Lignan Profiles of Indoor-Cultivated Anthriscus sylvestris

• Abstract
• Material and Methods
• Acknowledgments
• References

Abstract

In a previous study we have shown that different populations of Anthriscus sylvestris (L.) Hoffm. (Apiaceae) yield significantly different lignan profiles. In this study we collected the seeds of A. sylvestris from 4 locations, one in England and three in The Netherlands. The seeds germinated and were grown under identical laboratory conditions. After 5 months the plants were harvested and the lignan profile in the roots and aerial parts was analysed using GC-MS. The plants from the seeds of the 4 locations showed several significant phenotypic differences in, for instance, dry-weight. However the 4 groups did not differ significantly in their lignan profile in their roots. The lignans in the roots of A. sylvestris, deoxypodophyllotoxin, yatein and anhydropodorhizol reached, on average, equal amounts in all 4 groups. Within the 4 groups, however, a large quantitative variation in lignan profile between the individual plants was observed, pointing to within-population genetic differences between individual plants.

Podophyllotoxin is a lignan that can be used to treat genital warts, one of the most common sexually transmitted diseases [1]. The same lignan is also a unique starting compound for the production of anticancer drugs, e. g., etoposide. So far podophyllotoxin has been isolated from the rhizomes of Podophyllum peltatum and Podophyllum hexandrum (Berberidaceae) plants. This is not a very ideal production system. The supply of P. hexandrum rhizomes becomes increasingly limited due to both intensive collection and lack of cultivation [2], [3]. Therefore much research effort is devoted to a more economical production system.

Especially the use of biotechnological procedures would be an interesting alternative and Anthriscus sylvestris (L.) Hoffm. (Apiaceae; wild chervil) may play an important role in this context. Wild chervil is a common weed in Northwest Europe and its rhizomes contain considerable amounts of the lignan deoxypodophyllotoxin, which can be converted into podophyllotoxin by cultures of undifferentiated plant cells or fungi [4], [5].

In a previous study we showed that there is a large variation in the lignan profile between different populations of A. sylvestris. This raised the question, if these variations are caused by environmental factors or by genetic differences. Therefore we collected seeds of A. sylvestris from 4 different locations. The seeds of each location were weighed. The average seed weight ranged from 5.2 ± 0.6 mg for the Hoogerheide location to 5.6 ± 0.2 mg for the Groningen location (x ± st. dev., n = 5, weighed per 10 seeds).

The method used yielded 80 to 90 % germination. The shape and size of the plantlets seemed not to be different to those observed in the wild. The final dry weight of the aerial parts of the plants from the Groningen location (0.46 g ± 0.19; x ± st. dev., n = 6) was significantly lower than the Leicester (2.39 g ± 1.34) and Hoogerheide (3.57 g ± 1.43) location (see Table [1]). The shoot/root ratios (calculated as dry-weight aerial parts divided by dry-weight roots) differed significantly between the plants from the seeds of the Groningen location (1.13 ± 0.23) and the Hoogerheide location (8.80 ± 3.13). These data show that there were clear phenotypic differences in growth and allocation characteristics between the plants from the seeds originating from the 4 locations.

The populations originating from the 4 locations did not show any significant difference in lignan content of the roots. The mean concentrations of deoxypodophyllotoxin, yatein or anhydropodorhizol in the roots were similar.

There was a large variation in the total lignan content between the individual plants within each group. The maximum variation within one group was found for the roots from the seeds of the Leicester location. The lowest total lignan content was 0.074 % and the highest was 0.76 %. The CV (coefficient of variance) was 108 %. The smallest variation in total lignan content was found in the roots from the Diever location, with a highest content of 0.32 % and a lowest of 0.23 % leading to a CV of 18 %. The plants of the Groningen and Hoogerheide populations showed intermediate variance (CV = 54 % and 73 %, respectively). None of the analysed plants yielded any other detectable lignans.

In the previous study [6], A. sylvestris collected from the wild contained significantly less lignans in their aerial parts. In the aerial parts from the seeds of the Diever location the concentrations of all the three lignans were significantly higher than in the roots. For the Groningen location this difference was not significant, and for the other two locations the lignan content in the roots was higher than in the aerial parts.

For the aerial parts of the plants we did find a significant difference in total lignan content between the Diever location (0.510 ± 0.105) and the Leicester location (0.158 ± 0.05). It is not clear why the indoor-cultivated plants accumulate such high concentrations of lignans in their aerial parts.

The total lignan content of the roots of plants cultivated indoors was comparable to the concentrations found in the wild. The wild population of A. sylvestris of the Groningen location yielded a lignan content of 0.51 % ± 0.35 [6], which is not significantly different to the indoor cultivated plants from that location. In this previous study of wild populations, we already noted a large variation within each studied location.

In general, it can be concluded from these data that there are phenotypic differences between the four locations as far as growth characteristics are concerned. Because the conditions during the indoor cultivation were equal for all populations it must be assumed that there is a genotypic difference between these locations. This difference does not show in the lignan content of the roots, due to the large variations in lignan profile within 2 of the 4 groups, which points to within-population genetic variation between each of the individual plants. Apparently there is little selection pressure on the lignan concentration in these populations. It is generally assumed that there is a selection pressure of herbivores or pathogens on the profile of defence metabolites of a plant, but this theory is only supported with limited data [7], [8]. In Senecio jacobaea there also seems to be little or no selection pressure on their pyrrolizidine alkaloid profile, which is not yet understood [9], still it is assumed that these compounds function as a defence against herbivores. It is therefore too early to make any assumptions on the ecological function of lignans in A. sylvestris, based on these results.

The significant differences in lignan content between the populations in the wild point to either ecological factors or differences in development (or both) that influence the lignan profile in the roots. This study does not yet deliver enough data to select and breed for high lignan producing strain of A. sylvestris. Further research in this field is necessary.

Material and Methods

Ripe seeds of Anthriscus sylvestris (L.) Hoffm. (Apiaceae) were collected from plants from 4 locations (see Table [1]). Seeds were stored cold (4 °C), dry, and dark until the germination procedure started. A voucher specimen is present in our institute coded Asylv1999 - 1.

The seeds germinated after an exposure of two weeks to a day-night regime (16 hours light and 8 hours dark) at 4 °C on a humid soil, followed by an increase in temperature to 21 °C. Most seeds germinated within one week after the temperature increase. The seedlings were then transferred to potting soil and planted 1 cm deep, with 5 cm distance between the seedlings and 5 seeds in every pot. For every location we used 5 pots. The seedlings grew under day-night regime (16 hours light and 8 hours dark; 3,000 lux using L36W/10 daylight lamps (Osram, Germany). The temperature was kept between 23 and 25 °C, with 60 % humidity. The plants were watered every other day. The harvesting took place exactly 5 months after the transfer to the potting soil. After harvesting the plants were air-dried, divided in roots and aerial parts, weighted, and ground for analyses.

All the plant material was analysed with the previously reported GC and GC-MS methods [10]. The data obtained with the GC-MS analysis was subjected to statistical analysis using SPSS for Windows (version 11.0.1) (SPSS Inc. Illinois, USA).

Acknowledgments

We would like to thank Dr. Randolph Arroo for the seeds from Leicester, Prof. Manuel Medarde for providing us with different lignans as reference compounds, and Prof. Dr. Jelte van Andel for his helpful comments on an earlier draft of this manuscript.

References

• 1 Gross G. Clinical diagnosis and management of anogenital warts and papillomavirus-associated lesions.  Hautarzt. 2001;  52 6-17
  CrossrefPubMedSearch in Google Scholar
• 2 Choudhary D K, Kaul B L, Khan S. Cultivation and conservation of Podophyllum hexandrum  - An overview.  J Med Arom Plant Sci. 1998;  20 1071-3
  PubMedSearch in Google Scholar
• 3 Rai L K, Prasad P, Sharma E. Conservation threats to some important medicinal plants of the Sikkim Himalaya.  Biol Cons. 2000;  93 27-33
  PubMedSearch in Google Scholar
• 4 Van Uden W, Bos J A, Boeke G M, Woerdenbag H J, Pras N. The large scale isolation of deoxypodophyllotoxin from rhizomes of Anthriscus sylvestris followed by its bioconversion into 5-methoxypodophyllotoxin β-d-glucoside by cell cultures of Linum flavum .  J Nat Prod. 1997;  60 401-3
  CrossrefPubMedSearch in Google Scholar
• 5 Kondo K, Ogura M, Midorikawa Y, Kozawa M, Tsuijbo H. et al . Conversion of deoxypodophyllotoxin to podophyllotoxin-related compounds by microbes.  Agric Biol Chem. 1989;  53 777-82
  PubMedSearch in Google Scholar
• 6 Koulman A, Bos R, Medarde M, Pras N, Quax W J. A fast and simple GC-MS method for lignan profiling in Anthriscus sylvestris and biosynthetically related plant species.  Planta Med. 2001;  67 858-62
  Thieme ConnectPubMedSearch in Google Scholar
• 7 Carroll M J, Berenbaum M R. Behavioral responses of the parsnip webworm to host plant volatiles.  J Chem Ecol. 2002;  28 2191-201
  CrossrefPubMedSearch in Google Scholar
• 8 Talley S M, Coley P D, Kursar T A. Antifungal leaf-surface metabolites correlate with fungal abundance in sagebrush populations.  J Chem Ecol. 2002;  28 2141-68
  CrossrefPubMedSearch in Google Scholar
• 9 Macel M, Klinkhamer P GL, Vrieling K, van der Meijden E. Diversity of pyrrolizidine alkaloids in Senecio species does not affect the specialist herbivore Tyria jacobaeae .  Oecologia. 2002;  133 541-50
  CrossrefPubMedSearch in Google Scholar
• 10 Bernasconi C. Etoposide: fifteen years experience.  Bone Marrow Transplant. 1989;  4 52-5
  PubMedSearch in Google Scholar

Dr. Rein Bos

Department of Pharmaceutical Biology

GUIDE (Groningen University Institute for Drug Exploration)

University of Groningen

A. Deusinglaan 1

9713 AV Groningen

The Netherlands

Phone: +31-50-3633354

Fax: +31-50-3633000

Email: R.Bos@farm.rug.nl

References

• 1 Gross G. Clinical diagnosis and management of anogenital warts and papillomavirus-associated lesions.  Hautarzt. 2001;  52 6-17
  CrossrefPubMedSearch in Google Scholar
• 2 Choudhary D K, Kaul B L, Khan S. Cultivation and conservation of Podophyllum hexandrum  - An overview.  J Med Arom Plant Sci. 1998;  20 1071-3
  PubMedSearch in Google Scholar
• 3 Rai L K, Prasad P, Sharma E. Conservation threats to some important medicinal plants of the Sikkim Himalaya.  Biol Cons. 2000;  93 27-33
  PubMedSearch in Google Scholar
• 4 Van Uden W, Bos J A, Boeke G M, Woerdenbag H J, Pras N. The large scale isolation of deoxypodophyllotoxin from rhizomes of Anthriscus sylvestris followed by its bioconversion into 5-methoxypodophyllotoxin β-d-glucoside by cell cultures of Linum flavum .  J Nat Prod. 1997;  60 401-3
  CrossrefPubMedSearch in Google Scholar
• 5 Kondo K, Ogura M, Midorikawa Y, Kozawa M, Tsuijbo H. et al . Conversion of deoxypodophyllotoxin to podophyllotoxin-related compounds by microbes.  Agric Biol Chem. 1989;  53 777-82
  PubMedSearch in Google Scholar
• 6 Koulman A, Bos R, Medarde M, Pras N, Quax W J. A fast and simple GC-MS method for lignan profiling in Anthriscus sylvestris and biosynthetically related plant species.  Planta Med. 2001;  67 858-62
  Thieme ConnectPubMedSearch in Google Scholar
• 7 Carroll M J, Berenbaum M R. Behavioral responses of the parsnip webworm to host plant volatiles.  J Chem Ecol. 2002;  28 2191-201
  CrossrefPubMedSearch in Google Scholar
• 8 Talley S M, Coley P D, Kursar T A. Antifungal leaf-surface metabolites correlate with fungal abundance in sagebrush populations.  J Chem Ecol. 2002;  28 2141-68
  CrossrefPubMedSearch in Google Scholar
• 9 Macel M, Klinkhamer P GL, Vrieling K, van der Meijden E. Diversity of pyrrolizidine alkaloids in Senecio species does not affect the specialist herbivore Tyria jacobaeae .  Oecologia. 2002;  133 541-50
  CrossrefPubMedSearch in Google Scholar
• 10 Bernasconi C. Etoposide: fifteen years experience.  Bone Marrow Transplant. 1989;  4 52-5
  PubMedSearch in Google Scholar

Dr. Rein Bos

Department of Pharmaceutical Biology

GUIDE (Groningen University Institute for Drug Exploration)

University of Groningen

A. Deusinglaan 1

9713 AV Groningen

The Netherlands

Phone: +31-50-3633354

Fax: +31-50-3633000

Email: R.Bos@farm.rug.nl