity could be a useful tool in generating new communesin analogues biosynthetically. The position of fluorine in the tryptophan-ring
moiety of fluoro-communesins derived from fluoro-tryptophan was assumed to be the same as in the parent precursor. The absence of a fluorine atom on the prenylated indole moiety suggests enzyme specificity at some stage in the biosynthesis, e.g., specificity of the dimerase, methylase or enzymes involved in prenylation whereby DL-6-fluorotryptophan or 6-fluoro-tryptamine cannot be accepted as sole indolic substrates for communesin biosynthesis. Since many secondary metabolite biosynthetic pathways are catalysed by multienzyme complexes (Gutierrez et al., 1991; Kleinkauf and von Dohren, 1990; Beck et al., 1990), it is conceivable that a metabolite precursor remains bound to the enzyme complex during a number of sequential biosynthetic steps. For example, tryptophan, implicated early in the biosynthesis, could be the starter unit, enzyme-bound to a very specific active site via interactions involving the aromatic ring. It seems rational therefore that tryptophan analogues can not be tolerated at this stage due to enzyme specificity, but this does not eliminate the possible acceptance of a second, modified tryptophan. It is probable that the presence of a highly electronegative fluorine atom would alter some chemical properties of communesins. However, low enrichment of native communesins with co-chromatographing 3-fluorocommunesins generated by directed biosynthesis made accurate biological assessment impossible, a problem also encountered in the generation of some squalestatin analogues via directed biosynthesis (Cannell et al., 1993).

566

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

Fig. 6. Interpretation of EI mass spectrum fragmentation patterns of mono-fluorinated 1 and 2.

The incorporation of 6-fluoro-tryptamine into communesins provides further evidence that tryptamine is a biosynthetic precursor.

Mass spectra of communesins biosynthesised in the presence of DL-5-bromo-tryptophan, 5-chloro-tryptamine and DL-5-methyl-tryptophan only contained signals associated

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

with native communesins (Wigley, 1995), suggesting that these putative precursor analogues can not be incorporated into communesins. The preferential incorporation of fluorine reported in the directed biosynthesis of squalestatin analogues, was attributed to the almost isosteric nature of hydrogen and fluorine atoms (Cannell et al., 1993) and possibly accounts for the same phenomenon observed during communesin analogue generation.

3. Experimental 3.1. Fungal isolate, media and fermentation conditions Sporulating communesin-producing cultures of Penicillium sp. isolate N934-53 (from soil, Miyazaki, Japan; Pfizer culture collection) were stored at
20 °C following growth at 25 °C for 7–10 days on Bacto potato dextrose agar (PDA) slopes. Spores were transferred, as an opaque green suspension in a 0.01% Tween 80 solution, 2 ml, into baffled Erlenmeyer flasks (500 ml) containing seed-stage medium, (g/l: Corn starch 20, Pharmamedia 15, yeast extract 5, CaCO3 2, pH 7.0) 100 ml. Flasks were incubated at 27 °C on a rotary shaker (200 rpm, 10 cm eccentric throw) for approximately 36 h, giving a dense mycelial suspension, when 4% (v/v) transfer was made to production media, (ARM; g/l: Soluble starch 7.5, Trusoy flour 7.5, sodium chloride 2, yeast extract 0.05, Pharmamedia 2, pH 6.5), 100 ml in 500 ml un-baffled Erlenmeyer flasks, and incubated as above for between 96 and 144 h. 3.2. Isotopically-labelled-precursor feeding experiments 3.2.1. Radiolabelled putative precursor addition 14 L-[side chain (methylene)-3- C] tryptophan (aqueous solution containing 2% ethanol): specific activity, 53.80 mCi mmol
1 (New England Nuclear), 1 lCi; [1-14C] acetic acid, sodium salt (aqueous solution): specific activity, 60 mCi mmol
1 (Amersham International), 1 lCi; RS[2-14C] mevalonic acid lactone (benzene solution): specific activity 50–60 mCi mmol
1 (Amersham International), 3 DL-[2- H] mevalonic acid lactone (toluene solution): specific activity 1 Ci mmol
1, 8.5 lCi; L-[methyl-14C]met (freezedried, under nitrogen): specific activity, 50–60 mCi mmol
1 (Amersham International), 10 lCi; L-14C tryptamine, 0.14 lCi (see below for preparation): specific activity, ca. 7.2 lCi mmol
1. All compounds in sterile distilled water, 1 ml, were delivered to separate cultures of Penicillium sp. N934-53 in four equal sub-doses at 48, 64, 68 and 72 h. The tryptamine and tryptophan-fed cultures were harvested at 80 and 144 h respectively; all others were harvested at 96 h. 3.2.2. Preparation of 14C-tryptamine by decarboxylation of 14C L-tryptophan by L-phenylalanine decarboxylase Decarboxylation of trp was achieved using L-phenylalanine decarboxylase (EC 4.1.1.53) from Streptococcus fae-

567

calis (Sigma). Dry cells 1 mg, were added to DLtryptophan, 0.8 mg plus 14C-L tryptophan, 5 lCi, in 50 mM Tris–HCl, containing 0.1 mM pyridoxal phosphate, 1 ml (Taylor and Wightman, 1987), at pH 5.5, incubated at 37 °C for 48 h. Cells were removed by centrifugation and tryptamine extracted from the supernatant, basidified to pH > 12 with NaOH, by partition into diethyl ether and purified by thin layer chromatography (as in 3.2.X), mobile phase: methyl acetate: propan-2-ol: 25% ammonium hydroxide, 9:3:1. 14C-tryptamine: specific activity (7.2 lCi mmol
1), 0.36 lCi total, was isolated. 3.2.3. Addition of differentially labelled precursors at various stages in the submerged culture fermentation (a) Delivery of a mixture of 14C-tryptophan and 3HmethionineA solution of 14C-tryptophan and L[methyl-3H] methionine (aqueous solution containing 0.2% 2-mercaptoethanol, sterilised): specific activity, 70–85 Ci mmol
1 (Amersham International); in the ratio of 1:8 counts per minute (cpm), was delivered to production stage cultures at either time 0, 8, 24, 32, 48, 72 or 96 h. All cultures were harvested at 168 h. (b) Delivery of a mixture of 14C-tryptophan and RS(2-3H)mevalonic acidA solution of 14C-tryptophan and DL-[2-3H]mevalonic acid lactone (toluene solution): specific activity 1 Ci mmol
1, in the ratio of 1:30 (cpm), was delivered to production stage cultures at either time 0, 12, 24, 48, 72, 96, 120 or 144 h. All cultures were harvested at 192 h. The ratio (cpm) of 14 C: 3H in 1 and 2 (measured in the total of each compound extracted) was determined by scintillation counting (see the following section).

3.2.4. Metabolite extraction and analysis Fungal mycelium, separated under vacuum from broth by filtration (Whatman 541, 9 cm diameter) was extracted twice with methanol (25 ml per 100 ml of culture) for 12 h. Metabolites in broth (filtrate) that are soluble in organic solvent were isolated by partition extraction twice with an equal volume of ethyl acetate. Combined methanol and ethyl acetate extracts were taken to dryness under reduced pressure prior to analysis by HPLC and/or TLC. TLC employed silica gel, 0.25 mm thick on plastic backed plates, (Polygram SIL G/UV254, Camlab). In the mobile phase chloroform:acetone (6:1), 1 and 2 had Rf values 0.3 and 0.48, respectively. HPLC was performed on a Waters Z-module C-18 reverse phase ‘‘Novapak’’ (0.8 cm by 10 cm) HPLC column controlled by an Apple II GS computer using the Gilson HPLC system manager with detection at 268 nm. Communesins were purified and analysed using methanol:water 80:20, flow 0.75 ml min
1, Rt for 1 and 2 10 and 14 min, respectively. Radiolabel incorporation was determined quantitatively by scintillation counting (Kontron Intertechnique) of the

568

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

HPLC-purified communesins [sample:scintillant (Ecolumee ICN) ca. 8:1]. Counting efficiency for 14C and 3H was 90–93% and ca. 40%, respectively. Semi-quantitative/ qualitative analysis was by autoradiography of TLCresolved extracts (Fuji X-ray film NIF RX developed automatically in a Fuji FPM 2100 processor). Mass spectrometry (MS) used either a VG Micromass 7070E instrument, operated at 70 eV for electron impact (EI) or a VG Auto Spec instrument at 70 or 100 eV for EI and chemical ionisation (CI), respectively. Ammonia was the reagent gas for CI. 1 H and 13C NMR spectroscopy used a Brucker AM 500 MHz instrument. Chemical shift values were all relative to tetramethylsilane. 3.2.5. Dry cell weight Dry cell weight as a measure of fungal growth was determined by lyophilisation of filtered mycelium in pre-weighed flasks to constant weight. 3.2.6. Administration of DL-2-13C-tryptophan to submerged fermentations Preliminary experiments, measuring incorporation efficiency of 14C-tryptophan into communesins in the presence of significant additions of tryptophan (Wigley, 1995), showed less than a 3-fold reduction when 10 mg tryptophan was added to a 100 ml fermentation. With 20 mg the efficiency was correspondingly halved, and thus there was no advantage in adding more than 10 mg 13C-tryptophan for 13 C NMR enrichment. Further, maximum incorporation into communesins of a single probe-dose of tryptophan, added to a fermentation at 96 h was only reached after a further 24 h. Therefore, near-optimal administration of a 13 C-tryptophan probe was designed. DL-2-13C-tryptophan (99 atom% 13C, C/D/N Isotopes INC), 10 mg, was dissolved in distilled water, 25 ml, and filter sterilised (AcrodiscÒ, 0.2 lm). Aliquots were added to a 100 ml culture on 5 evenly-spaced occasions during the 3rd and 4th day of fermentation. The probe was therefore added between 46.5 and 75.5 h, and there were four replicate 100 ml cultures. The cultures were harvested at 78.5 h to prevent undue dilution of 13C-communesins by new unlabelled compound. 13C-tryptophan-enriched 1 and 2, 9.9 and 7.3 mg, respectively, were purified, dissolved in acetone-d6, analysed by 13C NMR and the spectra compared to the natural abundance 13C NMR spectra of 1 and 2. 3.2.7. Administration of indole-N-(methyl-13C) tryptophan to submerged fermentations Indole-N-(methyl-13C)tryptophan, 50 mg per 100 ml culture (pH 5.0) (prepared by G. Goodwin, Pfizer, Sandwich by the method of Yamada et al., 1965) was administered to ten production stage cultures, 100 ml, in eight 1 ml aqueous aliquots between 27.5 and 48 h of the fermentation. One culture flask also received 1 lCi 14C-tryptophan. Control cultures were fed distilled water (pH 5.0) or were not treated in any way. All cultures were harvested at

120 h and organic solvent extracts were analysed by HPLC and TLC. 3.2.8. Biosynthesis of tryptophan analogues Escherichia coli, W 3110 trpAEdel2, containing the plasmid pHP3 encoding tryptophan synthetase, can convert Lserine and indole to L-tryptophan (Matthews et al., 1992). Provision also of plasmid encoded marker genes for ampicillin and tetracycline resistance facilitates selection and maintenance of bacteria containing plasmids (Enger-Valk et al., 1980). The following protocol for bacterial growth and the incubation procedure for tryptophan synthesis was modified from Matthews et al. (1992). The E. coli strain was grown and maintained in medium 2TY (Maniatis et al., 1989); Bacto tryptone 16 g/l, Bacto yeast extract 10 g/l, NaCl 5 g/l. pH was adjusted to 7.0 with 5 M NaOH. Filter-sterilised ampicillin (Acrodisc, 0.2 lm), 50 lg/ml, was added to the sterilised medium. For solid medium Bacto agar, 15 g/l, was added. Liquid cultures were grown for 15 h at 37 °C. Sterile glycerol was added (10%) and 1 ml aliquots dispensed into sterile eppendorff tubes. Tubes were cooled slowly to
20 °C before storage in liquid nitrogen. For fermentations, the contents of one thawed tube was used to inoculate 200 ml of 2TY/ampicillin medium and incubated on a rotary shaker for 15 h at 37 °C. Cells were harvested by centrifugation and the pellet re-suspended in 100 mM potassium phosphate buffer, pH 7.8, supplemented with pyridoxal phosphate, 10 mg/l, (NH4)2SO4, 15.6 g/l, Na2SO3 Æ 7H2O, 1.25 g/l. To cells from ca. 50 ml culture, suspended in the supplemented buffer, serine (151 mg) and 4-bromo-indole (36 mg) were added. Since the bromo-indole is only slightly soluble in water it was trickle-fed
in 18 drops of DMSO during the first 3 h of incubation. Incubation was for 6 h at 37 °C and, after centrifugation, the supernatant was treated with 5 volumes of ethanol (96%) to precipitate salts. Evaporation of the supernatant left a residue rich in tryptophan which was purified by PLC on 1 mm silica gel plates (Polygram SIL G/UV254, Camlab) in n-butanol:acetic acid:water (12:3:5, Rf values for tryptophan and analogues in the range 0.7–0.8): yield 66%. 4-Bromotryptophan structure was confirmed (NMR, UV and EIMS) by Prof C. Moody (Loughborough University, UK). Similar protocol on a smaller scale was used to make radiolabelled tryptophan analogues from 4-bromo-indole, 4-fluoro-indole, 5-bromo-indole, 6-bromo-indole and appropriate L-serine by adding 5–10 lCi 14C-serine. Products, purified by TLC, were added (total 0.1–0.4 lCi according to availability) to Penicillium fermentations in aliquots at 52.5, 53, 69.5, 70.25, and 72 h and the cultures harvested at 96 h. Communesins were isolated, separated by TLC and the chromatograms autoradiographed (Fuji NIF RX). 3.2.9. Preparation of halogenated and methylated communesins by directed biosynthesis DL-6-Fluorotryptophan (Aldrich) 11.9 mg, DL-5-bromotryptophan (Sigma) 11.9 mg, 6-fluorotryptamine (Sigma)

L.J. Wigley et al. / Phytochemistry 67 (2006) 561–569

10 mg, 6-chlorotryptamine (Sigma) 10.6 mg and DL-5methyltryptophan (Aldrich) 10 mg, were trickle-fed
, each in 25 ml sterile distilled water, to separate culture flasks of production medium, 100 ml, frequently between 40 and 69 h of the fermentation, which continued until 96 h. The ethyl acetate broth extracts and methanol cell extracts were purified by TLC, and compounds co-chromatographing with 1, 2 and other selected bands were eluted from the silica and investigated by CIMS. Accurate mass measurements were obtained for any important ions in the spectra that did not correspond to either 1 or 2. 3.2.10. Probing for un-methylated communesins with 14Ctryptophan following ethionine addition to fermentations DL-Ethionine (Sigma), 50 mg in 10 ml sterile distilled water, was added to a 100 ml fermentation at 38 h. 10 ml water was added to a control fermentation. 30 min later 14 C-tryptophan (1 lCi) in 1 ml sterile water was added to each culture. Fermentations were harvested at 72 h and the broth extracted with ethyl acetate. Extract was chromatographed and the TLC plate autoradiographed for one month.

Acknowledgements We thank R. Shepherd and J. Barton (Chemistry Department, Imperial College) for obtaining NMR and MS data, G. Goodwin (Pfizer, UK) for synthesis of indole-N-(13Cmethyl) tryptophan, and Prof. C. Moody, Loughborough University, UK, for confirming our tryptophan synthesis. We thank BBSRC for a research studentship (LJW) in collaboration with Pfizer Central Research (Sandwich, Kent, UK) and acknowledge the advice of Dr. M. Haxell (Pfizer, UK) concerning communesin fermentations.

References Adjibade, Y., Weniger, B., Quirion, J.C., Kuballa, B., Cabalion, P., Anton, R., 1992. Dimeric alkaloids from Psychotria fosteriana. Phytochemistry 31, 317–319. Aerts, R.J., Alarco, A.-M., De Luca, V., 1992. Auxins induce tryptophan decarboxylase activity in radicles of Catharanthus seedlings. Plant Physiology 100, 1014–1019. Beck, J., Ripka, S., Siegner, A., Schiltz, E., Schweizer, E., 1990. The multifunctional 6-methylsalicylic acid synthase gene of Penicillium patulum. Its gene structure relative to that of other polyketide synthases. European Journal of Biochemistry 192, 487–498.

569

Cannell, R.J.P., Dawson, M.J., Hale, R.S., Hall, R.M., Nobel, D., Lynn, S., Taylor, N.L., 1993. The squalestatins, novel inhibitors of squaline synthase produced by a species of Phoma 1V. Preparation of fluorinated squalestatins by directed biosynthesis. Journal of Antibiotics 46, 1381–1389. Enger-Valk, B.E., Heyneker, H.L., Oosterbaan, R.A., Pouwells, P.H., 1980. Construction of new cloning vehicles with genes of the tryptophan operon of Escherichia coli as genetic markers. Gene 9, 69–85. Goddijn, O.J.M., De Kam, R.J., Zanetti, A., Schilperoort, R.A., Hoge, J.H.C., 1992. Auxin rapidly down-regulates transcription of the tryptophan decarboxylase gene from Cantharanthus roseus. Plant Molecular Biology 18, 1113–1120. Gutierrez, S., Diez, B., Montenegro, E., Martin, J.F., 1991. Characterisation of the Cephalosporium acremonium pcbAB gene encoding aaminoadipyl-cysteinyl-valine synthetase, a large multidomain peptide synthetase: Linkage to the pcbC gene as a cluster of early cephalosporin biosynthetic genes and evidence of multiple functional domains. Journal of Bacteriology 173, 2354–2365. Henrickson, J.B., Goschke, R., Rees, R., 1964. Total synthesis of the calycanthaceous alkaloids. Tetrahedron 20, 565–579. Kirby, G.W., Shah, S.W., Herbert, E.J., 1969. Biosynthesis of chimonanthine from [2-3H] tryptophan and [2-3H] tryptamine. Journal of the Chemical Society (C), 1916–1919. Kleinkauf, H., von Dohren, H., 1990. Nonribosomal biosynthesis of peptide antibiotics. European Journal of Biochemistry 192, 1–15. Maniatis, T., Fritsch, E.L., Sambrook, J., 1989. Molecular cloning, a laboratory manual. Cold Spring Harbor, New York. Mantle, P.G., Shipston, N.F., 1987. Temporal separation of steps in the biosynthesis of verruculogen. Biochemistry International 14, 1115– 1120. Matthews, S., Jandu, S.K., Leatherbarrow, R.J., 1992. 13C NMR study of the effects of mutation on the tryptophan dynamics in chymotrypsin inhibitor 2: correlations with structure and stability. Biochemical Journal 32, 657–662. Noe, W., Mollenschott, C., Berlin, J., 1984. Tryptophan decarboxylase from Catharanthus roseus cell suspension cultures: purification, molecular and kinetic data of the homogeneous protein. Plant Molecular Biology 3, 281–288. Numata, A., Takahashi, C., Ito, Y., Takada, T., Kawai, K., Usami, Y., Matsumura, E., Imachi, M., Ito, T., Hasegawa, T., 1993. Communesins, cytotoxic metabolites of a fungus isolated from a marine alga. Tetrahedron Letters 34, 2355–2358. Robinson, Sir R., Teuber, H.J., 1954. Calycanthine and calycanthidine. Chemistry and Industry, 783–784. Songstad, D.D., De Luca, V., Brisson, N., Kurz, W.G.W., Nessler, C.L., 1990. High levels of tryptamine accumulation in transgenic tobacco expressing tryptophan decarboxylase. Plant Physiology 94, 1410– 1413. Taylor, D.C., Wightman, F., 1987. Metabolism of D,L-chloro-phenylalanines by phenylalanine aminotransferase isozymes purified from bushbean shoots. Phytochemistry 26, 1279–1288. Wigley, L.J. 1995. Biosynthesis of communesin alkaloids. Ph D thesis, University of London. Yamada, S., Shioiri, T., Itaya, T., Hara, T., Matsueda, P., 1965. Ind.-Nalkylation of tryptophan and synthesis of 1-alkyltryptophan hydrazides. Chemical and Pharmaceutical Bulletin 13, 88–93.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 570–578 www.elsevier.com/locate/phytochem

In vitro shoot and root organogenesis, plant regeneration and production of tropane alkaloids in some species of Schizanthus Miguel Jordan

b

a,*

, Munir Humam b, Stefan Bieri b, Philippe Christen b, Estrella Poblete a, Orlando Mun˜oz c

a Pontificia Universidad Cato´lica de Chile, Facultad de Ciencias Biolo´gicas, Departamento de Ecologı´a, Alameda 340, Santiago, Chile University of Geneva, Laboratory of Pharmaceutical Analytical Chemistry, School of Pharmacy EPGL 20, bd d’Yvoy, Geneva 4, Switzerland c Universidad de Chile, Facultad de Ciencias, Departamento de Quı´mica, Casilla 653, Santiago, Chile

Received 22 July 2005; received in revised form 16 November 2005 Available online 20 January 2006

Abstract A rapid in vitro propagation system leading to formation of shoots from callus, roots, and plantlets was developed for Schizanthus hookeri Gill. (Solanaceae), an endemic Chilean plant. The genus Schizanthus is of particular interest due to the presence of several tropane alkaloids. So far, in vitro propagation of species of this genus has not been reported. Propagation of S. hookeri consisted of two phases, the first one for callus initiation and shoot formation and the second for rhizogenesis and plantlet regeneration. From a single callus that rapidly increased in cell biomass (from 50 mg to 460 mg/culture tube [25 · 130 mm] in 60 days) in the presence of 2.69 lM NAA and 2.22 lM BA, more than 10 shoots/callus explant were formed. From the latter, approx. twenty plantlets formed after 90–110 days shoot subculture in medium devoid of growth regulators that favored root formation. Ten alkaloids ranging from simple pyrrolidine derivatives to tropane esters derived from angelic, tiglic, senecioic or methylmesaconic acids were obtained from in vitro regenerated plantlets. One of them, 3a-methylmesaconyloxytropane was not previously described. The same growth conditions, as well as other growth regulator levels tested, were required to induce callus and root formation in S. grahamii Gill. Root organogenesis occurred despite a high level of BA vs. NAA used, (i.e., 4.44 lM BA and 0.54 lM NAA); however no shoot formation was achieved. In the case of S. tricolor Grau et Gronbach, only callus formation was obtained in the presence of various growth regulators.
2005 Elsevier Ltd. All rights reserved. Keywords: Schizanthus hookeri; S. grahamii; S. tricolor; Solanaceae; Organogenesis; In vitro propagation; Tropane alkaloid content

1. Introduction The genus Schizanthus (Solanaceae), which belongs to the tribe Salpiglossideae, includes 12 species (Grau and Gronbach, 1984; Marticorena, 1990; Grau, 1992) endemic to the south western slopes of the Chilean Andes. Previous chemical studies on Schizanthus have shown that this genus accumulates a number of tropane-derived alkaloids such as hydroxytropane esters, dimeric tropane diesters, cyclobuAbbreviations: BA, benzyladenine; IBA, indole-3-butyric acid; NAA, naphthaleneacetic acid; MS, Murashige and Skoog medium. * Corresponding author. Tel.: +56 2 6862637; fax: +56 2 6862621. E-mail address: [email protected] (M. Jordan). 0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.007

tane tricarboxylic acid triesters and pyrrolidine alkaloids (San-Martı´n et al., 1980, 1987; Gambaro et al., 1982, 1983; De la Fuente et al., 1988; Mun˜oz et al., 1991, 1994; Mun˜oz, 1992; Mun˜oz and Corte´s, 1998; Griffin and Lin, 2000). The well known effects of tropane alkaloids as anticholinergic, antiemetic, parasympatholytic and anesthesic agents (Hashimoto and Yamada, 1992; Fodor and Dharanipragada, 1994) have stimulated considerable interest during the last decades on the biosynthetic pathway leading to tropane alkaloids (Lounasmaa and Tamminen, 1993). Schizanthus hookeri, S. grahamii and S. tricolor (formerly S. litoralis) are three Andean species that accumulate several tropane bases, in particular schizanthines (SanMartı´n et al., 1987; De la Fuente et al., 1988; Mun˜oz

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

et al., 1991; Mun˜oz and Corte´s, 1998), and grahamine isolated from S. grahamii (Hartmann et al., 1990). Many efforts have been made to develop economically feasible methods for the production of tropane alkaloids by applying cell culture techniques (Za´rate et al., 1997; Christen, 2000; Khanam et al., 2001). The main efforts toward the industrial preparation of tropane alkaloids by cell culture methods have concentrated on hyoscyamine and scopolamine (Lounasmaa and Tamminen, 1993; Oksman-Caldentey and Hiltunen, 1996; Oksman-Caldentey and Arroo, 2000). The biosynthesis of tropane alkaloids has been extensively studied over the last decades (Robins et al., 1994). In particular, it has been demonstrated that the site of biosynthesis is the root, the alkaloids being translocated from the roots to the aerial parts of the plants (Hashimoto et al., 1991). Despite considerable efforts to produce secondary metabolites by undifferentiated plant cell cultures, it has become increasingly apparent that having a degree of morphological organization present greatly enhances the likelihood of successful product formation in vitro (Robins and Walton, 1993). The study of alkaloids in Schizanthus is fairly recent and in vitro morphogenic aspects and regeneration of these species from cell cultures have not been reported so far. Furthermore, there is no information on the secondary metabolite production of in vitro cultures in these species and its variation according to explant-types and growth regulators. The aim of this work was to evaluate the in vitro morphogenic potential of three Schizanthus species and the regeneration of selected plants. A second aim of the work was to evaluate the alkaloid patterns produced by the

571

various tissues, organs and plantlets of Schizanthus developed under in vitro conditions.

2. Results and discussion 2.1. In vitro propagation Callus initiation was the first and generalized response observed in internodal explants of all the species examined. Intense cell proliferation on the surface at the cut edges of the sections, leading to calli, was observed with high frequency after a period of 7–10 days in culture. The mass of the calli increased rapidly and they gradually turned green with the exception of S. tricolor whose calli turned brown after a period of 2–3 weeks. Shoot organogenesis was first observed after 2 months, occurring profusely in the calli of S. hookeri in the initiation medium supplemented with 2.69 lM NAA and 2.22 lM BA (Table 1) leading to several plantlets in a single culture tube. After the first subculture, multiple new shoots formed, developing roots and plantlets after 4–5 weeks in the same MS medium but devoid of sucrose and growth regulators. However, the same conditions, including treatment with 2.69 lM NAA and 2.22 lM BA did not trigger shoot formation in S. grahamii leading only to intensive callus growth. Although the calli developed remarkably dense green spots on the surface, these outgrowths never developed into shoots. However, in subculture the calli formed roots (more than 5/explant) in the presence of 0.54 lM NAA and 4.44 lM BA. Less intensive callus growth was observed with S. tricolor. In comparison with the other

Table 1 Morphogenic responses observed in 3 Schizanthus species after 2 months in culture and subculturea Species

Growth regulators NAA BA (lM)

Callus initiation (%) after 30 daysb

S. hookeri

2.69 2.22

100

460.2

5.37 2.22

100

365.3

2.69 2.22

100

5.37 2.22

S. grahamii

S. tricolor

a b

Callus FW (mg) after 60 days (aver.)

Morphogenic responses after 60 days

Morphogenic responses after 1st subculture (results after 30–50 days)

Green callus forming new shoots (approx. 10 shoots/ callus). Plantlet regeneration Green callus with roots only

Shoot elongation, root formation and regeneration of plantlets in medium devoid of growth regulators and sucrose Intense callus growth in presence of 26.85 lM NAA and 0.44 lM BA

1257.0

Dense green spots on basal callus tissue, no shoots or roots

100

594.0

Green spots on basal callus, no shoots or roots

Root formation and callus growth in presence of 0.54 lM NAA and 4.44 lM BA. Green callus; no shoots Root formation and callus growth in presence of 0.54 lM NAA and 4.44 lM BA. No shoots

2.69 2.22

75

175.0

5.37 2.22

80

160.3

No morphogenic responses; little callus growth; brown callus No morphogenic responses, little callus growth; brown callus

All treatments started with 20 tubes per culture using internodal sections. Internodal sections initiating callus.

Brown callus in presence of 26.85 lM NAA and 0.44 lM BA Brown callus in presence of 0.54 lM NAA and 4.44 lM BA.

572

Table 2 Characteristics of tropane alkaloids identified in in vitro plantlets and calli of Schizanthus hookeri, S. tricolor and S. grahamii Schizanthus species S. hookeri Callus

Alkaloid

Rt (min)

[M+] m/z

Peak area % (average values)

Tigloyloxy- or angeloyloxy- or seneciolyoxy- and methylmesaconyloxy or methylitaconyloxytropane

36.09

365

–

Reference mass spectrum

Structure

O ,

or

N tigloyl

angeloyl

O

O

R' =

O

methylmesaconyl

141

4.16

+

N O Hygroline A (or B) Hygroline B (or A)

7.66 8.00

143 143

2.46 0.97

+ +

N * OH

3a-Senecioyloxy-7bhydroxytropane

25.57

239

0.95

+

N HO O O

3a-Hydroxy-7btigloyloxytropane

26.13

239

0.40

+

N O O

Line missing

OH

O

O

O

7.43

or

methylitaconyl

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

O R

Hygrine

senecioyl

O

R'

Regenerated plantlets

O

O

R=

New alkaloid 3amethymesaconyloxytropane

27.1

267

1.41

+

N O O

O O

35.7

Cuscohygrine orNmethylpyrrolidinylhygrine

25.9

36.12

321

365

2.93

N R'

1.16

O

O R

or

tigloyl

365

0.97

senecioyl

angeloyl

O

37.00

O

O ,

R, R' =

O

O

O

R=

,

or

N tigloyl

O

R'

37.23

365

angeloyl

senecioyl

O

O

1.86

O R

R' =

or

O

O

O

O methylmesaconyl

224

–

methylitaconyl

+

N N

N

N

O

O

Cuscohygrine

S. grahamii Callus

Acetoxy and tigloyloxy or angeloyloxy or senecioyloxytropane

30.82

281

N-methylpyrrolidinylhygrine

–

O

N R'

R or R' =

O

tigloyl

O

O ,

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

S. tricolor Callus

3,7-Disubstituted tropane alkaloid with tiglic or angelic or senecioic acids Tigloyloxy- or angeloyloxy- or seneciolyoxy- and methylmesaconyloxy- or methylitaconyloxytropane Tigloyloxy- or angeloyloxy- or seneciolyoxy- and methylmesaconyloxy- or methylitaconyloxytropane Tigloyloxy- or angeloyloxy- or seneciolyoxy- and methylmesaconyloxy- or methylitaconyloxytropane

or angeloyl

senecioyl

O R R or R' =

O acetoxy

573

574

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

species the calli of S. tricolor, showed less chloroplast formation turning brown within a month. Also, contrary to the other species, the calli of S. tricolor were unable to recover in subculture under various levels of growth regulator combinations tested (data not shown). Growth and morphogenic responses of these three species are summarized in Table 1.

N

N

1

7 6

4

O

3

O 1

O O

O O

O 2

O

Fig. 1. Structures of the two isomers of 267 Da, 3a-methylmesacoyloxytropane (1) and 3a-methylitaconyloxytropane (2).

2.2. Alkaloid determination Preliminary determination of tropane alkaloids in tissues developed in vitro by weighing indicated their presence in callus tissue as well as in the regenerated plantlets. While the level detected in calli of the three species under study was similar (i.e., 0.9, 1.1, and 1.6 mg g 1 dry weight), this content increased fivefold in regenerated plantlets of S. hookeri (5.6 mg g 1 dry weight). 2.3. Alkaloid identification Different alkaloids were detected in the various plant parts studied. Only one alkaloid was detected in the calli of S. hookeri (Table 2). The [M]+ at 365 thomsons (Th), together with prominent ions at 238, 222, 138, 122 and 94 Th (base peak) suggested a 3,6- or 3,7-disubstituted tropane nucleus of molecular formula C19H27NO6 esterified with C5H8O2 (tiglic, senecioic or angelic acid) and C6H8O4 (methylmesaconic or methylitaconic acid) moieties. However, in the absence of reference compound, its unambiguous identification was not possible and only tentative assignments were made. From the in vitro regenerated plantlets of S. hookeri, ten alkaloids were detected by GC–MS (Table 2) including hygrine and hygrolines A and B. The latter compounds are not discussed in more details as they are not tropane alkaloids and are frequently encountered in solanaceous plants. According to the fragmentation patterns, seven alkaloids proved to belong to the tropane series. Two isomers of 239 Daltons (Da) were identified by comparison of their retention indices (Bieri et al., 2006). The first one was identified as 3a-senecioyloxy-7b-hydroxytropane (I = 1866.2) and the second one as 3a-hydroxy-7b-tigloyloxytropane (I = 1894.0). Another alkaloid of 267 Da was detected (I = 1944.8). This compound has not been previously identified and is a new alkaloid. Its characterization was a difficult task due to the very small amount present in the extract. According to its fragmentation determined by electron impact MS, two possible structures were suggested: 3a-methylmesaconyloxytropane (1) or 3a-methylitaconyloxytropane (2). These two compounds are configurational isomers (Fig. 1). In order to assign unambiguously a structure to this new alkaloid, it was necessary to synthesize both possible isomers. Fig. 2(a) and (b) show the mass spectra of the synthetic 3a-methylmesaconyloxytropane (I = 1942.4) and 3a-methylitaconyloxytropane (I = 1893.0), respectively. Fig. 2(c) shows the mass spectrum of the compound at 267 Th

detected in the regenerated plant extract which matched isomer (1). This compound was also detected in the stem–bark extract of the field-grown plants. In the same extract, other alkaloids were detected but could not be identified unambiguously and among them, an isomer of 321 Da (Table 2). The fragmentation pattern showed that it was a 3,6- or 3,7-disubstituted tropane alkaloid with tigloyl, senecioyl or angeloyl moiety. Three other isomers of 365 Da were also detected. Fragmentation patterns indicated that they were 3,6 - or 3,7-tropane diesters with tigloyl, senecioyl or angeloyl and methylmesaconyl or methylitaconyl moieties. However, for all of them, lack of reference material precluded their unambiguous identification. In order to compare the alkaloid pattern of the regenerated plantlets with that of the normal plants, GC– MS analysis of the stem–bark extract of field grown S. hookeri led to the identification of the following tropane alkaloids: hygrine; hygrolines A and B; tropine, tropinone, 3,7-tropane-diol, 3a-senecioyloxytropane, 3a-senecioyloxy7b-hydroxytropane, 3a-hydroxy-7b-angeloyloxytropane, 3a-hydroxy-7b-tigloyloyloxytropane, 3a-hydroxy-7b-senecioyloxytropane, 3a-methylmesaconyloxytropane and 6 isomers at 365 Da. In the absence of reference compounds, the identification of these isomers was not possible. The presence of the 3a-senecioyloxytropane derivative in the stem–bark of S. hookeri was confirmed thanks to the linear retention index calculated for the analyte detected in the extract (I = 1661.2) and from a reference compound (I = 1661.4). Previously, (San-Martı´n et al., 1980; Gambaro et al., 1982, 1983) the following alkaloids in S. hookeri were identified: The diastereoisomeric hygrolines, 3a-hydroxytropane (tropine), 3a-senecioyloxytropane, 3a,7b-dihydroxytropane, 3a-hydroxy-7b-tigloyloxytropane, 3a-hydroxy-7b-angeloyloxytropane and 3a-senecioyloxy7b-hydroxytropane. In the calli of S. tricolor, one alkaloid only, cuscohygrine or N-methylpyrrolidinylhygrine (224 Da) was detected. Similarly, in the green calli of S. grahamii (Table 2), one alkaloid only was identified (281 Da). Its fragmentation pattern showed that it is a 3,6- or 3,7-disubstituted tropane alkaloid with acetic and tiglic, senecioic or angelic substituents as esterifying acids. Previous chemical work on S. grahamii and S. tricolor aerial parts have shown that both species accumulate tropane alkaloids, i.e hydroxytropane esters, dimeric tropane diesters of methylmesaconic and methylitaconic acids, cyclobutane tricarboxylic triesters and ferulamides (San-Martı´n et al., 1987; De la Fuente

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

575

124

Abundance %

a

90 80 70

83

96

60 50 42

40

67

30 20 10

140

55

108

30

164 180 194

208

236

267

0 40

m/z-->

60

80

100

120

140

160

180

200

220

240

260

124

Abundance %

b

90 80 70 60 50 40

82

30

94

42

67

20 10

55

30

140

267 67

208

108

0 m/z-->

40

60

80

100

120

Abundance % 90

140

160

180

200

220

240

260

124

c

80 83

70

96

60 50 42

40

67

30 20 10

140 55

30

108

164 180

208

236

267

0 m/z-->

40

60

80

100

120

140

160

180

200

220

240

260

Fig. 2. MS spectra of: (a) 3a-methylmesaconyloxytropane (1); (b) 3a-methylitaconyloxytropane (2); (c) alkaloid of 267 Th present in the regenerated plantlet extract of S. hookeri.

et al., 1988; Hartmann et al., 1990; Bieri et al., 2006; Mun˜oz et al., 1996). 2.4. Synthesis of 3a-methylmesaconyloxytropane (1) This alkaloid of 267 g/mol (1) was obtained via a threestep synthesis (Fig. 3). First, 2-methylmesaconic acid (4) was obtained by Wittig reaction as the condensation product between the commercially available methyl 2-bromopropionate (3) and glyoxylic acid monohydrate

(Zumbrunn et al., 1985; Wolff et al., 2002). The synthetic strategy was to perform the selective formation of the methyl group on C-2 and carboxylic functional group on C-4 with E double bond configuration. The identity of the product was confirmed by HMBC NMR experiment. Methylmesaconic acid (4) was then converted to methylmesaconic acid chloride (5) by treatment with thionyl chloride. Finally, the 3a-methylmesaconyloxytropane (1) was obtained by the addition of 3a-tropine to methylmesaconic acid chloride (5).

576

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

(1st subculture). In the case of S. hookeri calli, which produced shoots, roots and plantlets directly in the first culture phase, further subculturing was carried out in MS medium devoid of growth regulators and sucrose. Later on, plantlets were transferred to non-sterile conditions using a substrate composed of equal volumes of sand/vermiculite/ organic soil in growth chambers for 1-month acclimatization followed by transfer to the greenhouse.

Br i MeO2C

CO2Me 3

CO2H

2 3

1

iii

4

4 N

O Cl

O

O

iii

O 5

3.2. Chemicals

O 11

O O

Fig. 3. Synthesis of 3a-methylmesaconyloxytropane (1). Reagent and conditions: (i) (1) PPh3, dry CH3CN, N2, 65 C, overnight, (2) glyoxylic acid monohydrate, N-diisopropylethylamine, dry CH3CN, N2, 0 C 2 h, room temp. overnight; (ii) Thionyl chloride, Ar, reflux; (iii) 3a-Tropine, Ar, 105 C, 4 h, room temp. overnight.

Ethyl acetate, diethyl ether, methanol, dichloromethane, chloroform, hexane, NaHCO3, Na2SO4, MgSO4, HCL, ammonia and thionyl chloride were purchased from Fluka (Buchs, Switzerland). Acetonitrile was from SDS (Peypin, France). 3.3. Extraction of tropane alkaloids

3. Experimental 3.1. Plant material Schizanthus plants growing in the wild were collected by O. Mun˜oz from November 2002 to January 2003 in the following sites: S. hookeri, Lagunillas at 2100 m altitude (SQF 22231), S. grahamii, Rengo, Laguna Los Cristales at 2200 m (SQF 0015), S. tricolor, Papudo at 10 m (SQF 0017); voucher specimens are kept in the Herbario de la Escuela de Quı´mica y Farmacia, Santiago. All samples were identified by Dr. F. Pe´rez, with the only exception of S. grahamii (J. San-Martı´n, Universidad de Talca). Internodes from plants collected in the field were washed and their surface disinfected with the fungicides Captan (Micro Flo Company LLC, Memphis, USA) and Benlate (Du Pont and Co. Inc. Nemours, France) (0.2% each) for 30 min under constant shaking, washed with sterile distilled H2O and sterilized with 3% NaOCl for 5 min. Internodal sections of approx. 1 cm (50 mg), were placed in tubes (25 · 130 mm) containing 12.5 ml MS liquid nutrient medium (Murashige and Skoog, 1962) supplemented with 3% sucrose on Whatman Nr 1 paper bridges in the presence of NAA at 2.69 and 5.37 lM combined with 2.22 lM BA, according to the best responses obtained in preliminary work (data not shown). Various growth regulator levels were used in subculture (0.54 lM NAA combined with 4.44 lM BA, or 0.49 lM IBA); supplemented with agar, but with poorer results. All plant tissue culture media, salts, vitamins, growth regulators and agar were obtained from Sigma (St. Louis, USA). Explants were maintained in a light regime of 14 h at 48 lmol m 2 s 1 provided by daylight fluorescent lamps (Philips TLT 40W/54 R.S.) and a temperature of 22 ± 1 C. Calli of S. grahamii and S. tricolor reaching approx. 2.0 cm in diameter after 2 months were transferred to the same MS medium in the presence of various combinations of NAA and BA to induce shoot and/or root formation

Tropane alkaloid extraction was performed as previously described (Kamada et al., 1986; Za´rate et al., 1997). Material derived from in vitro culture was lyophilized for 24 h, homogenized and extracted with a mixture of CHCl3:MeOH:NH3 15:15:1. Ten ml per 100 mg of sample was sonicated for 10 min and left at room temperature for 1 h. After filtration, the plant material was washed with CHCl3 (2 · 1 ml) and the pooled filtrate was evaporated to dryness. Thereafter, CHCl3 (5 ml) and 1 N H2SO4 (2 ml) were added to the residue and mixed thoroughly for alkaloid extraction. The organic phase was discarded and the aqueous phase was made basic (pH 10) with conc. ammonia. Alkaloids were extracted with CHCl3 (1 · 2; 2 · 1 ml). The combined organic phases were filtered after addition of anhydrous Na2SO4, filtered and evaporated to dryness (alkaloid crude extract). The extracts were taken up in MeOH (1 ml) (HPLC grade, Merck, Darmstadt, Germany), filtered through a 0.45 lm filter membrane (Lida Manufacturing Corp.) and kept at 20 C until required for analysis. 3.4. Gas chromatography–mass spectrometry GC–MS analyses were carried out using a Hewlett– Packard chromatograph 5890 series II coupled to a HP 5972 mass-selective detector (Agilent Technologies, Palo Alto, USA). The MS detector was used in the electron impact ionization (EI) mode with an ionization voltage of 70 eV. Mass spectra were recorded in the range 30–600 Th at 1.3 scan/s and the MS transfer line was set at 280 C. The capillary column (HP5-MS 30 m · 0.25 mm i.d., 0.25 lm film thickness) was used with He as carrier gas under the following conditions: an initial oven temperature of 70 C was maintained for 1 min then linearly increased at 5 C/min to a final temperature of 285 C, and held at this temperature for 15 min. A sample volume of 1 ll was injected in the splitless mode into a laminar liner at 250 C using a fast HP 6890 series autosampler.

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

Retention indices were determined in the split mode (15:1) and calculated using a linear gradient temperature: 70 C (0 min) linearly increased at 5 C/min to a final temperature of 285 C and held at this temperature for 15 min. 3.5. NMR 1

H and 13C NMR, COSY, NOESY, HMBC and HSQC spectra were recorded in CDCl3 using a 500 MHz Bruker DRX instrument (Bruker, Du¨bendorf, Switzerland) equipped with a QNP probehead. Chemical shift values (d) were reported in parts per million (ppm) related to tetramethylsilane as internal standard and coupling constants (J) are given in Herz. 3.6. HRMS High resolution mass spectra were obtained on a QStar XL TOF mass spectrometer (MDS Sciex, Concord, Ontario, Canada) by direct injection (5 ng/ml CH3CN + 0.1% HCO2H) using NanoMate 100 (Advion BioSciences, Ithaca, NY, USA). 3.7. IR IR spectra were recorded on a Perkin–Elmer FT-IR spectrometer (Boston, USA). 3.8. Synthesis and spectral data 2-Methylmesaconic acid (4): To a solution of methyl-2bromopropionate (3) (4.183 g, 24.3 mmol) in dry CH3CN (60 ml) was added triphenylphosphine (5.66 g, 21.6 mmol, 0.9 eq.), in a 100 ml flask. After stirring at 65 C overnight, N-diisopropylethylamine (3.82 ml, 21.9 mmol, 0.9 eq.) and glyoxylic acid monohydrate (2.077 g, 21.9 mmol, 0.9 eq.) dissolved in dry CH3CN (10 ml) were added to the reaction mixture at 0 C. The solution was further stirred at 0 C for 2 h and at room temperature overnight. Half of the solvent was removed under reduced pressure and EtOAc (20 ml) was added. The resulting solution was washed with saturated aqueous NaHCO3 (3 · 40 ml). The combined aqueous layers were extracted with EtOAc (2 · 20 ml), made acid (pH 1–2) at 0 C with conc. HCl and extracted with EtOAc (3 · 30 ml). The combined organic layers were dried over MgSO4, filtered and evaporated to dryness, yielding a pale yellow solid (1.487 g, 43%, Rf = 0.30, ethyl acetate:hexane 7:3), which was used in the next reaction without further purification. 1H NMR (CDCl3): d 2.30 (3H, d, J = 1.45 Hz, H-5), 3.82 (3H, s, H-6), 6.79 (1H, q, J = 1.5 Hz, H-3), 11.36 (1H, s, OH); 13C NMR (CDCl3): d 14.6 (C-5), 52.8 (C-6), 126.1 (C-3), 145.9 (C-2), 167.4 (C-1), 171.3 (C-4). 3-Methylmesaconic acid chloride (5): 1.43 g (9.92 mmol) of 2-methylmesaconic acid (4) was added to a two-necked round-bottomed flask equipped with a magnetic stirrer

577

and a condenser. Thionyl chloride (3.6 ml, 49.6 mmol, 5 eq.) was then added under argon. The reaction mixture was heated to reflux (70 C) during 2 h under stirring and argon. The excess of thionyl chloride was removed by distillation. The crude product was purified by distillation at reduced pressure yielding a colorless oil weighting 1.069 g (66%). 1H NMR (CDCl3): d 2.23 (3H, d, J = 1.64 Hz, H-4), 3.81 (3H, s, H-6), 6.98 (1H, q, J = 1.55 Hz, H-2); 13C NMR (CDCl3): d 14.7 (C-4), 52.8 (C-6), 131.2 (C2), 146.5 (C-3), 164.7 (C-5), 166.1 (C-1). 3a-Methylmesaconyloxytropane (1): 0.773 g (5.47 mmol, 1 eq.) of 3a-tropine was added to 0.890 g of (5). The mixture was heated to 105 C while stirring under Ar during 4 h and then left at room temperature overnight. The reaction mixture was washed with Et2O (3 · 25 ml). The aqueous layer was made basic with NaHCO3 (pH 8– 9) and extracted with CH2Cl2 (3 · 25 ml). The organic layers were dried over Na2SO4, filtered and the solvent removed. The crude brown oil was purified by silica CC (MeOH:CHCl3:NH3 (25%) 10:90:1.5) Rf = 0.4, yielding 0.396 g (27%); IR: mmax (KBr) cm 1: 3426, 2951, 1716, 1646, 1436, 1359, 1341, 1266, 1203, 1115, 1063, 1036, 983, 808, 774; GC–EI–MS (Rt 26.90 min, IPT = 1942.4) 70 eV: m/z 267 [M]+ (7), 236 (3), 208 (4), 140 (14), 124 (100%), 108 (4), 96 (67), 94 (53), 83 (63), 82 (56), 67 (28), 59 (18), 42 (35); 1H NMR (CDCl3): d 1.72 (2H, d, J = 14.7 Hz, Heq-2, Heq-4), 1.92 (2H, m, Hexo-6, Hexo-7), 2.00 (2H, m, Hendo-6, Hendo-7), 2.16 (2H, dt, J = 15.02 Hz, J = 4.25 Hz, Hax-2, Hax-4), 2.24 (3H, s, H8), 2.26 (3H, d, J = 1.5 Hz, H-12), 3.08 (2H, s, H-1, H-5), 3.78 (3H, s, H-14), 5.04 (1H, t, J = 5.35 Hz, H-3), 6.71 (1H, q, J = 1.5 Hz, H-10); 13C NMR (CDCl3): d 14.2 (C-12), 25.6 (C-6, C-7), 36.5 (C-2, C-4), 40.4 (C-8), 52.6 (C-14), 59.6 (C-1, C-5), 67.7 (C-3), 127.2 (C-10), 143.3 (C-11), 165.2 (C-9), 167.6 (C-13); HRMS-CI m/z calculated for C14H22NO4 268.1543 [M + H]+, found 268.1554.

References Bieri, S., Mun˜oz, O., Veuthey, J.L., Christen, P., 2006. Analysis of isomeric tropane alkaloids from Schizanthus grahamii by very fast gas chromatography. J. Sep. Sci. (2006). Christen, P., 2000. Tropane alkaloids: old drugs used in modern medicine. In: Rahman, A. (Ed.), Studies in Natural Product Chemistry, vol. 22. Elsevier, Amsterdam, pp. 717–749. De la Fuente, G., Reina, M., Mun˜oz, O., San-Martı´n, A., Girault, P.J., 1988. Tropane alkaloids from Schizanthus pinnatus. Heterocycles 27, 1887–1897. Fodor, G., Dharanipragada, R., 1994. Tropane alkaloids. Nat. Prod. Rep. 12, 443–450. Gambaro, V., Labbe´, C., Castillo, M., 1982. Tropane alkaloids from Schizanthus hookeri. Bol. Soc. Chil. Quı´m. 2, 296–298. Gambaro, V., Labbe´, C., Castillo, M., 1983. Angeloyl, tigloyl and senecioyloxytropane alkaloids from Schizanthus hookeri. Phytochemistry 22, 1838–1839. Grau, J., 1992. Particularidades de la flora Chilena. In: Grau, J., Zizka, G. (Eds.), Flora Silvestre de Chile. Palmengarten, Sonderheft 19, 117–122. Grau, J., Gronbach, A., 1984. Untersuchungen zur Variabilita¨t in der Gattung Schizanthus (Solanaceae). Mitt. Bot. Mu¨nchen 20, 111–203.

578

M. Jordan et al. / Phytochemistry 67 (2006) 570–578

Griffin, W.J., Lin, G.D., 2000. Chemotaxonomy and geographical distribution of tropane alkaloids. Phytochemistry 53, 623–637. Hartmann, R., San-Martı´n, A., Mun˜oz, O., Breitmaier, E., 1990. Grahamine, an unusual tropane alkaloid from Schizanthus grahamii. Angew. Chem. Int. Ed. 29, 385–386. Hashimoto, T., Yamada, Y., 1992. In: Singh, B.K., Flores, H.E., Shannon, J.C. (Eds.), Tropane Alkaloid Biosynthesis: Regulation and Application. American Society of Plant Physiology Press, Rockville, pp. 262–274. Hashimoto, T., Hayashi, A., Amano, Y., Kohno, J., Iwanari, H., Usuda, S., Yamada, Y., 1991. Hyoscyamine 6b-hydroxylase, an enzyme involved in tropane alkaloid biosynthesis, is localized at the pericycle of the root. J. Biol. Chem. 266, 4648–4653. Kamada, H., Okamura, N., Satake, M., Harada, H., Shimomura, K., 1986. Alkaloid production by hairy root cultures in Atropa belladona. Plant Cell Rep. 5, 239–242. Khanam, N., Khoo, C., Close, R., Khan, A.G., 2001. Tropane alkaloid production by shoot culture of Duboisia myoporoides R. Br.. Phytochemistry 56, 59–65. Lounasmaa, M., Tamminen, T., 1993. The tropane alkaloids. In: Brossi, A. (Ed.), The Alkaloids, vol. 44. Academic Press, New York, pp. 1–114. Marticorena, C., 1990. Contribution to the statistics of the vascular flora of Chile. Gayana Bot. 47, 85–113. Mun˜oz, O., 1992. Solanaceae. In: Mun˜oz, O. (Ed.), Quı´mica de la Flora de Chile., DTI. Universidad de Chile, Santiago, pp. 189–212. Mun˜oz, O., Corte´s, S., 1998. Tropane alkaloids from Schizanthus porrigens (Solanaceae). Pharm. Biol. 36, 1–5. Mun˜oz, O., Hartmann, R., Breitmaier, E., 1991. Schizanthine X, a new alkaloid from Schizanthus grahamii (Gill.). J. Nat. Prod. 54, 1094–1096. Mun˜oz, O., Piovano, M., Garbarino, J., Hellwig, V., Breitmaier, E., 1996. Tropane alkaloids from Schizanthus litoralis. Phytochemistry 43, 709– 713. Mun˜oz, O., Schneider, C., Breitmaier, E., 1994. A new pyrrolidine alkaloid from Schizanthus integrifolius Phil. Liebigs Ann. Chem., 521–522.

Murashige, T., Skoog, F., 1962. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15, 473–497. Oksman-Caldentey, K.M., Arroo, R., 2000. Regulation of tropane alkaloid metabolism in plants and plant cell cultures. In: Verpoorte, R., Alfermann, A.W. (Eds.), Metabolic Engineering of Plant Secondary Metabolism. Kluwer Academic Publishers, Dordrecht, pp. 253– 281. Oksman-Caldentey, K.M., Hiltunen, R., 1996. Transgenic crops for improved pharmaceutical products. Field Crop. Res. 45, 57–69. Robins, R.J., Walton, R.J., 1993. The biosynthesis of tropane alkaloids. In: Cordell, G.A. (Ed.), The Alkaloids, vol. 44. Academic Press, Orlando, pp. 115–187. Robins, R.J., Walton, N.J., Parr, A.J., Aird, E.L.H., Rhodes, M.J.C., Hamill, J.D., 1994. Progress in the genetic engineering of the pyridine and tropane alkaloid biosynthetic pathways of Solanaceous plants. In: Ellis, B.E., Kuroki, G.W., Stafford, H.A. (Eds.), Genetic Engineering of Plant Secondary Metabolism. Plenum Press, New York, pp. 1–33. San-Martı´n, A., Rovirosa, J., Gambaro, V., Castillo, M., 1980. Tropane alkaloids from Schizanthus hookeri. Phytochemistry 19, 2007–2008. San-Martı´n, A., Labbe´, C., Mun˜oz, O., Castillo, M., Reina, M., De la Fuente, G., Gonza´lez, A.G., 1987. New tropane alkaloids from Schizanthus grahamii. Phytochemistry 26, 819–822. Wolff, M., Seemann, M., Grosdemange-Billiard, C., Tritsch, D., Campos, N., Rodrı´guez-Concepcio´n, M., Boronat, A., Rohmer, M., 2002. Isoprenoid biosynthesis via the methylerythritol phosphate pathway. (E)-4-Hydroxy-3-methylbut-2-enyl diphosphate: chemical synthesis and formation from methylerythritol cyclodiphosphate by a cell-free system from Escherichia coli. Tetrahedron Lett. 43, 2555–2559. Za´rate, R., Hermosı´n, B., Cantos, M., Troncoso, A., 1997. Tropane alkaloid distribution in Atropa baetica plants. J. Chem. Ecol. 23, 2059– 2066. Zumbrunn, A., Uebelhart, P., Eugster, C.H., 1985. Synthesen von Carotinen mit w-Endgruppen und (Z)-Konfiguration an terminalen konjugierten Doppelbindungen. Helv. Chim. Acta 68, 1519–1539.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 579–583 www.elsevier.com/locate/phytochem

C-prolinylquercetins from the yellow cocoon shell of the silkworm, Bombyx mori Chikara Hirayama a

a,*

, Hiroshi Ono b, Yasumori Tamura a, Masatoshi Nakamura

a

National Institute of Agrobiological Sciences, 1-2 Owashi, Tsukuba, Ibaraki 305-8634, Japan b National Food Research Institute, Tsukuba, Ibaraki 305-8642, Japan Received 6 September 2005; accepted 26 November 2005 Available online 23 January 2006

Abstract Two flavonoids containing the L-proline moiety, 6-C-[(2S,5S)-prolin-5-yl] quercetin (prolinalin A) and 6-C-[(2S,5R)-prolin-5-yl] quercetin (prolinalin B), were isolated from the cocoon shell of the silkworm, Bombyx mori. Their structural elucidation was achieved by application of acid hydrolysis and spectroscopic methods. These compounds were not found in the leaves of mulberry (Morus alba L.), the host plant of the silkworm, suggesting that the flavonoids are metabolites of the insect. This is the first time that flavonoids with an amino acid moiety have been found as naturally occurring compounds.
2005 Elsevier Ltd. All rights reserved. Keywords: Silkworm; Bombyx mori; Mulberry; Morus alba; Moraceae; Flavonoid; Quercetin; L-Proline

*

OH OH O

HO

OH

HOOC

OH

O

1 OH OH O

HO N H

It has been reported that cocoon shells of some strains of the silkworm, Bombyx mori, contain flavonoid-like pigments (Oku, 1934; Hayashiya et al., 1959). Recently, Tamura et al. (2002) isolated quercetin 5-glucoside, quercetin 5,4 0 -diglucoside, and quercetin 5,7,4 0 -triglucoside from the yellow-green cocoon shell of a race Multi-Bi. These flavonoids were not found to be present in leaves of its host plant, mulberry tree (Morus alba), in which quercetin-3glycosides (rutin, isoquercitrin) and quercetin aglycon are naturally occurring (Zhishen et al., 1999). This suggests that flavonoids from the diet are modified within the silkworm by a glucosyltransferase that can transfer a glucose residue to the C-5 hydroxy position of quercetin. Such an enzyme has not been reported in animal tissues. Flavonoids modified by B. mori may be useful as medicinal or cosmetic materials because the ethanolic extracts of yellow-green colored cocoon shells of a range of strains of B. mori have potent antibacterial activity (Kurioka et al., 1999) and strong antioxidant activity (Yamazaki et al., 1999). In a

continuing search among silkworm strains for novel flavonoids with possible biological activity, we found that the aqueous MeOH extract of the yellow cocoon shell of a Chinese race of Daizo contain novel flavonoids with an amino acid moiety. In the present study, we describe the isolation and structural elucidation of two new compounds, designated as prolinalin A (1) and prolinalin B (2).

N H

1. Introduction

OH

HOOC

OH Corresponding author. Tel.: +81 29 838 6087; fax: +81 29 838 6028. E-mail address: [email protected]ffrc.go.jp (C. Hirayama).

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.11.030

O

2

580

C. Hirayama et al. / Phytochemistry 67 (2006) 579–583

2. Results and discussion The aqueous MeOH extract of the cocoon shell of the silkworm, B. mori, was purified as described in the experimental section to yield two new compounds. Prolinalin A (1) was obtained as a yellow amorphous solid. The UV absorption data of the compound in MeOH were similar to those of quercetin. The UV spectral changes with various shift reagents were indicative of the presence of a flavonol skeleton with free hydroxyl groups at C-3, C-5, C-7, and C-4 0 positions, and the presence of a free ortho-dihydroxyl group (Markham, 1982). The compound gave a red-purple pigment by reaction with ninhydrin reagent, indicating that it contains nitrogen. The compound was resistant to hydrolysis by 0.1 N HCl at 100 C, but afforded d-pyroline-5-carboxylic acid in 6 N HCl at 110 C. The D/L configuration of d-pyroline-5-carboxylic acid was determined to be L by the chemical correlation to L-proline. The high-resolution FTICR-MS displayed a protonated ion at m/z 416 (observed 416.0973, calculated 416.0976 for C20H18NO9), and product ions from [M + H]+ were observed at m/z 303 and 370 by MS/MS experiments. From these results, we considered 1 to be a quercetin derivative with an L-proline moiety. Finally, the structural elucidation was established by 1H and 13C NMR spectroscopic analysis. The 1H and 13C NMR spectroscopic data (Table 1) were similar to those of quercetin (Markham and Geiger, 1994; Markham et al., 1978). From the 1H NMR data, a 3 0 ,4 0 -dihydroxy group in ring-B was evidenced by three aromatic resonances located at d 6.88 (1H, d, J =

8.5 Hz), d 7.62 (1H, dd, J = 2.1 and 8.5 Hz), and d 7.73 (1H, d, J = 2.1 Hz), which were assigned to H-5 0 , H-6 0 , and H-2 0 , respectively. A one-proton singlet at d 6.45 suggested the presence of a substitution at C-6 or C-8. The 13 C NMR spectrum displayed 15 distinct carbon signals, which were assigned to the quercetin moiety, including a carbonyl carbon (d 177.3) and five oxygenated aromatic carbons (d 146.3, 149.0, 157.9, 160.2, 165.1). The absence of a methine carbon signal in ring-A, but the presence of a quaternary carbon signal at d 105.3, indicated a ring-A substitution. Residual 1H and 13C resonances of 1 were observed in the relatively up-field region of the spectra. The DQFCOSY, HSQC, and HMBC spectra showed that the residual moiety contained the fragment –CO–CH–CH2–CH2– CH–. Signals at d4.30 (dd, J = 7.7, 9.6) and d 5.27 (dd, J = 7.3, 10.8) were assigned to H-200 and H-500 , respectively. Furthermore, the signals at d 2.12 (dddd, J = 7.1, 9.6, 11.1, 13.1) and d 2.62 (dddd, J = 2.1, 7.4, 7.7, 13.1) were attributable to H-300 , while the signals at d 2.30 (dddd, J = 2.1, 7.1, 7.3, 12.8) and d 2.40 (dddd, J = 7.4, 10.8, 11.1, 12.8) could be assigned to H-400 . In addition, the five carbon resonances were assigned, respectively, to d 174.3 to the carboxyl carbon (C-600 ), and d 63.4, 31.6, 31.5, and 56.1 to the alicyclic carbons (C-200 , C-300 , C-400 , and C-500 ). The 1H and 13 C NMR spectroscopic data for the moiety were consistent with those of C-5-substituted proline derivatives (Severino et al., 2003). These NMR observations and the acidolysis results indicated the presence of an L-prolin-5yl substituent in 1.

Table 1 13 C NMR (125 MHz) and 1H NMR (500 MHz) Spectroscopic data for compounds 1 and 2 (d ppm, in CD3OD) Position

1

2 3 4 5 6 7 8 9 10 10 20 30 40 50 60 200 300

148.3 137.4 177.3 160.2 105.3 165.1 94.9 157.9 103.9 124.0 116.1 146.3 149.0 116.3 121.7 63.4 31.6

400

31.5

500 600

56.1 174.3

dC

a

J values in parentheses are recorded in Hz.

2 dH

6.45s

7.73d (2.1)a

6.88d (8.5) 7.62dd (2.1, 8.5) 4.30dd (7.7, 9.6) a: 2.12dddd (7.1, 9.6, 11.1, 13.1) b: 2.62dddd (2.1, 7.4, 7.7, 13.1) a: 2.30dddd (2.1, 7.1, 7.3, 12.8) b: 2.40dddd (7.4, 10.8, 11.1, 12.8) 5.27dd (7.3, 10.8)

dC 147.8 137.1 177.0 160.1 104.3 165.0 95.7 158.3 103.0 124.2 116.0 146.3 148.8 116.3 121.7 61.8 30.0

29.6

57.2 174.6

dH

6.38s

7.74br, s

6.88br, s 7.62br, s 4.08dd (5.1, 10.4) a: 2.33m b: 2.51m a: 2.24m b: 2.27m 5.05t (7.9)

C. Hirayama et al. / Phytochemistry 67 (2006) 579–583

Comparison of the carbon shifts of 1 with those of published data for quercetin (Markham et al., 1978) revealed a down-field shift of C-6 (Dd 7.0 ppm ), which indicated that the proline substitution occurred at C-6. The linkage of the L-prolin-5-yl group at C-6 was also deduced from an HMBC experiment, which showed correlations of the resonance at d 5.27 (H-500 ) to carbon signals at d 105.3 (C-6), 160.2 (C-5), and 165.1 (C-7), while the resonance at d 6.45 (H-8) showed correlations to the signals at d 103.9 (C-10),d 157.9 (C-9) and d 165.1 (C-7) (Fig. 1). On the basis of these observations, the structure 1 was determined as 6C-(prolin-5-yl) quercetin. The stereochemistry of 1 was determined by the 1,2-diaxial relationships between H-200 and Ha-300 (J = 9.6), Ha-300 and Hb-400 (J = 11.1), and Hb-400 and H-500 (J = 10.8), as shown in Fig. 1. These large coupling constants (9.6
11.1) suggested that H-200 , Ha-300 , Hb-400 , and H-500 , respectively, occupy pseudo-axial positions, with respect to the five-membered ring. This configuration should be observed when the five-membered ring adopts an envelope conformation, in which C-200 , C-400 , C-500 , and N atoms of the ring are coplanar, with C-300 displaced above that plane. In this case, the carboxyl group at C-200 and the quercetin at C-500 should be trans on the pyrrolidine ring. Based on the L-configuration of the d-pyroline-5-carboxylic acid, acidolysis product of 1, the absolute configuration of pyrrolidine ring was determined to be 200 S, 500 S. Thus, prolinalin A (1) was characterized as 6-C-[(2S,5S)-prolin-5-yl] quercetin. Prolinalin B (2) was observed to be a yellow amorphous powder as prolinalin A. The high-resolution FT-ICR-MS measurement of 2 revealed the elementary composition of protonated ion to be C20H18NO9 ([M + H]+), which is the same as that of 1. The 1H and 13C NMR (Table 1) and other spectroscopic data of 2 were very similar to those of 1, suggesting that the overall structures of 1 and 2 were the same. Upon acid hydrolysis, 2 also yielded L-d -pyroline-5-carboxylic acid, indicative of a 200 S configuration. Therefore, 2 was identified as the diastereoisomer of 1 at the C-500 position of the proline moiety. In the 1 H NMR spectrum, the H-500 signal of 2 was shifted 0.22 ppm upfield compared with that of 1, thus suggesting a 500 R config-

581

uration of 2 (Manfre´ and Pulicani, 1994; Severino et al., 2003). In association with the 200 S, 500 R configuration of 2, it is noteworthy that its lower solubility in MeOH and broadened NMR signals compared to 1 were possibly due to the presence of intra-molecular hydrogen bonds between the carboxyl group of the proline moiety and free hydroxyl groups at C-5 and C-7 of the quercetin moiety. Such intra-molecular interactions were not present in 1. From these observations, 2 was determined to be 6-C[(2S,5R)-prolin-5-yl] quercetin. Flavonoids containing nitrogen are very rare natural products. So far, only eight flavonoidal alkaloids have been reported: ficine and isoficine from Ficus pantoniana (Johns et al., 1965), phyllospadine from Phyllosphadix iwatensis (Takagi et al., 1980), vochsine from Vochysia guaianensis (Baudouin et al., 1983), lilaline from Lilium candidum (Masterova et al., 1987), aquiledine and isoaquiledine from Aquilegia ecalcarata (Chen et al., 2001), and lotthanongine from Trigonostemon reidioides (Kanchanapoom et al., 2002). To our knowledge, prolinalin A (1) and B (2) represent the first examples of a flavonoid conjugated with an amino acid. Using an LC mass spectrometer, we attempted to determine whether the novel flavonoids isolated from the silkworm naturally occur in its host plant, M. alba, but found no evidence of these compounds in the plant. In contrast, these flavonoids were found in the cocoon shells of the silkworm reared on a semi-synthetic diet containing quercetin (data not shown), indicating that prolinalin A (1) and B (2) are produced from dietary quercetin within the insect. A number of lepidopteran insects and some grasshoppers are known to sequester dietary flavonoids, and some of these insects have been shown to be able to glycosylate flavonoid aglycones (Harborne and Grayer, 1994). In larvae of several races of B. mori, in addition to glycosylation reactions, the dehydration reaction with L5-hydroxyproline may be a major pathway of sequestered flavonoids, however, the physiological roles for the pathway remain to be studied.

3. Experimental OH OH 1'

O

HO

1 7

H N 1''

9.6 Hz

3

5

H

OH

10.8 Hz

H

OH

HOOC 4''

Ha

O

Hb

Ha Hb

HMBC correlations 2.1 Hz

11.1Hz

3.1. General experimental procedure

3'

H

3

J H-H couplings

Fig. 1. Selected HMBC correlations and 3JH–H couplings of 1.

UV spectra were recorded on a Shimadzu UV-2500 PC spectrophotometer. High-resolution ESI mass spectra were obtained by a Bruker Apex II 70e Fourier-transform ion cyclotron resonance (FT-ICR) mass spectrometer. The MS/MS experiment was carried out with the FT-ICR mass spectrometer coupled with an infrared multiphoton dissociation (IRMPD) system. 1H NMR and 13C NMR spectra were recorded on a Bruker AVANCE 500 spectrometer (500 MHz for 1H; 125 MHz for 13C) or a Bruker AVANCE 800 spectrometer (800 MHz for 1H; 200 MHz for 13C) in CD3OD, with TMS as an internal standard. The preparative and analytical HPLC system consisted of a Shimadzu

582

C. Hirayama et al. / Phytochemistry 67 (2006) 579–583

LC-7A pump, a Shimadzu SPD-7AV UV–Visible detector, and a Shimadzu CTO-10A column oven. 3.2. Biological material Daizo, a Chinese race of B. mori is stocked in the National Institute of Agrobiological Sciences. The larvae of Daizo were reared on fresh leaves of mulberry M. alba L. cv. Shin-ichinose planted at the institute. Cocoon shells produced by the larvae were collected after pupation, then cut into small pieces. 3.3. Extraction and isolation Yellow pigments were extracted from the cocoon shell sample (160 g) by MeOH–H2O (2:1, v/v) at 60 C for 2 h. The extract was filtered and concentrated to a small volume under reduced pressure, after which H2O was added. The aqueous solution was applied to a solid-phase extraction cartridge (Oasis HLB, 35 ml, Waters). After washing with MeOH–H2O (50:50, V/V), the column was eluted with MeOH–H2O (80:20, V/V). The eluate was concentrated by evaporation and loaded on a column of Toyoperl HW-40F (TOSOH). The column was eluted at room temperature with a linear gradient from 50% to 80% solvent B (MeOH containing 0.1% formic acid) in solvent A (0.1% formic acid in H2O). Fractions containing flavonoids were concentrated in vacuo and further purified by reversed-phase HPLC using a Nova-Pak C18 column (19 · 300 mm, Waters) with a flow rate of 10 ml/min at 40 C. A 90-min gradient, from 12.5% to 25% solvent B (solvent A, 0.2% formic acid; solvent B, 0.2% formic acid in MeCN), was used to afford compounds 1 (tR = 55.2 min, 3.2 mg) and 2 (tR = 51.3 min, 1.5 mg).

were eluted from the column with a linear gradient of MeCN in 50 mM triethylamine–phosphate (pH 3.5), from 10% to 40%, for 45 min at a flow rate of 1 ml/min at 25 C. The FDAA-derivatized L-proline was clearly distinguishable from the D-isomer, and its retention time exactly corresponded to those for the FDAA derivatives of amino acids derived from 1 and 2. 3.5. Prolinalin A (1) Yellow amorphous solid. UV (MeOH) kmax nm: 267, 385; 283, 471 (+AlCl3); 279, 440 (+AlCl3 + HCl); 285, 332, 397 (+NaOAc); 275, 402 (+NaOAc + H3BO3). HRFT-ICR-MS: 416.0973 (calculated 416.0976). For 1H, and 13 C NMR spectroscopic data, see Table 1. 3.6. Prolinalin B (2) Yellow amorphous solid. UV (MeOH) kmax nm: 266, 384; 283, 470 (+AlCl3); 279, 444 (+AlCl3 + HCl); 285, 332, 397 (+NaOAc); 275, 401 (+NaOAc + H3BO3). HRFT-ICR-MS: 416.0962 (calculated 416.0976). For 1H, and 13 C NMR spectroscopic data, see Table 1.

Acknowledgements We thank Ms. Ikuko Maeda, Dr. Takashi Murata (Instrumental Analysis Center for Food Chemistry) for their technical help with the NMR and MS measurements. We also thank Ms. Mayumi Hazeyama (National Institute of Agrobiological Sciences) for her technical assistance.

References 3.4. Absolute configuration of amino acid derived from 1 and 2 Each purified compound (0.5 mg) was dissolved in DMSO (50 ll) and mixed with 6 M HCl (1 ml). After deaeration, the tube was sealed with a cap and the contents were heated at 110 C for 3 h. The hydrolysate was brought to dryness in vacuo to remove HCl, and the residue was dissolved in distilled water. The acid hydrolysis of the compounds gave d-pyrolline 5-carboxylic acid, which was confirmed by the reaction with o-amino benzaldehyde and by the analysis using an automatic amino acid analyzer (Hitachi 8500A). Since the absolute configuration of this amino acid could not be directly elucidated, d-pyrolline 5carboxylic acid was reduced to proline by NaBH4. The D-, L-configurations of proline derived by reduction of d-pyrolline 5-carboxylic acid were identified by an HPLC method using a chiral derivatizing reagent, 1-fluoro-2,4-dinitrophenyl-5-L-alanine amide (FDAA; Marfey’s reagent, Pierce, Co.), according to the literature (Marfey, 1984). The HPLC analysis was carried out using a reversed-phase column, Nova-Pak C-18 (3.9 · 150 mm, Waters). FDAA derivatives

Baudouin, G., Tillequin, F., Koch, M., Vuilhorgne, M., Lallemand, J.Y., Jacqumin, H., 1983. Isolement, structure et synthese de la Vochysine, pyrrolidinoflavanne de Vochysia guianensis. J. Nat. Prod. 46, 681–687. Chen, S.-B., Gao, G.-Y., Leung, H.-W., Yeung, H.-W., Yang, J.-S., Xiao, P.-G., 2001. Aquiledine and Isoaquiledine, novel flavonoid alkaloids from Aquilegia ecalcarata. J. Nat. Prod. 64, 85–87. Harborne, J.H., Grayer, R.J., 1994. Flavonoids and insects. In: Harbone, J.B. (Ed.), The Flavonoids, Advances in Research Since 1986. Chapman & Hall/CRC, pp. 589–618. Hayashiya, K., Sugimoto, S., Fujimoto, N., 1959. Studies on the pigments of cocoon. (III) The qualitative test of the pigments of green cocoon. J. Seric. Sci. Jpn. 28, 27–29. Johns, S.R., Russel, J.H., Hefferman, M.L., 1965. Ficine, a novel flavonoidal alkaloid from Ficus pantoniana. Tetrahedron Lett. 24, 1987–1991. Kanchanapoom, T., Kasai, R., Cumsri, P., Kraisintu, K., Yamasaki, K., 2002. Lotthanongine, an unprecedented flavonoidal indole alkaloid from the roots of Thai medicinal plant, Trigonostemon reidioides. Tetrahedron Lett. 43, 2941–2943. Kurioka, A., Ishizaka, H., Yamazaki, M., Endo, M., 1999. Antibacterial activity of cocoon shells. J. Silk Sci. Tech. Jpn. 8, 57–60. Manfre´, F., Pulicani, J.P., 1994. Enantiospecific synthesis and absolute configuration of (+)-RP 66803 a new non-peptide CCK antagonist. Tetrahedron-Asymmetr. 5, 235–238.

C. Hirayama et al. / Phytochemistry 67 (2006) 579–583 Marfey, P., 1984. Determination of D-amino acids. II. Use of a bifunctional reagent, 1,5-difloro-2,4-dinitrobenzene. Carlsberg Res. Commun. 49, 591–596. Markham, K.R., 1982. Techniques of Flavonoid Identification. Academic Press, London. Markham, K.R., Geiger, H., 1994. 1H nuclear magnetic resonance spectroscopy of flavonoids and their glycosides in hexadeuterodimethylsulfoxide. In: Harbone, J.B. (Ed.), The Flavonoids, Advances in Research Since 1986. Chapman & Hall/CRC, pp. 441–497. Markham, K.R., Ternai, B., Stanley, R., Geiger, H., Mabry, T.J., 1978. Carbon-13 NMR studies of flavonopids-III. Tetrahedron 34, 1389– 1397. Masterova, I., Uhrin, D., Tomko, J., 1987. Lilialine-A flavonoid alkaloid from Lilium candidum. Phytochemistry 26, 1844–1845. Oku, M., 1934. The chemical studies on the pigments in the cocoon filaments of Bombyx mori (VII). Nippon Nogeikagaku Kaishi 10, 1014–1028 (in Japanese).

583

Severino, E.A., Costenaro, E.R., Garcia, A.L.L., Correia, C.R.D., 2003. Probing the stereoselectivity of the Heck arylation of endocyclic enecarbamates with diazonium salts. Concise synthesis of (2S,5R)phenylproline methyl ester and Schramm’s C-azanucleotide. Org. Lett. 5, 305–308. Takagi, M., Funahashi, S., Ohta, K., Nakabayashi, T., 1980. Phyllospadine, a new flavonoidal alkaloid from the sea-grass Phyllosphadix iwatensis. Agr. Biol. Chem. 44, 3019–3020. Tamura, Y., Nakajima, K., Nagayasu, K., Takabayashi, C., 2002. Flavonoid 5-glucosides from the cocoon shell of the silkworm, Bombyx mori. Phytochemistry 59, 275–278. Yamazaki, M., Nakamura, N., Kurioka, K., Komatsu, K., 1999. Antioxidative activity of ethanolic extracts of cocoon shell. J. Seric. Sci. Jpn. 68, 167–169. Zhishen, J., Mengcheng, T., Jianming, W., 1999. The determination of flavonoid contents in mulberry and their scavenging effects on superoxide radicals. Food Chem. 64, 555–559.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 584–588 www.elsevier.com/locate/phytochem

A hemiterpene glucoside as a probing deterrent of the bean aphid, Megoura crassicauda, from a non-host vetch, Vicia hirsuta Naohiro Ohta, Naoki Mori, Yasumasa Kuwahara, Ritsuo Nishida

*

Laboratory of Chemical Ecology, Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan Received 15 September 2005; received in revised form 28 October 2005 Available online 24 January 2006

Abstract The bean aphid, Megoura crassicauda Mordvilko, feeds selectively on plants belonging to the genus Vicia (Fabaceae). However, it never infests the tiny vetch, V. hirsuta (L.) Gray. The aphid appeared to discriminate between host and non-host plants by tasting specific chemicals during penetration of its stylet into the plant tissues. The aphid, after being stimulated by specific probing stimulants, deposited characteristic proteinous stylet sheaths through a parafilm membrane, which has one side in contact with an extract solution of Vicia angustifolia. However, an addition of a V. hirsuta extract to the medium strongly inhibited the salivary sheath formation. A specific probing deterrent was isolated from a V. hirsuta extract by monitoring the inhibitory effect, and identified as (E)-2-methyl-2-butene-1,4-diol 4O-b-D-glucopyranoside. A mixture of the glycoside and the stimulatory V. angustifolia fraction in the same equivalency found in plants significantly decreased the probing activity in M. crassicauda. Since the stylet insertion process is a crucial step for the aphid’s settlement on a plant, the glycoside seems to act as an effective chemical barrier for V. hirsuta.
2006 Elsevier Ltd. All rights reserved. Keywords: Aphid; Megoura crassicauda; Vicia hirsuta; Fabaceae; Probing deterrent; Host selection; Allomone; (E)-2-methyl-2-butene-1,4-diol 4-O-b-Dglucopyranoside; Hemiterpene glycosides

1. Introduction The bean aphid, Megoura crassicauda Mordvilko, feeds selectively on a variety of fabaceous plants in the genus Vicia, such as broad bean (Vicia faba L.) and narrow leaf vetch (Vicia angustifolia L.). The host selectivity of oligophagous aphids is controlled by specific chemicals in their host plants (Van Emden, 1972; Klingauf, 1987). After landing on a plant, the aphid inserts its stylet into the internal tissues and senses the chemicals contained within (Montllor, 1991 and refs. therein). This information determines whether the aphid leaves the plant or continues to probe further (Van Emden, 1972). During the penetration into the host plant tissue towards the phloem, aphids produce proteinous salivary sheaths (Miles, 1965). The same stylet sheaths can be

*

Corresponding author. Tel.: +81 75 753 6313; fax: +81 75 753 6312. E-mail address: [email protected] (R. Nishida).

0031-9422/$ - see front matter
2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.013

observed on a stretched Parafilm M membrane, which is in contact with a solution of host plant extract (Mittler and Dadd, 1965). We have recently characterised two acylated flavonol glycosides [quercetin 3-O-a-L-arabinopyranosyl(1 ! 6)-(200 -O-(E)-p-coumaroyl)-b-D-galactopyranoside and quercetin 3-O-a-L-arabinopyranosyl-(1 ! 6)-(200 -O-(E)-pcoumaroyl)-b-D-glucopyranoside] as the specific probing stimulants of M. crassicauda, to induce the salivary sheath formation. This suggests that the two compounds are possible host finding cues prior to ingestion of the phloem sap (Takemura et al., 2002). In the course of this study, we noticed that the tiny vetch, Vicia hirsuta (L.) Gray, was never attacked by M. crassicauda, while dense colonies of the aphids were commonly observed on a very closely related species, V. angustifolia. Both the plant species are common vinous weeds that grow closely together, often tangling with each other during the spring season (April to early June) in Japan. Although the seedpods of V. hirsuta are more hairy

N. Ohta et al. / Phytochemistry 67 (2006) 584–588

than those of V. angustifolia, the surface structure of the stem, on which the aphids would colonise, did not appear different between the two species under a microscope. When the aphids were exposed to a V. hirsuta plant in the laboratory, they displayed a short-term probing on the stem, but eventually left the plant. This suggests that the aphids were capable of sensing some odd chemicals in the internal tissues of V. hirsuta. An extract of V. hirsuta was found to deter the probing behaviour of the aphids when added to the solution containing the positive stimulants in feeding tests using the parafilm membrane. We report the isolation and structural elucidation of a specific probing deterrent in V. hirsuta against M. crassicauda.

2. Results and discussion When freshly cut stems from both V. angustifolia (host) and V. hirsuta (non-host) were placed together in a cage, the introduced adult aphids (apterous or alatae) settled selectively on V. angustifolia and started feeding on the stems producing a large quantity of honeydew within few hours. The aphids seemed to conduct some trial probing on V. hirsuta as well, but they left the plant within a short period. M. crassicauda actively displayed probing behaviour toward the crude aqueous extracts of V. angustifolia, depositing a substantial number of thick stylet sheaths (thickness, approx. 10 lm; length > 100 lm) on a parafilm membrane at a concentration of 1.0 g fresh leaf equivalent/ml (gle/ml) (Takemura et al., 2002). However, the aphid deposited much fewer stylet sheaths on the parafilm in contact with a crude aqueous extract of V. hirsuta (Table 1). When a mixture of both V. angustifolia and V. hirsuta extracts was supplied in an equivalency (1.0 gle/ml each), the prob-

Table 1 Probing responses of Megoura crassicauda to solutions of Vicia plant extracts and mixtures with various fractions and compound 1 Sample

Probing activity (points) (N)

Distilled water V. angustifolia (Va) extract [host, control] V. hirsuta extract [non-host] Va ext. + V. hirsuta ext. Va ext. + Fr. 1 Va ext. + Fr. 2 Va ext. + Fr. 3 Va ext. + Fr. 4 Va ext. + Fr. 5 Va ext. + Fr. 6 Va ext. + Compound 1 (0.5 gle/ml) Va ext. + Compound 1 (1.0 gle/ml) Va ext. + Compound 1 (2.0 gle/ml)

10.5 ± 5.0 (10)a 48.2 ± 6.1 (43) 9.1 ± 4.3 (10)a 7.5 ± 3.3 (10)a 45.1 ± 9.0 (9) 12.0 ± 5.6 (9)a 20.9 ± 5.8 (9) 41.3± 10.4 (9) 34.2 ± 6.3 (9) 35.7 ± 12.3 (9) 22.4 ± 6.9 (10) 17.1 ± 5.0 (10)a 13.8 ± 3.4 (10)a

Probing activity represents the average frequency of stylet sheath formation. Dose (gle = gram leaf equivalent): all extracts and fractions were tested at 1.0 gle/ml, except for compound 1 (0.5–2.0 gle/ml). a Significantly different from control Vicia angustifolia crude extract (1.0 gle/ml) at p = 0.05 using the Mann–Whitney U test.

585

ing activity decreased significantly (Table 1). This suggests the presence of a deterrent (or masking) substance(s) that interfered with the production of salivary sheaths. The probing deterrent assay was performed by mixing V. hirsuta fractions (1.0 gle/ml unless otherwise specified) in a stimulative fraction of V. angustifolia (1.0 gle/ml) as shown in Table 1. An aqueous extract of V. hirsuta was applied to an ODS column with increasing concentrations of methanol in water as eluant to give six fractions. These were eluted with water, 10% methanol, 20% methanol, 40% methanol, 60% methanol and 80% methanol in water, respectively. The 10% methanol eluate exhibited a significant probing deterrent activity almost equivalent to that of a crude extract of V. hirsuta (no significant difference from a mixture of crude V. hirsuta extract + V. angustifolia extract, Mann–Whitney U test) (Table 1). The active eluate was further separated by a reversed-phase HPLC and a single deterrent compound 1 was isolated as an amorphous solid (yield: 370 lg/g leaf). Compound 1 clearly suppressed the stimulant activity of V. angustifolia extract at doses of 1.0 and 2.0 g.l.e, but not at a 0.5 g.l.e. dose (Table 1). Compound 1 was deduced to possess the molecular formula of C11H20O7 (mw 264) from the electro-spray ionization (ESI) mass spectra (positive m/z 287 [M + Na]+; 303 [M + K]+), although [M + H]+ was not observed. An enzymatic hydrolysis of compound 1 with b-glucosidase yielded 2-methyl-2-butene-1,4-diol (2) and D-glucose. Compound 2 was determined to possess the (E)-configuration by comparing both geometrical isomers derived from mesaconic acid (trans) and citraconic acid (cis), respectively. The 1H NMR spectrum of 1 (Table 2) exhibited signals arisen from a methyl (d 1.69), two methylenes (d 3.96, 2H; d 4.28 and 4.39, 2H) and an olefinic methine (d 5.36) of a 2-methyl-2-butene-1,4-diol moiety. A glucopyranosy moiety was found to be attached to one of the primary alcohol to form a b-configuration. The linkage of glucose to the hemiterpene diol

Table 2 1 H NMR spectral data for compound 1, (E)- and (Z)-2-methyl-2-butene1,4-diols (CD3OD, 400 MHz) Position Diol 1 3 4 5 Glucosyl 10 20 30 40 50 60

Compound 1

(E)-diol (2)

(Z)-diol

3.96 5.36 4.28 4.39 1.69

s br. t (6.8) m dd (12.1, 6.2) s

3.93 s 5.59 br. t (6.6) 4.13 d (6.3)

4.085 s 5.47 t (6.8) 4.11 d (8.0)

1.67 s

1.80 s

4.29 3.17 3.34 3.25 3.31 3.67 3.87

d (8.0) m m m m d (11.7, 4.9) br. d (11.7)

586

N. Ohta et al. / Phytochemistry 67 (2006) 584–588

was determined by the heteronuclear multiple bond correlation (HMBC) spectrum, where correlation from H-1 0 (d 4.29) of glucose to the C-4 (d 66.2) of diol, and from H-4 (d 4.39) to the C-1 0 (d 103.0) of glucose were observed. This indicated that the glucose moiety was linked to the C-4 of 2-methyl-2-butene-1,4-diol. Thus, compound 1 was determined as (E)-2-methyl-2-butene1,4-diol 4-O-b-D-glucopyranoside. Compound 1 appeared to be a novel compound from plants, although its positional isomer (E)-2-methyl-2-butene-1,4-diol 1-O-b-Dglucopyranoside (3) and (Z)-isomer of 3 have been known from other plants (Nicoletti et al., 1992; Kitajima et al., 1998; Zhu et al., 1998). Compounds 2 and 3 were also found in fraction 2 as minor components, but the compounds did not show any significant deterrent effect at the equivalent dosage found in the plant.

HO O

HO HO

O

OH

OH

Compound 1 HO OH

Compound 2

It is intriguing that a novel hemiterpene glycoside exhibits probing deterrent activity against the aphid. Since the stylet insertion process is a crucial step for the aphid’s settlement on a plant (Klingauf, 1987; Montllor, 1991), the glycoside 1 seems to act as an effective chemical barrier for V. hirsuta. At the initial step of probing, M. crassicauda inserted the stylet ‘intercellularly’ through the epidermis and mesophyll of host Vicia plant. The compound may block the stylet penetration toward the phloem during gustation, if the substance is detected along the intercellular path, although we do not know the distribution of 1 in the plant. Further support for the feeding preference of the aphids between these two closely related Vicia species is as follows: firstly, the specific allelochemical is provided in the absence of the probing deterrent 1 of the host V. angustifolia and secondly, the lack of the specific probing stimulants (acylated flavonol glycosides previously identified from the host V. angustifolia, Takemura et al., 2002) in the nonhost plant, V. hirsuta. A series of plant secondary metabolites, including terpenoids (Asakawa et al., 1988), phenolics (Montgomery and Arn, 1974; Dreyer and Jones, 1981; Dreyer et al., 1981; Jones and Klocke, 1987), alkaloids (Corcuera, 1984) and hydroxamic acid derivatives (Argandona et al., 1983; Givovich and Niemeyer, 1995), have been reported as feeding deterrents in several aphid species. However, the actual role of these allelochemicals as probing deterrents, ingestion

deterrents or toxicants in many cases remains unclear due to the complexity of host assessment process by aphids plus the lack of knowledge related to the distribution of the chemicals within the plant tissues (Montllor, 1991). A systematic study on the mechanisms of host and non-host recognition of aphids within closely related plant taxa (or cultivars) might provide us a new control technique against pest aphids of economic importance.

3. Experimental 3.1. General Optical rotation was measured with a JASCO DIP-370 spectropolarimeter. The MS data were recorded with a Shimadzu LCMS-2010A using a reversed phase column (Mightysil RP-18GP, 75 mm · 3 mm i.d.) eluted with 10100% MeOH in H2O (0.2 ml/min) with an electro-spray ionization (ESI)-positive mode. 1H and 13C NMR spectra were measured with a Bruker AC400 FT-NMR spectrometer (400 MHz) with TMS as an internal standard. The letters s, d, t and m represent singlet, doublet, triplet, and multiplet, respectively, and coupling constants are given in Hz. 3.2. Plant material Arial parts of V. hirsuta were collected in Sakyo-ku, Kyoto, Japan in April 2003, and identified by Dr. Reiichi Miura of Kyoto University. A voucher specimen is deposited at the Herbarium of Pesticide Research Institute, Kyoto University. 3.3. Extraction and isolation of deterrent Fresh leaves and stems of V. hirsuta (300 g) were extracted with EtOH–H2O (9:1, v/v) three times. The combined ethanolic solution was evaporated in vacuo, and the residue (30.30 g) was dissolved in H2O (600 ml) and partitioned four times with hexane (400 ml each). The crude aqueous extract (14.30 g) was applied to a reversed phase column (200 g of Cosmosil 140C18OPN, Nacalai tesque, 310 · 50 mm i.d.) eluted (2 L each) in sequence (yields of dry materials given in the parentheses) with H2O (8.69 g), MeOH–H2O (1:9) (0.40 g), MeOH–H2O (1:4) (0.17 g), MeOH–H2O (2:3) (0.30 g), MeOH–H2O (3:2) (0.41 g) and MeOH–H2O (4:1) (0.05 g). MeOH–H2O (1:9) eluate was subjected to preparative HPLC using a reverse phase column (YMCPack Pro C18 250 · 10 mm i.d.), eluting with MeOH– H2O (1:9) at flow rate of 2.0 ml/min. Active fraction was found at a retention time (Rt) range of 18.4–21.4 min, which was further subjected to the same column eluting with MeOH–H2O (7.5:92.5) at a flow rate of 2.0 ml/min. Compound 1 was isolated at a Rt = 26.0 min. The yield of compound 1 from 100 g of

N. Ohta et al. / Phytochemistry 67 (2006) 584–588

the leaves was 37 mg. Compound 2 was detected at Rt = 21.5 min in a small quantity under the latter HPLC condition. 3.4. (E)-2-Methyl-2-butene-1,4-diol 4-O-b-Dglucopyranoside (1) Colourless solid. ½a
17 D 41.8 (10% MeOH in water; c 1.00); LCMS (ESI-positive) m/z (relative intensity): 287 [M + Na]+ (100), 303 [M + K]+ (52). For 1H and 13C NMR spectroscopic data, see Tables 2 and 3. 3.5. Enzymatic hydrolysis of compound 1 Compound 1 (26.4 mg) and b-glucosidase (from almond, Wako Pure Chemical Industries Ltd., 193 units, 5.2 mg) was dissolved in 1 ml of AcOH–NaOAc buffer (1 ml) and incubated for 4 h at 37 C. The reaction mixture was acidified with 0.5 N HCl, filtered (syringe filter unit, Millex), and passed through a reversed phase column (10 g, Cosmosil 140C18-OPN, Nacalai tesque) eluted with water (100 ml). The water eluate was separated by HPLC (Cosmosil AR-II, 200 · 15 mm i.d.) eluting with MeOH– H2O (1:4) in H2O (2.0 ml/min) and D-glucose ð15:3 mg ½a
17 D ¼ þ43:4Þ and (E)-diol (2) (7.1 mg) were isolated at Rt = 6.6 and 16.8 min, respectively. 3.6. Preparation of (E)- and (Z)-2-methyl-2-butene-1,4diols (E)-2-methyl-2-butene-1,4-dioic acid dimethyl ester was prepared from mesaconic acid (Tokyo Chemical Industries Co. Ltd.) by refluxing in a mixture with methanol and benzene (3:1) in the presence of a catalytic amount of p-toluenesulfonic acid. The ester was reduced to (E)diol (2) by treating with LiAlH4 in Et2O. Likewise, (Z)diol was obtained from citraconic acid (Aldrich Chemical Co. Ltd.) via esterification and LiAlH4 reduction. The 1H and 13C NMR spectroscopic data are given in Tables 2 and 3. Table 3 13 C NMR spectral data for compound 1, (E)- and (Z)-2-methyl-2-butenel,4-diols (CD3OD, 100 MHz) Position

Compound 1

(E)-diol (2)

(Z)-diol

Diol 1 2 3 4 5

68.0 140.9 121.7 66.2 13.9

68.2 138.7 125.0 59.2 13.7

61.3 139.1 127.7 58.6 21.5

Glucosyl 10 20 30 40 50 60

103.0 75.1 78.2 71.7 78.1 62.8

587

3.7. Bioassay Adults of M. crassicauda (apterous viviparae) were starved for 2 h; and five insects were introduced in a glass tube fitted with a parafilm membrane in contact with either a test solution (0.4 ml) or distilled H2O (control), and allowed to probe through the membrane for 24 h under laboratory conditions (25 C, 16L:8D) (Takemura et al., 2002). The number of stylet sheaths deposited on the parafilm membrane was observed under a microscope after being stained with a red fuchsin basic solution. The probing activity was scored by frequency of stylet sheath formation as described previously (Takemura et al., 2002).

Acknowledgements We thank Dr. Reiichi Miura of Kyoto University for identification of the plant. This study was partly supported by a Grant-in-Aid for the 21st Century COE Program for Innovative Food and Environmental Studies Pioneered by Entomomimetic Sciences, from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References Argandona, V.H., Corcuera, L.J., Niemeyer, H.M., Campbell, B.C., 1983. Toxicity and feeding deterrency of hydroxamic acids from Gramineae in synthetic diets against the greenbug, Schizaphis graminum. Entomol. Exp. Appl. 34, 134–138. Asakawa, Y., Dawson, G.W., Griffiths, D.C., Lallemand, J.Y., Ley, S.V., Mori, K., Mudd, A., Pezechk-Leclaire, M., Pickett, J.A., Watanabe, H., Woodcock, C.H., Zhang, Z.N., 1988. Activity of drimane antifeedants and related compounds against aphids, and comparative biological effects and chemical reactivity of ()- and (+)-polygodial. J. Chem. Ecol. 14, 1845–1855. Corcuera, L.J., 1984. Effects of indole alkaloids from Gramineae on aphids. Phytochemistry 23, 539–541. Dreyer, D.L., Jones, K.C., 1981. Feeding deterrency of flavonoids and related phenolics towards Schizaphis graminum and Myzus persicae: Aphid feeding deterrents in wheat. Phytochemistry 20, 2489–2493. Dreyer, D.L., Reese, J.C., Jones, K.C., 1981. Aphid feeding deterrents in sorghum. Bioassay, isolation, and characterization. J. Chem. Ecol. 7, 273–284. Givovich, A., Niemeyer, H.M., 1995. Comparison of the effect of hydroxamic acids from wheat on five species of cereal aphids. Entomol. Exp. Appl. 74, 115–119. Jones, K.C., Klocke, J.A., 1987. Aphid feeding deterrency of ellagitannins, their phenolic hydrolysis products and related phenolic derivatives. Entomol. Exp. Appl. 44, 229–234. Kitajima, J., Suzuki, N., Ishikawa, T., Tanaka, Y., 1998. New hemiterpenoid pentol and monoterpenoid glycoside of Torilis japonica fruit, and consideration of the origin of apiose. Chem. Pharm. Bull. 46, 1583–1586. Klingauf, F., 1987. Host plant finding and acceptance. In: Minks, A.K., Harrewijn, P. (Eds.), Aphids: Their Biology, Natural Enemies, and Control. Elsevier, Amsterdam, pp. 209–223. Miles, P.W., 1965. Studies on the salivary physiology of plant-bugs: the salivary secretions of aphids. J. Insect Physiol. 11, 1261–1268. Mittler, T.E., Dadd, R.H., 1965. Differences in the probing responses of Myzus persicae (Sulzer) elicited by different feeding solutions behind a parafilm membrane. Entomol. Exp. Appl. 8, 107–122.

588

N. Ohta et al. / Phytochemistry 67 (2006) 584–588

Montgomery, M.E., Arn, H., 1974. Feeding response of Aphis pomi, Myzus persicae, and Amphorophora agathonica to phlorizin. J. Insect Physiol. 20, 413–421. Montllor, C.B., 1991. The influence of plant chemistry on aphid feeding behavior. In: Bernays, E.A. (Ed.), Insect–Plant Interactions III. CRC Press, Boca Ranton, pp. 125–173. Nicoletti, M., Tomassini, L., Foddai, S., 1992. A new hemiterpene glucoside from Ornithogalum montanum. Planta Med. 58, 472.

Takemura, M., Nishida, R., Mori, N., Kuwahara, Y., 2002. Acylated flavonol glycosides as probing stimulants of a bean aphid, Megoura crassicauda, from Vicia angustifolia. Phytochemistry 61, 135–140. Van Emden, H.F., 1972. Aphids as phytochemists. In: Harborne, J.B. (Ed.), Annual Proceeding of the Phytochemical Society Number 8, Phytochemical Ecology. Academic Press, London, pp. 34–36. Zhu, N.Q., Sharapin, N., Ziang, J.H., 1998. Three glucosides from Maytenus ilicifolia. Phytochemistry 47, 265–268.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 589–594 www.elsevier.com/locate/phytochem

Volatile oil from Guarea macrophylla ssp. tuberculata: Seasonal variation and electroantennographic detection by Hypsipyla grandella Joa˜o Henrique G. Lago a,b,*, Marisi G. Soares a,c, Luciane G. Batista-Pereira c, M. Fa´tima G.F. Silva c, Arlene G. Correˆa c, Joa˜o B. Fernandes c, Paulo C. Vieira c, Nı´dia F. Roque d b

a Instituto de Quı´mica, Universidade de Sa˜o Paulo, 05508-900 Sa˜o Paulo, SP, Brazil Faculdade de Cieˆncias Biolo´gicas, Exatas e Experimentais, Universidade Presbiteriana Mackenzie, 01302-907 Sa˜o Paulo, SP, Brazil c Departamento de Quı´mica, Universidade Federal de Sa˜o Carlos, 13565-905 Sa˜o Carlos, SP, Brazil d Instituto de Quı´mica, Universidade Federal da Bahia, 40170-290 Salvador, BA, Brazil

Received 28 July 2005; received in revised form 4 November 2005 Available online 24 January 2006

Abstract GC and GC–MS analyses of the volatile oils from Guarea macrophylla (Meliaceae) collected during three different periods in one year (February, June and October) indicated a seasonal variation in chemical composition. Whilst sesquiterpenes were the predominant class of components present in the leaf oil, a seasonal dependent variation in the degree of oxygenation of these compounds was detected, which seemed to be associated with phenological factors. The leaf oil, and fractions thereof, were subjected to GC coupled with electroantennographic detection employing antennae of females of Hypsipyla grandella, an insect pest that attacks several meliaceous species. Five compounds elicited significant responses and these were identified as ledol, 1-cubenol, guai-6-en-10b-ol, 1-epi-cubenol, and s-muurolol. The results suggest that these components could be responsible for the attraction of H. grandella to G. macrophylla.
2005 Elsevier Ltd. All rights reserved. Keywords: Guarea macrophylla tuberculata; Meliaceae; Hypsipyla grandella; Pyralidae; Mahogany shoot borer; Electroantennography; Volatile oils; Sesquiterpenes

1. Introduction The Meliaceae (mahogany family) consists of 51 genera (containing nearly 1400 species) of which Cabralea, Carapa, Cedrela, Trichilia, Guarea and Swietenia are represented in Brazil. Many of the ca. 150 species of the genus Guarea are widely distributed (Pennigton and Styles, 1975) throughout Latin America and Africa. Thus, within Brazil, Guarea macrophylla Vahl ssp. tuberculata occurs in the southern states of Rio Grande do Sul, Rio de Janeiro and Minas Gerais, in the central states of Mato Grosso and Brası´lia, and also in *

Corresponding author. Tel.: +55 11 30913813; fax: +55 11 30913875. E-mail address: [email protected] (Joa˜o Henrique G. Lago).

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.018

the Amazonian region to the north of the country. In the southern areas, the tree is confined to the lowland coastal rain forest and is often found growing alongside river banks, while in Mato Grosso and Goias it occurs mainly in gallery forests (Correa, 1984). The tree, known locally as ‘‘Atau´ba’’, blossoms in the summer months between October and February, but mainly in November and December, and bears fruit in the winter from June to October. Several phytochemical studies on G. macrophylla have been published, and a wide variety of secondary metabolites, including sequi-, di- and tri-terpenes, have been identified (Lago et al., 2000; Lago and Roque, 2002b). With respect to volatile compounds, one monoterpene,

590

Joa˜o Henrique G. Lago et al. / Phytochemistry 67 (2006) 589–594

sixteen sesquiterpenes and six diterpenes were reported in the leaf oil (Lago and Roque, 2002a), whilst 24 sesquiterpenes were identified in the oil from the fruit (Lago et al., 2005). However, no data concerning the dynamics of the composition of the volatile oils have been presented. The mahogany shoot borer, Hypsipyla grandella (Zeller, 1848) (Lepidoptera: Pyralidae), is one of the most economically important pests of Neotropical forests since it attacks the terminal shoots of plants of the family Meliaceae (Schabel et al., 1999). However, the volatile compounds of G. macrophylla have not been the subject of electroantennographic studies involving H. grandella, even though this tree is attacked by the insect larvae. Thus, with the aim of investigating the nature of the interaction between the shoot borer and G. macrophylla, we have studied the variation in the composition of the volatile oil of the leaves in three different periods (February, June and October) of one year, and established the selectivity and sensitivity of the antennae receptors of the female moths to the leaf volatiles.

2. Results and discussion The total yield of volatile oil obtained from each batch of G. macrophylla leaves harvested in different months of the year was approximately 0.05%. Table 1 presents the mean relative percentage of each of the components of the oil of leaves collected from the same tree at four different times during the days of 15th February, 15th June and 15th October 2000. These data indicate that the leaf oils were mainly composed of sesquiterpenes, but that the relative amounts of sesquiterpene hydrocarbons and oxygenated sesquiterpenes varied considerably. Thus, during February (the sterile period) the relative level of oxygenated sesquiterpenes was 38%, but this fell to 29.2% during fruiting (October). Over the same period, the relative level of sesquiterpene hydrocarbons was 24%, but this rose with fruit development to ca. 36% and 38.8%, respectively, in June and October, suggesting that some phenological effect could be involved in this variation (Lopes et al., 1997). In February, the oxygenated sesquiterpene guai-6-en-10b-ol (18) was the major component of the oil (16 ± 1%) (see

Table 1 Variation in the composition of the volatile oil of leaves of Guarea macrophylla during one year

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Total Total Total Total Total a

Compounds

Kovats retention index

Relative percentage compositiona February

June

October

a-Terpineol a-Cubebene a-Ylangene a-Copaene a-Gurjunene b-Caryophyllene a-Humulene allo-Aromadendrene Germacrene-D Bicyclogermacrene c-Cadinene a-Cadinene Palustrol Germacradien-4-ol Spathulenol Ledol 1-Cubenol Guai-6-en-10b-ol 1-epi-Cubenol s-Cadinol s-Muurolol Isopimara-7,15-diene Manoyl oxide Isopimara-7,15-dien-3b-ol Isopimara-7,15-dien-3-one Isopimara-7,15-dien-2a-ol Labda-8,13-(E)-dien-15-ol

1189 1351 1372 1376 1409 1418 1454 1461 1480 1496 1513 1524 1561 1574 1576 1629 1611 1654 1629 1637 1651 1926 1989 2161 2279 2284 2412

0.13 ± 0.05 0.25 ± 0.02 0.9 ± 0.1 0.54 ± 0.04 0.40 ± 0.07 2.0 ± 0.2 0.9 ± 0.2 2.3 ± 0.3 1.6 ± 0.2 6.5 ± 0.6 8±1 1.7 ± 0.2 1.8 ± 0.1 1.0 ± 0.3 1.8 ± 0.7 8.9 ± 0.4 3.3 ± 0.6 16 ± 1 1.8 ± 0.1 1.6 ± 0.2 1.64 ± 0.07 – 4.0 ± 0.6 0.07 ± 0.07 4.5 ± 0.3 0.4 ± 0.2 0.13 ± 0.08

0.22 ± 0.03 0.26 ± 0.02 1.4 ± 0.2 0.9 ± 0.4 0.50 ± 0.03 2.7 ± 0.4 1.5 ± 0.3 3.4 ± 0.2 3.1 ± 0.2 7.3 ± 0.3 12.8 ± 0.9 1.84 ± 0.08 1.71 ± 0.10 1.2 ± 0.1 1.0 ± 0.1 8.7 ± 0.6 3.1 ± 0.2 14 ± 1 1.4 ± 0.1 1.5 ± 0.1 1.4 ± 0.1 0.07 ± 0.04 3.9 ± 0.4 0.03 ± 0.03 2.6 ± 0.3 0.13 ± 0.08 0.07 ± 0.04

– 0.12 ± 0.07 1.56 ± 0.05 0.51 ± 0.05 0.16 ± 0.09 3.6 ± 0.2 1.82 ± 0.07 2.5 ± 0.1 3.5 ± 0.1 7.0 ± 0.6 16.1 ± 0.6 2.0 ± 0.2 1.5 ± 0.2 1.06 ± 0.07 0.77 ± 0.06 6.6 ± 0.2 4.7 ± 0.1 10.4 ± 0.6 1.9 ± 0.1 1.2 ± 0.1 1.1 ± 0.1 – 5.9 ± 0.4 – 3.8 ± 0.3 – –

0.13 ± 0.05 24 ± 1 38 ± 1 – 10 ± 1

0.22 ± 0.03 36 ± 2 34 ± 2 0.07 ± 0.04 6.8 ± 0.8

– 38.8 ± 0.8 29.2 ± 0.7 – 9.7 ± 0.6

monoterpenes sesquiterpene hydrocarbons oxygenated sesquiterpenes diterpene hydrocarbon oxygenated diterpenes

Mean values (±standard deviation) for leaf oils collected at four different times on the same day and analysed separately (n = 3).

Joa˜o Henrique G. Lago et al. / Phytochemistry 67 (2006) 589–594

Table 1, for compound numbers). In June, although the accumulation of sesquiterpene hydrocarbons in the oil had increased, 18 remained the major constituent (14 ± 1%). In October, sesquiterpene hydrocarbons were the predominant volatile compounds, with c-cadinene (11) becoming the main component present (16.1 ± 0.6%). During this period, the oil obtained from the fruit contained high concentrations of sesquiterpene hydrocarbons, with viridiflorene, c-cadinene and cadina1,4-diene being the main components (Lago et al., 2005). The effects of different concentrations of leaf oil (and fractions derived therefrom) of G. macrophylla on antennae from female and male specimens of H. grandella are shown in Figs. 1 and 2, respectively. At all concentrations tested the electroantennographic responses were significantly higher for the volatile oil samples than for the control (hexane). At concentrations of 10 and 100 mg/ml, females and males showed levels of antennal activity towards the sesquiterpene hydrocarbon fraction that were significantly lower than the responses obtained from the crude oil or the oxygenated sesquiterpene fraction (there being no statistical difference between the responses to the latter two samples). Similar differences have also been noted in studies carried out with other, unrelated, insects including Yponomeuta sp. (Lepidoptera: Yponomeutidae), Anthonomus grandis (Coleoptera: Curculionidae) and Megastigmus spermotrophus (Hymenoptera: Torymidae) (Van Der Pers, 1981; Dickens et al., 1984; Thie´ry and Marion-Poll, 1998). These results indicate that adults of H. grandella recognize some compounds present in the volatile oil of G. macrophylla; that is, the antennae of both sexes of this insect must possess receptor neurons (chemoreceptors) in sensillas that detect components of the leaf oil.

Fig. 1. Electroantennographic responses [expressed as mean percentage values with respect to the control (j hexane = 100%)] elicited from antennae of H. grandella females by the crude volatile oil , a sesquiterpene hydrocarbon fraction and an oxygenated sesquiterpene fraction from G. macrophylla at concentrations of 1, 10 and 100 mg/ml. [Mean values labelled with the same letter are not significantly different at P < 0.05 on the basis of the Tukey test (n = 10 antennae)].

591

Fig. 2. Electroantennographic responses [expressed as mean percentage values with respect to the control (j hexane = 100%)] elicited from antennae of H. grandella males by the crude volatile oil , a sesquiterpene hydrocarbon fraction and an oxygenated sesquiterpene fraction from G. macrophylla at concentrations of 1, 10 and 100 mg/ml. [Mean values labelled with the same letter are not significantly different at P < 0.05 on the basis of the Tukey test (n = 10 antennae)].

Figs. 1 and 2 also demonstrate that the responses to the volatile extracts obtained with antennae from females were almost twice as large as those attained with male antennae. Furthermore, the dose–response exhibited by female antennae was more pronounced than that observed with antennae from males. It is thus suggested that, whilst the volatile oils from G. macrophylla are general attractants, they may also play a specific role in host plant and oviposition selection in agreement with the study of Maia et al. (2000). From EAG experiments, these authors claimed that the high selectivity of antennae of H. grandella females to the essential oils from leaves of Cedrela odorata and Toona ciliata indicated their potential roles as chemical messengers for habitat location and oviposition. The oxygenated sesquiterpene fraction of the leaf oil of G. macrophylla was subjected to GC analysis with electroantennographic detection (GC–EAD) using an antenna from a H. grandella female. The antennal olfactory system clearly showed differential sensitivity to several compounds present in the volatile oil as depicted in Fig. 3. The main constituents producing significant antenna responses were identified as ledol (16), 1-cubenol (17), guai-6-en-10b-ol (18), 1- epi-cubenol (19), and s-muurolol (21). Since the relative amounts of oxygenated sesquiterpenes increases during the sterile period, it is suggested that these components may play a role in attracting H. grandella females to G. macrophylla leaves at this time for feeding and/or ovipositing. A number of studies have demonstrated the role of plant volatiles in the orientation of species of moths to their host plants. Host volatiles, particularly terpenoids, have potential importance in the feeding and/or mating behaviour of polyphagous insects, and exhibit various biological activities such as herbivore attractants, repellents, and feeding

592

Joa˜o Henrique G. Lago et al. / Phytochemistry 67 (2006) 589–594

Sa˜o Paulo, Sa˜o Paulo, SP, Brazil, were collected at 08.00, 12.00, 16.00 and 20.00 h on 15th February 2000 from the tree without fruit, and at the same times on 15th June and 15th October 2000 when the tree was bearing fruit. A voucher specimen of the plant has been deposited in the Instituto de Biocieˆncias, Universidade de Sa˜o Paulo. Volatile oil was obtained from fresh leaves (400 g) by hydrodistillation for 4 h using a Clevenger-type apparatus (yield ca. 200 mg). 3.2. Quantitative analysis

Fig. 3. GC–FID profile (upper trace) of the oxygenated sesquiterpene fraction (1.0 mg/ml) of the leaf oil of G. macrophylla, and the simultaneous GC–EAD response (lower trace) of antennae of H. grandella females. The numbered peaks elicited electrophysiological responses and were identified as 16, ledol; 17, 1-cubenol; 18, guai-6-en-10b-ol; 19, 1-epi-cubenol; and 21, s-muurolol.

stimulants (Gershenzon and Croteau, 1991; Harborne, 1991; Langenheim, 1994). Studies concerning the interaction between H. grandella and the mahogany tree, Swietenia macrophylla (Meliaceae), have shown that the sesquiterpene hydrocarbon b-caryophyllene is the main constituent responsible for the antennal response (Soares et al., 2003) and is associated with the attraction of H. grandella to oviposit on the leaves of this plant. The effects of environmental and phenologic factors should be related to the accumulation of different components in the plant. Such mechanisms are specific to each species and probably associated with the attack of the insect on the plant. In the case of G. macrophylla, since the amounts of oxygenated sesquiterpenes in the leaves decrease during the fruiting period, there may be a reduced interaction between the volatile oil compounds and H. grandella at this time. In summary, the present findings provide additional information concerning the chemistry of the Guarea–Hypsipyla relationship, and signify that the behavioural role of these allelochemicals require further detailed investigation.

Volatile oils were analysed using a Hewlett–Packard (HP) 5890 series II gas chromatograph equipped with an FID detector, an automatic injector (HP 7673) and an electronic integrator (HP3396A). An HP-5 capillary column (30 m · 0.32 mm, I.D.; 0.25 lm film thickness) of crosslinked 5% phenyl–methyl silicone was employed with helium as the carrier gas at a flow rate of 1 ml/min. Temperature programming was performed as follows: 100 C isothermal for 2 min, then increased from 100 to 240 C at 5 C/min, and finally isothermal at 240 C for 5 min. The injector and detector temperatures were 180 and 260 C, respectively. Samples were analysed in triplicate, using n-nonane (Sigma) as internal standard, and component concentrations were calculated from the relative GC peak areas as shown in Table 1. 3.3. Qualitative analysis GC–MS analyses were carried out using an HP model 5973 MS coupled to an HP-5890 gas chromatograph fitted with an HP-5 column (30 m · 0.25 mm, I.D., 0.25 lm film thickness). The chromatographic conditions outlined above were employed, and the EI/MS spectra were recorded at 70 eV. Components were identified on the basis of their retention times in comparison to those determined when the oil was analysed previously (Lago and Roque, 2002a) and by co-injection with appropriate authentic samples. Additionally, the Kovats retention index for each component was determined relative to the retention times of a series of n-alkanes (Table 1). 3.4. Fractionation of the volatile oil A portion (50 mg) of the crude volatile oil (collected at 12.00 h on 15th October 2000) was separated into two fractions by SiO2 prep. TLC using CH2Cl2 as eluent. The fractions were extracted with CH2Cl2 and shown (by GC and GC–MS) to contain, respectively, sesquiterpene hydrocarbons (33 mg) and oxygenated sesquiterpenes (7 mg).

3. Experimental 3.5. Insects 3.1. Plant material and extraction of volatile oil Fresh leaves of a single specimen of G. macrophylla, growing in the Instituto de Biocieˆncias, Universidade de

Specimens of H. grandella were obtained from the Entomology Laboratory of the Faculdade de Cieˆncias Agra´rias do Para´, Bele´m, PA, Brazil, and maintained in the labora-

Joa˜o Henrique G. Lago et al. / Phytochemistry 67 (2006) 589–594

593

tory on mahogany foliage. Pupae were established and sexed (Parra, 1986) in the Insect Bioassay Laboratory, Universidade Federal de Sa˜o Carlos, Sa˜o Carlos, SP, Brazil. Male and female pupae were placed separately into plastic vials (6 · 6 cm I.D.) and incubated in a chamber at 25 ± 1 C and 60 ± 5% relative humidity under a 12 light:12 dark illumination regime.

was split equally between FID and EAD detectors: for the latter, a mounted antenna was positioned such that a stream of humidified air could direct the output from the GC over the antenna/electrode assembly.

3.6. Electroantennographic analyses

The authors wish to thank Fundac¸a˜o de Amparo a Pesquisa do Estado de Sa˜o Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Cientifico e Tecnolo´gico (CNPq) for their financial support of this work.

Electroantennographic experiments (EAG) and analyses employing electroantennographic detection (EAD) were carried out using male and female moths 1–2 days after emergence. The complete antenna was excised using biological forceps, and a few segments removed from both the base and the tip (Bjostad, 1998). The antenna was then fixed between two stainless steel electrodes by inserting the base and tip into droplets of Spectra 360 electrode gel (Parker, Orange, NJ, USA) that had been applied to each of the metal surfaces. Antennal responses were amplified and recorded using a Syntech (Davis, CA, USA) intelligent data acquisition controller interfaced to an AT486PC running EAG software. In order to evaluate the EAG response, a filter paper (about 0.8 cm2) impregnated with 10 ll of a freshly prepared hexane solution of the test sample, or with 10 ll of hexane as negative control, was placed into a Pasteur pipette, the tip of which was be positioned within a continuous stream of humidified and purified air that passed over the mounted antenna at a flow rate of 1.2 l/min. EAGs were obtained by releasing the test sample or hexane control, in the form of a 0.3 s flush of the pipette, into the air stream. Control stimulations were made at the beginning and at the end of each series of EAG experiments, and the test samples were applied randomly to the antenna at intervals of 60 s. Owing to the large differences in overall sensitivities between individual antennae, and to compensate for the decline in the sensitivity of an antenna during a measuring session, EAG amplitudes recorded in response to the test samples were normalised with respect to the control responses. In this normalization process, performed automatically by the Syntech EAG software, responses to the initial and final controls were defined as 100%, and values obtained between these two reference points were calculated by linear interpolation. The essential oil or oil fractions were tested on ten antennae each from H. grandella females. The mean normalized responses of the different compounds were submitted to ANOVA for statistical analysis and compared by the Tukey test (P < 0.05). 3.7. GC–FID coupled with EAD GC analyses with electroantennographic detection (GC– EAD) were performed on a Shimadzu GC-17A instrument equipped with a Supelcowax 10 column (30 m · 0.25 mm I.D.; 0.25 lm film thickness). The chromatographic parameters outlined above were employed. The column effluent

Acknowledgements

References Bjostad, L.B., 1998. Electrophysiological methods. In: Haynes, K.F., Millar, J. (Eds.), Methods in Chemical Ecology: Chemical Methods, vol. 1. Chapman & Hall, London, pp. 339–369. ´ teis e das Exo´ticas CultivaCorrea, M.P., 1984. Diciona´rio de Plantas U das, vol. 1. Imprensa Nacional, Rio de Janeiro. Dickens, J.C., Payne, T.L., Ryker, L.C., Rudinsky, J.A., 1984. Single cell responses of Douglas fir beetle, Dendroctonus pseudotsugae Hopkins (Coleoptera: Scolytidae) to pheromones and host odours. Journal of Chemical Ecology 10, 583–600. Gershenzon, J., Croteau, R., 1991. Terpenoids. In: Rosenthal, G.A., Berenbaum, M. (Eds.), Herbivores: Their Interactions with Secondary Plant Metabolites, second ed, The Chemical Participants, vol. 1 Academic Press, San Diego, pp. 165–219. Harborne, J.B., 1991. Recent advances in the ecological chemistry of plant terpenoids. In: Harborne, J.B., Tomes-Barberan, F.A. (Eds.), Ecological Chemistry and Biochemistry of Plant Terpenoids. Clarendon Press, Oxford, pp. 399–426. Lago, J.H.G., Roque, N.F., 2002a. Terpenes from the essential oil of the leaves of Guarea macrophylla Vahl ssp. tuberculata Vellozo (Meliaceae). Journal of Essential Oil Research 14, 12–13. Lago, J.H.G., Roque, N.F., 2002b. Cycloartane triterpenoids from Guarea macrophylla. Phytochemistry 60, 329–332. Lago, J.H.G., Brochini, C.B., Roque, N.F., 2000. Terpenes from leaves of Guarea macrophylla (Meliaceae). Phytochemistry 55, 727–731. Lago, J.H.G., Corne´lio, M.L., Moreno, P.R.H., Apel, M.A., Limberger, R.P., Henriques, A.T., Roque, N.F., 2005. Sesquiterpenes from essential oil from fruits of Guarea macrophylla Vahl ssp. tuberculata (Meliaceae). Journal of Essential Oil Research 17, 84–85. Langenheim, J.H., 1994. Higher plant terpenoids: a phytocentric overview of their ecological roles. Journal of Chemical Ecology 20, 1223–1280. Lopes, N.P., Kato, M.J., Andrade, E.H.A., Maia, J.G.S., Yoshida, M., 1997. Circadian and seasonal variation in the essential oil from Virola surinamensis leaves. Phytochemistry 46, 689–693. Maia, B.H.L.N.S., Paula, J.R., Sant’ana, J., Silva, M.F.G.F., Fernandes, J.B., Vieira, P.C., Costa, M.S.S., Ohashi, O.S., Silva, J.N.M., 2000. Essential oils of Toona and Cedrela species (Meliaceae): taxonomic and ecological implications. Journal of the Brazilian Chemical Society 11, 629–639. Parra, J.R.P., 1986. Criac¸a˜o de insetos para estudos com pato´genos. In: Alves, S.B. (Ed.), Controle Microbiano de Insetos. Manole, Sa˜o Paulo, pp. 349–373. Pennigton, T.D., Styles, B.D., 1975. A generic monograph of Meliaceae. Blumea 22, 419–540. Schabel, H., Hilje, L., Nair, K.S.S., Varma, R.V., 1999. Economic entomology in tropical forest plantations: an update. Journal of Tropical Forest Science 11, 303–315. Soares, M.G., Batista-Pereira, L.G., Fernandes, J.B., Correa, A.G., Silva, M.F.G.F., Vieira, P.C., Rodrigues-Filho, E., Ohashi, O.S., 2003. Electrophysiological responses of female and male Hypsipyla grandella

594

Joa˜o Henrique G. Lago et al. / Phytochemistry 67 (2006) 589–594

(Zeller) to Swietenia macrophylla essential oils. Journal of Chemical Ecology 29, 2143–2151. Thie´ry, D., Marion-Poll, M., 1998. Electroantennogram responses of Douglas-fir seed chalcids to plant volatiles. Journal of Insect Physiology 44, 483–490.

Van Der Pers, J.N.C., 1981. Comparison of electroantennogram response spectra to plant volatiles in seven species of Yponomeuta and in the Tortricid Adoxophyes orana. Entomologia Experimentalis et Applicata 30, 181–192. Zeller, P.C., 1848. Exotische Phyciden. Isis van Oken 41, 857–890.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 595–604 www.elsevier.com/locate/phytochem

Gluconic acid: An antifungal agent produced by Pseudomonas species in biological control of take-all Rajvinder Kaur a, John Macleod b, William Foley c, Murali Nayudu

c,*

a

c

Department of Plant and Microbial Biology, University of California, 111 Koshland Hall, Berkeley, CA 94720-3102, USA b Research School of Chemistry, The Australian National University, Canberra, ACT 0200, Australia School of Botany and Zoology, Faculty of Science, The Australian National University, Daley Road, ANU Campus Acton, Canberra, ACT 0200, Australia Received 12 October 2005; received in revised form 7 December 2005 Available online 30 January 2006

Abstract Pseudomonas strain AN5 (Ps. str. AN5), a non-fluorescent Australian bacterial isolate, is an effective biological control (biocontrol) agent of the take-all disease of wheat caused by the fungus Gaeumannomyces graminis var. tritici (Ggt). Ps. str. AN5 controls Ggt by producing an antifungal compound which was purified by thin layer and column chromatography, and identified by NMR and mass spectroscopic analysis to be D-gluconic acid. Commercially bought pure gluconic acid strongly inhibited Ggt. Two different transposon mutants of Ps. str. AN5 which had lost take-all biocontrol did not produce D-gluconic acid. Gluconic acid production was restored, along with take-all biocontrol, when one of these transposon mutants was complemented with the corresponding open reading frame from wild-type genomic DNA. Gluconic acid was detected in the rhizosphere of wheat roots treated with the wild-type Ps. str. AN5, but not in untreated wheat or wheat treated with a transposon mutant strain which had lost biocontrol. The antifungal compounds phenazine-1-carboxylic acid and 2,4-diacetylphloroglucinol, produced by other Pseudomonads and previously shown to be effective in suppressing the take-all disease, were not detected in Ps. str. AN5 extracts. These results suggest that D-gluconic acid is the most significant antifungal agent produced by Ps. str. AN5 in biocontrol of take-all on wheat roots.
2005 Elsevier Ltd. All rights reserved. Keywords: Pseudomonas; Gaeumannomyces graminis; Triticum; Wheat; Take-all; Antifungal; Biological control; Gluconic acid; Root disease; Australia

1. Introduction Pseudomonas bacteria produce different metabolites that can suppress fungal plant pathogens. The production of antibiotics (compounds which at very low concentrations, lg levels, inhibit microorganisms) such as phenazine and phloroglucinol by symbiotic fluorescent Pseudomonas bacteria has been shown to provide a natural defense to the plant against fungal diseases such as take-all (Weller, 1988; Keel et al., 1992; Cook et al., 1995). Pseudomonas fluorescens 2–79 and Pseudomonas aureofaciens 30–84 produce the novel antifungal secondary metabolite phenazine-1-carboxylic acid (PCA). PCA produced by *

Corresponding author. Tel.: +61 2 61253643; fax: +61 2 61259738. E-mail address: [email protected] (M. Nayudu).

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.011

Pseudomonas on wheat roots has been shown to be a crucial factor in take-all disease suppression (Turner and Messenger, 1986; Thomashow and Weller, 1990; Thomashow et al., 1990; Pierson and Thomashow, 1992). The antibiotic 2,4-diacetylphloroglucinol (DPG), which has also been shown to suppress the take-all pathogen on plant roots, is produced by several Pseudomonads including Pseudomonas fluorescens strain CHA0 (Keel et al., 1992, 1996). Other antibiotics produced by Pseudomonads include pyoluteorin (Maurhofer et al., 1994), Pyrrolnitrin (Ligon et al., 2000) and oomycin A (James and Gutterson, 1986) which can suppress a range of different plant pathogenic fungi. Novel bacterial antifungal metabolites produced by Pseudomonas species have also been identified: 3-(1-hexenyl)5-methyl-2-(5H)-furanone (Paulitz et al., 2000); N-mercapto-4-formylcarbostyril (Fakhouri et al., 2001); aerugine

596

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

(Lee et al., 2003); cyclic lipodepsipeptides (Pedras et al., 2003); phenazine-1-carboxamide (Kumar et al., 2005). It is noteworthy that almost all of the antifungal agents identified to date have complicated structures and are hydrophobic in nature. For example, PCA and DPG show medium polarity in nature. The Australian continent contains unique soil microbes as it has been geographically isolated. Farming commenced in Australia approximately 200 years ago. The take-all disease of wheat is caused by the fungus Gaeumannomyces graminis var. tritici (Ggt), which was first identified in South Australia in 1870 (Rovira et al., 1991). Large wheat yield losses over a long period of time due to the take-all disease have instigated significant Australian research in this area. There has been extensive characterization of various bacterial and fungal isolates that show biocontrol against take-all (Rovira et al., 1991). However, little is known about either the precise mechanisms involved in biocontrol, or the nature of the antifungal metabolites that Australian bacteria produce. Pseudomonas strain AN5 (Ps. str. AN5) is a non-fluorescent bacterial soil isolate from the Cowra region of New South Wales, Australia. Ps. str. AN5 is able to effectively protect against Ggt in agar plate bioassays and pot experiments (Nayudu et al., 1994). Furthermore, it has been demonstrated that Ps. str. AN5 biocontrol of take-all in field trials at dryland sites induces significant increases in wheat yield (Nayudu et al., 1994). Transposon mutants of Ps. str. AN5 which exhibited either decreased or abolished biocontrol against Ggt were isolated previously (Nayudu et al., 1994). We have since endeavoured to determine the antifungal compound(s) that Ps. str. AN5 produces and the nature of take-all biocontrol protection conferred by this unique Australian isolate. In this paper, we report the discovery that D-gluconic acid, a simple sugar acid, is the most significant antifungal metabolite produced by Ps. str. AN5 against the take-all fungal pathogen in biocontrol protection.

2. Results and discussion Ps. str. AN5 extracts were made using methods developed for DPG (Keel et al., 1992) or PCA (Rosales et al., 1995) isolation for other Pseudomonads. These extracts were tested for their ability to suppress Ggt. DPG and PCA compounds are insoluble in water but soluble in organic solvents such as chloroform, methylene chloride, methanol and ethyl acetate (Gurusiddaiah et al., 1986). In the case of Ps. str. AN5, the biologically active compound(s) was soluble in water but insoluble in the organic solvents (Fig. 1a). Ps. str. AN5 extracts showed different Ggt inhibition patterns to those of PCA, DPG and pyoluteorin (Gurusiddaiah et al., 1986; Keel et al., 1992; Maurhofer et al., 1994) in agar overlay assays on thin layer chromatography (TLC) plates (data not shown). These assays were done using organic solvent systems that have successfully separated DPG, PCA or pyoluteorin previ-

ously. No organic solvent systems were able to move the biologically active compound(s) from the origin in Ps. str. AN5 extracts, as observed by Ggt inhibition in TLC agar overlay assays (Fig. 2a). The solvent systems previously used to identify hydrophobic antifungal agents which suppress plant pathogens (Gurusiddaiah et al., 1986; Keel et al., 1992; Maurhofer et al., 1994) could not resolve the biologically active compound(s) isolated from Ps. str. AN5. Solvent systems that have been previously used for isolation of hydrophilic compounds (Aszalos et al., 1968; Hellmut et al., 1990) were subsequently tested. The solvent system comprising n-propanol:ethyl acetate:water (5:2:3), previously used for separating carbohydrates (Hellmut et al., 1990), separated the biologically active compound(s) in Ps. str. AN5 extracts with Ggt inhibition being observed at approximately 0.7 Rf value in TLC agar overlay assays (Fig. 2b). Crude extracts of Ps. str. AN5 were differentiated on a silica column using n-propanol:ethyl acetate:water (5:2:3). The column fractions were collected and separated on TLC plates and also tested for biological activity against Ggt on potato dextrose agar (PDA) overlay bioassays. Fractions from 13 to 26 with Rf about 0.7 on TLC were found to be active in the case of Ps. str. AN5. These biologically active fractions were pooled for further analysis. The pooled fractions from Ps. str. AN5 were characterized by 1H and 13C nuclear magnetic resonance (NMR), and mass spectroscopy. NMR spectra of pooled fractions were dominated by the presence of signals which could be assigned to a- and b-glucopyranose (Horton et al., 1983). None of the signals present in the NMR spectra of Ps. str. AN5 (Fig. 3a) corresponded to resonances characteristic of either PCA or DPG. Due to the presence of a large quantity of glucose in biologically active Ps. str. AN5 fractions (as seen in NMR spectra) GC/MS of the silylated extract was carried out to separate glucose and other impurities from the biologically active component(s). The total ion current (TIC) trace showed the presence of a significant chromatographic peak in Ps. str. AN5 silylated pooled fractions. The spectra observed in Ps. str. AN5 was consistent with the spectra of a glucopyranose, and in particular, gluconic acid (Horton et al., 1983; Milson and Meers, 1985; Tsai et al., 1995). This was confirmed by comparing with a library of mass spectra (Merkey et al., 1974) which identified this peak as the trimethylsilyl (TMS) derivative of gluconic acid (data not shown). A retention time of 13.47 min in GC/MS is the same as for pure D-gluconic acid. It also had an identical mass spectrum to that of the hexa-TMS derivative of a pure sample of D-gluconic acid which was obtained commercially (Fig. 4). After comparison of 1H and 13C NMR spectra of pure D-glucose and D-gluconic acid with the partially purified active fraction from Ps. str. AN5, it was possible to reject the signals assigned to glucose. The remaining resonances matched those of D-gluconic acid exactly (data not shown). In agar plate bioassays, commercially bought pure D-gluconic acid strongly inhibited Ggt (Fig. 1b) but, as

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

597

Fig. 2. Thin layer chromatography (TLC) agar overlay bioassays showing antifungal activity of separated crude fractions of Pseudomonas strain AN5 (Ps. str. AN5) and Ps. str. AN5MN1 against the take-all fungus Gaeumannomyces graminis var. tritici (Ggt). (a) Activity of crude extracts of Ps. str. AN5 separated on TLC plates using the following solvent systems: (1) methanol:chloroform, 1:9; (2) methanol:chloroform, 3:7; (3) methanol:chloroform, 5:5. (b) Activity of crude extracts separated on TLC plates of Ps. str. AN5 (1) and mutant Ps. str. AN5MN1 (2) with a npropanol:ethyl acetate:water (5:2:3) solvent system. In all cases, a 100 ll suspension of crude extract was applied as a band at the origin of the TLC plate for development. The arrow indicates clear zones where Ggt growth has been inhibited.

Fig. 1. Agar overlay plate bioassays showing antifungal activity of crude bacterial extracts and pure gluconic acid against the take-all fungus Gaeumannomyces graminis var. tritici. (a) Activity of crude extracts of Pseudomonas strain AN5 grown on PDA using for extraction: (1) methanol; (2) chloroform; (3) ethyl acetate; (4) water. (b) Activity of commercially obtained pure gluconic acid at concentrations of mg/ml: (1) 50; (2) 25; (3) 12.5; (4) 6.25. (c) Activity of crude ethyl acetate extracts derived from the following bacterial strains on MA (DPG extraction): Pseudomonas fluorescens strain Pf-5 (1,3); Pseudomonas. strain AN5 (2,4).

expected, no inhibition was observed with D-glucose (data not shown). Mass spectroscopic analysis of pure D-gluconic acid showed the presence of D-gluconic acid and a small amount of D-gluconolactone. Gluconolactone is an intermediary in the conversion of glucose to gluconic acid (Goodwin and Anthony, 1998). In the isolation procedures used for Ps. str. AN5, the neutral extraction conditions would shift the D-gluconolactone/D-gluconic acid equilibrium towards the acid species (Milson and Meers, 1985). However, one would reasonably expect to detect D-gluconolactone, if present, as we observed it in the pure D-gluconic acid. D-Gluconolactone was not detected in any of the biologically active extracts from Ps. str. AN5, which

suggested that it is not an important compound in biocontrol of take-all by Ps. str. AN5. DPG or PCA producing Pseudomonas strains are known to produce coloured pigments in medium. Pseudomonas DPG producing strains are characterized by a red pigment in King’s B medium (Keel et al., 1996). Turner and Messenger (1986) reported that PCA production leads to golden yellow crystals in pigment production medium (PPM) and PDA. Ps. str. AN5 did not produce red pigments on King’s B agar or yellow crystals on PPM agar or PDA (data not shown). There is a good correlation between DPG and red pigment production in a large number of Pseudomonas species tested (Keel et al., 1996). PCA production by Pseudomonas is also consistent with golden yellow crystal formation (Gurusiddaiah et al., 1986; Turner and Messenger, 1986). Ps. str. AN5 extracts made from malt agar (MA), according to the methods of Keel et al. (1996) for isolation of DPG, did not suppress the take-all pathogen in agar overlay bioassays (Fig. 1c). As a control, DPG was also isolated from P. fluorescens strain Pf-5 (Nowakthompson et al., 1994) which did suppress takeall (Fig. 1c). There was no DPG detected in Ps. str. AN5 extracts but DPG was detected in P. fluorescens strain Pf5 extracts, by 1H NMR and mass spectroscopy (data not shown). Ps. str. AN5 extracts were made from PPM using established methods for PCA isolation (Rosales et al., 1995) but showed no take-all antifungal activity in bioassays. There was no PCA detected in Ps. str. AN5 extracts

598

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

Fig. 3. Nuclear magnetic resonance (NMR) of Pseudomonas. strain AN5 (Ps. str. AN5) and Ps. str. AN5MN1. 13C NMR spectra of semi-purified fractions from Ps. str. AN5 (a) and the mutant Ps. str. AN5MN1 (b). Arrows indicate gluconic acid peaks which are absent in the Ps. str. AN5MN1 but present at the same position in Ps. str. AN5. In the 13C NMR of Ps. str. AN5, additional peaks at 181.2, 76.67, 73.54, 75.16, 73.76 and 65.2 were observed compared to the mutant Ps. str. AN5MN1. The peak at 181.2 is for C1 and the remaining peaks are for C2–C6 of D-gluconic acid.

by mass spectroscopy (data not shown). These results suggest that Ps. str. AN5 does not produce the antifungal metabolites DPG or PCA.

Using the same methods described above for isolating gluconic acid from Ps. str. AN5, we were unable to detect any biologically active fraction in P. fluorescens strain Pf5 extracts from TLC agar overlay assays using n-propanol:ethyl acetate:water (5:2:3) solvent system at 0.7 Rf value. Furthermore, mass spectroscopy analysis of the same fractions from P. fluorescens strain Pf5 extracts (where gluconic acid was detected in Ps. str. AN5 extracts) did not detect the presence of gluconic acid. Two transposon mutants of Ps. str. AN5 (Ps. str. AN5MN1 and Ps. str. AN5MN2) which had lost biocontrol against the take-all disease were assayed for gluconic acid production. These two biocontrol deficient mutant strains each had a single Tn5gus insertion in a different region of the Ps. str. AN5 genome (Nayudu et al., 1994). There was no biocontrol activity detected against Ggt in any of the extracts from these strains. The crude extracts of these strains were separated on silica columns using an n-propanol:ethyl acetate:water (5:2:3) solvent system. Individual column fractions 13–26, with Rf about 0.7 on TLC (where the active compound was found for the parent strain) were found to be biologically inactive against Ggt in these mutants (Fig. 2b). Pooled fractions obtained for the two mutant strains were subject to 1H and 13C NMR, and mass spectroscopy. Glucose was present in these fractions but gluconic acid was not detected in 13C NMR (Fig. 3b). A cosmid containing the complementary wild-type region (pLAFR1-Mur 1) from the parent genome (Nayudu et al., 1994) was transferred into the mutant Ps. str. AN5MN1. This construct, Ps. str. AN5MN1(pLAFR1Mur 1), inhibits the take-all fungus in agar plate bioassays (data not shown). Using the isolation method described for gluconic acid (with agar strips extracted from next to where the strain was growing on PDA plates) it was shown that Ps. str. AN5MN1(pLAFR1-Mur 1) produced a biologically active fraction. The active fraction had a Rf value (identical to that of gluconic acid) of about 0.7 using a npropanol:ethyl acetate:water (5:2:3) solvent system in a TLC agar overlay bioassay. Analysis of a freeze dried biologically active fraction from this strain by mass spectroscopy confirmed the presence of gluconic acid. From the

Fig. 4. Mass spectra of TMS derivative of commercially obtained pure D-gluconic acid. Characteristic peaks for gluconic acid are observed at 205, 217 and 292.

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

location of the Tn5gus transposon insertion site in Ps. str. AN5MN1 we were able to find a complementary open reading frame (ORF) region on pLAFR1-Mur 1. A polymerase chain reaction (PCR) product of approximately 3.0 kb containing the putative wild-type gene (ORFMN1) was obtained using pLAFR1-Mur 1 DNA as the template. ORFMN1 was cloned into the EcoRI site located in the chloramphenicol resistance gene of the broad host range vector pSUP106 (Priefer et al., 1985). This plasmid (pSUP 106::ORFMN1) was then transferred into Ps. str. AN5MN1. This Ps. str. AN5MN1 (pSUP 106::ORFMN1) construct produced gluconic acid and inhibited the take-all fungus in agar plate bioassays (data not shown). Although a recombinant deficient strain of Ps. str. AN5 was not used for these experiments, it was found that all of the plasmids used for complementation were relatively stable in the Ps. str. AN5MN1 background as they exhibited the same restriction enzyme pattern when transferred back to Escherichia coli K-12 and digested with a number of six base pair recognizing restriction enzymes (PstI, SmaI, HindIII, SalI). These experiments show that a Ps. str. AN5 mutant which is unable to produce gluconic acid concomitantly loses take-all biocontrol, but biocontrol and gluconic acid production is restored when this mutant is complemented by the corresponding wild-type gene. This strongly suggests that the ability to produce gluconic acid is directly correlated with the ability of the bacterial strain to suppress the take-all pathogen. In a similar manner, complementation experiments with other antifungal agents, such as DPG (Keel et al., 1992) and PCA (Thomashow et al., 1990), disclosed their essential role in take-all biocontrol. Gluconic acid is produced by specialized bacteria such as Gluconobacter (Velizarov and Beschkov, 1994), and some filamentous fungi (Magnuson and Lasure, 2004). Production of D-gluconic acid has been generally reported to be transient in most cases in Pseudomonas bacteria (Haltrich et al., 1996). Schleissner et al. (1997) reported accumulation of low levels of D-gluconic acid in the extracellular medium in a Pseudomonas putida soil isolate. However, this P. putida isolate is not a biocontrol strain. This P. putida isolate requires D-glucose to be oxidized to D-gluconic acid for utilization in the glycolytic pathway, as it cannot use D-glucose. The reported gluconic acid production by this P. putida strain is a novel method used to metabolize glucose (which it cannot normally use as a carbon source). Pseudomonas cepacia isolates have been shown to produce gluconic acid. Gluconic acid production by these isolates has been linked to their ability to solubilize mineral phosphate (Babukhan et al., 1995). Gluconic acid producing bacteria or fungi have not been reported to be biocontrol strains. Using the titration method of Schleissner et al. (1997) we estimated that Ps. str. AN5 produced approximately 5.0 mg/ml of acid when glucose was provided as the sole carbon source in water. We were only able to detect very low levels of acid production (approximately 0.12 mg/ml of acid) with the mutants Ps. str. AN5MN1 and Ps. str.

599

AN5MN2. This suggests that most of the acid produced by Ps. str. AN5 is gluconic acid. There is significant production of D-gluconic acid by Ps. str. AN5 in medium where glucose is present, such as PDA. Ggt is inhibited by between 0.1 and 0.5 mg/ml concentration of pure D-gluconic acid when supplemented in PDA (data not shown). This indicates a good correlation between the concentration of D-gluconic acid that can be produced by Ps. str. AN5 and the concentration of gluconic acid required to inhibit Ggt. Strikingly, there are no reports of D-gluconic acid production at such a high level in any other biocontrol bacteria that demonstrate antifungal activity. James and Gutterson (1986) reported that antifungal metabolite synthesis can be regulated by glucose in P. fluorescens strain HV37A which shows biological activity against Pythium ultimum. P. fluorescens strain HV37A produces oomycin A, a complex protein antibiotic of molecular weight 500–600 (James and Gutterson, 1986). The reported complexity of glucosedependent regulation of antibiotic production in P. fluorescens strain HV37A (Gutterson et al., 1986; Gutterson et al., 1988) seems different to the simple production of gluconic acid by Ps. str. AN5. It is important to know if Ps. str. AN5 in the rhizosphere of the wheat root is capable of producing gluconic acid. Wheat grown on Herridge’s (H) agar plates were subject to one of the following three treatments: no treatment of wheat seed (control); treatment with Ps. str. AN5; treatment with Ps. str. AN5MN1. Ps. str. AN5 and Ps. str. AN5MN1 bacterial strains streaked alone on H agar plates supplemented with bromocresol purple pH indicator did not grow even after 10 days of incubation. As well, all plates were purple in colour (pH > 7) indicating that no acid was produced. This showed that H plates did not contain any carbon sources that Ps. str. AN5 could utilize for growth or gluconic acid production. For each of the three treatments, aqueous extracts were prepared using individual agar strips removed from plates with wheat roots growing on the surface of the agar. There was strong inhibition of Ggt observed in aqueous extracts obtained with Ps. str. AN5 treated wheat plants. However, there was no inhibition in aqueous extracts obtained with Ps. str. AN5MN1 treated wheat or untreated wheat. All of these extracts were freeze-dried and analyzed by mass spectroscopy. The mass spectra of aqueous extracts obtained using agar strips of Ps. str. AN5 treated plants disclosed peaks that are characteristic of gluconic acid (between 200 and 300 mass charge
1). These were absent in aqueous freeze-dried extracts of Ps. str. AN5MN1 treated wheat or untreated wheat. This suggests that gluconic acid is being produced in the rhizosphere of wheat plants treated with Ps. str. AN5 but not untreated wheat or wheat treated with the mutant Ps. str. AN5MN1. Evidence that gluconic acid is produced in the rhizosphere of wheat roots colonized by the parent Ps. str. AN5 but not the mutant Ps. str. AN5MN1 (which has lost biocontrol) supports the hypothesis that this antifungal

600

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

metabolite is crucial for biocontrol protection of take-all on wheat roots. In summary, Ps. str. AN5 was shown to produce the sugar acid D-gluconic acid. We were unable to detect any other putative antifungal agents (such as PCA or DPG) from Ps. str. AN5. Commercially bought pure gluconic acid strongly inhibited Ggt. Transposon mutants of Ps. str. AN5, which had lost take-all biocontrol, did not produce D-gluconic acid. However, when complemented with wild-type genomic DNA containing the corresponding ORF region, gluconic acid production was restored along with take-all biocontrol. Gluconic acid production was detected in the wheat rhizosphere with the wild-type parent strain but not with a transposon mutant strain which had lost biocontrol. These results suggest that D-gluconic acid produced by Ps. str. AN5 is a significant antifungal agent in the biocontrol of take-all on wheat roots. Gluconic acid is a very different antifungal metabolite to the antibiotics identified previously, such as PCA and DPG. Gluconic acid is known to be a strong chelating agent and a strong acid (Milson and Meers, 1985). This suggests that local changes in pH leading to a more acidic environment, caused by Ps. str. AN5 producing gluconic acid, is responsible for the ability of the bacteria to suppress the take-all pathogen in biocontrol. This is consistent with the results observed with PCA and DPG in inhibiting Ggt. The acidic form of PCA is active against Ggt. In contrast, the nonacidic form, phenazine-1-carboxylate, is inactive against Ggt (Brisbane et al., 1987; Chin-A-Woeng et al., 1998). DPG is more acidic in nature than monoacetylphloroglucinol, and also correspondingly shows significantly more antifungal activity against Ggt (Shanahan et al., 1993). The above proposition is also supported by the observation that the take-all disease is inhibited by low pH soils, and that incidence of take-all is high in alkaline soils (Murray et al., 1987). However, wheat root exudates do contain some acidic compounds (Wu et al., 2001), so further work is needed to determine the precise contribution that gluconic acid makes to pH of the wheat rhizosphere. Currently, the only known common property of the three take-all antifungal agents (PCA, DPG and D-gluconic acid) is that they are acidic and have the potential to lower pH in the environment they grow. Therefore, we suggest that the ability of these antifungal agents to inhibit the take-all fungus must be, at least in part, due to their ability to lower pH in the wheat rhizosphere.

3. Experimental 3.1. Media and growth conditions The following media were used for growth of bacterial strains. Luria agar (L agar) – 10 g tryptone, 5 g yeast extract, 5 g NaCl, 15 g agar in 1000 ml of distilled water. Pigment production medium (PPM) – 20 g peptone, 20 g glycerol, 5 g NaCl, 1 g KNO3, 15 g agar in 1000 ml of dis-

tilled water at pH 7.0. King’s B medium (KBM) – 20 g proteose peptone, 15 ml glycerol, 1.9 g K2HPO4 Æ 3H2O, 1.5 g MgSO4 Æ 7H2O in 1000 ml of distilled water. Potato dextrose agar (PDA) – 32 g of potato dextrose agar, 5 g agar in 1000 ml distilled water. Malt agar (MA) – 15 g of malt extract, 17 g of agar in 1000 ml distilled water. Nutrient agar (NA) – nutrient agar 20 g, yeast extract 5 g, agar 5 g in 1000 ml distilled water. Potato dextrose broth (PDB) – 24 g of potato dextrose broth in 1000 ml distilled water. Pontiac Broth (PB) – 400 g pontiac potatoes were washed. The potatoes were cut into approximately 1 cm2 cubes without peeling, and boiled in distilled water till soft. The solution was allowed to cool down slightly and then strained through four layers of cheesecloth and the extract was made up to 1 L. The filtered solution was then autoclaved before use. Bromocresol purple was added to media at a concentration of 0.0075% and pH adjusted to pH 7 by addition of 1 M Potassium hydroxide solution before autoclaving. Top agar – 5 g potato dextrose agar, in 300 ml distilled water was solidified with 1.3% agar for the agar overlay bioassay on PDA plates and thin layer chromatographic (TLC) plates. E. coli bacteria used in this study (refer to Section 3.2) were grown overnight in solid medium, or liquid medium with vigorous shaking, at 37 C. Pseudomonas bacteria used in this study (refer to Section 3.2) were grown for 2 d in solid medium, or liquid medium with vigorous shaking, at 25 C. The take-all fungi used for this study were Australian isolates C3 and QW1 (P.T.W. Wong collection). For identification of the antifungal compound, take-all strain C3 was used in agar overlay bioassays on plates and TLCs. For plate bioassays both C3 and QW1 strains were used independently and the results observed were identical so they are not differentiated in the results section. Take-all fungus was grown on PDA or PDB at 18 C for 8– 10 d and stored at 4 C. Media components were purchased from DifcoBacto Laboratories and Sigma–Aldrich. 3.2. Bacteria, fungi and plasmids Nayudu et al. (1994) previously generated a spontaneous rifampicin-resistant derivative of Ps. str. AN5 (AN5rif) and a spectinomycin-resistant derivative of Ps. str. AN5 (AN5sp) which were used for isolation of the antifungal agent in this study. Ps. str. AN5rif was grown on NA supplemented with 100 lg/ml rifampicin. Ps. str. AN5sp was grown on NA supplemented with 250 lg/ml spectinomycin. The Tn5gus mutant strains used in this study, Ps. str. AN5MN1 and Ps. str. AN5MN2, were derived from Ps. str. AN5 by suicide mutagenesis in tri-parental mating with E. coli K-12 using the suicide vector pRK600 Tn5gusA1 (Sharma and Signer, 1990). These strains have completely lost biocontrol against the take-all disease in plate assays and pot trials (Nayudu et al., 1994). Transposon Tn5gus mutants were grown on NA supplemented with kanamycin hydrochloride 250 lg/ml. The plasmid pLAFR1 (Friedman et al., 1982) encodes tetracycline resistance. Therefore,

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

601

selection for the cosmid was made by growing these strains on medium supplemented with tetracycline (E. coli K-12– 20 lg/ml oxytetracycline hydrochloride in L agar). The cosmid (present in E. coli K-12 DH5a) was transferred into Ps. str. AN5MN1 using triparental mating with E. coli K12 SM10 (helper strain) with selection being imposed for tetracycline resistance (Ps. str. AN5 – 60 lg/ml oxytetracycline hydrochloride in NA). Microbial genetic methods have been described previously (Nayudu and Holloway, 1981; Moore et al., 1983; Nayudu and Rolfe, 1987). Antibiotics were purchased from Sigma–Aldrich.

3.4. Plate bioassays to determine fungal inhibition

3.3. PCR and plasmid construction

This bioassay was specifically developed in this study to determine Ggt inhibition by bacterial extracts and pure compounds. Ggt was grown in 100 ml PDB for 10 d at 18 C without shaking. The resultant fungal growth was macerated in a Waring blender and then diluted 1 in 4 using potato dextrose top agar cooled to 42 C. For plate agar overlay bioassays, this top agar was poured on top of a normal PDA plate. A standard 10 ll aliquot of bacterial extract or pure compound was spotted onto the agar overlay bioassay plates and dried. Plates were incubated for 3–4 d at 18 C. Incubation of uninoculated plates for 3–4 d at 18 C led to confluent growth of the fungus in the top agar. Clearing zones where the extract or pure compound was spotted on the agar overlay plate indicted inhibition of the fungus. For TLC agar overlay bioassays, this top agar was poured on top of a TLC plate which had tape around it to hold the top agar. This was done on developed TLC plates which had been air dried to evaporate the solvent. TLC plates were incubated at 18 C for 3–4 d. Inhibition of Ggt led to a clearing zone on the TLC plate, compared to confluent growth of Ggt over the rest of the plate. All assay treatments were replicated. Each assay has been repeated independently on at least two separate occasions. The results reported were consistently observed on all occasions.

Nayudu et al. (1994) showed that there are two regions involved in antifungal agent production in Ps. str. AN5. They constructed a cosmid bank of Ps. str. AN5 and were able to complement transposon mutants in each of these regions with Ps. str. AN5 cosmids carrying wild-type regions leading to restoration of biocontrol. Cosmid pLAFR1-Mur1 has an approximate 23 kb insertion of the Ps. str. AN5 genome wild-type region corresponding to the site of insertion of the transposon in Ps. str. AN5MN1 (Nayudu et al., 1994). A 3.0 kb fragment encompassing an ORF region (corresponding to the site of insertion of Ps. str. AN5MN1) was generated by PCR using cosmid pLAFR3-Mur1 DNA with the primers 8 forward 1 (5 0 -AGCGGGTCAGCTTTTTACTG-3 0 ) and 8 reverse 1 (5 0 -GGAACGATCAACAAGCTC-3 0 ). Amplification was carried out using a Qiagen Mutiplex PCR kit (Cat. No. 206143) from Qiagen. There was an initial denaturation step at 95 C for 15 min. Amplification occurred at 95 C (30 s) for denaturation, 57 C (45 s) for annealing and 72 C (240 s) for extension. This was repeated for 30 cycles. A 3.0 kb PCR product was separated using 1.0% agarose gel electrophoresis, then the product was isolated from the gel and purified using a QIAquick gel purification kit (Cat. No. 28704) from Qiagen. The isolated 3.0 kb fragment was then cloned into a pGEM T-easy vector (Promega Corporation). This cloned insert did not have any EcoRI sites, so the EcoRI sites on either side of the 3.0 kb PCR cloned fragment (located on the pGEM T-easy vector) were used to liberate the fragment. EcoRI digestion products were separated by agarose gel electrophoresis. The 3.0 kb EcoRI–EcoRI digest product was isolated from the gel, purified using a QIAquick gel purification kit, and cloned into the EcoRI site, (located in the middle of the chloramphenicol gene) of plasmid pSUP106 (Priefer et al., 1985). Ligation was carried with T4 DNA ligase (Promega Corp.) at 16 C for 12 h and transformed into DH5a-E competent cells (Invitrogen). This plasmid, pSUP106::ORF MN1, was transferred into Ps. str. AN5MN1 using the same methods described above (i.e., triparental mating) using tetracycline resistance for selection. Molecular biology methods have been described previously (Nayudu and Holloway, 1981; Nayudu and Rolfe, 1987; Nayudu et al., 1994).

Inhibition of Ggt by different Pseudomonas strains was determined in agar plate bioassays using PDA (Poplawsky et al., 1988). Antifungal activity of pure compounds was tested by supplementing PDA with each compound and then carrying out an agar plate bioassay. Growth was observed after incubation at 18 C for approximately 8–10 d. 3.5. Agar overlay bioassays to test fungal inhibition

3.6. Quantification of acid production The amount of acid produced by Ps. str. AN5 and its mutant derivative strains was estimated using titration according to the methods of Schleissner et al. (1997). Ps. str. AN5 and the mutants Ps. str. AN5MN1 and Ps. str. AN5MN2 were grown in 100 ml PB for 2 d with shaking at 25 C. Then 100 ml of fresh PB broths were inoculated with 100 ll from these stationary phase PB cultures and grown for a further 15–18 h under the same conditions. Bacterial cells were pelleted in a bench centrifuge at 6000 rpm for 10 min. The bacterial pellets were washed twice with sterile water and resuspended in 100 ml glucose medium containing 5.0 g/L glucose at pH 7. This was incubated on a shaker at 25 C. Aliquots of 5 ml of the culture were collected at different time intervals and cells pelleted. The supernatant was used for titration. After measuring the initial pH of solutions, 5 ml of the supernatant

602

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

was titrated with 0.001 N sodium hydroxide (NaOH) using phenolphthalein as an indicator (Whistler and Schweiger, 1959; Vogel, 1978). The titration was repeated three times, and the amount of acid detected in each case is presented as an average of these results. 3.7. Extraction of antifungal compounds Extractions of antifungal metabolites from bacterial strains were from agar plates. Each extraction was done in duplicate and repeated at least twice. The results reported were consistently observed on all occasions. A modification of the method of Aszalos et al. (1968) was used to determine if Ps. str. AN5 produced hydrophilic antifungal compounds. Ps. str. AN5rif , Ps. str. AN5sp and its transposon mutant derivatives Ps. str. AN5MN1 and Ps. str. AN5MN2 were grown on PDA for 5 d at 25 C for extraction of antifungal compounds. Twenty agar plates of each strain (including bacterial growth) were cut into pieces of approximately 1 cm2 and extracted by shaking on a rotary shaker for 1 h with 1 L 60% aqueous isopropanol (60:40 water:isopropanol). The extracts were filtered through cheesecloth and reduced in vacuo to evaporate isopropanol. The reduced extract was centrifuged at 3832g for 20 min. An equal volume of acetone was added to the supernatant in a sealed bottle and left overnight at 4 C to precipitate proteins. The solution was again centrifuged and the supernatant was then reduced in vacuo to evaporate acetone. The remaining aqueous solution was freeze-dried and used for further analysis. To determine if Pseudomonas strains produced phenazine they were grown on PPM media and then extracted according to the method of Rosales et al. (1995). To detect phloroglucinol based metabolites, the bacterial strains were grown on MA and a similar method of extraction was carried out to that reported by Keel et al. (1992). P. fluorescens strain Pf-5 (Nowakthompson et al., 1994) was used to obtain DPG. PCA was purchased from Sigma–Aldrich (Cat. No. S998907). Comparison of the extracts in both cases was done with TLC, NMR and mass spectroscopy analysis. 3.8. Differentiation of antifungal compound(s) from crude extracts using thin layer chromatography Ten microliter of suspensions of crude extracts were applied to silica gel GF254 TLC plates at 2 cm intervals. TLC glass plates with silica gel F254 (0.25 mm thickness) used in this study were purchased from Merck (Cat. No. 1.05715). For DPG a chloroform:methanol solvent (9:1) and toluene:acetone (4:1) solvent system was used (Strunz et al., 1978). Ps. str. AN5 extracts were trialed with chloroform:methanol (19:1) and acetonitrile:methanol:water (1:1:1) solvent systems. Reverse-phase C18 TLC for PCA separation used an acetonitrile:methanol:water (1:1:1) solvent system (Pfender et al., 1993; Rosales et al., 1995). The solvent system n-propanol:ethyl acetate:water (5:2:3)

was able to separate the biologically active compound(s) of Ps. str. AN5. TLC plates were run until the solvent was about 1 cm from the upper rim. The plates were dried at room temperature and visualized under UV light. For isolation of antifungal compounds, the bands were scratched from the TLC plate and the compounds extracted using 50% aqueous methanol. The biological activity of extracted compounds was determined in agar overlay bioassays. Individual TLC runs were always done in duplicate and the experiment repeated at least twice in the initial screening experiments to determine the appropriate solvents. Further experiments were done at least five times on separate occasions. The results reported were consistently observed on all occasions. 3.9. Purification of antifungal compound(s) from crude extracts using silica columns Silica columns were prepared according to the method described by Still et al. (1978). 100 mg of freeze-dried crude extracts from Ps. str. AN5sp, Ps. str. AN5rif, Ps. str. AN5MN1 and Ps. str. AN5MN2 were dissolved in 5 ml n-propanol:ethyl acetate:water (5:2:3) and applied to silica columns separately. Fractions of 5 ml at a flow rate 1 ml/ min were collected. The column fractions were analyzed by spotting a 10 ll sample from each fraction on 20 cm · 20 cm TLC plates. Biological activity of these fractions was determined using agar overlay bioassays. The fractions containing similar active bands on TLC plates (i.e., Ps. str. AN5) were pooled for further analysis. This was repeated independently on two separate occasions. The results reported were consistently observed on all occasions. 3.10. Extraction of antifungal metabolite from wheat rhizosphere A sterile agar plate assay was used for growth of wheat plants. Herridge’s (H) medium was prepared as described by Delves et al. (1986). H medium does not contain any carbon source for normal bacterial growth. Wheat seeds were surface sterilized using 12% sodium hypochlorite solution with gentle shaking for 5 min. Seeds were thoroughly washed with sterile water and then arranged on H agar plates. The plates were incubated vertically in the dark at 25 C for two days. Germinated wheat seedlings were carefully removed and treated with the bacterial inoculum grown in PB, or left untreated. There is no gluconic acid (or other antifungal agents) produced by Ps. str. AN5 in PB (data not shown). Bacterial treatments used were Ps. str. AN5rif and the mutant strain Ps. str. AN5MN1. Four treated seedlings were fixed with 2% agar on H agar on a large Petri dish (20 cm diameter). These plates were then incubated vertically in a 12 h light/dark cycle and a 18 C/12 C regime. The wheat seedlings were grown for two weeks. The agar directly below and adjacent to where the wheat roots were growing was cut into strips and

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

extracted with 1 L 60% aqueous isopropanol as described above. The extracts from plates inoculated with Ps. str. AN5, Ps. str. AN5MN1 and uninoculated wheat were tested for take-all fungal inhibition in agar overlay bioassays. The extracts were freeze-dried, concentrated and subjected to further analysis by mass spectrometry. The results presented are from duplicate analyses. 3.11. Structure elucidation using NMR spectroscopy 1

H NMR spectra were obtained using a Varian Inova 500 MHz spectrometer in D2O. 3-Trimethyl silyl propane sulphonic acid (TSP) was used as an internal standard at 0 ppm. 13C NMR spectra were recorded on the same spectrometer at 100 MHz in D2O. 3.12. Structure elucidation using GC/MS analysis To a small amount of freeze-dried extract in a Reactivial was added dry pyridine (25 ll) and Regisil (10% chlorotrimethylsilane in N,O-bis(trimethylsilyl)trifluoroacetamide, Regis Chemical Company, USA; 25 ll). This was done in a dry incubator at 25 C. The vial was heated to 80 C for 30 min after sealing, then cooled and centrifuged. The same silylating procedure was used for pure compounds. The solution of the silylated compounds (1 ll) was injected directly onto the GC/MS system. GC/MS analyses were performed on a Hewlett-Packard 5890 gas chromatogram interfaced to a HP 5970 mass selective detector. Gas chromatography was carried out using a HP-1 fused silica capillary column (12.5 m · 0.20 mm i.d., 0.33 lm film thickness bonded methyl silicone stationary phase) in the splitless mode. Conditions used were: injector, 250 C; transfer line, 250 C; oven programme, 100–250 C at 10 degrees/min, hold 10 min. Mass spectra were obtained by electron impact (EI) ionization at 70 eV. Full scan spectra were recorded by scanning from m/z 100 to m/z 600.

Acknowledgements Rajvinder Kaur was supported by an ANU Women’s Re-entry Scholarship and a Grains Research and Development Corporation (GRDC) Scholarship. This research was only possible because of the long-term research support of the Australian Grain Growers Association (GGA) and the GRDC. We would like to acknowledge Christian Samundsett for help in providing bioassay gluconic acid inhibition data and Geraldine Pons for help in cloning. Dr. Percy Wong is thanked for providing the take-all fungi used in this study. We thank Terry Murphy and Christian Samundsett for editorial review of the manuscript. A special thanks to Dr. Andrew Franklin for his time and help in specific editorial review of the manuscript, which improved it significantly. We would like to thank Dr. Bart Eschler for help with chemical drawings.

603

References Aszalos, A., Davis, S., Frost, D., 1968. Classification of crude antibiotics by instant thin-layer chromatography (ITLC). Journal of Chromatography 37, 487–498. Babukhan, S., Yeo, T.C., Martin, W.L., Duron, M.R., Rogers, R.D., Goldstein, A.H., 1995. Cloning of a mineral phosphate-solubilizing gene from Pseudomonas cepacia. Applied and Environmental Microbiology 61, 972–978. Brisbane, P.G., Janik, L.J., Tate, M.E., Warren, R.O., 1987. Revised structure for the phenazine antibiotic from Pseudomonas fluorescens 2– 79 (NRRL B-15 132). Antimicrobial Agents and Chemotherapy 31, 1967–1971. Chin-A-Woeng, T.F.C., Bloemberg, G.V., van der Bij, A.J., van der Drift, K.M.G.M., Schripsema, J., Kroon, B., Scheffer, R.J., Keel, C., Bakker, P.A.H.M., Tichy, H., de Bruijn, F.J., Thomas-Oates, J.E., Lugtenberg, B.J.J., 1998. Biocontrol by phenazine-1-carboxamide-producing Pseudomonas chlororaphis PCL1391 of tomato root rot caused by Fusarium oxysporum f. sp. radicislycopersii. Molecular Plant–Microbe Interactions 11, 1069–1077. Cook, R.J., Thomashow, L.S., Weller, D.M., Fujimoto, D., Mazzola, M., Bangera, G., Kim, D., 1995. Molecular mechanisms of defense by rhizobacteria against root disease. Proceedings of the National Academy of Sciences of the United States of America 92, 4197– 4201. Delves, A.C., Mathews, A., Day, D.A., Carter, A.S., Carroll, B.J., Gresshoff, P.M., 1986. Regulation of the soybean-Rhizobium nodule symbiosis by shoot and root factors. Plant Physiology 82, 588–590. Fakhouri, W., Walker, F., Vogler, B., Armbruster, W., Buchenauer, H., 2001. Isolation and identification of N-mercapto-4-formylcarbostyril, an antibiotic produced by Pseudomonas fluorescens. Phytochemistry 58, 1297–1303. Friedman, A.M., Long, S.R., Brown, S.E., Buikema, W.J., Ausubel, F.M., 1982. Construction of a broad host range cosmid cloning vector and its use in the genetic analysis of Rhizobium mutants. Gene 18, 289–296. Goodwin, P.M., Anthony, C., 1998. The biochemistry, physiology and genetics of PQQ and PQQ-containing enzymes. Advances in Microbial Physiology 40, 1–80. Gurusiddaiah, S., Weller, D.M., Sarkar, A., Cook, R.J., 1986. Characterization of an antibiotic produced by a strain of Pseudomonas fluorescens inhibitory to Gaeumannomyces-graminis Var tritici and Pythium spp. Antimicrobial Agents and Chemotherapy 29, 488–495. Gutterson, N.I., Layton, T.J., Ziegle, J.S., Warren, G.J., 1986. Molecular cloning of genetic determinants for inhibition of fungal growth by a fluorescent Pseudomonad. Journal of Bacteriology 165, 696–703. Gutterson, N.I., Ziegle, J.S., Warren, G.J., Layton, T.J., 1988. Geneticdeterminants for catabolite induction of antibiotic biosynthesis in Pseudomonas fluorescens Hv37A. Journal of Bacteriology 170, 380– 385. Haltrich, D., Nidetzky, B., Miemietz, G., Gollhofer, D., Lutz, S., Stolz, P., Kulbe, K.D., 1996. Simultaneous enzymatic synthesis of mannitol and gluconic acid. 1. Characterization of the enzyme system. Biocatalysis and Biotransformation 14, 31–45. Hellmut, J., Werner, F., Walter, F., Hans, W., 1990. Thin Layer Chromatography: Reagents and Detection Methods. VCH, NY, USA. Horton, D., Walaszek, Z., Ekiel, I., 1983. Conformations of deuterium– gluconic, deuterium–mannonic, and deuterium–galactonic acids in solution, as determined by NMR-spectroscopy. Carbohydrate Research 119, 263–268. James, D.W., Gutterson, N.I., 1986. Multiple antibiotics produced by Pseudomonas fluorescens Hv37a and their differential regulation by glucose. Applied and Environmental Microbiology 52, 1183–1189. Keel, C., Schnider, U., Maurhofer, M., Voisard, C., Laville, J., Burger, U., Wirthner, P., Haas, D., Defago, G., 1992. Suppression of root diseases by Pseudomonas fluorescens CHA0 – importance of the bacterial secondary metabolite 2,4-diacetylphloroglucinol. Molecular Plant– Microbe Interactions 5, 4–13.

604

R. Kaur et al. / Phytochemistry 67 (2006) 595–604

Keel, C., Weller, D.M., Natsch, A., Defago, G., Cook, R.J., Thomashow, L.S., 1996. Conservation of the 2,4-diacetylphloroglucinol biosynthesis locus among fluorescent Pseudomonas strains from diverse geographic locations. Applied and Environmental Microbiology 62, 552–563. Kumar, R.S., Ayyadurai, N., Pandiaraja, P., Reddy, A.V., Venkateswarlu, Y., Prakash, O., Sakthivel, N., 2005. Characterization of antifungal metabolite produced by a new strain Pseudomonas aeruginosa PUPa3 that exhibits broad-spectrum antifungal activity and biofertilizing traits. Journal of Applied Microbiology 98, 145–154. Lee, J.Y., Moon, S.S., Hwang, B.K., 2003. Isolation and antifungal and antioomycete activities of aerugine produced by Pseudomonas fluorescens strain MM-B16. Applied and Environmental Microbiology 69, 2023–2031. Ligon, J.M., Hill, D.S., Hammer, P.E., Torkewitz, N.R., Hofmann, D., Kempf, H.J., van Pee, K.H., 2000. Natural products with antifungal activity from Pseudomonas biocontrol bacteria. Pest Management Science 56, 688–695. Magnuson, J.K., Lasure, L.L., 2004. Organic acid production by filamentous fungi. In: Tkacz, J. (Ed.), Advances in Fungal Bio/ technology for Industry, Agriculture and Medicine. Kluwer/Plenum Academic Publishers, Norwell, MA, USA. Maurhofer, M., Keel, C., Haas, D., De´fago, G., 1994. Pyoluteorin production by Pseudomonas fluorescens strain CHA0 is involved in the suppression of Pythium damping-off of cress but not of cucumber. European Journal of Plant Pathology 100, 221–232. Merkey, S.P., Urban, W.G., Levine, S.P., 1974. Mass Spectra of Compounds of Biological Interest. USAEC Technical Information Centre, Oak Ridge, TN, USA. Milson, P.E., Meers, J.L., 1985. Gluconic and itaconic acids. In: MooYoung, M.M. (Ed.), Comprehensive Biotechnology, vol. 3. Pergamon Press, Oxford, UK, pp. 681–700. Moore, A.T., Nayudu, M., Holloway, B.W., 1983. Genetic mapping in Methylophilus methylotrophus AS1. Journal of General Microbiology 129, 785–799. Murray, G.M., Scott, B.J., Hochman, Z., Butler, B.J., 1987. Failure of liming to increase grain-yield of wheat and Triticale in acid soils may be due to the associated increase in incidence of take-all (Gaeumannomyces graminis Var tritici). Australian Journal of Experimental Agriculture 27, 411–417. Nayudu, M., Holloway, B.W., 1981. Isolation and characterization of RPlasmid variants with enhanced chromosomal mobilization ability in Escherichia coli K-12. Plasmid 6, 53–66. Nayudu, M., Rolfe, B.G., 1987. Analysis of R-Primes demonstrates that genes for broad host range nodulation of Rhizobium strain NGR234 Are dispersed on the Sym plasmid. Molecular and General Genetics 206, 326–337. Nayudu, M., Groom, K.A.E., Fernance, J., Wong, P.T.W., Turnbull, K., 1994. The genetic nature of biological control of the take-all fungus by Pseudomonas. In: Ryder, M.H., Stephen, P.M., Bowen, G.D. (Eds.), Improving Plant Productivity with Rhizosphere Bacteria. CSIRO, Australia, pp. 122–124. Nowakthompson, B., Gould, S.J., Kraus, J., Loper, J.E., 1994. Production of 2,4-diacetylphloroglucinol by the biocontrol agent Pseudomonas fluorescens Pf-5. Canadian Journal of Microbiology 40, 1064–1066. Paulitz, T., Nowak-Thompson, B., Gamard, P., Tsang, E., Loper, J., 2000. A novel antifungal furanone from Pseudomonas aureofaciens, a biocontrol agent of fungal plant pathogens. Journal of Chemical Ecology 26, 1515–1524. Pedras, M.S.C., Ismail, N., Quail, J.W., Boyetchko, S.M., 2003. Structure, chemistry, and biological activity of Pseudophomins A and B, new cyclic lipodepsipeptides isolated from the biocontrol bacterium Pseudomonas fluorescens. Phytochemistry 62, 1105–1114.

Pfender, W.F., Kraus, J., Loper, J.E., 1993. A genomic region from Pseudomonas fluorescens Pf-5 required for pyrrolnitrin production and inhibition of Pyrenophora tritici repentis in wheat-straw. Phytopathology 83, 1223–1228. Pierson, L.S., Thomashow, L.S., 1992. Cloning and heterologous expression of the phenazine biosynthetic locus from Pseudomonas aureofaciens 30–84. Molecular Plant–Microbe Interactions 5, 330–339. Poplawsky, A.R., Peng, Y.F., Ellingboe, A.H., 1988. Genetics of antibiosis in bacterial strains suppressive to take-all. Phytopathology 78, 426–432. Priefer, U.B., Simon, R., Puhler, A., 1985. Extension of the host range of Escherichia coli vectors by incorporation of RSF1010 replication and mobilization functions. Journal of Bacteriology 163, 324–330. Rosales, A.M., Thomashow, L., Cook, R.J., Mew, T.W., 1995. Isolation and identification of antifungal metabolites produced by rice-associated antagonistic Pseudomonas spp. Phytopathology 85, 1028–1032. Rovira, A.D., Kollmorgen, J.F., MacNish, G.C., Sivasithamparam, K., Wong, P.T.W., 1991. Take-all. In: Burgess, L.W. (Ed.), Soilborne Diseases of Wheat in the Australian Environment. Oxford University, Sydney, Australia. Schleissner, C., Reglero, A., Luengo, J.M., 1997. Catabolism of D-glucose by Pseudomonas putida U occurs via extracellular transformation into D-gluconic acid and induction of a specific gluconate transport system. Microbiology-UK 143, 1595–1603. Shanahan, P., Glennon, J.D., Crowley, J.J., Donnelly, D.F., O’Gara, F., 1993. Liquid chromatographic assay of microbially derived phloroglucinol antibiotics for establishing the biosynthetic route to production and the factors affecting their regulation. Analytica Chimica Acta 272, 271–277. Sharma, S.B., Signer, E.R., 1990. Temporal and spatial regulation of the symbiotic genes of Rhizobium meliloti in planta revealed by transposon Tn5-gusA. Genes and Development 4, 344–356. Still, W.C., Kahn, M., Mitra, A., 1978. Rapid chromatographic technique for preparative separations with moderate resolution. Journal of Organic Chemistry 43, 2923–2925. Strunz, G.M., Wall, R.E., Kelly, D.J., 1978. Phloroglucinol derivatives from Aeromonas hydrophila. Journal of Antibiotic (Tokyo) 31, 1201–1202. Thomashow, L.S., Weller, D.M., 1990. Role of antibiotics and siderophores in biocontrol of take-all disease of wheat. Plant and Soil 129, 93–99. Thomashow, L.S., Weller, D.M., Bonsall, R.F., Pierson, L.S., 1990. Production of the antibiotic phenazine-1-carboxylic acid by fluorescent Pseudomonas species in the rhizosphere of wheat. Applied and Environmental Microbiology 56, 908–912. Tsai, C.S., Ye, H.G., Shi, J.L., 1995. C-13 NMR-studies and purification of gluconate pathway enzymes from Schizosaccharomyces Pombe. Archives of Biochemistry and Biophysics 316, 155–162. Turner, J.M., Messenger, A.J., 1986. Occurrence, biochemistry and physiology of phenazine pigment production. Advances in Microbial Physiology 27, 211–275. Velizarov, S., Beschkov, V., 1994. Production of free gluconic acid by cells of Gluconobacter oxydans. Biotechnology Letters 16, 715–720. Vogel, A.I., 1978. Titrimetric Analysis. Wiley, New York. Weller, D.M., 1988. Biological-control of soilborne plant-pathogens in the rhizosphere with bacteria. Annual Review of Phytopathology 26, 379– 407. Whistler, R.L., Schweiger, R., 1959. Preparation of D-arabinose from Dglucose with hypochlorite. Journal of American Chemical Society 81, 5190–5191. Wu, H.W., Haig, T., Pratley, J., Lemerle, D., An, M., 2001. Allelochemicals in wheat (Triticum aestivum L.): cultivar difference in the exudation of phenolic acids. Journal of Agricultural and Food Chemistry 49, 3742–3745.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 605–609 www.elsevier.com/locate/phytochem

Newbouldiaquinone A: A naphthoquinone–anthraquinone ether coupled pigment, as a potential antimicrobial and antimalarial agent from Newbouldia laevis Kenneth Oben Eyong a, Gabriel Ngosong Folefoc a,*, Victor Kuete c, Veronique Penlap Beng c, Karsten Krohn b,*, Hidayat Hussain b, Augustin Ephram Nkengfack a, Michael Saeftel d, Salem Ramadan Sarite d, Achim Hoerauf b

d

a Department of Organic Chemistry, Yaounde University I, P.O. Box 812, Yaounde, Cameroon Department of Chemistry, University of Paderborn, Warburger Straße 100, 33098 Paderborn, Germany c Department of Biochemistry, Yaounde University I, P.O. Box 812, Yaounde, Cameroon d Institute of Medical Parasitology, University of Bonn, Sigmund Freud Str. 25, 53105 Bonn, Germany

Received 12 September 2005; received in revised form 9 December 2005 Available online 26 January 2006

Abstract The study of the chemical constituents of the roots of Newbouldia laevis (Bignoniaceae) has resulted in the isolation and characterization of a naphthoquinone–anthraquinone coupled pigment named newbouldiaquinone A (1) together with 14 known compounds: apigenin, chrysoeriol, newbouldiaquinone, lapachol, 2-methylanthraquinone, 2-acetylfuro-1,4-naphthoquinone, 2,3-dimethoxy-1,4benzoquinone, oleanolic acid, canthic acid, 2-(4-hydroxyphenyl)ethyl triacontanoate, newbouldiamide, 5,7-dihydroxydehydroiso-alapachone, b-sitosterol, and b-sitosterol glucopyranoside. The structure elucidation of the isolated compounds was established based on spectroscopic studies, notably of the 2D NMR spectra. The antimalarial activity of compound (1) against Plasmodium falciparum in vitro shows moderate chemo suppression of parasitic growth. Its antimicrobial activity against a wide range of microorganisms was 13- and 24-fold more active against Candida gabrata and Enterobacter aerogens than the reference antibiotics nystatin and gentamycin. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Newbouldia laevis; Bignoniaceae; Ether-coupled naphthoquinone–anthraquinone; Antimalarial and antibacterial activity

1. Introduction Newbouldia laevis SEEM. or ‘‘Boundary Tree’’ is a medium sized angiosperm in the Bignoniaceae family. It is native to tropical Africa, and grows to a height of about 10 m (Okeka, 2003). The species N. laevis is widely used in African folk medicine for the treatment of several diseases such as an astringent in diarrhea and dysentery. It is also employed in the treatment against worms, malaria, sexually transmitted disease, and in the reduction of dental caries (Eyong et al., 2005). However, little is known of its *

Corresponding authors. Tel.: +237 7510281; fax: +237 2226018. E-mail address: [email protected] (G.N. Folefoc).

0031-9422/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.019

antimalarial properties despite the fact that this species is widely used by local healers to treat malaria. In an earlier study, we reported the isolation and structure elucidation of newbouldiaquinone, a naphthoquinone-anthraquinone C–C coupled pigment (Eyong et al., 2005). To obtain the minor compounds, the roots, seeds and leaves of Newbouldia laevis were again collected and extracted with CH2Cl2:MeOH (1:1). The root fraction was subjected to different chromatographic procedures resulting in the isolation and structure elucidation of newbouldiaquinone A (1), another naphthoquinone–anthraquinone pigment from N. laevis, but coupled via an ether bridge rather than a C–C bond (Eyong et al., 2005). We now report on the structure elucidation and antibacterial potency of 1 against a wide

606

K.O. Eyong et al. / Phytochemistry 67 (2006) 605–609

range of microorganisms and on antimalarial screening of this compound together with other quinones and derivatives from this plant.

2. Results and discussion The MeOH:CH2Cl2 (1:1) extract of the leaves, stem and root of N. laevis was fractionated by silica gel column chromatography to give several fractions, which were further chromatographed on silica gel to give a new naphthoquinone–anthraquinone dimeric quinone, newbouldiaquinone A (1) together with 14 known compounds. The latter group of compounds were identified by comparison of their spectroscopic data with the literature data as apigenin (Hiermann and Kartnig, 1978), chrysoeriol (Harrison and Kulshreshtha, 1984), newbouldiaquinone (Eyong et al., 2005), lapachol (Khan and Mlungwana, 1999), 2-methylanthraquinone (Boivert and Brussard, 1988), 2-acetylfuro-1,4-naphthoquinone (Lopes et al., 1984), 2,3-dimethoxy-1,4-benzoquinone (Matsumoto and Kobayashi, 1985), oleanolic acid (Ikuta and Hokawa, 1988), canthic acid (Chatterjee et al., 1979), 2-(4-hydroxyphenyl)ethyl triacontanoate (Ali and Houghton, 1999), newbouldiamide (Eyong et al., 2005), 5,7-dihydroxydehydroiso-a-lapachone (Gafner et al., 1996), b-sitosterol (Schuhr et al., 2003), and b-sitosterol glucopyranoside (Seo et al., 1978). The a-lapachone and b-lapachone were synthesized from lapachol using the Hooker procedure (Hooker, 1936). Newbouldiaquinone A (1) was obtained as a yellow powder with m.p. 260 °C. The UV spectrum of 1 exhibited absorption maxima at 270, 308 and 388 nm, suggesting a naphthoquinone derivative (Gorman et al., 2003). This was supported by IR bands at 1647 cm
1 for carbonyl absorption, a broad signal at 3404 cm
1 for a non-chelated hydroxyl group and also by a sharp signal at 1236 and 1036 cm
1 for an ether function. Analysis of the chemical ionisation mass spectrum (CI-MS) gave a molecular ion at m/z 411.1 [M
H]+, corresponding to the molecular formula C25H14O6, supported by the 1H NMR, 13C NMR and DEPT analysis. The mass spectrum also showed peaks at m/z 394.1 [M
O]+, 348 [M
O
CO
H2O]+ and 322 [M
H2O
2CO
CH2]+, exhibiting loss of methyl, water and multiple loss of carbon monoxide, suggesting the presence of hydroxyl, methyl, and carbonyl

O

groups. This fragmentation pattern is typical for hydroxyanthraquinone and/or napthoquinone (Aguinaldo et al., 1993; Singh and Singh, 1986). The 1H NMR in CDCl3 (see Section 3) of newbouldiaquinone A (1) had one singlet at d 2.51 of three protons, characteristic of a methyl group attached to an aromatic system (Aguinaldo et al., 1993). It also had 10 strongly deshielded protons, 6 of which appear above d 8.10, characteristic of protons peri to carbonyl groups, suggesting the presence of an anthraquinnone–anthraquinone, anthraquinone–naphthoquinone or naphthoquinone–naphthoquinone ring system. From the proton–proton correlation spectroscopy (COSY), two pairs of four protons were coupling to one another while two protons did not have any coupling suggesting the presence of two pairs of AA 0 BB 0 spin system of four aromatic protons each. The first AA 0 BB 0 4spin system of four aromatic protons at d 8.37–8.32 (2H, m, H-5, H-8), 7.95 (2H, td, J = 8.5, 1.5 Hz, H-6, H-7) along with 8.27 (1H, s, H-4) and 8.18 (1H, s, H-1) indicated that compound 1 contains an anthraquinone, possessing an unsubstituted ring A and a di-substituted ring C at positions 2 and 3 (Chart 1, partial structure A). A second AA 0 BB 0 spin system of four aromatic protons at 8.25 (1H, dd, J = 8.5, 1.5 Hz, H-8 0 ), 8.22 (1H, dd, J = 8.5, 1.5 Hz, H-5 0 ), 7.90 (2H, td, J = 8.5, 1.5 Hz, H-6 0 , H-7 0 ) is attributed to a second partial structure B, i.e., a naphthoquinone, possessing an unsubstituted ring A. The structures of the two fragments were similar to those in newbouldiaquinone (Eyong et al., 2005) and were also confirmed by the analysis of the 13C NMR spectral data (see Section 3). The acetylation of compound 1 gave a monoacetate 2 with M+ at m/z 452 with the addition of one acetyl unit suggesting that it contains only one free hydroxyl group, probably at position 2 0 of the naphthoquinone moiety from biogenetic consideration as well as from HMBC interactions (Fig. 1). Moreover, the signals at d 2.30 in 1H H

H

O

O O

H H O

H

Fig. 1. Important HMBC data for compound 1.

O 1

8 7

9

6

10

2 4

3 O 3'

O O fragment B

H

H H O

5

O

H CH3

H

O OH

fragment A

O

H

1: R = H 2: R = Ac

RO 2'

O 5' 4'

6'

1'

7' 8'

O

Chart 1. Structures of fragments A and B and newbouldiaquinone A (1) and its acetate 2.

K.O. Eyong et al. / Phytochemistry 67 (2006) 605–609

607

Table 1 In vitro activity (parasitemia %) of lapachol, newbouldiaquinone A (1), b-lapachone, and a-lapachone against Plasmodium falciparum Days

Control

Lapachol

Newbouldiaquinone A

b-Lapachone

a-Lapachone

0 1 2 3 4

0.55 0.79 1 1.65 2.17

0.55 0.52 0.937 1.3 1.97

0.55 0.41 0.78 0.9 1.7

0.55 0.37 0.66 0.84 1

0.55 0.5 0.5 0.86 0.957

NMR and at d 21.5 and 170.1 in 13C NMR further confirm the monoacetylation of 1. Thus, the remaining oxygen must be incorporated in an ether linkage. The absence of any quinoidal protons for H-2 0 or H-3 0 that usually occurs at ca. d 6.87 (Hassanean et al., 2000), for naphthoquinones and/or benzoquinones confirms our partial structures. The ether, therefore, links the position C-3 0 of fragment B and position 2 or 3 of fragment A since positions 1 and 4 show signals for peri-hydrogens in the 1H NMR spectrum. The methyl group is thus attached at C-2 of ring C in the anthraquinone due to strong HMBC correlations with carbon C-1, suggesting a coupled anthraquinone–naphthoquinone skeleton at C-3 0 and C-3 by ether linkage. Consequently, the structure of compound 1 was established as 3-(2-hydroxyl-naphthoquinon-3-O-yl)-2-methyl-anthracen-9,10-dione, named newbouldiaquinone A (1). Newbouldiaquinone A (1), lapachol, a-lapachone and blapachone were tested against Plasmodium falciparum in vitro. All of them showed moderate suppression of parasitic growth (Table 1). Table 2 Antimicrobial activity of newbouldiaquinone A (1) and of reference antibiotics Microbial strains

(1) lg/ml (lM)a

RAblg/ml (lM)a

Gram-negative bacteria Entrobacter freundii Enterobacter aerogens Enterobacter cloacae Escherichia coli Klebsiella pneumoniae Morganella morgani Proteus mirabilis Proteus vulgaris Pseudomonas aeruginosa Shigella dysenteriae Shigella flexneri Salmonella typhi

0.31 0.31 0.31 0.31 0.31 0.61 0.31 4.88 9.76 4.88 1.22 4.88

(0.75) (0.75) (0.75) (0.75) (0.75) (1.49) (0.75) (11.9) (23.8) (11.9) (2.97) (11.9)

4.88 9.76 4.88 1.22 2.44 2.44 2.44 1.22 4.88 2.44 2.44 2.44

(9.0) (18) (9.0) (2.25) (4.5) (4.5) (4.5) (2.25) (9.0) (4.5) (4.5) (4.5)

Gram-positive bacteria Bacillus cereus Bacillus megaterium Bacillus stearothermophilus Bacillus subtilis Staphylococcus aureus Streptococcus faecalis

9.76 9.76 9.76 4.88 9.76 9.76

(23.8) (23.8) (23.8) (11.9) (23.8) (23.8)

2.44 4.88 4.88 2.44 4.88 4.88

(4.5) (9.0) (9.0) (4.5) (9.0) (9.0)

Yeasts Candida albicans Candida krusei Candida gabrata

4.88 (11.9) 4.88 (11.9) 0.31 (0.75)

a

4.88 (5.21) 4.88 (5.21) 9.76 (10.4)

MIC: minimal inhibition concentration or the lowest concentration that prevents the growth of the tested pathogens. b RA: reference antibiotics (gentamycin for bacteria, nystatin for yeast).

Newbouldiaquinone A (1) is a powerful antimicrobial agent (Table 2). Very pronounced activities were observed against Gram-negative bacteria with the minimal inhibition concentration (MIC) varying from 0.31 to 9.76 lg/ml (0.75–23.8 lM). The inhibition effect observed was greater than that of the reference antibiotics (RA). However, it appeared to be less active against Gram-positive bacteria. Its antifungal activity was also important with Candida gabrata being the most sensitive yeast. Newbouldiaquinone A (1) is 13- and 24-fold more active against Candida gabrata and Enterobacter aerogens than nystatin and gentamycin, respectively.

3. Experimental 3.1. General experimental procedures 1

H, 2D 1H–1H COSY, 13C, 2D HMQC and HMBC spectra were recorded with a Bruker Avance 500 MHz spectrometer. Chemical shifts are referenced to internal TMS (d = 0) and coupling constants J are reported in Hz. Optical spectra were recorded with a NICOLET 510P FT-IR spectrometer, a UV-2101PC spectrometer, and Perkin–Elmer 241 polarimeter. 3.2. Plant material The plant Newbouldia laevis SEEM. (Bignoniaceae) was collected at Mamfe, South West province of the Republic of Cameroon in December 2004, earlier identified by Mr. Ndive Elias (Plant taxonomist), Botanical Garden, Limbe Cameroon. A voucher specimen (No. 1754/SRFK) has been deposited at the National Herbarium, Yaounde, Cameroon. 3.3. Extraction and isolation Dried and powdered leaves (1 kg), seeds (1 kg), root bark (2.5 kg), and stem bark (3 kg) of N. laevis were separately extracted with a mixture of MeOH:CH2Cl2 (1:1) at room temperature for 24 h. The suspensions were filtered and each filtrate was concentrated under vacuum to give 300, 80, 80 and 90 g of crude residue, respectively. 50 g of the 300 g crude extract of the leaves was separated using sepadex LH-50 to give yellow eluents that were regrouped based on their TLC pattern and purified by column chromatography on silica gel eluting with hexane:EtOAc

608

K.O. Eyong et al. / Phytochemistry 67 (2006) 605–609

(7.5:2.5) to afford Apigenin (40 mg). The crude extract from the seeds (80 g) was subjected to column chromatography (silica gel, hexane, hexane–EtOAc and EtOAc, in order of increasing polarity) yielding 205 fractions (F1–205). Fractions F35–40, from elution with a mixture of hexane–EtOAc (9:1) yielded b-sitosterol (200 mg), and fractions F107–112, which were eluted with hexane:EtOAc (5.5:4.5) and subjected to a second CC, afforded newbouldiamide (80 mg). Column fractions F151–162 [hexane:EtOAc (5.5:4.5)], F170–172 [hexane:EtOAc (4:6)] that was similarly subjected to a second CC yielded chrysoeriol (40 mg) and fractions F190–200 hexane–EtOAc (3.0:7.0) afforded b-sitosterol glucopyranoside (500 mg). Similarly, the crude extract of the root bark (80 g) was also chromatographed on a silica gel column and eluted with gradient mixtures of hexane:EtOAc yielding 200 fractions (F1–200). Fractions F20– F30 were eluted with hexane:EtOAc (9.5:0.5) and afforded 2-acetylfuro-1,4-naphthoquinone (6.5 mg) and 2-methylanthraquinone (80 mg). Fractions F31–35, eluted with a mixture of hexane:EtOAc (9:1), gave lapachol (200 mg); fractions F96–106 (hexane:EtOAc 8.5:1.5) gave newbouldiaquinone (30 mg). Fractions F136–138 on CC using hexane:EtOAc (8:2), gave canthic acid (50 mg), F145–150 hexane:EtOAc (7.5:2.5) afforded a 5,7-dihydroxydehydroiso-a-lapachone while F165-170 eluted with hexane:EtOAc (7:3) afforded newbouldiaquinone A (1). Finally, the crude extract of stem bark (90 g) was subjected to CC using hexane–EtOAc yielding 151 fractions (F1–151). Fractions F17–20 eluted with hexane afforded 2,3-dimethoxy-1,4-benzoquinone (8 mg) and fractions F48–50 gave 2-(4-hydroxyphenyl)ethyl triacontanoate on subjecting to CC using hexane:EtOAc (9.5:0.5) while F90–100 (Hex:EtOAc 8.0:2.0) gave oleanolic acid (80 mg). 3.3.1. Newbouldiaquinone A (1) Yellow powder, m.p. 260 °C; UV kmax, nm (log e): 270 (2.9), 308 (4.80), 388 (4.50); IR mmax CHCl3 cm
1: 3404, 1647, 1236, 1036;. 1H NMR (500 MHz, CDCl3): d 8.37– 8.32 (2H, m, H-5, H-8), 8.27 (1H, s, H-4), 8.25 (1H, dd, J = 8.5, 1.5 Hz, H-8 0 ), 8.22 (1H, dd, J = 8.5, 1.5 Hz, H5 0 ), 8.18 (1H, s, H-1), 7.95 (2H, td, J = 8.5, 1.5 Hz, H-6, H-7) and 7.90 (2H, td, J = 8.5, 1.5 Hz, H-6 0 , H-7 0 ), 2.51 (3H, s, CH3-2); 13C NMR (125 MHz, CDCl3): d 183.5 (C-10), 182.9 (C-1 0 ), 182.7 (C-9), 181.6 (C-4 0 ), 145.4(C-3), 140.1 (C-3 0 ), 139.9 (C-2 0 ), 135.2 (C-6 0 ), 134.9 (C-7 0 ), 134.8 (C-6), 133.9 (C-7), 133.6 (C-8a), 132.7(C-10a), 132.6 (C9a), 132.5 (C-4a), 130.9 (C-4a 0 ), 130.8 (C-8a 0 ), 129.8 (C1), 128.2 (C-4), 127.2 (C-5), 127.1 (C-8), 126.4 (C-8 0 ), 126.3 (C-5 0 ), 121.3 (C-2), 20.4 (CH3); CI–MS (CH4): m/z 411.1 [M + 1]; EIMS m/z (rel. int.): 410.1 [M]+ (41), 394.1 [M
O]+ (100), 348.1 [M
O
CO
H2O]+ (30), 322.1 [M
H2O
2CO
CH2]+(20), 252.1 (20), 176.1 (14), 126.1 (8), 105.0 (16), 76.0 (21), 50.0 (9). 3.3.2. Acetylation A solution of dry pyridine (0.5 ml) and Ac2O (0.5 ml) were added to compound 1 (5 mg), and left overnight. After usual

hydrolytic (HCl) workup, compound 2 was isolated and purified by filtration over a short batch of silical gel (CH2Cl2) (3 mg); m.p.: 205 °C; 1H NMR (500 MHz, CDCl3): d 8.37– 8.33 (2H, m, H-5, H-8), 8.27 (1H, s, H-4), 8.24 (1H, dd, J = 8.5, 1.5 Hz, H-8 0 ), 8.21 (1H, dd, J = 8.5, 1.5 Hz, H-5 0 ), 8.15 (1H, s, H-1), 7.88–7.83 (4H, m, H-6, H-7, H-6 0 , H-7 0 ), 2.50 (3H, s, CH3-2) and 2.30 (3H, s, CH3COO); 13C NMR (125 MHz, CDCl3):d 183.3 (C-10), 183.0 (C-1 0 ), 182.7 (C9), 181.6 (C-4 0 ), 170.1 (CH3COO)145.3(C-3), 141.9 (C-3 0 ), 140.8 (C-2 0 ), 135.5 (C-6 0 ), 134.2 (C-7 0 ), 134.8 (C-6), 133.9 (C-7), 133.6 (C-8a), 132.7(C-10a), 132.5 (C-9a), 132.4 (C4a), 130.8 (C-4a 0 ), 130.7 (C-8a 0 ), 129.6 (C-1), 128.1 (C-4), 127.4 (C-5), 127.0 (C-8), 126.7 (C-8 0 ), 126.5 (C-5 0 ), 121.4 (C-2), 21.5 (CH3COO), 20.4 (CH3-2); EI–MS m/z (rel. int.): 452.1 [M]+ (30), 392.1 [M
HOAc]+ (25), 410.1 [M
CH3CO + H]+ (30), 176.1 (12), 126.1 (11), 105.1 (16), 76.2 (29), 50.1 (12). 3.4. Antimalarial test Newbouldiaquinone A (1), lapachol, a-lapachone and blapachone were dissolved in water + DMSO 0.02% v/v (Andrade-Neto et al., 2004). The compounds were administered over a period of four days to the culture and the number of parasites was determined daily. An untreated culture of plasmodia served as a control (for results see Table 1). 3.4.1. Culturing of P. falciparum NF54 strain P. falciparum isolate NF54 and R strain were maintained in small Petri dishes (5 cm) according to a protocol from Moloney (Moloney et al., 1990) and Trager (Trager and Williams, 1992) in a gaseous phase of 90% N2, 5% CO2 and 5% O2. Parasites were cultured in human erythrocytes (blood group A+) in RPM1640 medium (Sigma) supplemented with 25 lM HEPES, 20 mM sodium bicarbonate, and 10% heat inactivated human A+ plasma at 10% (v/v) hematocrit. The parasitemia of infected erythrocytes was determined by light microscopy and estimated by Giemsa-stained smears. Parasitemia detected in the cultures were scored visually with a 100-fold oil immersion objective, counting at least 1000 infected erythrocytes to determine the parasitemia. 3.4.2. Inhibitor experiments by monitoring multiplication and growth of plasmodia Cultures were adjusted to a parasitemia of 0.5%. Aliquots were diluted 1:10-fold in RPMI-medium, dispensed into 12-well microculture trays and incubated at 37 °C in a candle jar. Thereafter, growth medium was changed once a day for four days and inhibitors were added to the media in concentration of 20 lM as indicated. Each substance was analyzed in four independent wells of the microculture tray. Parasitemia was estimated as triplicates daily in each of the four independent wells from Giemsastained smears by counting 1000 erythrocytes (for results see Table 1).

K.O. Eyong et al. / Phytochemistry 67 (2006) 605–609

3.5. Antimicrobial test 3.5.1. Microbial strains A total of 21 microbial cultures belonging to six Gram positive bacterial species (Bacillus cereus, Bacillus megaterium, Bacillus substilis, Bacillus stearothermophilus, Staphylococcus aureus, Streptococcus faecalis), 12 Gram negative bacteria (Escherichia coli, Shigella dysenteriae, Proteus vulgaris, Proteus mirabilis, Shigella flexneri, Klebsiella pneumoniae, Pseudomonas aeruginosa, Salmonella typhi, Morganella morgani, Enterobacter aerogens, Citrobacter freundii, Enterobacter cloacae), and three yeasts from Candida species (Candida albicans, Candida krusei and Candida gabrata) were used in this study. Three of the four Bacillus species were provided by ‘‘l’institut Appert de Paris’’ while, Bacillus cereus was provided by the A.F.R.C Reading Laboratory of Great Britain. The other strains were clinically isolated from patients in the Centre Pasteur de YaoundeCameroon (health institution). They were then maintained on agar slant at 4 °C in the Laboratory of the Applied Microbiology and Molecular Pharmacology (Faculty of Science, University of Yaounde I) where the antimicrobial tests were performed. The strains were activated at 37 °C for 24 h on nutrient agar (NA) (bacteria) or Sabouraud glucose agar (yeasts). The nutrient broth (NB) was used to determine the minimal inhibition concentration of compound (1) against the tested pathogens. 3.5.2. Antimicrobial assays MICs of compound (1) were evaluated against the pathogens. The inocula of micro organisms were prepared from 12 h broth culture and the suspensions were adjusted to 0.5 Mc Farland turbidity. Compound (1) was first dissolved in dimethyl sulfoxide (DMSO) 10% v/v to the highest dilution (39.06 lg/ml), and serial twofold dilutions were made in a concentration ranged from 0.031 to 39.06 lg/ml in the 96 wells microplate containing NB. MIC values of the tested compounds against pathogens were determined based on the microdilution method, as the lowest concentration at which there was 100% growth inhibition of the tested pathogens (Zgoda and Porter, 2001). Gentamycin (bacteria) and nystatin (yeasts) diluted prior in water were also used as reference antibiotics. Negative control was made with DMSO 10% v/v. Acknowledgements K.O. Eyong thanks DAAD, Germany, for financial support. We thank J.A. Efim, P.K, Pemha, S. Bessong for harvesting N. laevis from Mamfe and to E.B. Betek for revealing several antimalarial plants among which was N. laevis. References Aguinaldo, A.M., Ocampo, O.P.M., Bowden, B.F., Gray, A.I., Waterman, P.G., 1993. Tectograndone, an anthraquinone-naphthoquinone

609

pigment from the leaves of Tectona grandis. Phytochemistry 33, 933– 935. Ali, R.M., Houghton, P.J., 1999. A new phenolic fatty acid ester with lipoxygenase inhibitory activity from Jacaranda filicifolia. Plant. Medic. 65, 455–457. Andrade-Neto, V.F., Goulart, M.O.F., Filho, J.F.S., Silva, M.J., Pinto, M.C.F.R., Zalis, M.G., Carvalho, L.H., Krettli, A.U., 2004. Antimalarial activity of phenazines from lapachol, b-lapachone and its derivatives against Plasmodium falciparum in vitro and Plasmodium berghei in vivo. Bioorg. Med. Chem. Lett. 14, 1145–1149. Boivert, L., Brussard, P., 1988. Regiospecific addition of monooxygenated dienes to halo-quinones. J. Org. Chem. 53, 4052–4059. Chatterjee, T.K., Basak, A., Borua, A.K., Mukherjee, K., Roy, L.N., 1979. Studies on the structure and stereochemistry of canthic acid- a new triterpene acid sapogenin from Canthium dicoccum. Trans. Bose Res. Inst. 42, 85–87. Eyong, O.K., Krohn, K., Hussain, H., Folefoc, G.N., Nkengfack, A.E., Schulz, B., Hu, Q., 2005. Newbouldiaquinone and newbouldiamide: a new naphthoquinone–anthraquinone coupled pigment and a new ceramide from Newbouldia laevis. Chem. Pharm. Bull. 53 (6), 616–619. Gafner, S., Wolfender, J.L., Nianga, M., Stoeckli, E.H., Hostetmann, K., 1996. Antifungal and antibacterial naphthoquinones from Newbouldia laevis roots. Phytochemistry 42 (5), 1315–1320. Gorman, R., Kaloga, M., Li, X.-C., Ferreira, D., Bergenthal, D., Kolodziej, H., 2003. Furanonaphthoquinones, atraric acid and a benzofuran from the stem barks of Newbouldia laevis. Phytochemistry 64, 583–587. Harrison, A.D., Kulshreshtha, D.K., 1984. Chemical constituents of Amberboa ramose. Fitoterapia 55, 189–192. Hassanean, H.A., Ibraheim, Z.Z., Takeya, K., Itorawa, H., 2000. Further quinoidal derivatives from Rubia cordifolia L. Pharmazie 55, 317–319. Hiermann, A., Kartnig, T., 1978. Flavonoids in the leaves of Digitalis lanata (Ehrart). Plant. Medic. 34 (2), 225–226. Hooker, S.C., 1936. The constitution of lapachol and its derivatives. Part V. The structure of Paterno’s ‘‘Isolapachone’’. J. Am. Chem. Soc. 58, 1190–1197. Ikuta, A., Hokawa, H., 1988. Triterpenoids of Paeonia japonica callus tissue. Phytochemistry 27, 2813–2815. Khan, R.M., Mlungwana, S.M., 1999. 5-Hydroxylapachol: a cytotoxic agent from Tectona grandis. Phytochemistry 50, 439–442. Lopes, C.C., Lopes, R.S.C., Pinto, A.V., Costa, P.R.R., 1984. Efficient synthesis of cytotoxic quinones: 2-acetyl-4H,9H-naphtho[2,3-b]furan4,9-dione and (±)-2-(1-hydroxyethyl)-4H,9H-naphtho[2,3-b]furan-4,9dione. J. Heter. Chem. 21, 621–622. Matsumoto, M., Kobayashi, H., 1985. Hexacyanoferrate-catalyzed oxidation of trimethoxybenzenes to dimethoxy-p-benzoquinones with hydrogen peroxide. J. Org. Chem. 50, 1766–1768. Moloney, M.B., Pawluk, A.R., Ackland, N.R., 1990. Plasmodium falciparum growth in deep culture. Trans R. Soc. Trop. Med. Hyg. 84 (4), 516–518. Okeka, A.O., 2003. Three-minute herbal treatment to reduce dental caries with a Newbouldia laevis based on extract. Am. J. Undergrad. Res. 2 (2), 1–4. Schuhr, C.A., Radykewicz, T., Sagner, S., Latzel, C., Zenk, M.H., Arigoni, D., Bacher, A., Rohdlich, F., Eisenreich, W., 2003. Quantitative assessment of crosstalk between the two isoprenoid biosynthesis pathways in plants by NMR spectroscopy. Phytochem. Rev. 2 (1–2), 3–16. Seo, S., Tomita, Y., Tori, K., Yoshimura, Y., 1978. Determination of the absolute configuration of a secondary hydroxyl group in a chiral secondary alcohol using glycosidation shifts in carbon-13 nuclear magnetic resonance spectroscopy. J. Am. Chem. Soc. 100, 3331–3339. Singh, J., Singh, J., 1986. A bianthraquinone and a triterpenoid from the seeds of Cassia hirsute. Phytochemistry 25, 1985–1987. Trager, W., Williams, J., 1992. Extracellular (axenic) development in vitro of the erythrocytic cycle of Plasmodium falciparum. Proc. Natl. Acad. Sci. USA 89 (12), 5351–5355. Zgoda, J.R., Porter, J.R., 2001. A Convenient microdilution method for screening natural products against bacteria and fungi. Pharmaceut. Biol. 39 (3), 221–225.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 610–617 www.elsevier.com/locate/phytochem

Vasodilatory and hypoglycaemic effects of two pyrano-isoflavone extractives from Eriosema kraussianum N. E. Br. [Fabaceae] rootstock in experimental rat models John A.O. Ojewole a

a,*

, Siegfried E. Drewes b, Fatima Khan

b

Department of Pharmacology, School of Pharmacy and Pharmacology, Faculty of Health Sciences, University of KwaZulu-Natal, Private Bag X54001, Scottsville, Durban 4000, South Africa b School of Chemistry, University of KwaZulu-Natal, Private Bag X01, Scottsville, Pietermaritzburg, South Africa Received 18 August 2005; received in revised form 15 November 2005 Available online 24 January 2006

Abstract Zulu traditional health practitioners have claimed that the roots of Eriosema kraussianum N. E. Br. (Fabaceae) and other Eriosema species (Zulu indigenous umbrella name of ‘‘uBangalala’’) are effective remedies for the treatment of erectile dysfunction (ED) and/or impotence. In order to scientifically appraise the significance and contribution of Eriosema kraussianum to its ethnomedical use as ‘‘uBangalala’’ and ‘‘VIAGRAe substitute’’, the present study was undertaken to investigate the vasodilatory and hypoglycaemic properties of the two main bioactive chemical compounds [Kraussianone-1 (K1), and Kraussianone-2 (K2), Drewes, S.E., Horn, M.M., Munro, O.Q., Dhlamini, J.T.B., Meyer, J.J.M., Rakuambo, N.C., 2002. Pyrano-isoflavones with erectile-dysfunction activity from Eriosema kraussianum. Phytochemistry 59 739–747.] obtained from E. kraussianum, in experimental rat models, using sildenafil citrate (VIAGRAe) as the reference drug for comparison. The two E. kraussianum rootstock constituents (K1 and K2, 20–80 mg/kg p.o.) caused dose-dependent and significant (P < 0.05–0.001) hypoglycaemia in rats. Relatively low to high concentrations of the plant’s extracts (K1 and K2, 100– 2000 lg/ml) always produced biphasic effects on rat isolated portal veins. K1- and K2-provoked responses of the isolated portal veins always consisted of concentration-related initial transient, but significant (P < 0.05), contractions of the venous muscle preparations, followed by secondary, longer-lasting, highly significant (P < 0.01–0.001) relaxations of the venous muscle strips. Sildenafil citrate (VIAGRAe, 5–100 lg/ml) always produced concentration-related and highly significant relaxations of the rat isolated portal veins. Unlike K1 and K2 (20–80 mg/kg p.o.), however, sildenafil citrate (VIAGRAe, 100 mg/kg p. o.) only caused slight and insignificant (P > 0.05) reductions in the blood glucose levels of the experimental animals used. On the other hand, glibenclamide (10 mg/kg p.o.) induced highly significant (P < 0.05–0.001), marked reductions in the blood glucose concentrations of the rats. The findings of this laboratory animal study indicate that the two hydro-ethanol extractives of E. kraussianum (K1 and K2) possess hypoglycaemic and secondary, vasorelaxant effects in the experimental paradigms used. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Eriosema kraussianum; Pyrano-isoflavones; K1 and K2; Hypoglycaemic and vasodilatory properties; Rat model

1. Introduction

* Corresponding author. Tel.: +27 31 260 7767/7356; fax: +27 31 260 7907. E-mail addresses: [email protected], [email protected] (J.A.O. Ojewole).

0031-9422/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.11.019

In South Africa, the genus Eriosema consists of annual herbs or shrublets of 10–18 cm in height, and the plants in this genus are found mainly in grassland areas of the country. The species in the genus possess well-developed root system, and it is this morphological part of the plants that is mainly used in South African traditional medicine.

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

The Zulu people of South Africa traditionally use the roots of the genus Eriosema for a variety of human ailments, including management, control and/or treatment of impotence and urinary disorders (Hulme, 1954; Bryant, 1966; Hutchings et al., 1996). The plant’s roots are also used as expectorants and diuretics (Watt and Breyer-Brandwijk, 1962). Generally, the genus Eriosema contains plants which come under the isiZulu indigenous umbrella name of ‘‘uBangalala’’, and most of the plant species listed under this name (‘‘uBangalala’’) are used mainly for the purpose of curing or alleviating impotence (Bryant, 1966; Hutchings et al., 1996). In this regard, hot milk infusions of the plant’s roots and/or pounded boiled root decoctions are taken in small doses in the morning and at night for impotence (Hulme, 1954; Bryant, 1966). Previous studies in our laboratories (Drewes et al., 2002) have shown that Eriosema kraussianum N. E. Br. (Fabaceae) is one of the frequently used Eriosema species for the treatment of impotence, an ailment otherwise frequently referred to as ‘‘erectile dysfunction’’ (ED). Erectile dysfunction has been described as ‘‘a consistent inability to achieve and maintain an erection sufficient for satisfactory sexual activity’’ (Goldstein et al., 1998). Drewes et al. (2002, 2003) have shown in a rabbit experimental model, the beneficial effects of bioactive compounds of E. kraussianum in the management of erectile dysfunction. Cardiovascular disorders and diabetes mellitus are known to contribute significantly to ED of organic origin (Zusman et al., 1999; Rendell et al., 1999). Since men with ED of organic, psychogenic and mixed aetiologies are known to benefit from VIAGRAe therapy, it is speculated that E. kraussianum extractives may also be effective as ‘‘VIAGRAe substitutes’’ in the treatment of ED of organic, psychogenic and mixed origins. Although Drewes et al. (2002) used sildenafil citrate (VIAGRAe) as the reference compound for comparison in their ED studies on the two new pyrano-isoflavones, unlike sildenafil citrate, the exact mechanisms of action of the two extractives from E. kraussianum still remain unknown. The two new compounds tested positive in experimental rabbit ED treatment, and attained values of 85% and 65%, respectively, compared with VIAGRAe, in relaxing rabbit corpus cavernosum smooth muscles (Drewes et al., 2002, 2003). The Zulu people of South Africa traditionally employ hot milk infusions and pounded decoctions of the roots of E. kraussianum and other species of Eriosema as substitutes for VIAGRAe in the treatment of erectile dysfunction (ED) and/or impotence. For this reason, we have compared the effects of E. kraussianum extractives (K1 and K2) with that of VIAGRAe in the experimental animal models used in the present study. In order to throw some light on the plausible mechanisms of action of the extractives, the present study was undertaken to investigate the hypoglycaemic and vasodilatory properties of the two new pyrano-isoflavones [Kraussianone-1 (K1) and Kraussianone-2 (K2)] from the roots of E. kraussianum in experimental rat models, using sildenafil citrate (VIAGRAe) as the reference sexual stimulant drug for comparison (see Fig. 1).

611 O

Me N

OEt HN

N N

SO2

Me N N

Me

Viagra (Sildenafil) Me

O

Me Me HO

O

Me

O

H OH O HO

Me O

Kraussianone 1(K1)

Me

OH O HO

Me O

Me

Kraussianone 2 (K2)

Fig. 1. Structural formulae of Viagrae (sildenafil citrate), Kraussianone-1 (K1) and Kraussianone-2 (K2).

2. Results and discussion 2.1. Effects of K1 and K2 on blood glucose levels of healthy, normal rats In a separate set of experiments involving 16-h fasted healthy, normal, male rats, the baseline blood glucose levels were found to vary between 4.10 ± 0.13 and 4.36 ± 0.14 mmol/l. In our ‘control’ set of experiments, acute treatment of the animals with the hydro-ethanol vehicle (3 ml/kg p.o.) alone did not significantly modify (P > 0.05) the blood glucose concentrations of the fasted normal rats. In these animals, pretreatment with the hydro-ethanol vehicle (3 ml/kg p.o.) for 1, 2, 4 and 8 h either slightly but insignificantly (P > 0.05) decreased, increased, or did not affect at all, the blood glucose concentrations of the fasted ‘control’ animals. The vehicle-induced changes in the blood glucose levels of the fasted rats varied by values ranging between 0.1% and 1.0% of the mean baseline values (Table 1). Similarly, pretreatment of the normal rats with sildenafil citrate (VIAGRAe, 100 mg/ kg p.o.) for 1, 2, 4 and 8 h did not reduce the baseline blood glucose levels significantly (P > 0.05) after 21/2 h. However, compared with the vehicle-treated ‘control’ rats, pretreatment of the fasted rats with relatively moderate to high doses of K1 or K2 extract (20, 40 and 80 mg/kg p.o.) for 1, 2, 4 and 8 h produced significant reductions (P < 0.05– 0.001) in the blood glucose concentrations of the fasted normal rats (Table 1). Maximal reductions in the blood glucose concentrations of the fasted ‘test’ rats occurred at the plant’s extract (K1 or K2) dose of 80 mg/kg (p.o.). Compared with the vehicle-treated fasted ‘control’ rats, pretreatment of fasted normal rats with glibenclamide (10 mg/kg p.o.) for 1, 2, 4 and 8 h also produced significant reductions (P < 0.05–0.001) in the blood glucose concentrations of the animals (Table 1). The hypoglycaemic effects of the plant’s extracts (K1 and K2) became significant

612

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

Table 1 Effects of Krausianone-1 (K1, 80 mg/kg p.o.), Krausianone-2 (K2, 80 mg/kg p.o.), VIAGRAe (SC, 100 mg/kg p.o.) and glibenclamide (GBC, 10 mg/kg p.o.) on blood glucose concentrations (mmol/l) of normal (normoglycaemic) rats Treatment

Before treatment

After treatment

0h

1h

2h

4h

8h

Control (3 ml/kg hydro-ethanol vehicle p.o.) Kraussianone-1 (K1, 80 mg/kg p.o.) Kraussianone-2 (K2, 80 mg/kg p.o.) VIAGRAe (SC, 100 mg/kg p.o.) Glibenclamide (GBC, 10 mg/kg p.o.)

4.35 ± 0.11 4.28 ± 0.15 4.30 ± 0.12 4.12 ± 0.13 4.18 ± 0.14

4.34 ± 0.12 4.03 ± 0.13 4.01 ± 0.10 4.06 ± 0.12 3.53 ± 0.12*

4.36 ± 0.10 3.67 ± 0.12* 3.54 ± 0.13* 4.03 ± 0.10 3.04 ± 0.10**

4.35 ± 0.13 3.18 ± 0.14* 3.08 ± 0.11* 4.01 ± 0.14 2.24 ± 0.12***

4.34 ± 0.10 3.84 ± 0.11* 3.71 ± 0.10* 4.06 ± 0.13 3.08 ± 0.12**

Maximal reduction

% Maximal reduction

0.00 0.26 0.28 0.03 0.46

0.23 NS 25.70* 28.37* 2.67 NS 46.41***

Values given represent the mean (±SEM) of 8 observations. NS = P > 0.05. * P < 0.05. ** P < 0.01. *** P < 0.001 vs control.

(P < 0.05) 2 h following oral administration, reaching the peak of their hypoglycaemic effects 4 h after administration. However, the hypoglycaemic effects of the extracts were still significant 8 h after oral administration (Table 1). Thereafter, the blood glucose concentrations of the animals gradually returned to baseline levels at the end of the 24th h. 2.2. Effects K1 and K2 on oral glucose tolerance test (OGTT) in normal rats The effects of K1, K2 (80 mg/kg p.o.), VIAGRAe (100 mg/kg p.o.) and glibenclamide (10 mg/kg p.o.) on blood glucose levels of rats following oral glucose load

are shown in Fig. 2. This figure shows that following oral glucose load, the blood glucose concentrations of the 16h fasted normal rats increased to 8.7 ± 0.6 mmol/l from a baseline value of 4.36 ± 0.14 mmol/l; before gradually declining to 5.8 ± 0.6 mmol/l after 21/2 h. Pretreatment of the normal rats with K1 or K2 (80 mg/kg p.o.) for 20 min prior to oral glucose load, significantly reduced (P < 0.05) the peak blood glucose levels from 8.7 ± 0.6 to 5.1 ± 0.6 and 5.3 ± 0.7 mmol/l, respectively, after 21/2 h. This observation would appear to suggest that the plant’s extracts (K1 and K2) facilitate or promote the clearance of postprandial blood glucose in rats. The effects of K1 and K2 on the blood glucose concentrations of the rats were not significantly different (P > 0.05) after 21/2 h. Pre-

10 9 Control Viagra K1 K2 Glibenclamide

[Blood glucose] (mmol/L)

8 7 6 5 4 FBG

3 2 1 0 0

15

30

45

60

75

90

105

120

180

240

300

Time (minutes) Fig. 2. Effects of Kraussianone-1 (K1, 80 mg/kg p.o.), Kraussianone-2 (K2, 80 mg/kg p.o.), Viagrae (100 mg/kg p.o.) and glibenclamide (10 mg/kg p.o.) on blood glucose concentrations (mmol/l) of normal rats following oral glucose load (2 g/kg p.o.). Each point represents the mean (±SEM) of eight determinations, while the vertical bars represent standard errors of the means. FBG denotes ‘Fasting Blood Glucose’.

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

treatment of the normal rats with glibenclamide (10 mg/kg p.o.) for 20 min prior to oral glucose load, also significantly reduced (P < 0.001) the peak blood glucose levels from 8.7 ± 0.6 mmol/l to 4.8 ± 0.7 mmol/l after 21/2 h. However, pretreatment of the normal rats with sildenafil citrate (VIAGRAe, 100 mg/kg p.o.) for 20 min prior to oral glucose load, did not reduce the baseline blood glucose concentrations significantly (P > 0.05) after 21/2 h. 2.3. Effects of K1 and K2 on rat isolated portal vein Relatively low to high concentrations of K1 or K2 (100– 2000 lg/ml) always produced biphasic effects on rat isolated portal veins. The K1- and K2-induced responses of the isolated portal veins always consisted of dose-related initial transient, but significant (P < 0.05) contractions of the venous muscle preparations, followed by secondary, longer-lasting, highly significant (P < 0.01–0.001) relaxations of the muscle strips. During the initial transient, contractile phase, K1 or K2 (100–2000 lg/ml) usually increased the contractile frequency, and inhibited the amplitude of the spontaneous, myogenic contractions of the isolated portal veins in a concentration-dependent manner. Fig. 3 summarizes the results obtained with K1 and K2 (1000 lg/ml) and VIAGRAe (500 lg/ml). Sildenafil citrate (VIAGRAe, 5– 1000 lg/ml) always produced concentration-related, significant relaxations (P < 0.05–0.001) of the isolated portal veins. The possibility that the K1, K2 and VIAGRAe–induced responses of the isolated portal veins might involve interaction with Ca2+ at the cell membrane was also investigated. In these experiments, the concentration of Ca2+ in the bathing normal Krebs–Henseleit physiological solution [of composition, in g/l: NaCl, 6.92; KCl, 0.34; NaH2PO4, 0.15; NaHCO3, 2.10; MgCl2, 0.11; CaCl2, 0.26; and glucose,

*

20

*

Tension Change (%)

0

*

Viagra K2 K1

*

* -20

** -40 **

** -60

*** -80

*** ***

-100 1

2

3

4

5

30

Time (minutes)

Fig. 3. Effects of Krausianone-1 (K1, 1000 lg/ml), Krausianone-2 (K2, 1000 lg/ml) and sildenafil citrate (VIAGRAe, 500 lg/ml) on spontaneous contractions of rat isolated portal veins. Each point represents the mean (±SEM) of 8–10 preparations, while the vertical bars denote standard errors of the means. *P < 0.05; **P < 0.01; ***P < 0.001 vs. control.

613

1.00 – pH adjusted to 7.4 maintained at 34 ± 1 °C and continuously aerated with carbogen (i.e., 95% O2 + 5% CO2 gas mixture)] was either reduced from 0.26 to 0.13 g/l, or raised from 0.26 to 0.52 g/l, respectively. The initial transient, contractile responses of the isolated muscle strips induced by relatively low to high concentrations K1 and K2 (100–2000 lg/ml) were reduced and/or abolished in the presence of low calcium concentration [Ca2+ = 0.13 g/l] in the bathing Krebs–Henseleit physiological solution. However, the secondary, relaxant responses of the isolated venous muscle strips produced by low to high concentrations of the extracts (K1 and K2, 100–2000 lg/ml) increased as the concentration of the external Ca2+ was reduced. Raising the bathing fluid Ca2+ concentration from 0.26 to 0.52 g/l increased and/or enhanced K1 and K2 (100–2000 lg/ml)– induced initial transient, contractile responses of the isolated venous muscle preparations. However, the secondary, relaxant responses of the isolated venous muscle strips induced by low to high concentrations of the plant’s extracts (K1 and K2, 100–2000 lg/ml) decreased as the Ca2+ concentration of the external bathing fluid was increased. In all cases, washing of the isolated venous muscle preparations with fresh, normal Krebs–Henseleit physiological solution 3–5 times usually restored physiological activities of the isolated muscle strips to control, baseline values. 2.4. Possible mechanisms of action Experimental evidence obtained in the present study show that the hydro-ethanol rootstock extractives of E. kraussianum (K1 and K2, 20–80 mg/kg p.o.) dose-dependently and significantly (P < 0.05–0.001) reduced blood glucose levels in rats. While sildenafil citrate (VIAGRAe) only produced slight and non-significant (P > 0.05) hypoglycaemic effects in the experimental animal model used, glibenclamide induced highly significant hypoglycaemia (P < 0.01–0.001) in the experimental animals. The major classes of synthetic oral hypoglycaemic agents currently available for the management and/or control of adult-onset, type-2, non-insulin-dependent diabetes mellitus (NIDDM) include the sulphonylureas, biguanides, thiazolidinediones, a-glucosidase inhibitors, and so on. Glibenclamide, used as the reference hypoglycaemic agent in this study, is a member of the ‘first generation’ sulphonylureas. As a class, sulphonylureas enhance and increase the release of endogenous insulin from pancreatic b-cells. They also promote and facilitate peripheral tissue uptake and utilization of glucose. It has been proposed (Jackson and Bressler, 1981) that sulphonylureas produce their hypoglycaemic effects via three main mechanisms, viz: (i) increased insulin release from pancreatic b-cells; (ii) potentiation of insulin’s action on target tissues and increased glucose removal from the blood, and (iii) reduction of blood glucagon levels. Thus, any plant secondary metabolite or chemical compound which is capable of affecting the pancreatic b- or a-cell secretion in any of the three ways illustrated above will be a good mimicker of sulphonylureas,

614

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

and will produce hypoglycaemic effects in mammals via mechanisms similar to those of sulphonylureas. Although the precise mechanisms of the hypoglycaemic actions of K1 and K2 remain speculative at present, experimental evidence obtained in this study tends to suggest that the plant’s extracts dose-dependently enhance or promote the clearance of postprandial blood glucose in rats. Therefore, the possibility exists that the plant’s extracts (K1 and K2) may act like sulphonylureas, and mimic or improve insulin action at cellular level. VIAGRAe always produced concentration-dependent, highly significant (P < 0.001) relaxations of the rat isolated portal veins. However, K1 and K2 always caused concentration-related, initial slight contractions, followed by secondary, longer-lasting, significant (P < 0.05–0.001) relaxations of the isolated portal veins. The initial transient, contractile effects of K1 and K2 may be due to the ability of the extracts to transiently release noradrenaline (NA) from NA tissue stores. The precise mechanism of the secondary, pronounced vasorelaxant effects of the plant’s extractives (K1 and K2) on the rat isolated portal veins is unknown at the moment. However, because the secondary venous relaxant effects of the plant’s extracts were resistant to blockade by standard, receptor specific antagonists such as atropine sulphate (2 lg/ml) and mepyramine maleate (2 lg/ml) in all the muscle preparations tested, it is speculated that the secondary, longer-lasting venorelaxant effects of K1 and K2 on the isolated venous muscle strips may be non-specific in nature. Furthermore, the finding that changes (decrease or increase) in calcium ion concentrations [Ca2+] of the bathing physiological solution modified the responses of the isolated venous tissue preparations to bath-applied concentrations of E. kraussianum extracts (K1 and K2) would appear to suggest that the extracts affected calcium mobilization and/or sequestration, and possibly also, calcium release from its various tissue stores (Webb et al., 1999). Further studies are certainly needed to shed more light on the plausible mechanisms of action of the two E. kraussianum extracts. The data obtained in this study do not allow any definite conclusions to be drawn on the mechanisms of hypoglycaemic and vasorelaxant actions of E. kraussianum root hydro-alcohol extracts in the experimental animal models used. However, a number of investigators have shown that tannins and other polyphenolic compounds (e.g., coumarins), flavonoids, triterpenoids, and a host of other secondary plant metabolites possess anti-inflammatory, hypoglycaemic, spasmolytic and analgesic effects in various experimental animal models (Dongmo et al., 2003; Taesotiku et al., 2003; Ojewole, 2002, 2003; Adzu et al., 2003; Akah and Okafor, 1992; Li et al., 2003; Marles and Farnsworth, 1995). We have compared the effects of E. kraussianum extractives (K1 and K2) with that of VIAGRAe in the experimental animal models used in the present study. Although biomedical literature is limited to two publications (Drewes et al., 2002, 2003) on the extractives of E.

kraussianum as effective remedies for erectile dysfunction, literature abounds on the effectiveness of VIAGRAe in the treatment of erectile dysfunction. Erectile dysfunction (ED) is a common medical condition in men with cardiovascular disorders such as ischaemic heart disease, hypertension and peripheral vascular diseases (Feldman et al., 1994; Conti et al., 1999), and is also common in men with diabetes mellitus (Rendell et al., 1999), probably because of the shared factors that impair haemodynamic mechanisms in the penile and ischaemic vasculature. Erectile dysfunction is also caused by spinal cord injury [cord level range, T6–L5] (Derry et al., 1998), and other factors such as radical prostatectomy, long-term use of certain medications (e.g., antidepressants, antipsychotics, antihypertensives and diuretics), indeces of anger and depression, and cigarette smoking (Feldman et al., 1994). Sildenafil citrate (VIAGRAe) has been widely studied for its efficacy in the treatment of erectile dysfunction (ED) in a variety of patient populations. In men, oral sildenafil citrate is generally known to be effective in erectile dysfunctions of organic, psychogenic or mixed origins. However, the aetiology of erectile dysfunction has been shown to have a significant impact on treatment success and satisfaction rates, with neurogenic causes of erectile dysfunction (e.g., diabetes mellitus, prostate surgery, and so forth) having significantly lower treatment success rates than psychogenic or vasculogenic erectile dysfunction (Jarow et al., 1999). Although VIAGRAe has transient vasodilatory properties in vivo, it has no direct relaxant effect on isolated human corpus cavernosum in vitro (Conti et al., 1999). The main mechanism of action of Viagrae in erectile dysfunction is well-known and, therefore, needs no repetition here. Suffice it to say that it facilitates the patient’s ability to achieve and maintain an erection for satisfactory sexual activity after penetration. This it does by its ability to selectively inhibit cyclic guanosine monophosphate (cGMP)-specific phosphadiesterase type 5 (PDE5). According to Pfizer’s data on file, ‘‘VIAGRAe is rapidly absorbed from gastro-intestinal smooth muscles following oral administration, with an absolute bioavailability of about 40%. Its pharmacokinetics are dose-related over recommended dose ranges (25, 50 or 100 mg/kg p.o.). It is eliminated predominantly by hepatic metabolism (mainly by cytochrome P450 3A4), and is converted to an active metabolite with properties similar to that of the parent, sildenafil. Both sildenafil and its metabolite have terminal half-lives of about 4 h. Maximum observed plasma concentrations are reached within 30–120 min (median 60 min) of oral dosing in the fasted state. When VIAGRAe is taken with a high fat meal, the rate of absorption is reduced with a mean delay in Tmax of 60 min, and a mean reduction in Cmax of 29%’’. K1 and K2 caused significant hypoglycaemia in rats, and like VIAGRAe, produced relaxations of isolated portal veins. Unlike VIAGRAe, however, human knowledge on the benefit, usefulness and/or efficacy of the extractives of E. kraussianum (and other species of Eriosema) in the

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

treatment of erectile dysfunction is limited. Nonetheless, as with oral VIAGRAe and fat meal, oral administration of infusion and/or decoction of Eriosema species roots with milk probably delays and/or reduces the rate of absorption of the compounds from the patient’s gastro-intestinal tract (GIT), and thereby prolongs the duration of action of the extractives. It has been reported that for maximum benefit, milk infusions and/or decoctions of the plant’s roots are to be taken 2–4 h before any anticipated sexual intercourse, and the sexual activity effects (achievement and maintenance of penile erection sufficient for satisfactory sexual intercourse after penetration) of the extractives have also been reported to last for 4–6 h following oral dosing of the milk infusion or decoction of the root extracts (Drewes, Personal Communication). Unlike VIAGRAe, the bioavailability, half-life (t1/2), Tmax, Cmax and other pharmacokinetic parameters of the extracts are unknown at the moment. The effects of the plant’s extracts (K1 and K2) on the biochemical activities of cGMP and PDE5 are also obscure at present. Further studies are certainly warranted to throw more light on the pharmacokinetic profiles of K1 and K2, as well as on the biochemical effects of the extractives on cGMP and PDE5. The experimental evidence obtained in this laboratory animal study indicates that the two hydro-alcoholic extracts of E. kraussianum (K1 and K2) possess hypoglycaemic and secondary vasorelaxant effects. These pharmacological properties of the plant’s extracts may contribute significantly to the effectiveness of the herb in the management and/or treatment of erectile dysfunction among the Zulu men of South Africa.

3. Experimental The experimental protocol used in this study was approved by the Ethics Committee of the University of Durban-Westville, Durban 4000, South Africa; and conforms with the ‘‘Guide to the care and use of animals in research and teaching’’ [published by the Univerity of Durban-Westville, Durban 4000, South Africa]. 3.1. Plant material E. kraussianum N. E. Br. (Fabaceae) was collected in November 2004 from an open grassland on the northern boundary of the National Botanical Garden in Pietermaritzburg, South Africa. Identification of the plant was done by Professor Trevor Edwards (Curator of the Bews Herbarium at the University of KwaZulu-Natal, Pietermaritzburg). A voucher specimen of the plant (S.E.D. No. 7) has been deposited in the Herbarium. 3.2. Preparation of K1 and K2 hydro-ethanol solutions The plant material, E. kraussianum rootstock (700 g), was milled and extracted with CH2Cl2 for 3 weeks to give

615

a brown powder (7.1 g). On a TLC plate (with CH2Cl2 as solvent), five fluorescent bands were clearly seen. K1 and K2 migrated at Rf values of 0.65 and 0.20, respectively. Using the process described previously (Drewes et al., 2002), K1 (73 mg) and K2 (230 mg) were isolated in crystalline form from 6.2 g of the starting material. The structures of both compounds were verified by spectroscopic techniques (described earlier by Drewes et al., 2002). The hydro-ethanol solutions of the plant’s extracts were prepared using 70:30 ethanol:water mixtures, and had an initial concentration of 1 mg/ml. 3.3. Animal material Healthy, male, young adult, Wistar rats (Rattus norvegicus) weighing 250–300 g, were used. The animals were kept and maintained under laboratory conditions of temperature, humidity, and light, and were allowed free access to food (standard pellet diet) and water ad libitum. The animals were divided into plant extract- and drug-treated ‘test’, and hydroethanol-treated ‘control’ groups (of 8 animals per group) in the in vivo experiments. All the animals were fasted for 16 h, but still allowed free access to water, before the commencement of our experiments. 3.4. Assessment of hypoglycaemic activity 3.4.1. Effects of K1 and K2 on blood glucose levels of normal rats Seven groups of 8 rats each were used for this assessment. Group 1 rats served as ‘untreated control’ for the other six groups of rats. After a four-day acclimatization period, during which time the animals were allowed free access to food (standard pellet diet) and water ad libitum, all the rats were fasted for 16 h. Initial fasting blood glucose levels (G0) of the Groups 1 to 7 rats were determined and recorded. Thereafter, Group 2 rats received 3 ml/kg (p.o.) of hydro-ethanol [distilled water:ethanol (3:7) mixture] vehicle, and Groups 3, 4 and 5 rats received graded doses of the plant’s extracts (K1 or K2, 20, 40 and 80 mg/kg p.o., respectively). Group 6 rats received glibenclamide (10 mg/kg p.o.), while Group 7 rats received sildenafil citrate (VIAGRAe, 100 mg/kg) by gastric intubation. The effects of the hydro-ethanol vehicle, graded doses of the pure compounds (K1 and K2), sildenafil citrate (VIAGRAe) and glibenclamide on the fasting blood glucose concentrations of the animals were then monitored and measured hourly for eight hours. 1, 2, 4 and 8 h following administration of the ‘test’ compounds to the animals, blood glucose concentrations (Gt) were determined. In each case and for each dose, the rats were restrained in a cage, and blood samples (0.02 ml) were collected from the tail tip vein of each rat for blood glucose analysis. Blood samples were obtained by repeated needle puncture of the same tail tip vein. Blood glucose concentrations were determined by means of Bayer’s Glucometer EliteÒ and compatible blood glucose test strips. Percentage glycaemic variation

616

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

was calculated as a function of time (t) by applying the formula: % glycaemic change ¼ G0
Gt  100=G0 ; where G0 and Gt represent glycaemic values before (i.e., zero time or 0 h glycaemic values), and glycaemic values at 1, 2, 4 and 8 h after, oral administrations of the ‘test’ compounds, respectively. Under the same experimental conditions, rats treated with hydro-ethanol vehicle (3 ml/ kg p.o.) alone were used as ‘control’ animals. 3.4.2. Oral glucose tolerance test (OGTT) Seven Groups of healthy, 16-h fasted, normal, young adult, male Wistar rats (R. norvegicus) were used. After measuring the initial blood glucose levels (G0) in all the animals in each of the seven groups, each of the rats in Group 7 received sildenafil citrate (VIAGRAe, 100 mg/kg p.o.), while each rat in Group 6 was treated with glibenclamide (10 mg/kg p.o.). Each rat in Groups 5, 4 and 3 received 80, 40 and 20 mg/kg (p.o.) of either K1 or K2, respectively, while each of the animals in Group 2 was treated with hydro-ethanol vehicle (3 ml/kg p.o.). Group 1 rats were not treated with anything at all, and the 8 rats in this Group 1 served as ‘untreated control’ for the rats in Groups 2–7. Twenty minutes following pretreatment with either VIAGRAe (100 mg/kg p.o.), glibenclamide (10 mg/kg p. o.), K1 or K2 (20, 40 and 80 mg/kg p.o.) or the vehicle (3 ml/kg p.o.), glucose (2 g/kg body weight) was orally administered (by gastric intubation) into each of the rats in Groups 2 to 7. Postprandial (i.e., post-glucose administration) blood glucose levels (Gt) were then monitored in, and recorded for, each of the rats in Groups 2– 7 at 15 min-intervals for the first two-and-a-half hours, and hourly thereafter for another 2 h following the oral glucose load. In each case and for each treatment dose, the rats were restrained in a cage, and blood samples (0.02 ml) were collected from the tail tip vein of each rat for blood glucose analysis. Blood samples were obtained by repeated needle puncture of the same tail tip vein. Blood glucose concentrations were determined by means of Bayer’s Glucometer EliteÒ and compatible blood glucose test strips. Percentage glycaemic variation was calculated as a function of time (t) by applying the formula:

Ojewole (1977). Healthy, male, Wistar rats weighing 200– 350 g were used. Each rat was sacrificed by decapitation. The abdomen of each animal was quickly opened by a midline incision, and the intestines were pulled to the left side of the animal. Portal veins with in situ lengths of approximately 20 mm were carefully cleaned free of connective, extraneous and fatty tissues, and then removed. The portal veins were separately suspended in 30-ml Ugo Basile TwoChambered Organ Baths (model 4050) containing Krebs– Henseleit physiological solution (of composition, in g/l: NaCl, 6.92; KCl, 0.34; NaH2PO4, 0.15; NaHCO3, 2.10; MgCl2, 0.11; CaCl2, 0.26; and glucose, 1.00 – pH adjusted to 7.4) maintained at 34 ± 1 °C and continuously aerated with carbogen (i.e., 5% carbon-dioxide + 95% oxygen gas mixture). Two isolated portal veins (one used as ‘control’ and the other one used as K1-, K2- or sildenafil citratetreated ‘test’ preparation) were always set-up to allow for changes in the venous muscle sensitivity. Each of the isolated venous muscle preparations was allowed to equilibrate for a period of 30–45 min under an applied resting tension of 0.5 g, before it was challenged with graded concentrations of K1, K2 or sildenafil citrate. Stepwise concentrations of K1, K2 or sildenafil citrate were applied to the bath-fluid either cumulatively or sequentially, and washed out 3–5 times after the maximum responses of the tissues were attained. Concentrations of K1, K2 or sildenafil citrate were repeated where appropriate and/or possible, at regular intervals of 20–30 min after the last washing. Because acetylcholine and histamine have been shown to relax vascular blood vessels through endothelium-dependent nitric oxide (NO) production and release (Yin et al., 2005; Luscher and Vanhoutte, 1986; Zawadzki et al., 1980), the effects of atropine sulphate (2 lg/ml) and mepyramine maleate (2 lg/ml) were examined on K1- and K2induced secondary relaxant responses of the isolated portal veins. The amplitude and frequency (rate) of the spontaneous, myogenic contractions, as well as the K1-, K2- or sildenafil citrate-induced responses of the isolated portal veins were recorded isometrically with the aid of Ugo Basile force–displacement transducers, a 2-Channel ‘‘Gemini’’ Recorder, and pen-writing microdynamometers (model 7070).

% glycaemic change ¼ G0
Gt  100=G0

3.6. Data analysis

where G0 and Gt represent glycaemic values before (i.e., zero time or 0 h glycaemic values), and glycaemic values at 1/4, 1/2, 3/4, 1, 11/4, 11/2, 13/4, 2, 4 and 8 h after, oral glucose load. Under the same experimental conditions, rats treated with hydro-ethanol vehicle (3 ml/kg p.o.) alone were used as ‘control’ animals.

Experimental data obtained from ‘test’ rats treated with the two pure compounds (K1 and K2, isolated from E. kraussianum), glibenclamide and VIAGRAe, as well as those obtained from hydroethanol-treated ‘control’ rats, or isolated portal vein preparations, were pooled and expressed as means (±SEM). Differences between the plant’s extract-, sildenafil citrate (VIAGRAe)- or glibenclamide-treated ‘test’ means, and the hydroethanol-treated ‘control’ means, were analysed statistically by one-way analysis o f variance (ANOVA), followed by Scheffe’s multiple comparison test. Values of P 6 0.05 were taken to imply statistical significance.

3.5. Evaluation of vasodilatory activity 3.5.1. Effects of K1 and K2 on rat isolated portal vein The experimental procedure used for the rat isolated portal vein was adopted from that described in detail by

J.A.O. Ojewole et al. / Phytochemistry 67 (2006) 610–617

Acknowledgements The technical assistance of Ms Kogi Moodley is thankfully acknowledged. Dr. Fatima Khan thanks the University of KwaZulu-Natal’s Research Office for the award of a Post-Doctoral Fellowship.

References Adzu, B., Amos, S., Kapu, S.D., Gamaniel, K.S., 2003. Anti-inflammatory and anti-nociceptive effects of Sphaeranthus senegalensis. Journal of Ethnopharmacology 84, 169–173. Akah, P.A., Okafor, C.L., 1992. Blood sugar lowering effect of Vernonia amygdalina (Del) in an experimental rabbit model. Phytotherapy Research 6, 171–173. Bryant, A.T., 1966. Zulu Medicine and Medicine-Men. Centaur Struik, Cape Town. Conti, C.R., Pepine, C.J., Sweeney, M., 1999. Efficacy and safety of sildenafil citrate in the treatment of erectile dysfunction in patients with ischaemic heart disease. American Journal of Cardiology 83, 29C–34C. Derry, F.A., Dinsmore, W.W., Fraser, M., et al., 1998. Efficacy and safety of oral sildenafil (VIAGRAe) in men with erectile dysfunction caused by spinal cord injury. Neurology 51, 1629–1633. Drewes, S.E., Horn, M.M., Munro, O.Q., Dhlamini, J.T.B., Meyer, J.J.M., Rakuambo, N.C., 2002. Pyrano-isoflavones with erectiledysfunction activity from Eriosema kraussianum. Phytochemistry 59, 739–747. Drewes, S.E., George, J., Khan, F., 2003. Recent findings on natural products with erectile-dysfunction activity. Phytochemistry 62, 1019– 1025. Dongmo, A.B., Kamanyi, A., Dzikouk, G., Chungag-Anye Nkeh, B., Tan, P.V., Nguelefack, T., Nole, T., Bopelet, M., Wagner, H., 2003. Anti-inflammatory and analgesic properties of the stem-bark extract of Mitragyna ciliata (Rubiaceae) Aubrev. and Pellegr. Journal of Ethnopharmacology 84, 17–21. Feldman, H.A., Goldstein, I., Hatzichristou, D.G., Krane, R.I., McKinlay, J.B., 1994. Impotence and its medical and psychological correlates: results of the Massachusetts male aging study. Journal of Urology 151, 54–61. Goldstein, I., Lue, T.F., Padma-Nathan, H., Rosen, R.C., Steers, W.D., Wicker, P.A., 1998. Oral sildenafil in the treatment of erectile dysfunction. New England Journal of Medicine 339, 59. Hulme, M.M., 1954. Wild Flowers of Natal. Shuter & Shooter, Pietermaritzburg.

617

Hutchings, A., Scott, A.H., Lewis, G., Cunningham, A., 1996. Zulu Medicinal Plants – An Inventory. University of Natal Press, Pietermaritzburg, pp. 145–146. Jackson, J.E., Bressler, R., 1981. Clinical pharmacology of sulphonylurea hypoglycaemic agents. Part I. Drugs 22, 211–245. Jarow, J.P., Burnett, A.L., Geringer, A.M., 1999. Clinical efficacy of sildenafil citrate based on aetiology and response to prior treatment. American Journal of Urology 162, 722–725. Li, R.W., Lin, G.D., Myers, S.P., Leach, D.N., 2003. Anti-inflammatory activity of Chinese medicinal vine plants. Journal of Ethnopharmacology 85, 61–67. Luscher, T.F., Vanhoutte, P.M., 1986. Endothelium-dependent contractions to acetylcholine in the aorta of the spontaneously hypertensive rat. Hypertension 8, 344–348. Marles, R.J., Farnsworth, N.R., 1995. Antidiabetic plants and their active constituents. Phytomedicine 2, 137–189. Ojewole, J.A.O., 1977. Studies on the Pharmacology of Some Antimalarial Drugs. PhD Thesis, University of Strathclyde, Glasgow, Scotland, UK. Ojewole, J.A.O., 2002. Hypoglycaemic effect of Clausena anisata (Willd) Hook methanolic root extract in rats. Journal of Ethnopharmacology 81, 231–237. Ojewole, J.A.O., 2003. Evaluation of the anti-inflammatory properties of Sclerocarya birrea (A. Rich.) Hochst. (family: Anacardiaceae) stem– bark extracts in rats. Journal of Ethnopharmacology 85, 217–220. Rendell, M.S., Rajfer, J., Wicker, P.A., Smith, M.D., 1999. Sildenafil for treatment of erectile dysfunction in men with diabetes. Journal of American Medical Association 281, 421–426. Taesotiku, T., Panthong, A., Kanjanapothi, D., Verpoorte, R., Scheffer, J.J.C., 2003. Anti-inflammatory, antipyretic and antinociceptive activities of Tabernaemontana pandacaqui Poir. Journal of Ethnopharmacology 84, 31–35. Watt, J.M., Breyer-Brandwijk, M.G., 1962. The Medicinal and Poisonous Plants of Southern and Eastern Africa, Second ed. E. & S Livingstone Ltd., Edinburgh and London, pp. 600–601. Webb, D.J., Freestone, S., Allen, M.J., Muirhead, G.J., 1999. Sildenafil citrate and blood-pressure-lowering drugs: results of drug interaction studies with an organic nitrate and a calcium antagonist. American Journal of Cardiology 83, 21C–28C. Yin, M.H., Kang, D.G., Choi, D.H., Kwon, T.O., Lee, H.S., 2005. Screening of vasorelaxant activity of some medicinal plants used in Oriental medicines. Journal of Ethnopharmacology 99, 113–117. Zawadzki, J.V., Cherry, P.D., Furchgott, R.F., 1980. Comparison of endothelium-dependent relaxations of rabbit aorta by A23187 and acetylcholine. Pharmacologist 22, 271. Zusman, R.M., Morales, A., Glasser, D.B., Osterloh, I.H., 1999. Overall cardiovascular profile of sildenafil citrate. American Journal of Cardiology 83, 35C–44C.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 618–621 www.elsevier.com/locate/phytochem

An angiotensin-I converting enzyme inhibitor from buckwheat (Fagopyrum esculentum Moench) flour Yasuo Aoyagi

*

Kagawa Nutrition University, 3-9-21 Chiyoda, Sakado, Saitama 350-0288, Japan Received 25 February 2005; received in revised form 23 November 2005 Available online 3 February 2006

Abstract A compound that inhibited angiotensin-I converting enzyme (ACE) activity was isolated from buckwheat powder. This compound is thought to be the hydroxy derivative of nicotianamine and its chemical structure is 200 -hydroxynicotianamine. This compound showed a very high inhibitory activity toward ACE, and the IC50 was 0.08 lM. Only this hydroxy analog was found in buckwheat powder, at about 30 mg/100 g, and no nicotianamine was detected. However, nicotianamine was detected in the buckwheat plant body. 200 -hydroxynicotianamine was also found in other polygonaceous plants.
2005 Elsevier Ltd. All rights reserved. Keywords: Buckwheat; Fagopyrum esculentum; Polygonaceae; ACE inhibitor; Hypotension; Non-protein amino acid; Phytosiderophore; Nicotianamine; 200 -Hydroxynicotianamine

1. Introduction Buckwheat (Fagopyrum esculentum Moench) flour is made into noodles and commonly eaten in Japan. Many nutraceutical compounds such as proteins, flavonoids, and other rare compounds have been reported in buckwheat seeds (Kayashita et al., 1996, 1999; Tomotake et al., 2001; Ikeda, 2002). Buckwheat protein in humans shows serum cholesterol-lowering activity, hypotensive activity, constipation-improvement and other effects. These physiological effects are due to its nature as a dietary fiber and the low digestibility of buckwheat protein (Kayashita et al., 1997). Furthermore, the anti-oxidative functions of its polyphenols protect against renal dysfunction caused by ischemia-reperfusion in rats (Yokozawa et al., 2001). Angiotensin-I converting enzyme (ACE) plays an important role in the renin-angiotensin system, which regulates blood pressure. Inhibitors of this enzyme lower blood pressure and many antihypertensive drugs such as

*

Tel./fax: +81 49 282 3709. E-mail address: [email protected].

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.022

captopril and enapril are potent ACE inhibitors. It has been reported that the water extract of buckwheat flour strongly inhibits ACE (Suzuki et al., 1983). Furthermore, it has been shown that the hydrolyzate of buckwheat protein by a protease shows strong ACE interference (Kawakami et al., 1995), and biologically active peptides have been identified (Li et al., 2002). However, the bioactive plant compound in the aqueous extract has still not been identified. This study was performed to determine its identity.

2. Results 2.1. Isolation of ACE inhibitor The results of a preliminary experiment showed that the ACE-inhibitory principle of buckwheat was soluble in water and the major ACE-interference activity was in the dialysis extracellular fluid. Therefore, it was thought that the active components were of comparatively low molecular weight. Extraction with 70% ethanol was then carried out to avoid elution of the starch by gelatinization.

Y. Aoyagi / Phytochemistry 67 (2006) 618–621

The ACE inhibitor in the extract was not extracted by hexane or EtOAc. Most of the inhibitory activity remained in the water layer, and the IC50 value was estimated to be 31.5 lg ml
1. Based on separation of this aqueous fraction by a coupled column of cation- and anion-exchange resins, the bioactive compound was found to exist in the positive ion fraction, and not in other fractions. The IC50 value of the fraction was estimated to be 3.1 lg ml
1. In addition, this positive ion fraction was fractionated into P-1, P-2, P-3 and P-4 using an anion-exchange resin [Dowex 1 · 4 (AcO
)] column. While ACE-inhibitory activity was observed in both P-1 (IC50 = 23.6 lg ml
1) and P-2 (IC50 = 0.37 lg ml
1), most of the activity was in P-2. The active compound 1 (57 mg from 2 kg of buckwheat flour) was isolated from P-2 following several more purifications by chromatography and recrystallization. Compound 1 showed a single ninhydrin-positive spot in TLC, and a single peak in the vicinity of valine on the chromatogram in an automatic amino acid analyzer. 2.2. Chemical structure of 1 Compound 1 was shown to be C12H21O7N3, with a molecular weight of 319 based on MS analysis. The ninhydrin reaction of 1 was completely masked by Cu2+, suggesting that it had an a-amino group. In addition, 1 was estimated to be an amino acid which is easily decomposed by heating in 6 N HCl, since a complicated fragment was present without the formation of a usual amino acid by hydrolysis. The 13C NMR spectrum including DEPT measurement showed that 1 had three carbonyl, four methine and five methylene carbon atoms. A detailed examination of the 1H NMR spectra by measurements of the H–H COSY and C–H COSY spectra, showed that 1 had one – CH–CH–CH2– and two –CH2–CH2–CH– moieties. These data suggested that 1 was a hydroxy derivative of nicotianamine (Sugiura and Nomoto, 1984). The positive product ion spectrum of the MS/MS measurement of 1 is shown in Fig. 1. The prominent ions at m/z 56 and 114 were

619

attributable to an azetidine ring moiety, as seen in the spectra of mugineic acid and nicotianamine (Kenny and Nomoto, 1994). The position of the hydroxyl group was deduced from ions at m/z 128, 186, 201, 219 and 245, and was determined to be at the 200 position. Therefore, 1 was shown to be 200 -hydroxynicotianamine. 2.3. Quantitative analysis of 1 in buckwheat and other polygonaceae plants Nicotianamine and 1 were well-separated by ion-pair HPLC on an ODS column using sodium dodecyl sulfate as the counter ion.While only 1 was found in buckwheat powder, both 1 and nicotianamine were found in the buds of buckwheat and other polygonaceous plants (Table 1). 2.4. ACE-inhibitory activities of 1 and some related compounds The ACE-inhibitory activities of 1, azetidin-2-carboxylic acid, 2(S),3 0 (S)-N-(3-amino-3-carboxylpropyl)azetidine-2carboxylic acid, nicotianamine, mugineic acid (Takemoto et al., 1978) and 2 0 -deoxymugineic acid (Sugiura and Nomoto, 1984) are shown in Table 2. The IC50 value of 1 was as low as 0.08 lM, and its inhibitory effect was equivalent to that of nicotianamine. The interference activities of mugineic acid, 3 0 2(S),3 0 (S)-N-(3-amino-3-carboxylpropyl)azetidine-2-carboxylic acid and 2 0 -deoxymugineic acid were considerably lower and azetidine-2-carboxylic acid did not show any inhibition.

Table 1 1 and nicotianamine contents of the polygonaceous plants

Fagopyrum esculentum, flour Fagopyrum esculentum, sprout Rheum rhaponticum, stem Rumex acetosa, young bud Polygonum hydropiper, sprout

1 (mg/100 g dry)

Nicotianamine (mg/100 g dry)

30 48 6 3 52

– 18 11 34 –

Table 2 Angiotensin-I converting enzyme inhibition of 1 and azetidin-2-carboxiric acid derivatives

Fig. 1. Product ion spectra and proposed fragmentation scheme of the [M + H]+ ion at m/z 320 of 1.

Compound

ACE inhibition IC50 (lM)

(2S)-azetidine-2-carboxylic acid (2S)-1-((3S)-3-amino-3-carboxypropyl) azetidine-2-carboxylic acid Nicotianamine 1 Mugineic acid 2 0 -Deoxymugineic acid

(–) 3.3 0.085 0.08 0.28 >99

620

Y. Aoyagi / Phytochemistry 67 (2006) 618–621

3. Discussion

4.3. Assay for ACE-inhibitory activity

Nicotianamine occurs in all plants and chelates metal cations. It is thought to play a role in the internal transport of Fe and other metals in plants (Scholz et al., 1992). Concerning the hydroxyl derivative of nicotianamine, although there has been a report of the synthesis of 2 0 -hydroxynicotianamine (Matuura et al., 1994), this is the first report of 200 -hydroxynicotianamine (1) in a natural product. The physiological role of 1 in plants, while intriguing, is still unclear. Compound 1 was isolated from buckwheat powder, and has also been shown to exist in some polygonaceous plants. It is thought to be widespread in polygonaceous plants. Some other families plants so far analyzed contained no detectable amounts of 1. Thus, 1 may be a useful chemotaxonomic marker of polygonaceous plants. Attempts to detect 1 in other plants using she LCMS technique are also in progress from the viewpoint chemotaxonomic interest. Nicotianamine shows a strong ACE-inhibitory effect, and has been shown to have a blood pressure-lowering effect in SHR (Kinoshita et al., 1993; Shimizu et al., 1999). Compound 1 shows an almost equal ACE-inhibitory effect, and a similar blood pressure-lowering effect can be expected. Studies along these lines are now in progress, and the results will be reported soon.

ACE-inhibitory activity was assayed by the modified method of Cushman and Cheung (1971). A reaction mixture containing 500 ll of 7 mM HHL in pH 8.3 borate buffer (200 mM), 400 ll of 2 M NaCl, 40 ll of H2O, 30 ll of the test sample solution or H2O and 30 ll of 150 U/ml of ACE (in the pH 8.3 buffer) was incubated at 37 C for 30 min. The reaction was stopped by adding 500 ll of 1 N HCl. The hippuric acid liberated from the HHL by ACE was extracted with EtOAc (3 ml). An aliquot of the extract (2 ml) was evaporated to dryness, and the residue was dissolved in H2O (1 ml). The hippuric acid concentration was determined by measuring the UV absorbance at 228 nm. The inhibitory activity was calculated as [100(Ec
Es)/(Ec
Eb)], where Es is the absorbance of the test sample added to the reaction mixture, Ec is with water (instead of the test sample), and Eb is a blank without ACE. The activity of the ACE-inhibitory principle was shown by the molarity in the reaction mixture, which showed interference of 50% (IC50) under these conditions.

4. Experimental 4.1. Materials Buckwheat flour was made in Hokkaido, Japan. ACE (from rabbit lung) was purchased from Sigma Chemical Co. (St. Louis, MO, USA) and hippuryl-L-histidyl-L-leucine (HHL) was obtained from the Peptide Institute (Osaka, Japan). Azetidine-2-carboxylic acid was isolated from the fruiting bodies of Clavulinopsis pulchra (Peck) Corner. 2(S),3 0 (S)-N-(3-amino-3-carboxylpropyl)azetidine2-carboxylic acid was prepared from the azetidin-2-carboxylic acid according to the method of Kristensen and Larsen (1974). Nicotianamine was separated by the method reported for Morokheiya (Corchorus olitorius) (Kimoto et al., 1998). Mugineic acid and 2 0 -deoxymugineic acid were donated by Dr. S. Mori and Dr. N.-K. Nishizawa of The University of Tokyo. 4.2. General TLC was performed on a Merck silica gel 60 F254 plate in n-BuOH–HOAc–H2O (4:1:2) (solvent 1) and n-PrOH– conc. NH3 (7:3) (solvent 2). MS spectra were measured using a JEOL MS 700V spectrometer. NMR spectra were recorded using a JEOL EX-270 spectrometer with TMSP as the internal standard. AA analysis was recorded using a Hitachi 835 automatic amino acid analyzer.

4.4. Preliminary survey of ACE inhibitor Boiling H2O (100 ml) was added to buckwheat flour (10 g), and the mixture was heated in a bath at 100 C 30 min. The aqueous extract was obtained by centrifuging at 8000 rpm for 15 min. The ACE-inhibitory activity was measured using part of this extract. The remaining extract was transferred to a dialysis tube, and dialyzed against 1 l of H2O for 24 h. The dialysis internal fluid and extracellular fluid were concentrated, and then diluted to the original amount. The ACE-inhibitory activity of both of these liquids was also measured. 4.5. Isolation of ACE inhibitor Buckwheat flour (2.1 kg) was homogenized with EtOH–H2O (8 l, 7:3), and stored overnight. The homogenate was filtered through filter paper, and the residue was similarly extracted two times. These filtrates were collected and conc. to ca. 2 l (165 g) using a rotary evaporator. This concd. extract was defatted with hexane (3 · 500 ml). The aqueous layer was extracted three more times with 500 ml EtOAc, and separated into an EtOAc layer (6.75 g) and an aqueous layer (95.5 g). The aqueous layer was then passed through columns of a cationexchange resin Amberlite CG-120 (H+) column (8 cm · 60 cm) and an anion-exchange resin Amberlite CG-400 (OH
) column (5 cm · 70 cm) in series. The columns were washed with distilled H2O (10 l) and the wash liquid was conc. to 100 ml as the non-ionic fraction (35.2 g). From the separation column with the cationexchange resin, the positive-ion fraction was eluted using 2 N aqueous ammonia (101, 15.0 g). From the anionexchange resin column, the negative-ion fraction was

Y. Aoyagi / Phytochemistry 67 (2006) 618–621

eluted with 2 N aqueous ammonia (71, 7.8 g). These fractions were each conc. to 100 ml and the ACE-inhibitory activity was determined. The positive-ion fraction, which only showed ACE-inhibitory activity, was applied to a column of Dowex 1 · 4 resin (AcO
, 3 · 120 cm). This column was eluted in sequence with 0.1 N AcOH (1 l) to give a fraction of basic and neutral amino acids (P1, 7.9 g), with 0.5 N AcOH (1 l) to give a fraction of neutral amino acids (P-2, 1.9 g), with 1.0 N AcOH (1 l) to give a fraction of weakly acidic amino acids (P-3, 4.0 g) and with 2 N AcOH (2 l) to give a fraction of strong acidic amino acids (P-4, 1.4 g). Only the P-2 fraction showed ACE-inhibitory activity, and it was then applied to a column of Dowex 1 · 4 (AcO
, 3 · 120 cm). Elution was performed with 0.3 N AcOH and 10 ml fractions were collected. Each fraction was monitored by the ninhydrin reaction and assay of the ACE-inhibitory activity. The active fractions were collected and conc. to dryness using a rotary evaporator. Recrystallization of the residues from aqueous EtOH gave a colorless crystalline 1 (57 mg). 4.6. Identification of ACE inhibitor Compound 1: m.p. 275–280 C (decomp.), ½a235 D
40 (c 235 0.2, H2O), ½aD
34.5 (c 0.1, 1 N HCl), 1H NMR (D2O) d 2.05–2.3 (2H, m, H-2 0 ), 2.4–2.85 (2H, m, H-3), 3.25– 3.53 (4H, m, H-100 and H-1 0 ), 3.82–3.87 (1H, m, H-3 0 ), 3.92–4.15 (2H, m, H-4), 4.01 (1H, d, 3.3 Hz, H-300 ), 4.42– 4.48 (1H, m, H-200 ), 4.77 (1H, t, 9.57 Hz, H-2), 13C NMR (D2O) d 23.9 (C-3, CH2), 27.6 (C-2 0 , CH2), 51.2 (C100 ,CH2), 53.4 (C-4, CH2), 54.1 (C-1 0 , CH2), 60.5 (C-300 , CH), 62.4 (C-3 0 , CH), 68.3 (C-200 , CH), 69.7 (C-2, CH), 173.3 (C-400 , CO), 174.9 (C-4 0 , CO), 176.0 (C-1, CO), ESI-MS m/z: 320 (M + H), HRFABMS m/z: 320.1457 (M + H); Calcd. for C12H22O7N3: 320.1458. 4.7. Quantitative analysis of the active compound A sample (about 2 g) was extracted three times with 20 vol. of EtOH–H2O (7:3) by refluxing. The combined extract was concentrated to a small volume and defatted by EtOEt. The aqueous layer was applied to a column (2 · 10 cm) of Dowex 1 · 4 (AcO
) resin, washed with H2O (50 ml) and eluted with 0.5 N AcOH (100 ml). The eluate was conc. to dryness and dissolved in a suitable amount of HPLC eluting solvent. The HPLC conditions were as follows. Equipment: Shimadzu LC-7 with a fluorescence detector and a post-column reaction instrument, Column: Toso ODS 80TM, solvent: 0.05 M sodium dodecyl sulfate (pH 2.3) containing 30% CH3CN, flow rate: 0.75 ml min
1; post-column labeling agent: 900 ml of 0.1 M borate buffer (pH 9.0), 100 ml of EtOH containing 100 mg of o-phthalaldehyde and 0.5 ml of 2-mercaptoethanol; flow rate 1 ml min
1. Fluorescence detector: Ex 365 nm, Em 455 nm.

621

References Cushman, D.W., Cheung, H.S., 1971. Spectrophotometric assay and properties of the angiotensin-converting enzyme of rabbit lung. Biochem. Pharmacol. 20, 1637–1648. Ikeda, K., 2002. Buckwheat: composition, chemistry, and processing. Adv. Food Nutr. Res. 44, 395–434. Kawakami, A., Isobe, T., Kayahara, H., Horii, A., 1995. Preparations of enzymatic hydrolysates of buckwheat globulin and their angiotensinconverting enzyme inhibitory activities. Curr. Adv. Buckwheat Res., 927–934. Kayashita, J., Nagai, H., Kato, N., 1996. Buckwheat protein extract suppression of the growth depression in rats induced by feeding amaranth (Food Red No. 2). Biosci. Biotechnol. Biochem. 60, 1530– 1531. Kayashita, J., Shimaoka, I., Nakajyo, M., Yamazaki, M., Kato, N., 1997. Consumption of buckwheat protein lowers plasma cholesterol and raises fecal neutral sterol in cholesterol-fed rats because of its low digestibility. J. Nutr. 127, 1395–1400. Kayashita, J., Shimaoka, I., Nakajoh, M., Kondoh, M., Hayashi, K., Kato, N., 1999. Muscle hypertrophy in rats fed on a buckwheat protein extract. Biosci. Biotechnol. Biochem. 63, 1242–1245. Kenny, P.T.M., Nomoto, K., 1994. Tandem mass spectrometric investigation of the phytosiderophores mugineic acid, deoxymugineic acid and nicotianamine. Analyst 119, 891–895. Kimoto, K., Kuroda, Y., Saito, Y., Yamamoto, J., Murakami, T., Aoyagi, Y., 1998. Purification and identification of angiotensin-converting enzyme inhibitor from morokheiya (Corchorus olitorius). Food Sci. Technol. Int. Tokyo 4, 223–226. Kinoshita, E., Yamakoshi, J., Kikuchi, M., 1993. Purification and identification of an angiotensin-converting enzyme inhibitor from soy sauce. Biosci. Biotechnol. Biochem. 57, 1107–1110. Kristensen, I., Larsen, P.O., 1974. Azetidine-2-carboxylic acid derivatives from seed of Fagus silvatica L. and a revised structure for nicotianamine. Phytochemistry 13, 2791–2798. Li, C.H., Matsui, T., Matsumoto, K., Yamasaki, R., Kawasaki, T., 2002. Latent production of angiotensin I-converting enzyme inhibitors from buckwheat. J. Pept. Sci. 8, 267–274. Matuura, F., Hamada, Y., Shioiri, T., 1994. Total syntheses of phytosiderophores, 3-epi-hydroxymugineic acid, distichonic acid A, and 2 0 hydroxynicotianamine. Tetrahedron 50, 265–274. Scholz, G., Becker, R., Pichi, A., Stephan, U.W., 1992. Nicotianamine-A common constituent of strategies and of iron acquisition by plants: a review. J. Plant Nutr. 15, 1647–1665. Shimizu, E., Hayashi, A., Takahashi, R., Aoyagi, Y., Murakami, T., Kimoto, K., 1999. Effects of angiotensin I-converting enzyme inhibitor from ashitaba (Angelica keiskei) on blood pressure of spontaneously hypertensive rats. J. Nutr. Sci. Vitaminol. 45, 375–383. Sugiura, Y., Nomoto, K., 1984. Phytosiderophores structure and properties of mugineic acids and their metal complexesStructure and Bonding, vol. 58. Springer, Heidelberg, pp. 107–135. Suzuki, T., Ishikawa, N., Meguro, H., 1983. Angiotensin-converting enzyme inhibiting activity in foods. Nippon Nogeikagaku Kaishi 57, 1143–1146. Takemoto, T., Nomoto, K., Fushiya, S., Ouchio, R., Kusano, G., Hikino, H., Takagi, S., Matsuura, Y., Kakudo, M., 1978. Structure of mugineic acid, a new amino acid possessing an iron-chelating activity from roots washings of water-cultured Hordeum vulgare L. Proc. Jpn. Acad. 54, B469–B473. Tomotake, H., Shimaoka, I., Kayashita, J., Yokoyama, F., Nakajoh, M., Kato, N., 2001. Stronger suppression of plasma cholesterol and enhancement of the fecal excretion of steroids by a buckwheat protein product than by a soy protein isolate in rats fed on a cholesterol-free diet. Biosci. Biotechnol. Biochem. 63, 1412–1414. Yokozawa, T., Fujii, H., Kosuna, K., Nonaka, G., 2001. Effects of buckwheat in a renal ischemia-reperfusion model. Biosci. Biotechnol. Biochem. 65, 396–400.

PHYTOCHEMISTRY Phytochemistry 67 (2006) 622–629 www.elsevier.com/locate/phytochem

Structure of anthocyanin from the blue petals of Phacelia campanularia and its blue flower color development Mihoko Mori a, Tadao Kondo b, Kenjiro Toki c, Kumi Yoshida

d,*

a

b

Graduate School of Human Informatics, Nagoya University, Chikusa, Nagoya 464-8601, Japan Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya 464-8601, Japan c Laboratory of Floriculture, Minami-Kyusyu University, Takanabe, Miyazaki 844-0003, Japan d Graduate School of Information Science, Nagoya University, Chikusa, Nagoya 464-8601, Japan Received 21 September 2005; received in revised form 16 November 2005 Available online 7 February 2006

Abstract The dicaffeoyl anthocyanin, phacelianin, was isolated from blue petals of Phacelia campanularia. Its structure was determined to be 3-O-(6-O-(4 0 -O-(6-O-(4 0 -O-b-D-glucopyranosyl-(E)-caffeoyl)-b-D-glucopyranosyl)-(E)-caffeoyl)-b-D-glucopyranosyl)-5-O-(6-O-malonylb-D-glucopyranosyl)delphinidin. The CD of the blue petals of the phacelia showed a strong negative Cotton effect and that of the suspension of the colored protoplasts was the same, indicating that the chromophores of phacelianin may stack intermolecularly in an anti-clockwise stacking manner in the blue-colored vacuoles. In a weakly acidic aqueous solution, phacelianin displayed the same blue color and negative Cotton effect in CD as those of the petals. However, blue-black colored precipitates gradually formed without metal ions. A very small amount of Al3+ or Fe3+ may be required to stabilize the blue solution. Phacelianin may take both an inter- and intramolecular stacking form and shows the blue petal color by molecular association and the co-existence of a small amount of metal ions. We also isolated a major anthocyanin from the blue petals of Evolvulus pilosus and revised the structure identical to phacelianin.
2005 Elsevier Ltd. All rights reserved. Keywords: Phacelia campanularia; Evolvulus pilosus; Hydrophyllaceae; Convolvulaceae; Phacelianin; Diacylated anthocyanin; Blue flower color development; Circular dichroism

1. Introduction Most flower colors, especially blue colors, are due to anthocyanins (Goto and Kondo, 1991; Brouillard and Dangles, 1994). However, little is known about the mechanism of blue color development of anthocyanins; the blue dayflower (Kondo et al., 1992) and the blue cornflower (Kondo et al., 1994, 1998) acquire their color by a supramolecular metal complex, and the blue morning glory by an increase in vacuolar pH to 7.7 (Yoshida et al., 1995). Since Goto et al. (1982) reported the structure of gentiodelphin from the blue petals of Gentiana makinoi, many polyacylated anthocyanins, which contain two or more aromatic acyl residues, were isolated and structurally eluci*

Corresponding author. Fax: +81 52 789 5638. E-mail address: [email protected] (K. Yoshida).

0031-9422/$ - see front matter
2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.phytochem.2005.12.024

dated (Goto, 1987; Goto and Kondo, 1991). Now, many hundreds of polyacylated anthocyanins are known, and a fair percentage of them is from blue petals; therefore, polyacylated anthocyanins may contribute greatly to blue flower color development. Although the stability and blue color development of polyacylated anthocyanins are expected to be established with the intramolecular stacking of aromatic acyl residues to the anthocyanidin nucleus (Goto and Kondo, 1991; Yoshida et al., 1992, 2000; Brouillard and Dangles, 1994), their stacking structure and the mechanism of intramolecular charge-transfer, which is thought to derive from the bathochromic shift in the absorption spectrum, are obscure. Recently, we have been focusing our research on flower color development in colored cells and established several methodologies such as the simultaneous measurement of the absorption spectrum of a single cell (Yoshida et al.,

M. Mori et al. / Phytochemistry 67 (2006) 622–629

623

2003a,b) and vacuolar pH (Yoshida et al., 1995, 2003a,b), and the reproduction of cell color by mixing components in colored cells (Yoshida et al., 1995, 2003a,b). In this report, we describe the structure of a newly isolated polyacylated anthocyanin, phacelianin (1), in the blue petals of Phacelia campanularia. An interesting mechanism for blue color development of 1 is also discussed. We revised the structure of the major anthocyanin in blue petals of Evolvulus pilosus to be the same as 1.

2. Results and discussion 2.1. Isolation and structure of anthocyanins in petals of P. campanularia Blue petals of P. campanularia were extracted with aqueous acetonitrile (CH3CN) containing trifluoroacetic acid (TFA), and the extract was analyzed by HPLC. As shown in Fig. 1, the composition of the petal extract is relatively simple, containing only two anthocyanins (1 and 2) with a small amount of a UV-absorbing compound (3). The results of photodiode array detection-HPLC indicated that 1 and 2 may have the same chromophore and two aromatic acyl residues. Compound 3 may not be a flavonoid but a cinnamic acid derivative. The components from the petals (300 g) were purified according to our general procedure (Yoshida et al., 1992, 1996, 2002, 2003b) to give 1 (170 mg), 2 (24 mg) and 3 (24 mg). The structures of 1 and 2 were determined by FABMS and various 1D and 2D NMR experiments combined with degradation reactions (see Scheme 1). FABMS of 1 gave a molecular ion peak at m/z = 1361. The 1H NMR spectrum of 1 showed the existence of a delphinidin nucleus, four sugars and two E-caffeoyl residues (Table 1). The molecular weight was larger by 86 Da than that estimated by 1H NMR analysis, strongly suggesting that the remaining residue is malonyl group. The molecular ion peak of 2 was 1275, which was 86 mass units smaller than that of 1, corresponding to a loss of a malonyl residue. The 1H NMR spectrum of 2 was very similar to that of 1, but only the signals of one sugar residue (glc-2) were different. Furthermore, acidic methanol treatment of 1 gave 2 via methyl

Scheme 1. Structure of 1 and 2.

ester of 1 being very similar in behavior to malonylshisonin (Kondo et al., 1989; Yoshida et al., 1997). These data strongly suggest that 2 is a demalonylated pigment from 1. By 1D- and 2D TOCSY experiments (Kondo et al., 1990), the signals of all sugar residues of 1 were assigned so that all were b-glucopyranosides (Table 1). The linkage of glc-1 to the 3 position and glc-2 to the 5 position of the delphinidin nucleus was confirmed by the NOEs between the anomeric proton of glc-1 and H-4, and that of glc-2 and H-6, respectively (Fig. 2). The methylene protons of the sugar residues (glc-1, 2 and 3) showing a downfield shift indicated that the 6-positions of glc-1, 2 and 3 were acylated. To determine the linkage of the acyl groups, an HMBC experiment was conducted (Fig. 2). One of the Ecaffeoyl residues (C-1) was esterified to the 6-OH of the glc-1 because of the correlation between the a proton of C-1 and H-6 of glc-1, and the other caffeoyl residue (C-2) was connected to 6-OH of the glc-3 because of the correlation between the a proton of C-2 and H-6 of glc-3 (Fig. 2). Therefore, the substituted position of the malonic acid was determined to be the 6-OH of glc-2. This was confirmed from the correlation between H-6 of glc-2 and the 13C signal of 168.5 ppm assigned to be the carbonyl carbon of the malonyl residue.

Fig. 1. HPLC chromatograms (detection: 280 nm) of the extract from the blue petals of Phacelia campanularia (left) and the spectra of each peak obtained with 3D-detection HPLC (right).

624

M. Mori et al. / Phytochemistry 67 (2006) 622–629

Table 1 Assignment of the 1H and

13

C NMR spectra of 1 and 2 (1H: 600 MHz,13C: 150 MHz, 23 C, 10% TFA d-CD3OD) 1

2

1

H d (ppm)

2 3 4 5 6 7 8 9 10

8.70 s

10 2 0 ,6 0 3 0 ,5 0 40 C-1

C-2

J (Hz)

1 2 3 4 5 6 a b C@O

H d (ppm)

J (Hz)

8.66 s

7.00 d

2.0

7.00 d

2.0

6.76 d

2.0

6.68 d

2.0

7.64 s

1 2 3 4 5 6 a b C@O

1

1

7.57 s

6.91 d

2.0

6.98 d

2.0

7.03 6.75 6.37 7.31

8.5 8.5, 2.0 16.0 16.0

7.03 6.69 6.35 7.27

8.5 8.5, 2.0 16.0 16.0

6.81 d

2.0

6.79 d

2.0

6.84 6.43 6.17 7.25

d dd d d

8.5 8.5, 2.0 16.0 16.0

6.82 6.36 6.15 7.20

d dd d d

8.0 8.0, 2.0 16.0 16.0

d dd d d

d dd d d

13

2 C d (ppm)

13

C d (ppm)

164.3 146.2 132.8 156.4 106.2 168.7 97.6 156.3 112.8

164.1 146.3 133.2 156.6 105.9 168.9 97.6 156.3 112.7

119.7 113.3 147.5 146.4

119.6 113.3 147.4 146.1

130.9 117.5 148.7 148.5 118.6 121.5 117.3 146.5 168.8

130.8 116.8 148.6 148.5 119.1 122.3 117.2 146.4 168.7

130.4 115.3 148.3 148.7 117.2 122.0 117.2 145.8 168.5

130.4 115.3 148.3 148.7 117.2 122.0 117.2 145.8 168.5

glc-1

1 2 3 4 5 6a 6b

5.36 3.90 3.63 3.54 4.03 4.58 4.53

d dd t dd td dd dd

8.0 9.0, 8.0 9.0 9.5, 9.0 9.5, 2.5 11.5, 9.5 11.5, 2.5

5.37 3.90 3.64 3.56 4.01 4.64 4.48

d dd t t td dd dd

8.0 9.0, 8.0 9.0 9.0 9.0, 3.0 11.5, 9.0 11.5, 3.0

102.6 74.1 78.3 72.7 75.4 65.0

102.8 74.1 78.3 72.6 75.5 64.8

glc-2

1 2 3 4 5 6a 6b

5.09 3.78 3.53 3.42 3.73 4.53 4.23

d dd t t ddd dd dd

7.5 9.0, 7.5 9.0 9.0 9.0, 6.5, 2.5 11.5, 2.5 11.5, 6.5

5.12 3.78 3.55 3.37 3.61 4.01 3.64

d dd t t ddd dd dd

7.5 9.0, 7.5 9.0 9.0 9.0, 7.0, 1.5 11.5, 1.5 11.5, 7.0

103.4 74.9 78.1 70.9 76.1 65.1

103.5 74.9 78.0 71.4 79.2 62.8

glc-3

1 2 3 4 5 6a 6b

4.92 3.57 3.56 3.40 3.84 4.67 4.21

d dd dd dd td dd dd

7.5 8.5, 7.5 9.0, 8.5 9.5, 9.0 9.5, 2.0 11.5, 2.0 11.5, 9.5

4.92 3.58 3.55 3.39 3.82 4.65 4.19

d dd t dd td dd dd

7.5 9.0, 7.5 9.0 9.5, 9.0 9.5, 2.0 11.5, 2.0 11.5, 9.5

103.2 74.7 77.7 72.4 75.5 65.4

103.2 74.7 77.8 72.5 75.5 65.4

glc-4

1 2 3 4 5 6a 6b

4.90 3.54 3.55 3.44 3.56 3.89 3.74

d t dd dd m dd dd

8.5 8.5 9.0, 8.5 9.5, 9.0

4.91 3.53 3.57 3.47 3.58 3.89 3.75

d dd t dd m dd dd

7.5 9.0, 7.5 9.0 9.5, 9.0

102.6 74.7 77.8 71.2 77.6 62.4

102.5 74.7 77.5 71.1 78.0 62.3

Malonyl

12.5, 2.5 12.5, 5.0

C@O (carboxylate) C@O (carboxylic acid)

The connections between glc-3 to 4-OH of C-1 and glc-4 to 4-OH of C-2 were strongly indicated by the NOEs between the anomeric proton of glc-3 to 5-H of C-1 and

12.5, 2.0 12.5, 5.0

168.5 170.6

that of glc-4 to 5-H of C-2 (Fig. 2). However, the latter connection was not confirmed by the HMBC experiments because it was difficult to differentiate the 13C signals of

M. Mori et al. / Phytochemistry 67 (2006) 622–629

625

In the survey of polyacylated anthocyanins in blue flower petals, we found that the HPLC data of the major pigment of the blue petals of E. pilosus were identical to those of 1. However, Toki et al. (1994) reported a different structure for the petal anthocyanin. Therefore, we isolated anthocyanin from the blue petals of E. pilosus and compared it with 1 and the authentic pigment. The HPLC data and 1H NMR spectral data were identical to those of 1, resulting in the structure of the pigment in E. pilosus and P. campanularia being identical to 1. The structure of 3 was deduced to be 3-O-(E)-caffeoylquinic acid (chlorogenic acid) by comparison with an authentic sample. 2.2. VIS absorption spectra and CD of petals and colored cells of P. campanularia

Fig. 2. NOE and HMBC correlation of 1.

the 3 and 4 positions of C-2. Therefore, we conducted alkaline hydrolysis of 1. Reaction of 1 in aq. NaOH-methanol under an argon (Ar) atmosphere gave a mixture of 4-O-glucosyl-(E)-caffeic acid (6) and methyl 4-O-glucosyl-(E)-caffeate (7). The structure was confirmed by NMR, MS experiment and comparison with authentic sample (Goto et al., 1981). The combined yield of 6 and 7 was quantitative (195%) to 1; therefore, the linkage of glc-3 and 4 was determined to be at the 4-OH of caffeoyl residues, C-1 and C-2, respectively. Thus, 1 was deduced to be 3-O-(6O-(4 0 -O-(6-O-(4 0 -O-b-D-glucopyranosyl-(E)-caffeoyl)-b-Dglucopyranosyl)-(E)-caffeoyl)-b-D-glucopyranosyl)-5-O-(6O-malonylb-D-glucopyranosyl)delphinidin. The structure of 2 was determined by using the same MS and NMR techniques (Table 1) to be 3-O-(6-O-(4 0 -O-(6-O(4 0 -O-b-D-glucopyranosyl-(E)-caffeoyl)-b-D-glucopyranosyl)(E)-caffeoyl)-b-D-glucopyranosyl)-5-O-b-D-glucopyranosyldelphinidin. By NOESY and ROESY experiments long range NOEs between the anomeric proton of glc-4 and the protons at 2 0 and 6 0 of the delphinidin nucleus were observed (Fig. 2) Furthermore, the protons at 2, 5 and 6 positions in the both caffeoyl residues (C-1 and C-2) shifted toward 0.16– 0.61 ppm upfield as compared with the corresponding protons of methyl 4-O-glucosyl-(E)-caffeate (7) (Table 1 and data in Section 3.2.2). These data indicate that the caffeoyl residues of 1 and 2 stack to the anthocyanidin nucleus intramolecularly in TFA d-CD3OD just as the same as gentiodelphin (Yoshida et al., 1991, 2000). Since intramolecular stacking of acyl residues of polyacylated anthocyanins to the anthocyanidin nucleus becomes stronger in neutral aqueous condition than that in acidic methanol (Yoshida et al., 1991, 2000), the caffeoyl residues of phacelianin (1) and its demalonylated pigment (2) should stack to the delphinidin nucleus in the petal cells.

To help understand how pigment molecules are present in a living petal cell, we measured the VIS absorption spectra and CD of a living petal and colored cells. The CD can provide very important information on molecular association, especially the stacking manner of the chromophores (Goto and Kondo, 1991; Kondo et al., 1992). Thus, the CD recording of intact flower petals were also conducted by Hoshino (1986). However, CD measurement of intact petals has some difficulties; the colorless intercellular space filled with air causes diffuse reflections following low sensitivity with a noisy spectrum. To overcome these problems, we applied a single cell measurement using the microspectrophotometric method for recording the VIS spectrum. As another solution to reduce scattered reflection, we evacuated the petals, filling the intercellular space with water. The obtained petals became more transparent, and the VIS spectrum and CD measurements could then be conducted more noiselessly with higher sensitivity. We also purified the blue-colored cells and CD of the gathered cells was recorded in a quartz cuvette. Fig. 3 shows the VIS absorption spectra and CD of fresh petals and colored cells from the blue petals of P. campanularia. The absorption spectrum of the petals showed three kmax around 638, 578 and 546 nm, coinciding with that of a single protoplast. In the CD, a negative Cotton effect was observed at the absorption maximum at the VIS region, and there was no difference between the CD of the evacuated petal and the suspension of the colored protoplasts (Fig. 3). The negative Cotton effect strongly suggests chiral self-association of the anthocyanidin nuclei of 1 in an anti-clockwise stacking manner. Until now, many polyacylated anthocyanins have been isolated and their CD spectra recorded. However, no pigment showed such a CD with Cotton effects. Generally, aromatic acyl residues of polyacylated anthocyanins are stacked from both sides of the anthocyanidin nucleus to stabilize the chromophores, preventing hydration of the nucleus. Only simple anthocyanins without aromatic acyl residues (Hoshino et al., 1981a,b; Goto et al., 1987) or with one acyl residue (Yoshida et al., 1991; Kondo et al., 1992; Kondo et al.,

626

M. Mori et al. / Phytochemistry 67 (2006) 622–629

Fig. 3. VIS absorption spectra (lower) and CD (upper) of the blue petal and suspension of the blue cells of phacelia. - - - - -: petal, ——: suspension of protoplasts.

1994; Kondo et al., 2001) are supposed to be able to stack together with an anthocyanidin nuclei. Therefore, the Cotton effect of the blue petal and cells is very interesting and important information on blue color development. 2.3. Reproduction of blue color with combined petal components To clarify the mechanism of the blue flower color development of P. campanularia, we first analyzed the components of colored cells. The HPLC detected at 280 nm showed that only 1 with a small amount of 2 and 3 existed in the colored cells. The content of metals was analyzed, and Mg, Al and Fe were detected. The molar ratio of Mg, Al and Fe to the combined anthocyanins (1 and 2) was 0.2–0.5 eq., 0.01–0.02 eq. and 0.01–0.02 eq., respectively. By referring to the above-mentioned data, we tried to reproduce the same blue color of phacelia petals by mixing petal components in various pH aqueous solutions. As shown in Fig. 4, 1 (5 · 10
4 M) in aqueous solution at pH 5.5–6.0 gave the same VIS spectrum and CD as those of petals. However, the stability of the blue solution was low at this pH and after 24 h, blue precipitates were formed. By the addition of Al3+ or Fe3+ (>1/3 eq. to 1), a blue solution with different VIS spectra and CD from that of petals and protoplasts was obtained. The addition of an excessive amount of Mg2+ (1–10 eq.) to 1 in buffer (pH 5.5) affects neither the spectrum nor the stability of the solution. However, by the addition of a small amount of Al3+ or Fe3+ to 1 (