- Email: [email protected]
Variability in chemical constituents in Petasites hybridus from Austria
Biochemical Systematics and Ecology 28 (2000) 421}432
Variability in chemical constituents in Petasites hybridus from Austria Remigius Chizzola*, Bernhard Ozelsberger, Theodor Langer Institute for Applied Botany, University of Veterinary Medicine Vienna, Veterina( rplatz 1, A-1210 Vienna, Austria Received 25 February 1999; accepted 11 June 1999
Abstract Petasites hybridus (Asteraceae), butter bur, is an ancient medicinal plant with spasmolytic sesquiterpene esters. Two chemotypes, the petasine and the furanopetasine chemotype, occur in Austria. The "rst one is considered as pharmaceutically useful due to its spasmolytic constituents, but it is restricted to the northern parts of the Alps. This use, however, is impaired by the presence of low amounts of toxic pyrrolizidine alkaloids (PA), mainly senecionine and intergerrimine. PA are usually concentrated in the metabolically active parts of the complex rhizome which are the thickenings just below the leaves. They are also present in #ower stalks but are almost absent in leaf buds, the petioles and the leaf blades. The alkaloids showed a great variability within and between populations; the values recorded ranged from less than 2 to 500 mg kg~1 PA, median PA of 77 populations varied from 2 to 191 mg kg~1 in the rhizomes. In nearly 25% of the samples analysed the PA content was below 10 mg kg~1, another 25% had between 10 and 20 mg kg~1 PA. Histograms of PA concentrations in a population often showed a distinct skewness toward lower alkaloid contents. Alkaloid content was independent of sesquiterpene chemotype. The seasonal in#uence on PA content of rhizomes was little in comparison to the variability within the population or within the rhizome itself. Nevertheless, when comparable rhizome parts within a population were considered, the PA content may remain stable over several years. Although plants totally free of PA could not yet be found, it is possible to select populations low in alkaloids. Several populations of the petasine chemotype containing less than 10 mg kg~1 in the rhizomes could be found in the area investigated. ( 2000 Elsevier Science Ltd. All rights reserved. Keywords: Petasites hybridus; Asteraceae; Pyrrolizidine alkaloids; Senecionine; Sesquiterpene esters; Petasine; Furanopetasine; Population variability; Austria
* Corresponding author. Tel.: #43-1-25077-3104; fax: #43-1-25077-3190. E-mail address: [email protected] (R. Chizzola). 0305-1978/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 3 0 5 - 1 9 7 8 ( 9 9 ) 0 0 0 7 7 - 0
422
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
1. Introduction Petasites hybridus (L.) G.M. ET SCH. (Asteraceae), butter bur, is widespread on moist habitats along riversides in Central Europe. Besides its use in folk medicine several phytopharmaceutical preparations are made from the rhizomes and the leaves. The active principles are sesquiterpene esters (Fig. 1) with spasmolytic properties on the smooth musculature (Aebi et al., 1958). A gastro-protective e!ect of ethanolic Petasites extract could be attributed to the inhibition of leucotriene biosynthesis (Brune et al., 1993). An overview of the uses and the pharmaceutical quality requirements is given by Meier and Meier-Liebi (1994). The sesquiterpenes occur in two chemotypes, the petasine and the furanopetasine chemovariety (Novotny et al., 1966), which cannot be distinguished by morphological characters. Only the petasine chemotype with the main compound petasine is considered as pharmaceutically useful. The compouds of this chemotype have been studied in detail by Debrunner et al. (1995a,b). Furthermore, the rhizomes of P. hybridus contain trace amounts of toxic pyrrolizidine alkaloids (PA) (Fig. 1) (LuK thy et al., 1983) which have demonstrated a carcinogenic and mutagenic potential (RoK der, 1995). Therefore, plants used for pharmaceutical purposes must be from the petasine chemotype and low in PA. Another requirement is good vegetative growth for future agricultural production. The present work reports the natural variability in PA content in Petasites hybridus rhizomes and the distribution of the two chemotypes from populations of the Eastern Alps. To evaluate this variability the distribution of PA within the di!erent parts of the plant and the seasonal variability in alkaloid content has to be taken in account. The present data complete the report of Chizzola (1993a). 2. Materials and methods 2.1. Plant material Petasites plants were dug out at the natural growing sites and pieces of the rhizome were taken for analysis. The plant material originated from various locations in Austria, southern Bavaria and South-Tyrol. The sampling was done during six years to study the following aspects: f Approximately 1500 samples were taken from plants out of 77 populations. In 754 samples PA were investigated by capillary GC, 739 samples were analysed by enzyme immuno assay. f From further 90 di!erent sites 2}3 plants were taken and analysed. f From one large population growing near Reutte (Tyrol) 195 samples were analysed. f In one population of the petasine chemotype low in PA the alkaloids were analysed in three di!erent years. f From a population high in PA several plants were dug out as complete as possible. The rhizome was divided into obvious parts and the partitioning of the alkaloids in the individual fragments was studied.
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
423
Fig. 1. Pyrrolizidine alkaloids and sesquiterpene esters from Petasites hybridus.
f In two consecutive years samples were taken from a population in the southern Wienerwald at monthly intervals to study the variability of PA during the year. 2.2. Chemical analysis The plant material was freeze dried and "nely powdered. For PA analysis 1 g powdered rhizome was extracted at 653C 1 h with methanol. The reduction of the alkaloid N-oxides, which may represent more than 80% of the total alkloid content of the plant, was achieved with Serdoxit', an oxygen absorber resin. This reduction step was necessary since in the extracts the PA were mainly in the N-oxide form and the analytical methods used detect only free alkaloids. In the case of GC analysis the extracts were further puri"ed with an strong cation exchanger solid phase extraction column. GC was performed on a Carlo Erba Vega Series 6000 equipped with an Rtx-5 column (30 m]0.32 mm ID, 0.1 lm "lm thickness) and a temperature program from 120 to 2703C (63C/min). The main PA senecionine and integerrimine (Fig. 1) were quanti"ed using ca!eine as internal standard (Chizzola, 1994). Alternatively, these two PA were quanti"ed using a double-antibody solid-phase immuno assay according to Langer et al. (1996), a test system speci"c for the isomeres senecionine and integerrimine. In this case the methanolic extracts were diluted after reduction of the N-oxides with the assay bu!er. The results obtained were in good accordance with the GC analysis. The detection limit of PA was 2 and 0.1 mg kg~1 in the GC and immunoassay method, respectively. Senkirkine, a minor PA of the otonecine type in Petasites (Mauz et al., 1985) was not evaluated. For the determinaton of the sesquiterpene type we used a qualitative TLC method. The plant powder was extracted with dichloromethane, 5}10 ll of the resulting extract were applied to Kieselgel 60 F TLC plates (Merck 1.05715) and the 254 chromatogram was developed with xylol : acetone 9 : 1. The spots were visualized in the UV light at 254 nm or by dipping in an anisaldehyde}acetic acid}methanol reagent (Wagner et al., 1983). The petasine chemotype gave UV absorbing spots which develop after dipping and heating an yellow to light purple colour.
424
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Statistical analysis of the data was performed using SPSS for Windows version 6.0.1.
3. Results 3.1. Partitioning of PA in the plant Fig. 2 shows two examples of the complex structure of the rhizome which consists of aggregated tuberlike thickenings and long runners which may reach several meters in length. The distinct parts di!ered clearly in PA content, and the comparison of several plants even harvested at di!erent seasons allows the following observations: The parts exhibiting the highest alkaloid content were the young thickenings just below the emerging leaves. Older thickening, which may also become hollow and were generally farther away from leaf bases, were lower in alkaloids. Furthermore, the center of such a tuber contained more alkaloids than the cortex (Fig. 2b). Older runners connecting well established groups of tubers were generally low in alkaloids, whereas young runners, which by further growth will di!erentiate new leaves at their tips, might be
Fig. 2. Two examples of the complex structure of the Petasites hybridus rhizome and the repartition of pyrrolizidine alkaloids (mg kg~1) within di!erent parts of the rhizome. A. a. young runner with terminal bud, b. old runner, c. leaf, d. leaf bud, e. roots. B. b. border and c. centre of a thickening.
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
425
high in PA. The alkaloids were also present in the roots, but at lower concentrations than in most rhizome parts. In addition, the PA could be found in spring in the #ower stalks, but were absent or very low in leaf buds or leaf stalk and blades (Langer et al., 1996). 3.2. Seasonal variation of pyrrolizidine alkaloids within a population The PA content of the Petasites rhizomes parts just below the emerging leaves was followed during two consecutive years in monthly intervals. In both years comparable alkaloid contents could be recorded. Mean alkaloid content varied from 35 to 62 mg kg~1, but as can be seen from Fig. 3, all over the year there was always a great variability in PA content between the individuals of the population. Altogether the alkaloid levels appear to be higher in the "rst part of the year than in summer and autumn. 3.3. Variability of PA within a population The variability in PA content of the Petasites rhizomes within a population could be studied during the monitoring of a population over two years (Fig. 3) and in nearly 1500 individuals sampled from 77 populations throughout northern Austria (Fig. 4). In the "rst case, the alkaloids manifested high variability at any sampling date. In the latter case, the PA content of the samples collected during 4 years varied between less than 2 and 500 mg kg~1 PA. The variation coe$cients of the populations ranged from 20 to 146%. Variability within a population was not dependent upon mean or median PA content. The histogram of all samples put together (Fig. 4) shows a distinct skewness toward lower PA values. When the samples were arranged in four classes below 10 mg kg~1, between 10 and 20 mg kg~1, between 20 and 40 mg kg~1 and above 40 mg kg~1 PA, each of these classes would contain about 25% of the samples.
Fig. 3. Boxplot of the variability of PA (mg kg~1) within a population during two consecutive years. Box and whiskers represent the quartiles, the bold central horizontal bar marks the median. Black boxes: "rst year, empty boxes: 2nd year.
426
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Fig. 4. Histogram showing the PA concentration of 1493 samples of Petasites hybridus rhizomes analysed from 77 populations. Class width 10 mg kg~1 PA, the middle of each class is marked on the x-axis. The curve of the theoretical normal distribution is inscribed.
The high proportion of samples low in PA came partly from those populations generally low in alkaloids and partly from the fact that in most populations low alkaloid samples may occur. Plants of the petasine and furanopetasine chemotype showed comparable variability in the PA, but although approximately an equal number of individuals was tested from both chemotypes, the furanopetasine chemotype prevailed with 225 to 159 cases in the class below 10 mg kg~1 PA. Also many particular populations showed a skewness toward lower PA content (Fig. 5). This was also the case within a large population where in 195 individuals tested the mean PA content and the median were 26.6 and 23.3 mg kg~1 respectively, and 15 individuals had less than 10 mg kg~1 PA in their rhizomes. To test the homogeneity of populations several larger populations were divided into two to "ve sections and the variability of PA was considered separately in each section. In eight out of 22 populations tested, signi"cant di!erences in PA content could be found in the individual sections of these populations. The stability of the PA content in Petasites rhizomes over a longer period can be deduced from the comparable alkaloid levels monitored during two years. Furthermore, one population with very low alkaloid content in the rhizomes could be harvested three times in three di!erent years (Table 1). In most individuals, at each harvest the PA content remaind below 10 mg kg~1 in the rhizomes. 3.4. Geographical variation Petasites is widespread througout Central Europe. In Austria it is concentrated in the Northern and Southern Prealps and it is lacking in the great alpine valleys and basins (Niklfeld, personal communication). The geographical maps in Fig. 6 a}d present the PA content and chemotype of Petasites rhizomes from 167 sites. From 77 populations a larger number of samples has been taken, the median PA content has
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
427
Fig. 5. Boxplot representation of the variability of pyrrolizidine alkaloids in the rhizomes from 77 Petasites hybridus populations arranged according increasing median pyrrolizidine alkaloid content (mg kg~1) of population. Box and whiskers represent the quartiles, the bold central horizontal bar marks the median.
been evaluated; the data from further 90 populations are based on the mean PA content of 2}3 individuals. Most of the populations examined originated from the Northern Alps and Prealps. It can be seen that both chemotypes were present; the petasine and the furanopetasine chemotype could be found at 70 and 118 sites,
428
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Table 1 PA content (mg kg~1) in the rhizomes of a population low in alkaloids harvested in three di!erent years
Analysed by N Mean Median Minimum Maximum
Harvest 1
Harvest 2
Harvest 3
GC 15 3.8 2.0 (2 15.5
GC 13 10.1 9.6 (2 20.9
EIA 17 6.0 4.9 0.6 16.2
respectively. They occured mainly in separated populations, but there were also population made up by both chemotypes. In Austria the furanopetasine chemotype seems cover a wider range than the petasine chemotype; the latter appears to be restricted to the regions north of the Central Alps (Fig. 6 a}d) and lacking in the most eastern parts (Chizzola, 1993a). The di!erences of median or mean PA content of populations did not correlate with the various regions or sampling sites. Populations generally low in alkaloids were of either chemotype (Fig. 6a) and might be found throughout the area investigated.
4. Discussion The motive to study the chemical variability of Petasites hybridus was the need of plant material of the petasine chemotype poor in PA for pharmaceutical use. Petasites hybridus is a plant with vigorous clonal growth; tuberlike thickenings are connected by long runners attaining several meters in length. With further growth these connections between the thickenings might be lost, so that the original plant becomes fragmented and it will not be possible to decide of how many original indivmH duals, i.e genotypes, a dense Petasites population is composed. As in the Scandinavian territories populations are uniquely composed by androdynamic, i.e. functionally male plants, it is suggested that they were all derived by clonal growth from several plants introduced in the middle ages as medicinal plant by monasteries in Norway (Faegri, 1992). On the other hand, populations of plants building long runners which enable them to invade and colonize a given territory by a kind of guerilla strategy, are likely to be composed of more than one genotype (StoK cklin, 1992). In Austria presumably both cases occur. As the #ower capitules of Petasites are c Fig. 6. Geographic distribution of pyrrolizidine alkaloids and sesquiterpene chemotype in Petasites hybridus rhizomes from 167 locations in Austria and adjacent regions. legend: w Petasine chemtype o Furanopetasine chemotype d Both chemotypes in population A. Less than 10 mg kg~1 pyrrolizidine alkaloids in the population. B. 10}20 mg kg~1 pyrrolizidine alkaloids in the population. C. 20}40 mg kg~1 pyrrolizidine alkaloids in the population. D. More than 40 mg kg~1 pyrrolizidine alkaloids in the population.
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
429
430
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
composed of tubular #orets producing either pollen or ovules, the plant is functionally dioeceous. The plant #owers early in spring, so that sex distribution within the populations was recorded in a few cases only. Populations containing plants of both chemotypes or both androdynamic (male) and gynodynamic (female) individuals are certainly genetically heterogenous. The Petasites samples analysed showed a considerable variety in PA content. These di!erences have to be carefully evaluated. Genetic di!erences cannot be directly deduced from the varying amounts of a secondary plant product (Schratz, 1963). Plants growing in di!erent environments grow ordinarily at di!erent rates; they di!er in size and developmental stage (Coleman et al., 1994). A major di$culty in assessing the variability in PA is their high variability within a plant rhizome. Samples low in PA could be found in the majority of the populations, but they might have included old rhizome parts whose PA content is not representative for the whole plant, as it was not possible to analyse whole rhizomes. This circumstance was not completely avoidable due to the complex structure of the rhizome. If comparable rhizome pieces were considered, the PA content appeared to be stable for a longer period since similar alkaloid levels were recorded in the one population monitored during two years and in the other sampled three times. The in#uence of the season seems to be of minor importance. Clonal growth of Amphibromus responded readily to water and fertilizer supply, factors which may also be important within a population where they may vary within a small area (Cheplick, 1995). However, PA as secondary products do not react so dramatically to environmental factors. In Senecio vulgaris PA concentration was barely in#uenced by lack of water or nutrients (Brown & Molyneux, 1996). In this context populations showing a great di!erences in PA concentration are likely to be genetically heterogenous. Petasites belongs to the tribe of the Senecioneae, a group within the Asteraceae capable of synthesizing toxic PA. In an ecological context these alkaloids act as feeding deterrent agents against herbivores (Harborne, 1989). PA are mainly concentrated in young fast growing parts. The peaks in PA were recorded in young rosette leaves of Cynoglossum (Knight et al., 1984; Van Dam et al., 1994) or young buds emerging from the rhizomes in Symphytum (Chizzola, 1993b). In the case of Senecio vulgaris, the PA are synthetized in the roots and transported mainly to #owers to protect these reproductive organs (Hartmann and Zimmer, 1986). Petasites is a plant building up only low PA levels, therefore, it is hardly conceivable that these alkaloids might a!ord an e$cient protection against herbivory. Nevertheless, the highest PA concentrations were also found in Petasites in the young metabolic active parts of the rhizome, which are essential for the further growth of the plant. The presence of PA in Petasites appeared to re#ect the phylogenetic origin from a PA producing group. On the other hand, totally PA free individuals have not yet been found, so that a bene"cial e!ect of the PA on plant viability and survival cannot be ruled out de"nitely. With regard to PA distribution in the plant, P. hybridus is similar to Tussilago farfara, another member of the Senecioneae, where the alkaloids prevail also in the rhizomes. In this plant it was possible to select a variety without PA in the leaves, which are pharmaceutically used (Kopp et al., 1998). In accordance with the very low
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
431
PA content recorded in P. hybridus leaves (Langer et al., 1996) it seems feasible to also get PA free leaves from this species. The use of such leaves might be of interest, if they are high in the active sesquiterpenes (Wildi et al., 1998). The wide variability in PA in the rhizomes allows to select plants of the petasine chemotype low in alkaloids. In Germany the daily intake of toxic PA must not exceed 1 lg and the daily exposure in case of external application is limited to 10 lg (Meier and Meier-Liebi, 1994). Whether the plants low in PA meet these requirements depends on the production technique and the applied dose of the resulting phytopharmaceutical preparation. Actually, PA are removed from Petasites extracts by passing them several times through cation exchanger resins (Mauz et al., 1985; Meier and Meier-Liebi, 1994). The use of plants poor in alkaloids would considerably simplify these procedure. References Aebi, A., Waaler, T., BuK chi, J., 1958. Petasin und S-Petasin, die spasmolytisch wirksamen Inhaltssto!e von Petasites ozcinalis (L.) Fl. Wett. Pharm. Weekblad 93, 397}406. Brown, M.S., Molyneux, R.J., 1996. E!ects of water and mineral nutrient de"ciencies on pyrrolizidine alkaloid content of Senecio vulgaris #owers. J. Sci. Food Agric. 70, 209}211. Brune, K., Bickel, D., Peskar, B.A., 1993. Gastro-protective e!ects by extracts of Petasites hybridus: The role of inhibition of peptido-leukotriene synthesis. Planta Med. 59, 494}496. Cheplick, G.P., 1995. Genotypic variation and plasticity of clonal growth in relation to nutrient availability in Amphibromus scabrivalvis. J. Ecol. 83, 459}468. Chizzola, R., 1993a. The main pyrrolizidine alkaloids of Petasites hybridus: variation within and between populations. Acta Hortic. 333, 143}150. Chizzola, R., 1993b. Repartition of the pyrrolizidine alkaloids acetyl-intermedine and acetyl lycopsamine within the Symphytum ozcinale plant. Planta Med. 59, A644. Chizzola, R., 1994. Rapid sample preparation technique for the determination of pyrrolizidine alkaloids in plant extracts. J. Chromatogr. A 668, 427}433. Coleman, J.S., McConnaughay, K.M.D., Ackerly, D.D., 1994. Interpreting phenotypic variation in plants. Trends Ecol. Evol. 9, 187}191. Debrunner, B., Neuenschwander, M., Brenneisen, R., 1995a. Sesquiterpenes of Petasites hybridus (L.) G.M. et Sch.: distribution of sesquiterpenes over plant organs. Pharm. Acta Helv. 70, 167}173. Debrunner, B., Neuenschwander, M., 1995b. Sesquiterpenes of Petasites hybridus (L.) G.M. et Sch.: in#uence of locations and seasons on sesquiterpene distribution. Pharm. Acta Helv. 70, 315}323. Faegri, A.K., 1992. Butterbur (Petasites hybridus) * a monastery plant. Blyttia 50, 115}119. Harborne, J.B., 1989. In: Introduction to Ecological Biochemistry, 3rd edition. Academic Press, London. Hartmann, T., Zimmer, M., 1986. Organ-speci"c distribution and accumulation of pyrrolizidine alkaloids during the life history of two annual Senecio species. J. Plant Physiol. 122, 67}80. Knight, A.P., Kimberling, C.V., Stermitz, F.R., Roby, M.R., 1984. Cynoglossum ozcinale (hound's-tongue) * a cause of pyrrolizidine alkaloid poisoning in horses. J. Am. Vet. Med. Assoc. 185, 647}650. Kopp, B., Wawrosch, C., Lebada, R., 1998. PA-freie Hu#attichblaK tter. Teil I: In-vitro-Kultivierung und SelektionszuK chtung. Dt. Apoth. Ztg. 137, 4067}4069. Langer, Th., MoK stl, E., Chizzola, R., Gutleb, R., 1996. A competitive enzyme immunoassay for the pyrrolizidine alkaloids of the senecionine type. Planta Med. 62, 267}271. LuK thy, J., Zweifel, U., Schmid, P., Schlatter, Ch., 1983. Pyrrolizidinalkaloide in Petasites hybridus L. und P. albus L. Pharm. Acta Helv. 58, 98}100. Mauz, Ch., Candrian, U., LuK thy, J., Schlatter, Ch., Sery, V., Kuhn, G., Kade, F., 1985. Methode zur Entfernung von Pyrrolizidin-Alkaloiden aus Arzneip#anzenextrakten. Pharm. Acta. Helv. 60, 256}259.
432
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Meier, B., Meier-Liebi, M., 1994. Petasites. In: HaK nsel, R., Keller, K., Rimpler, H., Schneider, G. (Eds.) Hagers Handbuch der pharmazeutischen Praxis. Vol. 6: Drogen P-Z, 5th edition, Springer, Berlin. Novotny, L., Toman, J., Stary, F., Marquez, A.D., Herout, V., Sorm, F., 1966. Contribution to the chemotaxonomy of some european Petasites species. Phytochemistry 5, 1271}1287. RoK der, E., 1995. Medicinal plants in Europe containing pyrrolizidine alkaloids. Pharmazie 50, 83}98. Schratz, E., 1963. Die physiologischen GruK nde der chemischen VariabilitaK t. Planta Med. 11, 278}286. StoK cklin, J., 1992. Umwelt, Morphologie und Wachstumsmuster klonaler P#anzen * eine UG bersicht. Bot. Helv. 102, 3}21. Van Dam, N., Verpoorte, R., van der Meijden, E., 1994. Extreme di!erences in pyrrolizidine alkaloid levels between leaves of Cynoglossum ozcinale. Phytochemistry 37, 1013}1016. Wagner, H., Bladt, S., Zgainski, E.M., 1983. In: Drogenanalyse. DuK nnschichtchromatographische Analyse von Arzneidrogen. Springer, Berlin. Wildi, E., Langer, T., Scha!ner, W., Berger-BuK ter, K., 1998. Quantitative analysis of petasin and pyrrolizidine alkaloids in leaves and rhizomes of in situ grown Petasites hybridus plants. Planta Med. 64, 264}267.
Variability in chemical constituents in Petasites hybridus from Austria Remigius Chizzola*, Bernhard Ozelsberger, Theodor Langer Institute for Applied Botany, University of Veterinary Medicine Vienna, Veterina( rplatz 1, A-1210 Vienna, Austria Received 25 February 1999; accepted 11 June 1999
Abstract Petasites hybridus (Asteraceae), butter bur, is an ancient medicinal plant with spasmolytic sesquiterpene esters. Two chemotypes, the petasine and the furanopetasine chemotype, occur in Austria. The "rst one is considered as pharmaceutically useful due to its spasmolytic constituents, but it is restricted to the northern parts of the Alps. This use, however, is impaired by the presence of low amounts of toxic pyrrolizidine alkaloids (PA), mainly senecionine and intergerrimine. PA are usually concentrated in the metabolically active parts of the complex rhizome which are the thickenings just below the leaves. They are also present in #ower stalks but are almost absent in leaf buds, the petioles and the leaf blades. The alkaloids showed a great variability within and between populations; the values recorded ranged from less than 2 to 500 mg kg~1 PA, median PA of 77 populations varied from 2 to 191 mg kg~1 in the rhizomes. In nearly 25% of the samples analysed the PA content was below 10 mg kg~1, another 25% had between 10 and 20 mg kg~1 PA. Histograms of PA concentrations in a population often showed a distinct skewness toward lower alkaloid contents. Alkaloid content was independent of sesquiterpene chemotype. The seasonal in#uence on PA content of rhizomes was little in comparison to the variability within the population or within the rhizome itself. Nevertheless, when comparable rhizome parts within a population were considered, the PA content may remain stable over several years. Although plants totally free of PA could not yet be found, it is possible to select populations low in alkaloids. Several populations of the petasine chemotype containing less than 10 mg kg~1 in the rhizomes could be found in the area investigated. ( 2000 Elsevier Science Ltd. All rights reserved. Keywords: Petasites hybridus; Asteraceae; Pyrrolizidine alkaloids; Senecionine; Sesquiterpene esters; Petasine; Furanopetasine; Population variability; Austria
* Corresponding author. Tel.: #43-1-25077-3104; fax: #43-1-25077-3190. E-mail address: [email protected] (R. Chizzola). 0305-1978/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 3 0 5 - 1 9 7 8 ( 9 9 ) 0 0 0 7 7 - 0
422
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
1. Introduction Petasites hybridus (L.) G.M. ET SCH. (Asteraceae), butter bur, is widespread on moist habitats along riversides in Central Europe. Besides its use in folk medicine several phytopharmaceutical preparations are made from the rhizomes and the leaves. The active principles are sesquiterpene esters (Fig. 1) with spasmolytic properties on the smooth musculature (Aebi et al., 1958). A gastro-protective e!ect of ethanolic Petasites extract could be attributed to the inhibition of leucotriene biosynthesis (Brune et al., 1993). An overview of the uses and the pharmaceutical quality requirements is given by Meier and Meier-Liebi (1994). The sesquiterpenes occur in two chemotypes, the petasine and the furanopetasine chemovariety (Novotny et al., 1966), which cannot be distinguished by morphological characters. Only the petasine chemotype with the main compound petasine is considered as pharmaceutically useful. The compouds of this chemotype have been studied in detail by Debrunner et al. (1995a,b). Furthermore, the rhizomes of P. hybridus contain trace amounts of toxic pyrrolizidine alkaloids (PA) (Fig. 1) (LuK thy et al., 1983) which have demonstrated a carcinogenic and mutagenic potential (RoK der, 1995). Therefore, plants used for pharmaceutical purposes must be from the petasine chemotype and low in PA. Another requirement is good vegetative growth for future agricultural production. The present work reports the natural variability in PA content in Petasites hybridus rhizomes and the distribution of the two chemotypes from populations of the Eastern Alps. To evaluate this variability the distribution of PA within the di!erent parts of the plant and the seasonal variability in alkaloid content has to be taken in account. The present data complete the report of Chizzola (1993a). 2. Materials and methods 2.1. Plant material Petasites plants were dug out at the natural growing sites and pieces of the rhizome were taken for analysis. The plant material originated from various locations in Austria, southern Bavaria and South-Tyrol. The sampling was done during six years to study the following aspects: f Approximately 1500 samples were taken from plants out of 77 populations. In 754 samples PA were investigated by capillary GC, 739 samples were analysed by enzyme immuno assay. f From further 90 di!erent sites 2}3 plants were taken and analysed. f From one large population growing near Reutte (Tyrol) 195 samples were analysed. f In one population of the petasine chemotype low in PA the alkaloids were analysed in three di!erent years. f From a population high in PA several plants were dug out as complete as possible. The rhizome was divided into obvious parts and the partitioning of the alkaloids in the individual fragments was studied.
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
423
Fig. 1. Pyrrolizidine alkaloids and sesquiterpene esters from Petasites hybridus.
f In two consecutive years samples were taken from a population in the southern Wienerwald at monthly intervals to study the variability of PA during the year. 2.2. Chemical analysis The plant material was freeze dried and "nely powdered. For PA analysis 1 g powdered rhizome was extracted at 653C 1 h with methanol. The reduction of the alkaloid N-oxides, which may represent more than 80% of the total alkloid content of the plant, was achieved with Serdoxit', an oxygen absorber resin. This reduction step was necessary since in the extracts the PA were mainly in the N-oxide form and the analytical methods used detect only free alkaloids. In the case of GC analysis the extracts were further puri"ed with an strong cation exchanger solid phase extraction column. GC was performed on a Carlo Erba Vega Series 6000 equipped with an Rtx-5 column (30 m]0.32 mm ID, 0.1 lm "lm thickness) and a temperature program from 120 to 2703C (63C/min). The main PA senecionine and integerrimine (Fig. 1) were quanti"ed using ca!eine as internal standard (Chizzola, 1994). Alternatively, these two PA were quanti"ed using a double-antibody solid-phase immuno assay according to Langer et al. (1996), a test system speci"c for the isomeres senecionine and integerrimine. In this case the methanolic extracts were diluted after reduction of the N-oxides with the assay bu!er. The results obtained were in good accordance with the GC analysis. The detection limit of PA was 2 and 0.1 mg kg~1 in the GC and immunoassay method, respectively. Senkirkine, a minor PA of the otonecine type in Petasites (Mauz et al., 1985) was not evaluated. For the determinaton of the sesquiterpene type we used a qualitative TLC method. The plant powder was extracted with dichloromethane, 5}10 ll of the resulting extract were applied to Kieselgel 60 F TLC plates (Merck 1.05715) and the 254 chromatogram was developed with xylol : acetone 9 : 1. The spots were visualized in the UV light at 254 nm or by dipping in an anisaldehyde}acetic acid}methanol reagent (Wagner et al., 1983). The petasine chemotype gave UV absorbing spots which develop after dipping and heating an yellow to light purple colour.
424
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Statistical analysis of the data was performed using SPSS for Windows version 6.0.1.
3. Results 3.1. Partitioning of PA in the plant Fig. 2 shows two examples of the complex structure of the rhizome which consists of aggregated tuberlike thickenings and long runners which may reach several meters in length. The distinct parts di!ered clearly in PA content, and the comparison of several plants even harvested at di!erent seasons allows the following observations: The parts exhibiting the highest alkaloid content were the young thickenings just below the emerging leaves. Older thickening, which may also become hollow and were generally farther away from leaf bases, were lower in alkaloids. Furthermore, the center of such a tuber contained more alkaloids than the cortex (Fig. 2b). Older runners connecting well established groups of tubers were generally low in alkaloids, whereas young runners, which by further growth will di!erentiate new leaves at their tips, might be
Fig. 2. Two examples of the complex structure of the Petasites hybridus rhizome and the repartition of pyrrolizidine alkaloids (mg kg~1) within di!erent parts of the rhizome. A. a. young runner with terminal bud, b. old runner, c. leaf, d. leaf bud, e. roots. B. b. border and c. centre of a thickening.
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
425
high in PA. The alkaloids were also present in the roots, but at lower concentrations than in most rhizome parts. In addition, the PA could be found in spring in the #ower stalks, but were absent or very low in leaf buds or leaf stalk and blades (Langer et al., 1996). 3.2. Seasonal variation of pyrrolizidine alkaloids within a population The PA content of the Petasites rhizomes parts just below the emerging leaves was followed during two consecutive years in monthly intervals. In both years comparable alkaloid contents could be recorded. Mean alkaloid content varied from 35 to 62 mg kg~1, but as can be seen from Fig. 3, all over the year there was always a great variability in PA content between the individuals of the population. Altogether the alkaloid levels appear to be higher in the "rst part of the year than in summer and autumn. 3.3. Variability of PA within a population The variability in PA content of the Petasites rhizomes within a population could be studied during the monitoring of a population over two years (Fig. 3) and in nearly 1500 individuals sampled from 77 populations throughout northern Austria (Fig. 4). In the "rst case, the alkaloids manifested high variability at any sampling date. In the latter case, the PA content of the samples collected during 4 years varied between less than 2 and 500 mg kg~1 PA. The variation coe$cients of the populations ranged from 20 to 146%. Variability within a population was not dependent upon mean or median PA content. The histogram of all samples put together (Fig. 4) shows a distinct skewness toward lower PA values. When the samples were arranged in four classes below 10 mg kg~1, between 10 and 20 mg kg~1, between 20 and 40 mg kg~1 and above 40 mg kg~1 PA, each of these classes would contain about 25% of the samples.
Fig. 3. Boxplot of the variability of PA (mg kg~1) within a population during two consecutive years. Box and whiskers represent the quartiles, the bold central horizontal bar marks the median. Black boxes: "rst year, empty boxes: 2nd year.
426
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Fig. 4. Histogram showing the PA concentration of 1493 samples of Petasites hybridus rhizomes analysed from 77 populations. Class width 10 mg kg~1 PA, the middle of each class is marked on the x-axis. The curve of the theoretical normal distribution is inscribed.
The high proportion of samples low in PA came partly from those populations generally low in alkaloids and partly from the fact that in most populations low alkaloid samples may occur. Plants of the petasine and furanopetasine chemotype showed comparable variability in the PA, but although approximately an equal number of individuals was tested from both chemotypes, the furanopetasine chemotype prevailed with 225 to 159 cases in the class below 10 mg kg~1 PA. Also many particular populations showed a skewness toward lower PA content (Fig. 5). This was also the case within a large population where in 195 individuals tested the mean PA content and the median were 26.6 and 23.3 mg kg~1 respectively, and 15 individuals had less than 10 mg kg~1 PA in their rhizomes. To test the homogeneity of populations several larger populations were divided into two to "ve sections and the variability of PA was considered separately in each section. In eight out of 22 populations tested, signi"cant di!erences in PA content could be found in the individual sections of these populations. The stability of the PA content in Petasites rhizomes over a longer period can be deduced from the comparable alkaloid levels monitored during two years. Furthermore, one population with very low alkaloid content in the rhizomes could be harvested three times in three di!erent years (Table 1). In most individuals, at each harvest the PA content remaind below 10 mg kg~1 in the rhizomes. 3.4. Geographical variation Petasites is widespread througout Central Europe. In Austria it is concentrated in the Northern and Southern Prealps and it is lacking in the great alpine valleys and basins (Niklfeld, personal communication). The geographical maps in Fig. 6 a}d present the PA content and chemotype of Petasites rhizomes from 167 sites. From 77 populations a larger number of samples has been taken, the median PA content has
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
427
Fig. 5. Boxplot representation of the variability of pyrrolizidine alkaloids in the rhizomes from 77 Petasites hybridus populations arranged according increasing median pyrrolizidine alkaloid content (mg kg~1) of population. Box and whiskers represent the quartiles, the bold central horizontal bar marks the median.
been evaluated; the data from further 90 populations are based on the mean PA content of 2}3 individuals. Most of the populations examined originated from the Northern Alps and Prealps. It can be seen that both chemotypes were present; the petasine and the furanopetasine chemotype could be found at 70 and 118 sites,
428
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Table 1 PA content (mg kg~1) in the rhizomes of a population low in alkaloids harvested in three di!erent years
Analysed by N Mean Median Minimum Maximum
Harvest 1
Harvest 2
Harvest 3
GC 15 3.8 2.0 (2 15.5
GC 13 10.1 9.6 (2 20.9
EIA 17 6.0 4.9 0.6 16.2
respectively. They occured mainly in separated populations, but there were also population made up by both chemotypes. In Austria the furanopetasine chemotype seems cover a wider range than the petasine chemotype; the latter appears to be restricted to the regions north of the Central Alps (Fig. 6 a}d) and lacking in the most eastern parts (Chizzola, 1993a). The di!erences of median or mean PA content of populations did not correlate with the various regions or sampling sites. Populations generally low in alkaloids were of either chemotype (Fig. 6a) and might be found throughout the area investigated.
4. Discussion The motive to study the chemical variability of Petasites hybridus was the need of plant material of the petasine chemotype poor in PA for pharmaceutical use. Petasites hybridus is a plant with vigorous clonal growth; tuberlike thickenings are connected by long runners attaining several meters in length. With further growth these connections between the thickenings might be lost, so that the original plant becomes fragmented and it will not be possible to decide of how many original indivmH duals, i.e genotypes, a dense Petasites population is composed. As in the Scandinavian territories populations are uniquely composed by androdynamic, i.e. functionally male plants, it is suggested that they were all derived by clonal growth from several plants introduced in the middle ages as medicinal plant by monasteries in Norway (Faegri, 1992). On the other hand, populations of plants building long runners which enable them to invade and colonize a given territory by a kind of guerilla strategy, are likely to be composed of more than one genotype (StoK cklin, 1992). In Austria presumably both cases occur. As the #ower capitules of Petasites are c Fig. 6. Geographic distribution of pyrrolizidine alkaloids and sesquiterpene chemotype in Petasites hybridus rhizomes from 167 locations in Austria and adjacent regions. legend: w Petasine chemtype o Furanopetasine chemotype d Both chemotypes in population A. Less than 10 mg kg~1 pyrrolizidine alkaloids in the population. B. 10}20 mg kg~1 pyrrolizidine alkaloids in the population. C. 20}40 mg kg~1 pyrrolizidine alkaloids in the population. D. More than 40 mg kg~1 pyrrolizidine alkaloids in the population.
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
429
430
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
composed of tubular #orets producing either pollen or ovules, the plant is functionally dioeceous. The plant #owers early in spring, so that sex distribution within the populations was recorded in a few cases only. Populations containing plants of both chemotypes or both androdynamic (male) and gynodynamic (female) individuals are certainly genetically heterogenous. The Petasites samples analysed showed a considerable variety in PA content. These di!erences have to be carefully evaluated. Genetic di!erences cannot be directly deduced from the varying amounts of a secondary plant product (Schratz, 1963). Plants growing in di!erent environments grow ordinarily at di!erent rates; they di!er in size and developmental stage (Coleman et al., 1994). A major di$culty in assessing the variability in PA is their high variability within a plant rhizome. Samples low in PA could be found in the majority of the populations, but they might have included old rhizome parts whose PA content is not representative for the whole plant, as it was not possible to analyse whole rhizomes. This circumstance was not completely avoidable due to the complex structure of the rhizome. If comparable rhizome pieces were considered, the PA content appeared to be stable for a longer period since similar alkaloid levels were recorded in the one population monitored during two years and in the other sampled three times. The in#uence of the season seems to be of minor importance. Clonal growth of Amphibromus responded readily to water and fertilizer supply, factors which may also be important within a population where they may vary within a small area (Cheplick, 1995). However, PA as secondary products do not react so dramatically to environmental factors. In Senecio vulgaris PA concentration was barely in#uenced by lack of water or nutrients (Brown & Molyneux, 1996). In this context populations showing a great di!erences in PA concentration are likely to be genetically heterogenous. Petasites belongs to the tribe of the Senecioneae, a group within the Asteraceae capable of synthesizing toxic PA. In an ecological context these alkaloids act as feeding deterrent agents against herbivores (Harborne, 1989). PA are mainly concentrated in young fast growing parts. The peaks in PA were recorded in young rosette leaves of Cynoglossum (Knight et al., 1984; Van Dam et al., 1994) or young buds emerging from the rhizomes in Symphytum (Chizzola, 1993b). In the case of Senecio vulgaris, the PA are synthetized in the roots and transported mainly to #owers to protect these reproductive organs (Hartmann and Zimmer, 1986). Petasites is a plant building up only low PA levels, therefore, it is hardly conceivable that these alkaloids might a!ord an e$cient protection against herbivory. Nevertheless, the highest PA concentrations were also found in Petasites in the young metabolic active parts of the rhizome, which are essential for the further growth of the plant. The presence of PA in Petasites appeared to re#ect the phylogenetic origin from a PA producing group. On the other hand, totally PA free individuals have not yet been found, so that a bene"cial e!ect of the PA on plant viability and survival cannot be ruled out de"nitely. With regard to PA distribution in the plant, P. hybridus is similar to Tussilago farfara, another member of the Senecioneae, where the alkaloids prevail also in the rhizomes. In this plant it was possible to select a variety without PA in the leaves, which are pharmaceutically used (Kopp et al., 1998). In accordance with the very low
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
431
PA content recorded in P. hybridus leaves (Langer et al., 1996) it seems feasible to also get PA free leaves from this species. The use of such leaves might be of interest, if they are high in the active sesquiterpenes (Wildi et al., 1998). The wide variability in PA in the rhizomes allows to select plants of the petasine chemotype low in alkaloids. In Germany the daily intake of toxic PA must not exceed 1 lg and the daily exposure in case of external application is limited to 10 lg (Meier and Meier-Liebi, 1994). Whether the plants low in PA meet these requirements depends on the production technique and the applied dose of the resulting phytopharmaceutical preparation. Actually, PA are removed from Petasites extracts by passing them several times through cation exchanger resins (Mauz et al., 1985; Meier and Meier-Liebi, 1994). The use of plants poor in alkaloids would considerably simplify these procedure. References Aebi, A., Waaler, T., BuK chi, J., 1958. Petasin und S-Petasin, die spasmolytisch wirksamen Inhaltssto!e von Petasites ozcinalis (L.) Fl. Wett. Pharm. Weekblad 93, 397}406. Brown, M.S., Molyneux, R.J., 1996. E!ects of water and mineral nutrient de"ciencies on pyrrolizidine alkaloid content of Senecio vulgaris #owers. J. Sci. Food Agric. 70, 209}211. Brune, K., Bickel, D., Peskar, B.A., 1993. Gastro-protective e!ects by extracts of Petasites hybridus: The role of inhibition of peptido-leukotriene synthesis. Planta Med. 59, 494}496. Cheplick, G.P., 1995. Genotypic variation and plasticity of clonal growth in relation to nutrient availability in Amphibromus scabrivalvis. J. Ecol. 83, 459}468. Chizzola, R., 1993a. The main pyrrolizidine alkaloids of Petasites hybridus: variation within and between populations. Acta Hortic. 333, 143}150. Chizzola, R., 1993b. Repartition of the pyrrolizidine alkaloids acetyl-intermedine and acetyl lycopsamine within the Symphytum ozcinale plant. Planta Med. 59, A644. Chizzola, R., 1994. Rapid sample preparation technique for the determination of pyrrolizidine alkaloids in plant extracts. J. Chromatogr. A 668, 427}433. Coleman, J.S., McConnaughay, K.M.D., Ackerly, D.D., 1994. Interpreting phenotypic variation in plants. Trends Ecol. Evol. 9, 187}191. Debrunner, B., Neuenschwander, M., Brenneisen, R., 1995a. Sesquiterpenes of Petasites hybridus (L.) G.M. et Sch.: distribution of sesquiterpenes over plant organs. Pharm. Acta Helv. 70, 167}173. Debrunner, B., Neuenschwander, M., 1995b. Sesquiterpenes of Petasites hybridus (L.) G.M. et Sch.: in#uence of locations and seasons on sesquiterpene distribution. Pharm. Acta Helv. 70, 315}323. Faegri, A.K., 1992. Butterbur (Petasites hybridus) * a monastery plant. Blyttia 50, 115}119. Harborne, J.B., 1989. In: Introduction to Ecological Biochemistry, 3rd edition. Academic Press, London. Hartmann, T., Zimmer, M., 1986. Organ-speci"c distribution and accumulation of pyrrolizidine alkaloids during the life history of two annual Senecio species. J. Plant Physiol. 122, 67}80. Knight, A.P., Kimberling, C.V., Stermitz, F.R., Roby, M.R., 1984. Cynoglossum ozcinale (hound's-tongue) * a cause of pyrrolizidine alkaloid poisoning in horses. J. Am. Vet. Med. Assoc. 185, 647}650. Kopp, B., Wawrosch, C., Lebada, R., 1998. PA-freie Hu#attichblaK tter. Teil I: In-vitro-Kultivierung und SelektionszuK chtung. Dt. Apoth. Ztg. 137, 4067}4069. Langer, Th., MoK stl, E., Chizzola, R., Gutleb, R., 1996. A competitive enzyme immunoassay for the pyrrolizidine alkaloids of the senecionine type. Planta Med. 62, 267}271. LuK thy, J., Zweifel, U., Schmid, P., Schlatter, Ch., 1983. Pyrrolizidinalkaloide in Petasites hybridus L. und P. albus L. Pharm. Acta Helv. 58, 98}100. Mauz, Ch., Candrian, U., LuK thy, J., Schlatter, Ch., Sery, V., Kuhn, G., Kade, F., 1985. Methode zur Entfernung von Pyrrolizidin-Alkaloiden aus Arzneip#anzenextrakten. Pharm. Acta. Helv. 60, 256}259.
432
R. Chizzola et al. / Biochemical Systematics and Ecology 28 (2000) 421}432
Meier, B., Meier-Liebi, M., 1994. Petasites. In: HaK nsel, R., Keller, K., Rimpler, H., Schneider, G. (Eds.) Hagers Handbuch der pharmazeutischen Praxis. Vol. 6: Drogen P-Z, 5th edition, Springer, Berlin. Novotny, L., Toman, J., Stary, F., Marquez, A.D., Herout, V., Sorm, F., 1966. Contribution to the chemotaxonomy of some european Petasites species. Phytochemistry 5, 1271}1287. RoK der, E., 1995. Medicinal plants in Europe containing pyrrolizidine alkaloids. Pharmazie 50, 83}98. Schratz, E., 1963. Die physiologischen GruK nde der chemischen VariabilitaK t. Planta Med. 11, 278}286. StoK cklin, J., 1992. Umwelt, Morphologie und Wachstumsmuster klonaler P#anzen * eine UG bersicht. Bot. Helv. 102, 3}21. Van Dam, N., Verpoorte, R., van der Meijden, E., 1994. Extreme di!erences in pyrrolizidine alkaloid levels between leaves of Cynoglossum ozcinale. Phytochemistry 37, 1013}1016. Wagner, H., Bladt, S., Zgainski, E.M., 1983. In: Drogenanalyse. DuK nnschichtchromatographische Analyse von Arzneidrogen. Springer, Berlin. Wildi, E., Langer, T., Scha!ner, W., Berger-BuK ter, K., 1998. Quantitative analysis of petasin and pyrrolizidine alkaloids in leaves and rhizomes of in situ grown Petasites hybridus plants. Planta Med. 64, 264}267.