Doi:10.1016/j.phytochem.2004.02.009

Phytochemistry 65 (2004) 865–873 Variation in pyrrolizidine alkaloid patterns of Senecio jacobaea Mirka Macel, Klaas Vrieling, Peter G.L. Klinkhamer Leiden University, Institute for Biology, Plant Ecology, PO Box 95162300 RA Leiden, The Netherlands Received 30 July 2003; accepted 10 February 2004 We studied the variation in pyrrolizidine alkaloid (PA) patterns of lab-grown vegetative plants of 11 European Senecio jacobaea populations. Plants were classified as jacobine, erucifoline, mixed or senecionine chemotypes based on presence and absence of thePAs jacobine or erucifoline. Due to the presence of jacobine, total PA concentration in jacobine chemotypes was higher than inerucifoline chemotypes. Both relative and absolute concentrations of individual PAs differed between half-sib and clonal families,which showed that variation in PA patterns had a genetic basis. Within most populations relative abundance of PAs varied con-siderably between individual plants. Most populations consisted either of the jacobine chemotype or of the erucifoline chemotype,sometimes in combination with mixed or senecionine chemotypes.
# 2004 Elsevier Ltd. All rights reserved.
Keywords: Senecio jacobaea; Genetic variation; Geographic distribution; Pyrrolizidine alkaloids; Chemotypes; Jacobine; Erucifoline the jacobine type and the erucifoline type, based on PAcontent in inflorescences of plants collected in the field.
Plants from the genus Senecio (Asteraceae) are known Jacobine types were characterized by the PAs jacobine for the production of a wide variety of pyrrolizidine and jacozine and lacked erucifoline, while the erucifo- alkaloids (PAs) ) that have line types contained erucifoline and acetylerucifoline but hepatotoxic and carcinogenic properties hardly any jacobine (see for structures). The two In Senecio species PAs are pro- chemotypes did not differ in concentrations of other duced in the roots as senecionine N-oxides ( PAs such as senecivernine, senecionine, integerrimine and seneciphylline. Only rarely an intermediate chemo- N-oxide is transported via the phloem to the above type was found that contained both jacobine and eruci- ground plant organs (where it is foline. The patterns described by are transformed into several related PAs (Hartmann and interesting in relation to the evolution of diversity of Dierich, 1998). The transformation of senecionine into alkaloids. However, generally only one plant per popu- related PAs differs between Senecio species and hence lation was sampled and collected in its natural habitat.
species-specific PA patterns are produced (Hartmann Therefore, it is not possible to disentangle environmen- and Dierich, 1998). Senecio jacobaea can contain more tally induced patterns and genetic effects. Here, we than 10senecionine related alkaloids investigated the PA patterns in leaves of vegetative PA concentration in S. jacobaea is partly geneti- plants from 10European populations of S. jacobaea. To cally determined ( determine whether the different PA patterns have a described two PA chemotypes for S. jacobaea, genetic basis, we studied these patterns in cloned S.
jacobaea genotypes and half-sib families grown underidentical controlled conditions to minimize a possible * Present address: University of Fribourg, Department of Biology, effect of environment. To determine the variation within Ecology and Evolution, Chemin du Muse´e 10, CH-1700 Fribourg, and between the different populations, PAs from 5 Switzerland. Tel.: +41-26-3008849; fax: +41-26-3009698.
E-mail address: (M. Macel).
plants of every population were analyzed.
0031-9422/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2004.02.009 M. Macel et al. / Phytochemistry 65 (2004) 865–873 Fig. 1. Structures and codes of PAs found in S. jacobaea. sv=senecivernine; sn=senecionine; sp=seneciphylline; in=integerrimine; jb=jacobine;jz=jacozine; jl=jacoline; jn=jaconine; er=erucifoline; fl=eruciflorine.
2. Results and discussion erucifoline (KW, df=9, for all PAs: P < 0.005). Varia-tion in PA composition within clonal families was small 2.1. Variation in PA composition within clonal families compared to variation among these families indicating astrong genetic component. Inheritance of PA patterns To show that variation in PA composition among could further be investigated by crossing genotypes with plants is genetically determined, we studied the varia- a different PA composition.
tion in PA profiles within 10clonal families of S. jaco-baea (16–34 plants per family). Each clonal family was 2.2. Variation in PA composition between half-sib from a different population (). In some plants 6 senecionine related PAs were found while other plantsonly contained two PAs. The PA composition (relative We analyzed the PA composition of offspring from 25 abundance) of each clonal family is shown in plants in two areas (in total 50half-sib families) of the Clonal families differed in relative percentages of sene- population in Meijendel (The Netherlands). All 413 cionine, seneciphylline, integerrimine, jacobine and eru- plants analyzed, except for one plant, contained mainly cifoline (KW, df=9, for all PAs: P < 0.005). The clonal jacobine, the percentage of jacobine ranged from 41 to families also differed in absolute concentrations of 100% of total PA. One plant contained mainly jacoline senecionine, seneciphylline, integerrimine, jacobine and (45%). The percentage of erucifoline ranged from 0to Table 1Origin of S. jacobaea plants used in this study. Source indicates number of plants of which seeds were used in our study Amsterdam, The Netherlands Burghaamstede, The Netherlands Driel, The Netherlands Slenaken, The Netherlands Meijendel, The Netherlands Mavellier, Switzerland Sundsvall, Sweden Ejby Adal, Denmark Schiermonnikoog, The Netherlands

M. Macel et al. / Phytochemistry 65 (2004) 865–873 Fig. 2. PA composition of S. jacobaea clonal families in average relative abundance of PAs (  S.D.). Most clonal families that contained jacobinealso contained small amounts of erucifoline. Origin is indicated in the right corner. Average PA concentration per clonal family (  SE) is given.
n=16–34.

M. Macel et al. / Phytochemistry 65 (2004) 865–873 plants from the same population that we analyzed weresometimes half-sibs (and therefore the actualvariation in PA composition in a population might beeven greater than is shown by our data.
All plants contained the PAs senecionine or seneci- phylline. In contrast to the chemotypes described by, plants that contained jacobine had Fig. 3. PA composition of S. jacobaea from Meijendel in relativeabundance of PAs (  S.D.). n=413.
often also small amounts of erucifoline (Thejacobine types described by com- 19% of total PA. shows the mean relative abun- pletely lacked erucifoline. Here, we will consider (1) dance (%) of individual PAs over all 413 plants. The plants with mainly jacobine and no or little erucifoline percentage of senecionine, seneciphylline, integerrimine, as jacobine chemotypes, (2) plants that contained both jacobine and jacozine differed among the families (KW, jacobine and erucifoline in similar amounts as mixed df=49, for all PAs P < 0.005). Percentages erucifoline chemotypes and (3) plants without jacobine but with and jacoline did not differ among the families (er: KW erucifoline as erucifoline chemotypes. In addition, some w2=61.51, df=49, ns ; jl: KW w2=45.93, df=49, ns). In plants had only trace amounts of erucifoline and con- absolute concentrations senecionine, seneciphylline, tained no jacobine. We classified these plants as sene- integerrimine, jacobine and jacozine also differed cionine chemotypes. shows the PA composition between the families (KW, df=49, for all PAs of these chemotypes.
P < 0.002). Jacoline and erucifoline concentrations did Most populations consisted of either erucifoline che- not differ between the families (KW, ns). Families also motypes or jacobine chemotypes, sometimes in combi- did differ in their total PA concentration ranging from nation with mixed or senecionine chemotypes (, 1.57 mg/g (se  0.30) to 5.51 mg/g (se  0.97) (ANOVA ). In five of the ten populations we sampled (DRL, F=2.64, df=49, P < 0.001). PA concentration did not SUN, FIL, CHE, BUR) mixed chemotypes were found differ between plants from the two areas in Meijendel (Only in one population, CHE, plants of (ANOVA, F=1.79, df=1, P=0.18). These results show pure jacobine, mixed and pure erucifoline chemotype that although all plants contained mainly jacobine, were found It is possible, however, that more considerable (genetic) variation in both PA composition ‘mixed' populations exists because we may not have a and concentration existed within the Meijendel popula-tion.
Table 2PA composition of plants from three S. jacobaea populations as an 2.3. Variation within and between populations example for the variation within and between populations From 10European S. jacobaea populations Alkaloid (% relative abundance) 4–5 plants were analyzed for their PA composition. Ten senecionine related PAs were found in some plants whileother plants contained only two or three of these PAs.
shows the average PA composition of the popu- lations. Populations differed significantly in the relative abundance of seneciphylline (KW w2=24.53, df=9, P < 0.001). Relative abundance of all other PAs did not differ between populations (KW, df=9, all P > 0.05). In absolute concentrations only jacobine differed among the populations (KW w2=24.98, df=9, P < 0.005), con- centrations of the other PAs did not differ between the populations (KW, df=9, all P > 0.05). Although a rela- tively small number of plants was used of each popula- tion, variation in PA composition within populationswas considerable and this variation confirms the more The AMS population consists of jacobine and senecionine chemo-types, SLK population consists of erucifoline chemotypes only and the extensive data from the Meijendel population above. As CHE popopulation consists of jacobine, erucifoline and mixed che- an example, shows the variation in PA compo- motypes based on the relative abundance of individual PAs. tr=trace sition of individual plants in three populations. The amounts < 1% of total PA

M. Macel et al. / Phytochemistry 65 (2004) 865–873 Fig. 4. PA compositions of S. jacobaea from European populations given as average relative abundance of PAs (  S.D.). Population is indicated inthe right corner. Average PA concentration per populations in mg/g dry weight (  SE) is given. Letters indicate chemotypes of individual plants:S=senecionine type, E=erucifoline type, J=jacobine type, M=mixed type n=4–5.

M. Macel et al. / Phytochemistry 65 (2004) 865–873 Fig. 5. PA composition of the chemotypes in relative abundance of PAs (  SE).
complete overview of all the chemotypes present in a types (). Among the plants from the populations, population due to the limited number of genotypes of the mixed chemotype and senecionine chemotype had most populations. suggested that the intermediate PA concentrations (In the clonal distribution of the erucifoline chemotype was limited to families, senecionine chemotypes also had a lower total the Eastern and Southern Europe. However, in our PA concentration compared to the jacobine chemotypes study, plants of Northern European populations (Swe- (). The higher total PA concentration of the jaco- den, Denmark) were erucifoline chemotypes ( bine chemotype is due to the ‘‘addition'' of jacobine Furthermore, plants of some populations from The (If we compared the total concentration minus Netherlands also were erucifoline chemotypes. The PA jacobine we found no significant difference between the patterns found under uniform conditions may perhaps chemotypes for the populations (ANOVA, F=1.45, not accurately reflect differences in PA patterns between df=4, P=0.281). For the clones the total PA concen- populations in the field because plants may show differ- tration without jacobine was significantly higher in the ent norms of reaction for e.g. nutrient concentration, erucifoline and senecionine chemotypes (KW w2=23.93, water availability, day length. Our data do show that df=2, P < 0.001).
not only South-Eastern populations are geneticallyequipped to produce erucifoline chemotypes.
2.5. Natural selection? The total PA concentration of the jacobine chemo- types in both the clones and the populations was higher Our data for the clonal families show that differences than the PA concentration of the erucifoline chemo- in PA composition between populations of S. jacobaea

M. Macel et al. / Phytochemistry 65 (2004) 865–873 have a genetic basis. The variation in PA composition chemical range as studied in this paper, showed that between individual plants within the populations was there was considerable variation among different popu- considerable. In the population of Meijendel, half-sib lations. This molecular variation was not linked to the families differed in PA composition and concentration.
distribution of alkaloid patterns, although one marker Molecular genetic analyses (AFLP) of European S.
was directly linked with presence of jacobine (K. van jacobaea populations, with similar geographic and bio- den Hof, J. Joshi and K. Vrieling, unpublished data).
Fig. 6. Geographic distribution for the chemotypes of generative plants and vegetative plants of S. jacobaea in Europe.
Fig. 7. Average total PA concentrations of the chemotypes from plants of (A) clones or (B) populations. Letters indicate significant differences intotal PA concentration, Populations: ANOVA, Bonferroni post-hoc tests, Clonal families: Kruskal–Wallis, post-hoc Mann–Whitney tests withBonferroni correction.
M. Macel et al. / Phytochemistry 65 (2004) 865–873 This seems to indicate that the geographical distribution DB-17 80kPa, split mode 1-20 , detectors: FID and of the chemotypes in Europe is not caused by common NPD. Structure of the PAs was verified by GC–MS ancestry but by similar selection pressures in a certain (Total PA concentration was deter- region. The genetic variation in PA composition pro- mined spectrophotometrically, by a method modified vides a basis upon which natural selection, by e.g. her- after Heliotrine (Latoxan) was used as bivores or pathogens, may act. It is unlikely that the ref. compound. The concentration of individual PAs specialist herbivore Tyria jacobaeae plays a role in the was calculated by total PA concentrationfraction of evolution of the different chemotypes because larval individual PA measured by GC analysis.
performance and oviposition preference of this specia-list moth was not affected by chemotype of S. jacobaea 3.4. Statistical analysis However, other herbivores may Statistics were performed in SPSS 8.0(SPSS Inc, have been sensitive to difference in PA composition. It 1998). Total PA concentrations of the 413 plants from has been shown that the relative effect of a PA can differ the half-sib families in Meijendel were log-transformed, between insect species (and therefore dif- after log-transformation differences in PA concentration ferences in herbivore community in S. jacobaea popula- among families were tested with a one-way ANOVA.
tions could have led to the evolution of the different Differences between individual PAs were tested with a Kruskal–Wallis test. Differences in PA composition andconcentrations among the clonal families and popula-tions were also tested with a Kruskal–Wallis test.
3.1. Growth conditions All plants were grown in 50/50 dune sand/peat mix- We are most grateful to Dr. Jens Hagen, Dr. L. Witte ture in 11 cm diameter pots in a growth chamber: pho- and Prof. Dr. T. Hartmann for their help with the GC toperiod 8 h light: 16 h dark, 20 C day/15 C night, and GC–MS analysis. We thank Helene de Vos and relative humidity 70%. After two months in the pots the Karin van Veen- van Wijk for their technical assistance plants were given ample nutrients.
and Martin Brittijn for the artwork. Urs Schaffner,Stefan Andersson, Els Schlatmann, Henk van Dijk, 3.2. Clonal families, half-sib families and populations Nico de Boer and the Botanical Garden in Copenhagengenerously provided us with seeds from the different Clonal families: We used 16–34 plants per clonal populations. We thank Nicole van Dam, Ed van der family. A clonal family consisted of one randomly Meijden and two anonymous referees for their valuable selected genotype of a population. This genotype was comments on earlier drafts of the manuscript.
propagated via tissue culture. Plants were grown for twomonths in pots before analyzing the PA patterns.
Half-sib families: We collected seeds from S. jacobaea plants in two areas of the Meijendel population. In each Cheeke, P.R., 1988. Toxicity and metabolism of pyrrolizidine alka- area we sampled 25 plants. We analyzed 4–10seedlings loids. J. Anim. Sci. 66, 2343–2350.
per plant for their PA profile, in total 413 plants. Plants Hartmann, T., Dierich, B., 1998. Chemical diversity and variation of were grown from seed for two months before analysis.
pyrrolizidine alkaloids of the senecionine type: biological need or Populations: Plants were grown from seed for 5 coincidence? Planta 206, 443–451.
Hartmann, T., Ehmke, A., Eilert, U., Von Borstel, K., Theuring, C., months and then harvested for PA analysis.
1989. Sites of synthesis, translocation and accumulation of pyrroli-zidine alkaloid N-oxides in Senecio vulgaris. Planta 177, 98–107.
Hartmann, T., Toppel, G., 1987. Senecionine N-oxide, the primary product of pyrrolizidine alkaloid biosynthesis in root cultures of The fifth youngest leaf of each plant was harvested to Senecio vulgaris. Phytochemistry 26, 1639–1643.
Hartmann, T., Witte, L., 1995. Chemistry, biology and chemoecology determine PA composition. The leaves were dried at of the pyrrolizidine alkaloids. In: Pelletier, S.W. (Ed.), Alkaloids: 50 C for 3 days and then stored at Chemical and Biological Perspectives. Vol. 9. Pergamon Press, were extracted by acid-base extraction ( Elmsford, New York, pp. 156–233.
PA composition of the plants was Hartmann, T., Zimmer, M., 1986. Organ-specific distribution and determined with GC (WCOT, 15 m0.25 mm; DB-1 accumulation of pyrrolizidine alkaloids during the life history oftwo annual Senecio species. J. Plant Physiol. 122, 67–80.
and DB-17, J&W Scientific) with following conditions: Macel, M., 2003. On the evolution of the diversity of pyrrolizidine injector: 250 C, temperature program 150–300 C, 6 C alkaloids. PhD thesis. Leiden University, Leiden, The Netherlands.
per minute, carrier gas He, pressure DB-1 100 kPa and Macel, M., Klinkhamer, P.G.L., Vrieling, K., van der Meijden, E., 2002.
M. Macel et al. / Phytochemistry 65 (2004) 865–873 Diversity of pyrrolizidine alkaloids in Senecio species does not affect from some pyrrolizidine alkaloid producing Senecio species. Plant the specialist herbivore Tyria jacobaeae. Oecologia 133, 541–555.
Cell. Rep 6, 466–469.
Macel, M., Vrieling, K., 2003. Pyrrolizidine alkaloids as oviposition Vrieling, K., de Boer, N.J., 1999. Host plant choice and larval growth stimulants for the cinnabar moth, Tyria jacobaeae. J. Chem. Ecol.
in the cinnabar moth: do pyrrolizidine alkaloids play a role? Ento- 29, 1435–1446.
mol. Exp. Appl 91, 251–257.
Mattocks, A.R., 1967. Spectrophotometric determination of unsatu- Vrieling, K., de Vos, H., van Wijk, C.A.M., 1993. Genetic analysis of rated pyrrolizidine alkaloids. Anal. Chem. 34, 443–447.
the concentration of pyrrolizidine alkaloids of Senecio jacobaea.
Mattocks, A.R., 1968. Toxicity of pyrrolizidine alkaloids. Nature 217, Phytochemistry 32, 1141–1144.
Witte, L., Ernst, L., Adam, H., Hartmann, T., 1992. Chemotypes of Toppel, G., Witte, L., Riebesehl, B., von Borstel, K., Hartmann, T., two pyrrolizidine alkaloid-containing Senecio species. Phytochem- 1987. Alkaloid patterns and biosynthetic capacity of root cultures istry 31, 559–565.

Source: https://miekevd.home.xs4all.nl/PDF/81.pdf