J. ChiL Chern. Soc., 52, N 2 (2007)

SECONDARY METABOLITES IN THE FLOWER HEADS OF HAPLOPAPPUS BERTERII(ASTERACEAE) AND ITS RELATION WITH INSECT-ATTRACTING MECHANISMS

ALEJANDRO URZUA", Rocfo SANTANDER, JAVIER ECHEVERRiA, MARCOS C. REZENDEUniversidad de Santiago de Chile, Facultad de Quimica y Biologia, Departamento de Ciencias del Ambiente.

Laboratorio de Quimica Ecologiea, Casilla- 40, Correo- 33, Santiago, Chile.

ABSTRACT

The chemical compositions of the flower heads of H. bertedi and of Chrysantemum coronarium 1., visited by a varied entomofauna, were compared in thesearch of possible correlations that might explain why different plants are visited by the same insects. Though some similarities were observed in the flavonoidcontents of both species, their overall composition was dramatically different, pointing to the existence of rather complex mechanisms of insect attraction by thesespecies. OUf results thus represent a cautionary remark to interpretations of such mechanisms based solely on the chemical composition of the volatile componentsof flowers in the field.

keywords: Haplopappus berterii; Asteraccae; Monoterpcnes; Sesquiterpenes; Flavonoids; Insect-attracting stimuli

INTRODUCTION

The Chilean littoral rock formations (limit of IV and V Regions), LosMoUes (V Region, Chile 32" 30'S, 71 0 30'W), is the habitat of Haplopappusbcrterii Phil. (Astcraceae) an endemic evergreen shrub with yellow flowers of2.5 em ofdiameter I).

Flower heads of H. berterii are visited by a varicd entomofauoa andalthough no syslematic studics have been pcrformed, it is known that theyare the host of Trupanea Schrank (Diptera: Tephritidae)2l species. In additionArthrobracus sp. (Coleoptera: Melyridac) use the flower heads as diet, and oneChilean social bee Diasiadae sp. (Himenoptera: Apiadeae) and the butterflyVanessa carye (Lepidoptera: Nymphalinae) also visit the flowers.

These species have also been identified as visitors of Chrysanthemumcoronarium L. another Asteracca with yellow flowers like H.berterii, thatgrows near this Haplopappus species.

In this communication we report the composition oflhe volatile compounds,epicuticular chemistry and flavonoids of H. berterii flower heads, and make acomparison with the chemistry of Chrysanthemum coronarium L. flowers inorder IQ establish ifthere are some chemical similarities that might explain whythese two different species are visited by the same insects.

EXPERfMENTAL

Plant materialHaplopappus berterii Phil. (Asteraceae) flower heads, were collected in

November 2004 in Los MoUes (V Region, Chile 3230'S, 71 0 30'W). Voucherspecimens were deposited in the Hcrbarium onhe National Museum ofNaturalHistory, Santiago, Chile.

Plant extractionFresh flower heads of H. be11erii (275 g) were extracted by dipping the

plant material in 1.5 L of cold CH2Cl2 for 60 s. The extraction was repeatedtwice. The material exhausted with CH2Cl2 was dried in an oven aL50, milledand submitted to percolation in 95 % cold EtOH (1.5 L) for 24 h. The procedurewas repeated twice. The EtOH extracts were concentrated and partitionedbetween Water and CHCl]. The organic layer was discarded and the watcr layerwas extracted with AcOEt.

Column chromalography separation of the extractsThe CH2Cl2 extract (1.2 g, 0.44 %) was fractionated by CC (silica gel)

using pentane - CH1Cl2 and CH2CI2 - McOH step gradients to afford 4fractions. The AcOEI extract (1.4 g, 0.51 %) was fractionated by CC (silicagel) using a CHC~ - MeOH step gradienllo afford 80 fractions.

TLC study of the extracts and fractionsTLC of me extracts and fractions was performed on silica gel 60 F25-1 pre-

coated plares from ~lerck. Specific spray reagents were used for detection ofdiffetenl families of compounds 3l : anisaldehyde-~SO~, phosphomolibdicacid znd \"3illiI.I.in-H.!O... for terpenoids and diphenylboric acid-~thylamino

~~EG I fur ih"llDOids.

GC-EM analysis ofthe CH1CI2 extractThe fraction eluted with pentane, from the CH1.C!2 extract, werc analyzed

in triplicate in a GC-MS ( gas chromatograph: Hewlett-Packard model usinga HP5891; mass spectrometric detector with integrated data system: Hew1eU-Packard model HP5972). Separation was performed using an Ultra-2 H.P.capillary column (IS m x 0.25 mm). The temperature of tile injector was 295C, and the temperature of the column was programmed, starting at 45C, for2 min, followed by a rise to 200 C at 10C /min and to 300 C at 20 C / min l .The temperature was kept constant al 300C for 20 min. Helium was the carriergas at 10 lb. psi. Detection was done using QJ and EI.

SpectroscopyAll NMR experiments were perfonned on a Bruker-400 AV3nce

spectrometer using DMSO-d6 Two-dimensional spectra were obtained usingstandard Bruker software. FTIR spectra were obtained on a Perkin Elmerspectrophotometer in KEr.

Nomenclature of compoundsNames ofmonoterpenes, sesquiterpenes, and flavonoids are given according

Lo the Handbook of terpenoids 4} and Flavonoids Chemistry, Biochemistry andApplications'l.

Yields of fractions and compoundsThe yield ofextracts and compounds were calculated in relation to the fresh

plant material. The percentage of different families and individual compoundswas calculated from the peak areas of the chromatograms.

RESULTS AND DISCUSSfON

Chemical compositionThe CH2CI2 extract (1.2 g, 0.44 %) was fractionated by CC (silica gel)

using pentane - C~CI2 and CH2Cl2 - MeOH stcp gradients. Fraction A elutedwith pentane, (377 mg, 0.137%) was submiUed to extensive GC-MS analysis.The AcOEt extract (1.4 g, 0.51 %) was fractionated by CC (silica gel) using aCHell - MeOH step gradient to afford 80 fractions regrouped after TLC in fournew fraction" B (0.025 g), C: (0.027 g), 0 (0,12 g) and E (0,31 g).

Monoterpenes (0.0015 %): a.-pinene (1), p-pinene (2), p-myrcene (3),limonene (4).

Sesquiterpenes (0.013 %): 3-cubebene (5), 4(15)-cubebene (6), 1(10)-aristolene (7), copaene (8), isocaryophyllene (9), a.-caryophyllene (10), 4,9-bulgaradiene (II), 4,1O(14)-bulgaradiene (12), 4,11-amorphadiene (13),1(I 0),4-cadinadiene (14), 2,5,5-trimethyl-1 ,3,4,5,6,7-hexahydro-2H-2,4a-ethanonaphthalene (15).

Miscellaneous alkanes (0.00 I %): 2-methyldacaline; 2,4,6-trimethyloctanej

J. Chi!. Chern. Soc., 52, N2 (2007)

2,6-dimethylundecane; 4,6-dimethylundecane and 2,IO-dimethylundecane.

The compounds identified in the fractions obtained from the CC separationof the AcOEt extract were: Flavonoids (0.51 %): diosmetin (5,7,3'- trihydroxy-4'-methoxyflavone) (0.004 %) (16), tamirexin (3,5,7,3'- tetrahydroxy-4'-methoxyflavone) (0.006 %) (17), luteolin (5,7,3'4'-tetrahydroxyflavone) (0.043%)(18) and quercetin (5,6,7,3',4'-pentahydroxyflavone) (0.089 %) (19).

Identification oftbe compoundsThe identificalion of compounds in the chromatographic profiles was

achieved by comparison of their mass spectra with a library data base (NIST1998) using a reverse search technique which verified that main peaks in thereference spectrum were present in the unknown spectrum 6). Spectra wereconsidered coincident if the similarity index was higher than 95% 6).

Preliminary identifications were confinned by the obsel\'ation of peakenhancements upon coinjection of standards. When standards were notavailable, the mass spectra were compared with published spectra of authenticcompounds 7,~.9). Also, Kovats index of the peaks were compared with valuesfrom the literature.

R,

;?' R, 1" RI =: R2 =: H , _3 =: OC::=:.::

HO 0 ~ 17 R. =:R2 =OE, R]=0CE3

""" 18 Rl=H; R2=R3=OE:#' R, R1 =R2 = R3 =OE1.

OH 0

Compound C, yellow amorphous solid (17 mg) shows an IHNr-,.mspectrum that indicate the structure of a methyl derivative of quercetin(3,5,7,3',4'-pentahydroxyflavone) (l9). Bidimentional NMR experiments(HMBC and HSQC) provide support for the determination of position of theIllcthoxy group at C-4' in ring B. The compound was identified as tamirexin(3,5,7,3'- tetrahydroxy-4'-melhoxyflavone) (17) and data were in agreementwith those reported in the literature 10),

Fractions B, C, 0 and E were respectively crystallised from CHCI3-EtOH(85:15). Compound E, yellow crystals (230 mg) and compound D, yellowamorphous solid (89 mg),were respectively identified as qucrcctin (19) and[uteolin (18) by direct comparison (FTIR, TLC, IH-NMR) with authenticsamples obtained from Aldrich,

Compound B, yellow amorphous solid (14 mg) shows an IH_NMRspectrum that indicate the structure of a methyl derivative ofluteolin (5,7,3'4'-tetrahydroxyflavone) (i8). Bidimentional NMR experiments (HMBC andHSQC) provide support for the determination of position of the methoxygroup at C-4' in ring B. The compound was identified as diosmetin (5,7,3'-trihydroxy-4'-merhoxyflavone) (16) and data were in agreement with thosereported in the literature 10).

4

The composition of the flower extracts of H. bertcrii may be comparedwith that of other Chilean Haplopappus. The hydrocarbon fraction accountedfor 93.39 % of the total chromatogram peak area with nalkanes (81.71 %)comprising the major group of compounds, represented mainly by C29H60(42 %), C3lH62 (18.6 %) and C27H56 (13.1 %). Although Ihis hydrocarbonprofile is similar to thai of other Haplopappus species, a remarkable contrast inthe yields of this fraction is observed. The epicuticular compounds of speciesgrowing in the mountains comprise a 25-50% coment of hydrocarbons l,9)whereas H. bertel'ii only yielded 0.11 %, a value even lower than that of H.foliosus DC.(8%) growing in coast areas ofthe V region 5J and H. bustillosianusRemy (1.9 %) growing in the coast rock formations ofVillarrica lake II).

Only trace amounts (0.0015%) of monoterpencs 1-4 were found in H.berterii. Although in the epicuticular compounds ofHaplopappus, the presenceofmonoterpenes seems to be erratic, compounds 1-4 have been identified in H.velutinus Remy and H. illinitus Phil. while H. foliosus DC. contains limonene(4) 8.9) and H, bustillosianus Remy contains a-pinene (i) and ~-pinene (2) II).By contrast, no monolcrpenoids have been identified in: H. cuneifolius (Ness),H. uncinatus Phil. and H. shumalUli (OK.) Br. etClark 9).

In the specific case ofH. berterii, a large entomofauna may be observed onits flower heads. Some of the visitors include the Trupanea Schrank. (Diptera:Tephritidae) species, whose larvae have been found within the florets eatingaway the ovaries Z) .The Artbrobracus sp. (Coleoptera: Melyridae) uses theflower heads as part of their diet, eating the disc florets. Finally, one Chileansocial bee, Diasiadae sp. (Hymenoptera: Apiadeae) and the butterfly Vanessacarye (Lepidoptera: Nymphalinae) are also regular visitors to the flower heads.Bees visitors play only a minor role as pollinators and are mostly pollen robbersin Asteraceae 121.

Finally, an overview of the sesquiterpene composition of H. berterii isin agreement with the result'> found for the epicuticular chemistry of otherHaplopappus species. Even though some of these molecules and structuralfamilies are repeated among species, a clear sesquiterpene pattern common rothe genus Haplopappus eouid not be found.

The flower heads of Asteraceae are visited by various insects. The visitorsobtain shelter, abundant food and are found everywhere on the flower heads.Even in the same eco-system, the Asteraceae insect visitors differ during thecourse of a day. Oi ffercnt species visit the flower heads in early hours of themorning, mid-morning, noon and afternoon.

These four species were also identified on the yellow flower heads ofChrysanthemum coronarium L., another Asteracea that grew near the H. berteriiflowers, Our field observations took place during mid-morning, between 11 :00and 13:00 h, for two days a week if! November and December, for nine weeks.

10

IJ

15

"

14

II

1143

1. Chi!. Chern. Soc., 52, N' 2 (2007)

AJthougb field conditions prevented a rigorous quantification of the insectvisilS, there was apparently little difference regarding the insect preferencesbetween the two plants.

Our observations led us to search for a common set ofstimuli in both plantsthat might be responsible for lheattraction aCthe same insects. We hypothesizedthat the flower heads of both Asternceae might elicit similar chemical stimuli,and that this could be verified by a comparison of the compositions of theirvolatile constituents.

In a recent communication, the surface and volatile compounds of flowerheads of Chrysanl.hemum coronarium was described Il). In the metion ofvolatile compounds the only sesquiterpene identified was a-famesene. ThemonOlcrpenes camphor, hornyl acetate and tnms-chrysanlhenyl acetateaccounted for more than 60010 of the volatile components. None of thesecompounds were present in H.berterii. The only common compound in bothplants was limonene (4).

As regards flavonoids, large amounts of luteolin (I8) and quercetin(19) have also been found in the flower heads ofe. coronarium 14) .Thuswe may conclude that, with the exception of the common presence of lowconcentrations of limonene (4), the compositions of the volatile componentsof H. berterii and C. coronarium, responsible for eventual odorous stimuli toinsects, are dramatically different. Flowers of these species share in common,besides similar shape and size, the yellow flavonoids luteolin (18) and quercetin(19), responsible for their common colour. This might bc taken as an indicationthat optical cues are more import'ant than chemical stimuli in detcnniningthe choice of these insects for the two flowers. Because of the limitations offield observation, this conclusion should be regarded with caution. Chemicaland optical attractants to insects are difficult to measure and quantify in thefield. It has been suggested that there is linle orientation when an insect is atsome distance from its host, and that searching for a host plant is essentiallya random process 151. Once the insect finds a suitable host, this random searchis followed by a learning process after which the insect is able to respondpositively to optical and chemical stimuli by the plant, and recognize it to feedand for oviposition [61.

These considerations only emphasize the difficulties involved in fieldstudies, and the cautious nature of any conclusions drawn from them. Ourresults revealed little or no correlation between the chemical compositionof the volatile components of H. berterii and C. coronarium flowers and thepreferences of some of their guest insects. Because of the complex interplayof factors detennining this preference in the field, such absence ofcorrelationsdoes not rule out the existence of odorous stimuli in the interaction betweenthe studied insects and these plants. Such correlation's might be obscured bythe essentially random learning process undergone by an insect in the field.Nevertheless, our results also represent a word of caution to interpretationsof insect-plant interactions based solely on the presence of volatile chemicalconstituents in the latter.

1144

ACKNOWLEDGEMENTS

Financial support from DICYT (USACH) is gratefully acknowledge.

REFERENCES

1.- Hoffmann, 1. A., 1995. Flora Silvestre de Chile, Zona Central. 3 Ed.Edicioncs Claudio Gay, Santiago Chile pp. 238-239.

2.- Frias, D., 1985. Rev. Bras. Zool. 2, 363.3.- Wagner H., Bladt S., Zgainski E.M., 1984. "Plant Drug Analysis"

Springer-Verlag, Berlin.4.- Connolly J.D., Hill R.A., 1992. "Dictionary of Terpenoids Vol I. Mono

and Sesquiterpenoids" Chapman and Hall, London, UK.5.- Andersen O. M., Markham, K. R., 2006." Flavonoids: Chemistry,

Biochemistry and Applications" Taylor and Francis Group, Boca Raton,FL, USA.

6.- Pesyna, G. M., Venkataraghavan, R., Dayringer H. E., McLafferty, F. W.,1976 Anal. Chem. 48, 1362.

7.- Swingar, A. A., Silverstein, R. M., 1981. Monoterpenes Infrared, Mass,I H NMR, and U C NMR Spectra, and Kovats Indices. Aldrich ChemicalCompany, Inc., Milwaukee USA.

8.- Urz\ia, A., 2004. J. Chil. Chern. Soc. 49, 137.9.- Urzua, A., Contreras, R., Jara, P., Avila, F., Suazo, M., 2004 Biochem.

Syst. Ecol. 32, 215.10.- Harbome, J., B., 1996 " The fIavonoids Advances in Research Since

1986" Chapman and Hall, London, UK., pp. 295-330.11.- UrziJa, A., Iturra, 8., Sebastian, B., Munoz, M., 2007 Biochem. SySL

Eoo l. in press.12.- Mani, M., S., Saravanan, J., M., 1999. "Pollination Ecology and Evolution

in Compositac (Aslcraceae). Science Publishers, Inc. New Hampshire,USA, pp. 12-23.

13.- Urzita, A., Sebastian, 8., Vines, M., 2006 Flavour and Fragrance J. 21,783.

14.- Urzua, A., Wlpublished results.15.- Schoonhoven, L. M., A van Loon J. J., Dicke, M., 2005. "Insect-Plant

Biology" (Chapler 6, Host-plant selection: how to find a host plant).Oxford University Press, New York, USA, pp. 135-160.

16. Lewis, A. C., Lipani, G., 1990. Learning and flower use in bUllcrfiies:hypothesis from honey bees. In Bcmays, E. A., "Insect-plant interaction,vol 2". CRC Press, Boca Raton, FL., pp. 95-110

escanear0003escanear0004escanear0005