This is an author version of the contribution published on: [Rubiolo, P., Matteodo, M., Bicchi, C., Appendino, G., Gnavi, G., Bertea, C., Maffei, M.] Chemical and Biomolecular Characterization of Artemisia umbelliformis Lam., an Important Ingredient of the Alpine Liqueur "Genepi", JOURNAL OF AGRICULTURAL AND FOOD CHEMISTRY, Volume: 57 Issue: 9 Pages: 3436-3443 Published: MAY 13 2009, DOI: 10.1021/jf803915v, ACS] The definitive version is available at: [http:// pubs.acs.org/doi/abs/10.1021/jf803915v]
Chemical and Biomolecular Characterization of Artemisia umbelliformis Lam., an Important
Ingredient of the Alpine Liqueur “Genepì”
APPENDINO,‡ GIORGIO GNAVI,§ CINZIA BERTEA§ AND MASSIMO MAFFEI §
Dipartimento di Scienza e Tecnologia del Farmaco, Università di Torino, Via P. Giuria 9, I-10125
Torino, Italy; Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche,
Università del Piemonte Orientale, Via Bovio 6, 28100 Novara, Italy; and Unità di Fisiologia
Vegetale, Dipartimento di Biologia Vegetale, Università di Torino, Via Quarello 11/A, 10135
Running head: Chemical and Biomolecular Characterization of Artemisia umbelliformis Lam.
14 15 16
(Tel:+390116707661; Fax:+390116707687; E-mail: [email protected]
Dipartimento di Scienza e Tecnologia del Farmaco, Università di Torino
Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, Università del
Piemonte Orientale, Novara
Dipartimento di Biologia Vegetale, Università di Torino.
A. umbelliformis Lam., an important Alpine plant used for preparation of flavoured beverages
showed a remarkable intraspecific variability, both at genomic and gene products (secondary
metabolites) level. The variability of A. umbelliformis Lam. currently cultivated in Piedmont (Italy,
Au1) and in Switzerland (Au2) was investigated by combining the chemical analysis of essential oil
and sesquiterpene lactones and the molecular characterization of the 5S-rRNA-NTS gene by PCR
and PCR-RFLP. Marked differences were observed between the two plants. Au1 essential oil
contained α- and β-thujone as main components while Au2 1,8-cineole, borneol and β-pinene. Au1
sesquiterpene lactone fractions contained cis-8-eudesmanolide derivatives and Au2 the trans-6-
germacranolide costunolide. Specific A. umbelliformis Au1 and Au2 primers were designed on the
sequence of the 5S-rRNA gene spacer region. Furthermore, a PCR–restriction fragment length
polymorphism (PCR–RFLP) method was applied using RsaI and TaqI restriction enzymes.
Chemical and biomolecular data contributed to characterize A. umbeliformis chemotypes.
36 37 38
Keywords: Artemisia umbelliformis Lam.; alpine liqueurs; chemotype discrimination; chemical
analysis; biomolecular analysis.
Artemisia umbelliformis Lam. is an Alpine species used to prepare “genepì”, a highly praised
liqueur characterized by a bitter taste and a peculiar flavour (1). These properties have been traced
to the volatile constituents and to the sesquiterpene lactone fraction of the plant, that are
characterized by a high contents of α- and β-thujone (2, 3) and by the presence of the cis-
eudesmanolide sesquiterpene lactones 1-3, respectively (4-6).
Thujone is a natural terpenoid also associated with common wormwood (Artemisia absinthium L.)
and Roman wormwood (Artemisia pontica L.), absinthe’s most widely used ingredients (7 and
reference cited there in ). There is currently a heated debate on the toxicity of absinthe and thujones
(7 and reference cited there in), but the EU legislation have imposed a limit of 35 ppm to the total
amount of these compounds in alcoholic beverages (8). To overcome this issue, thujones-free
chemotypes of A. umbelliformis have been selected by horticultural techniques (9). Remarkably, an
investigation on the sesquiterpene lactone fraction of one of these thujones-free chemotypes showed
dramatic differences from the wild plant. Thus, the C-8 cis sesquiterpene lactones typical of A.
umbelliformis from Western Alps (1-3) were replaced by the C-6 trans lactones 4-6a,b, while a
structurally unique sesterpene lactone (7) was also detected (10) (Figure 1). Chemotypes (or
chemical phenotypes) are generally considered the phenotypical expression of a genotype, although
different chemotypes may derive from the same genotype. This means that, according to
environmental conditions, the same genotype may express different chemical patterns, or, on the
other side, that different genotype may respond to the same environmental pressure with the same
phenotypic expression. In this context, molecular genetic methods have recently been shown to be
very effective in genotypic discrimination. Genetic methods focus on genotype rather than
phenotype, and DNA based experiments are now widely used for a rapid identification (and
therefore autentication) of medicinal plants. Bertea et al. (11) recently showed that molecular
approaches are a powerful tool to distinguish the Acorus calamus diploid β-asarone-free cytotype
from the other cytotypes containing it. The same group also used specific Salvia divinorum primers 3
designed on the sequence of the 5S-rRNA gene spacer region (12) to develop a Real-Time PCR-
based mathematical model to quantify S. divinorum in commercial plant samples or hallucinogenic
preparation (13). Given the potential of this approach, it seemed interesting to apply a combination
of biomolecular and chemical techniques to characterise the chemotypes of A. umbelliformis
currently cultivated in Piedmont (Italy, Au1) and in Switzerland (Au2), complementing the analysis
of their essential oil and sesquiterpene lactones with a molecular characterization by PCR and PCR-
RFLP of the 5S-rRNA-NTS region of their genome.
MATERIALS AND METHODS
Chemicals - Thujones standard mixture (mixture of α-thujone and β-thujone purity 99.9%) and all
other pure reference compounds were from Sigma-Aldrich (St. Louis MO, USA). Sabinene was
France); sabinyl ester homologous series were synthesized in the authors’ laboratory (2). HPLC and
analytical grade solvents were from Carlo Erba Reagenti, Rodano, Italy. The sesquiterpene lactones
1-7 were available from previous studies (4-6, 10)
Plant Material - Forty three samples of Artemisia umbelliformis Lam aerial parts from
experimental cultivations run in different Alpine valleys [Val Grana (latitude 44°25’N, longitude
7°20’E), Valle Stura (44°21’N, 7°26’E), Valle Maira (44°28’N, 7°22’E), Val Chisone (44°57’N,
6°52’E)] at an height of least 1300 m a.s.l. Fresh plant material was directly indoor-dried by the
farmers under controlled temperature and humidity up to a constant weight, in agreement with the
WHO’s guidelines on Good Agricultural and Collection Practices (GACP) for medicinal plants.
Voucher specimen representative of the two chemotypes (native, defined Au1 and selected in
Switzerland, defined Au2) are deposited at the Dipartimento di Scienza e Tecnologia del Farmaco
(n°231 for Au1 and n°232 for Au2). For each chemotype, batches of 1 kg of aerial dried parts (see
above) were supplied by the “Associazione Genepì Occitan” (Cuneo - Italy)
(Irvine CA, USA); sabinol was kindly supplied by Robertet SA (Grasse,
Essential oils and Headspace solid phase-microextraction (HS-SPME) sample preparation -
Essential oils (EOs) were prepared according to the method of the European Pharmacopoeia (14).
Ten grams of dried aerial parts was suspended in 250 mL of water in a 500 mL flask for 1 h, and
then submitted to hydrodistillation in a Clevenger micro-apparatus for 2 h (2). The resulting EO was
left to stabilize for 1 h, then recovered with hexane and then analyzed by GC-MS.
The SPME device and the three component CAR/PDMS/DVB fused silica fiber (2 cm long, coating
volume: 1.000 µm3) were purchased from Supelco (Bellafonte, PA, USA), (15). Before use, the
fiber was conditioned as recommended by the manufacturer.
Each sample (200 mg of A. umbelliformis dried aerial parts) hermetically sealed in a 2.0 mL vial
was introduced in a thermostatic bath at 80 °C for 15 min; the SPME device was inserted in the
sealed vial containing the sample, and the CAR/PDMS/DVB fiber exposed to the matrix headspace
(30 min). The vial was vibrated for 10 s every 5 min with an electric engraver (Vibro-Graver V74),
(Burgess Vibrocrafters Inc., Brayslake, IL). After sampling, the SPME device was immediately
inserted into the GC injector and the fiber thermally desorbed. A desorption time of 5 min at 230 °C
was used. Before sampling, each fiber was reconditioned for 20 min in the GC injection port at 230
Sesquiterpene lactone extraction - One gram of dried aerial parts of both chemotypes were
sonicated three times with ethanol 96% (50 mL) for 10 min. The resulting total extract (150 mL)
was filtered, and evaporated to dryness under vacuum; the weighed solid residue was dissolved in
acetonitrile/water 20/80 at a concentration of 0.1 mg/mL and analyzed by high-performance liquid
chromatography-diode array-ultraviolet detection-mass spectrometry (HPLC-DAD-UV-MS).
GC and GC-MS Analyses - GC analyses were carried out on a Shimadzu QP2010 system provided
with a FID and a MS detector, and the results processed by GC Solution software and GC-MS
solution software (2.51version) (Shimadzu Italia, Milano Italy). Capillary GC-FID-MS analyses
were carried out on two 25 m, 0.25 mm i.d., 0.25 µm columns from MEGA (Milano – Italy), i.e.
Mega5 (95% polydimethyl-siloxane, 5% phenyl) and MegaWax (polyethyleneglycol, PEG20M). 5
GC and GC-MS conditions: injection mode: split; split ratio: 1: 20 Temperatures: injector: 230 °C,
transfer line: 230 °C; ion source: 200 °C; carrier gas: He flow-rate: 1.0 mL /min in constant flow-
mode. MS detector was in electron impact ionization mode (EI) at 70 eV, scan rate was 1111 amu/s
and mass range of 35–350 m/z. Temperature program: from 50 °C (1 min) to 220 °C (5 min) at 3 °C
EOs and headspace components were identified by comparison of both their linear retention indices,
calculated versus a C8-C25 hydrocarbon mixture, and their mass spectra with those of authentic
samples or with data in the literature.
Quantitative analysis. Suitable amounts of α + β-thujone commercial standard were diluted with
cyclohexane to obtain six concentration levels ranging from 0.5 ng /µL to 6 ng/µL for α-thujone
and from 0.04 ng/µL to 0.5 ng/µL for β-thujone. Calibration curves were obtained by analyzing the
resulting standard solutions three times by GC-FID using n-nonane as internal standard.
HPLC-DAD-UV analysis - HPLC-DAD-UV analyses were carried out on a Shimadzu 2010EV
system provided with a PDA detector (Shimadzu, Dusseldorf Germany). A 150 x 4.6 mm i.d., 5
μm, Zorbax Stable Bond column (Agilent, Waldbronn Germany) was used. Analysis conditions
were: mobile phase: eluent A: 20% acetonitrile/water; eluent B: 100% acetonitrile; mobile phase
gradient: from 100% A to 100% B in 25 min. Injection volume: 10 μL, flow rate: 1 mL/min. UV
detection wavelengths: 210nm.
Quantitative analysis – Suitable amounts of custonolide and umbellifolide were diluted with
methanol to obtain concentrations ranging from 0.5, 5, 10, 25, 50 and 100 ng/μL of each marker,
respectively. Umbellifolide (3) was adopted as standard for quantitation also for the
hydroperoxytelekins 1 and 2, because of the similarity of structures, their trace abundance and
chemical instability; the results are expressed as the sum umbellifolide + hydroperoxytelekins. A
calibration curve was made by analyzing the resulting standard solutions three times by HPLC-
DAD-UV at 210 nm.
HPLC-MSD analysis - HPLC-MSD analyses were carried out with a single quadrupole Shimadzu
2010EV system (Shimadzu, Dusseldorf, Germany) equipped with an orthogonal atmospheric
pressure chemical ionization (APCI) and electrospray ionization (ESI) sources. The same column
and mobile phase as for HPLC-DAD-UV analysis was used. Flow rate: 0.8 mL/min. MSD
conditions: MS-APCI+; temperature: 400 °C; nebulizer’s flow: 2.5mL/min; CDL voltage: 250 °C;
Q-Array: MS-ESI+; temp:250 °C; nebulizer’s flow: 1,5mL/min; CDL voltage: 250 °C.
MSD analysis conditions were optimized by direct flow injection of pure standards of costunolide
(4) and umbellifolide (3) and of a fraction containing a mixture of hydroperoxytelekines
Genomic DNA extraction - Plant material employed for the chemical analyses was also used for
genomic DNA extraction. . Fifty mg of dried material was frozen in liquid nitrogen and ground to a
fine powder with Tissue Lyser (Qiagen, Hilden, Germany). Genomic DNA was extracted from the
ground powder by using the Nucleospin Plant Kit (Macherey Nagel, Düren, Germany) following
manufacturer’s instruction. The quantity and quality of the DNA were assessed by
spectrophotometric analysis by using the Nanodrop ND-1000 (Thermo Scientific, Wilmington, DE,
USA) from several samples of the two chemotypes.
PCR amplification, subcloning and sequencing - Approximately 20 ng of genomic DNA isolated
from powdered leaf material of Au1 and Au2 were used as a template for PCR amplification with
forward primer 5S-P1 (5′ - GTGCTTGGGCGAGAGTAGTA-3′) and reverse primer 5S-P2 (5′ -
TTAGTGCTGGTATGATCGCA-3′) flanking the NTS of 5S-rRNA gene (11-13, 16). The
amplification was carried out in a 50 µL reaction mixture containing 5 µL 10X PCR reaction buffer
(Fermentas), 0.2 mM dNTPs, 20 pmol forward and reverse primers and 0.5 U of Taq DNA
polymerase (Fermentas, Glen Burnie, MA, USA). The PCR reactions were carried out in a
Whatman Biometra T-Gradient Thermalcycler (Whatman Biometra, Goettingen, Germany).
Cycling conditions consisted of an initial 4 min at 94 °C, followed by 30 sec denaturing at 94 °C, 1
min annealing at 52 °C, and 1 min elongation at 72 °C repeated for 30 cycles and with 5 min final
extension at 72 °C. 7
One microliter of the amplification reaction was analyzed by capillary gel electrophoresis (CGE)
using the Agilent 2100 Bioanalyzer (Agilent Technologies) and the DNA 1000 LabChip Kit
(Agilent Technologies) following manufacturer’s instructions. The DNA 1000 LabChip Kit
provides sizing and quantitation of dsDNA fragments ranging from 25 to 1000 bp. PCR products
were also analyzed by a 2% agarose gel electrophoresis and visualized by ethidium bromide
staining under UV. From this gel a band of about 220 bp for Au1 and about 320 bp for Au2 was
purified by using the Nucleospin Extract II Kit (Macherey Nagel) and then subcloned into pGEM-T
Easy vector (Promega). The ligated products were transformed into the Escherichia coli Subcloning
DH5α Efficiency Competent Cells (Invitrogen). Colonies containing DNA inserts of the correct size
were picked and grown overnight in 3 mL of Luria–Bertani (LB) liquid medium. The mini-
preparation of plasmid DNAs were performed using QIAprep Spin Miniprep Kit (Qiagen),
following manufacturer’s instructions. The plasmid DNAs were employed as a template for
sequencing. Both strands of DNA were sequenced at least twice and the sequences were aligned by
using ClustalX software.
PCR amplification using specific primers for Au1 and Au2 - Sequences derived from powdered
leaf material of Au1 and Au2 were aligned in a unique sequence that allowed the design of two
forward primers AuF 5′ -CTAGGATGGGTGACCTCCTG -3′(which is common to both
chemotypes) and Au2F 5′-GCGGTGACAGAGTCGGTAAA-3′ and two reverse specific primers:
TCCTTTCTCATTGCCTATTTTTC -3′, which corresponded respectively to nucleotides 21-40,
168-187, 212-231, 253-275 of Au2 non-transcribed spacer (NTS) sequence. The internal primers
were used for amplification also in combination with primer 5S-P1 and 5S-P2.
The conditions of the PCR reactions were the same as mentioned above. One microliter of the
amplification products were separated by CGE with the Agilent 2100 Bioanalyzer (Agilent
Technologies) and DNA 1000 LabChip Kit (Agilent Technologies) following manufacturer’s
PCR–RFLP - The purified PCR products of the 5S-rRNA gene spacer region of both Au1 and Au2
chemotypes were either digested with 10 U of RsaI (Amersham Biosciences) at 37 °C for 1h or, in a
separate reaction, with 10 U of TaqI (Sigma) at 65 °C for 1 h. One microliter of both digestion
reactions was fractionated by CGE using the Agilent 2100 Bioanalyzer (Agilent Technologies) and
DNA 1000 LabChip Kit (Agilent Technologies) following manufacturer’s instructions.
RESULTS AND DISCUSSION
This study aims to characterize the two chemotypes of A. umbelliformis Lam. under investigation
by combining results from chemical and genomic analyses of 43 samples from experimental
cultivation (Au1, Italian native and Au2, selected in Switzerland).
Chemical analyses - Chemical analyses investigated the fractions responsible for plant odor and
taste i.e. the composition of the volatile fraction including quantitation of α- and β-thujone, and that
of the sesquiterpene lactone fraction (i.e. the components responsible for the liqueur bitter taste).
The volatile fraction was studied by analyzing both the essential oil (EO) obtained by
hydrodistillation and the headspace (HS) sampled by solid phase microextraction (HS-SPME)
combined with GC and GC-MS. HS-SPME sampling was applied in view of developing a fully
automatic control method, to be run in combination with GC and GC-profile multivariate analysis
(Principal Component Analysis (PCA). Table 1 reports the average percent areas normalized vs. n-
nonane as internal standard and percent range of the characteristic components of the EOs of the
samples investigated together with their Linear Retention Indices (LRI) on both GC columns. Table
2 reports calibration curves, mean and range amounts of α- and β-thujone in the samples
investigated. This Table considers only ten samples of chemotype Au2 on seventeen because the
alpha and beta-thujone peak areas of the remaining seven samples were too low to be correctly used
for quantitative determination. From these results it is clear that the two chemotypes are
characterized by a different composition: Au1 was found to contain α- and β-thujone and an
homologous series of sabinyl esters as main components, whereas in the chemotype Au2, 1,89
cineole, borneol and β-pinene were the major compounds. Moreover, thujones were almost absent
from the Au2 chemotype, their total amount accounting from 0.2 to 0.4g/100g of EO, while in the
Au1 chemotype thujones ranged from 18 to about 58 g/100g of EO. The results obtained by HS-
SPME-GC analysis, although not directly comparable, were in full agreement with those of the
EOs, as shown by the PCA scatterplot of Figure 2. Each chemotype is clearly discriminated and the
samples belonging to the same chemotype analysed by HS-SPME-GC and through their EO are
coherently positioned in the PCA scatterplot (Figure 2).
Significant differences can also be found in the composition of the non-volatile bitter fraction.
Figure 3 reports the HPLC-DAD-UV profiles of two samples belonging to Au1 and Au2
chemotypes, respectively. The bitter taste of the native Au1 chemotype is mainly due to
sesquiterpene lactones of the cis-8-eudesmanolide type [5-deoxy-5-hydroperoxy-5-epitelekin (1), 5-
deoxy-5-hydroperoxytelekin (2), umbellifolide (3) (4-6)]. On the other hand, the Au2 chemotype is
characterized by high amounts of costunolide (4), a germacranolide typical of A. genipi Weber (4),
and by the presence of an unusual sesterpene lactone, named genepolide (7) (10). An in depth
investigation of the Au2 ethanolic extract composition after fractionation by column
chromatography in combination with NMR, and analysis by HPLC-UV and HPLC-single-
quadrupole-MS revealed that costunolide (4) was accompanied by a series of related oxygenated
sesquiterpene lactones (artemorine (5b), santamarine (6a),and reynosine(6b)) (4). On the other
hand, in the Au1 chemotype the presence of both telekine hydroperoxides and umbellifolide was
Costunolide and the sum of hydroperoxytelekins and umbellifolide were adopted as markers of the
two chemotypes to evaluate quantitatively the bitter fraction of the 43 samples under investigation.
These analyses showed an average amount of umbellifolide + hydroperoxytelekins expressed as
umbellifolide (y = 29.9392x -1.6795; R2: 0.99999) of 0.11 g/100g of dried plant material in a range
varying between 0.03 and 0.37 g/100g for the Au1 chemotype, and an average amount of
costunolide (y = 101.1709x-31.0592; R2 : 0.99975) of 0.56 g/100g of dried plant material in a range 10
between 0.20 and 0.93 g/100g for the Au2 chemotype. Costunolide was also detected in very low
amounts in some samples of the Au1 chemotype, its percentage never exceeding 0.05 %
Molecular characterization of the two A. umbelliformis chemotypes - In higher eukaryotes, the
5S-rRNA gene occurs in tandemly repeated units consisting of an 120 bp coding region separated
by a non-transcribed spacer varies from species to species (16). Thus the diversity of the spacer
region can be used as an identification basis (17).
Here, two primers flanking the spacer region of 5S-rRNA, already successfully employed for
differing Acorus calamus chemotypes (16), A. calamus cytotypes (11) and Salvia divinorum both as
pure plants (12) and in plant mixtures (13) were used in PCR analysis of genomic DNA isolated
from the two chemotypes Au1 and Au2.
A single fragment of approximately 220 bp was produced by Au1 (Figure 4, lane 1) and a single
fragment of about 320 bp was produced by Au2 (Figure 4, lane 2). Fragments derived from both
chemotypes were ligated into pGEM-T Easy vector and the nucleotide sequence was determined.
The sequenced region spans 224 bp for Au1 (NCBI GenBank Accession No. EU816950) and 327
bp for Au2 (NCBI GenBank Accession No. EU816951).
Sequence alignment of the 5S-rRNA spacer region flanked by the 3′-and 5′-ends of the coding
region is shown in Figure 5. Surprisingly, Au1 presented a difference of 103 nucleotides with
respect to Au2. This difference is quite consistent but not uncommon between chemotypes or
cytotypes, as it has been previously demonstrated with other plant species (11, 16).
In order to characterize better the two chemotypes and to simplify the identification method,
nucleotide sequences of the 5S-rRNA gene spacer region were used to design four specific primers
PCR products derived from all possible combinations of Au1 and Au2 specific primers also used
with the primers designed on the coding regions of the plant 5S-rRNA gene, were analyzed. In the
chemotype Au1 a single fragment of 204 bp in length was amplified using the primer AuF in
combination with the primer 5S-P2 (Figure 4, lane 3). The same strategy used with the chemotype 11
Au2 produced a single fragment of 307 bp in length (Figure 4, lane 4). The three additional specific
primers designed for Au2 gave a combination of single fragments as follows: 231 bp with 5S-P1
and Au2R1 (Figure 4, lane 5), 275 bp with 5S-P1 and Au2R2 (Figure 4, lane 6), 64 bp with Au2F1
and Au2R1 (Figure 4, lane 7), 108 bp with Au2F1 and Au2R1 (Figure 4, lane 8), 213 bp with AuF
and Au2R1 (Figure 4, lane 9) and 255 bp with AuF and Au2R2 (Figure 4, lane 10). All
amplifications occurred only in Au2 when the Au2 specific primers were used and no PCR products
were detected when Au1 DNA was employed as a template.
In addition, a PCR–RFLP method was applied. From the identified sequences, two RsaI sites could
be found in both chemotype 5S-rRNA spacer regions at 18 bp and 141 bp position in Au1 and at 18
and 301 bp in Au2 (Figure 5). Purified PCR products obtained by using 5S-P1 and 5S-P2 primers
were digested with Rsa I. As expected, PCR products from the chemotype Au1 could be digested by
RsaI giving two major fragments of 123 bp and 82 bp respectively, and a minor fragment of 19 bp
not visible in the gel because out of the resolution capacity of the instrument (Figure 4, lane 11).
When purified PCR products from the chemotype Au2 were digested using RsaI, a completely
different RFLP profile was observed. RsaI cleaved Au2 5S-rRNA spacer region giving a major
fragment of 283 bp and two minor fragments of 25 bp (barely visible in the gel) and 19 bp (not
visible in the gel) (Figure 4, lane 12). A TaqI site was also identified in the sequence of the
chemotype Au2 (see Figure 5). As expected, PCR products from the chemotype Au1 could not be
digested by Taq I (Figure 4, lane 13), whereas purified PCR products from the chemotype Au2.
Thus, our biomolecular characterization provides a useful tool for the unequivocal characterization
of the two chemotypes. If some intermediate chemotypes due to meiotic rearrangements were
already present in the population, they would be detected by using universal primers (different
fragment size or different nucleotide composition), but this was not the case, since all samples gave
the same results. Besides, this marker has been successfully used for Acorus calamus chemotype
determination (16) and it represents a powerful tool to deduce genetic relationship of medicinal
plants, especially at the intra-specific level. 12
In conclusion, these results clearly support the view that A. umbelliformis, a valuable plant for
Alpine agriculture, shows a remarkable intraspecific variabilty, both at the genomic and gene
products (secondary metabolites) levels. This multidisciplinary study, by showing remarkable
chemical variation in the terpenoid profile, and consistent genomic difference in the 5S-rRNA
spacer regions, has identified two chemotypes of A. umbelliformis. A multidisciplinary approach
based on the combination of metabolome- and genome-derived product analysis enabled the
unequivocal chemical and biomolecular fingerprinting of these two chemotypes. Combined “omics”
approaches are becoming a useful tool not only for basic science but also for industrial plant
characterization. Owing to the commercial relevance of A. umbelliformis and to the regulatory
issues related to the presence of thujones, the identification of a RsaI and Taq sites can be used for a
rapid and precise chemotype identification of the plant chemotypes, complementing the chemical
analysis of the essential oil and the sesquiterpene lactones.
1. Mucciarelli, M.; Maffei, M. Introduction to the genus. In Medicinal and Aromatic Plants -
Industrial Profiles: Artemisia, Wright, C. W., Ed.; Taylor&Francis: London, 2002; pp 1-50.
2. Bicchi, C.; Nano, G. M.; Frattini, C. On the composition of the essential oils of Artemisia
genipi Weber and Artemisia umbelliformis Lam. Z. Lebensm.-Unters.-Forsch. 1982, 175, 182-
3. Bicchi, C.; D'Amato, A.; Nano, G. M.; Frattini, C. Capillary GLC controls of some alpine Artemisiae and related liqueurs. Chromatographia 1985, 18, 560-566.
4. Appendino, G.; Belliardo, G. M.; Nano, G. M.; Stefenelli, S. Sesquiterpene lactones from
Artemisia genepi Weber: isolation and determination in plant material and in liqueurs. J. Agr.
Food Chem. 1982, 30, 518-521. 13
5. Appendino, G.; Gariboldi, P.; Nano, G. M. Isomeric hydroperoxy eudesmanolides from Artemisia umbelliformis. Phytochemistry 1983, 22, 2767-2772.
6. Appendino, G.; Gariboldi, P.; Calleri, M.; Chiari, G.; Viterbo, D. The structure and
conformation of umbellifolide, a 4,5-secoeudesmane derivative. J. Chem. Soc. Perkin Trans. I
1983, (1), 2705-2709.
7. Lachenmeier, D. W.; Nathan-Maister, D.; Breaux, T. A.; Sohnius, E. M.; Schoeberl, K.;
Kuballa, T. Chemical composition of vintage preban absinthe with special reference to
thujone, fenchone, pinocamphone, methanol, copper, and antimony concentrations. J. Agr.
Food Chem. 2008, 56 (9), 3073-3081.
8. Council Directive (EEC) No 88/388 on the approximation of the laws of the Member States
relating to flavourings for use in foodstuffs and to source materials for their production. Off. J.
Europ. Comm. 1988, L184, 61-66.
336 337 338
9. Rey, C.; Slacanin, I. Domestication du genépi blanc. Revue Suisse Vitic, Arboric, Hortic. 1997, 39, 1-7.
10. Appendino, G.; Taglialatela-Scafati, O.; Romano, A.; Pollastro, F.; Avonto, C.; Rubiolo, P.
Genepolide, a sesterpene γ−lactone with a novel carbon skeleton from mountain wormwood
(Artemisia umbelliformis). J. Nat. Prod. Publication Date on Web: November 20, 2008
11. Bertea, C. M.; Azzolin, C. M. M.; Bossi, S.; Doglia, G.; Maffei, M. E. Identification of an
EcoRI restriction site for a rapid and precise determination of beta-asarone-free Acorus
calamus cytotypes. Phytochemistry 2005, 66 (5), 507-514.
12. Bertea, C. M.; Luciano, P.; Bossi, S.; Leoni, F.; Baiocchi, C.; Medana, C.; Azzolin, C. M.;
Temporale, G.; Lombardozzi, M. A.; Maffei, M. E. PCR and PCR-RFLP of the 5S-rRNA-
NTS region and salvinorin A analyses for the rapid and unequivocal determination of Salvia
divinorum. Phytochemistry 2006, 67 (4), 371-378.
13. Luciano, P.; Bertea, C. M.; Temporale, G.; Maffei, M. E. DNA internal standard for the
quantitative determination of hallucinogenic plants in plant mixtures. Forensic Sci. Int.:
Genetics 2007, 1, 262-266.
European Directorate for the Quality of Medicines (EDQM) European Pharmacopoeia 6th Edition; 2008, 251-252.
15. Rubiolo, P.; Belliardo, F.; Cordero, C.; Liberto, E.; Sgorbini, B.; Bicchi, C. Headspace-solid-
phase microextraction fast GC in combination with principal component analysis as a tool to
classify different chemotypes of chamomile flower-heads (Matricaria recutita L.).
Phytochem. Anal. 2006, 17 (4), 217-225.
16. Sugimoto, N.; Kiuchi, F.; Mikage, M.; Mori, M.; Mizukami, H.; Tsuda, Y. DNA profiling of
Acorus calamus chemotypes differing in essential oil composition. Biol. Pharm. Bull. 1999,
22 (5), 481-485.
17. Cai, Z. H.; Li, P.; Dong, T. T. X.; Tsim, K. W. K. Molecular diversity of 5S-rRNA spacer
362 363 364
domain in Fritillaria species revealed by PCR analysis. Planta Med. 1999, 65 (4), 360-364. 18.
Adams RP. In Identification of Essential oil Components by Gas Chromatography/Mass Spectrometry 4th edn.; Allured: Carol Stream IL, 2007.
Note: P. Rubiolo, M. Matteodo and C. Bicchi are indebted with Regione Piemonte (Italy) for the
financial support to this study carried out within the project “Progetto genepì - Sviluppo di tecniche
innovative a supporto della coltivazione e della trasformazione del genepì in Piemonte”. C .M.
Bertea and M.E. Maffei are indebted with Centre of Excellence CEBIOVEM of the University of
Figure 1 - Structures of the sesquiterpene lactones identified in both chemotypes of A.umbelliformis
Lam. 1: 5-desoxy-5hydroperoxy-5-epitelekin; 2: 5-desoxy-5-hydroperoxytelekin;
umbellifolide; 4: costunolide; 5a: verlotorin; 5b: artemorine; 6a: santamarine; 6b: reynosine; 7:
Figure 2 - PCA scatterplot of the cumulative elaboration of both EOs and headspaces sampled by
HS-SPME of both Artemisia umbelliformis chemotypes. Capital letters: EO GC analysis, lower-
case letters: HS-SPME GC analysis.
Figure 3 - HPLC-DAD-UV profiles at 210 nm of Artemisia umbelliformis chemotypes Au1 and
Figure 4 - PCR products derived from all possible combinations of Au1 and Au2 specific primers
designed on the coding and non-transcribing regions of the plant 5S-rRNA gene. Lane 1, a
single fragment of approximately 224 bp is produced by Au1. Lane 2, a single fragment of
about 327 bp is produced by Au2. Lane 3, a single fragment of 204 bp in length amplified using
the primer AuF in combination with the primer 5S-P2 in the chemotype Au1. Lane 4, a single
fragment of 307 bp in length using the primer AuF in combination with the primer 5S-P2 in the
chemotype Au2. Lane 5, a single fragments of 231 bp with primers 5S-P1 and Au2R1 in the
chemotype Au2. Lane 6, a single fragments of 275 bp with primers 5S-P1 and Au2R2 in the
chemotype Au2. Lane 7, a single fragments of 64 bp with primers Au2F1 and Au2R1 in the
chemotype Au2. Lane 8, a single fragments of 108 bp with primers Au2F1 and Au2R1 in the
chemotype Au2. Lane 9, a single fragments of 213 bp with primers AuF and Au2R1 in the
chemotype Au2. Lane 10, a single fragments of and 255 bp with primers AuF and Au2R2 in the
chemotype Au2. Lane 11, purified PCR products obtained by using 5S-P1 and 5S-P2 primers 17
digested with Rsa I give two major fragments of 123 bp and 82 bp respectively, and a minor
fragment of 19 bp not visible in the gel because out of the resolution capacity of the instrument
in the chemotype Au1. Lane 12, RsaI cleaved Au2 5S-rRNA spacer region gives a major
fragment of 283 bp and two minor fragments of 25 bp (barely visible in the gel) and 19 bp (not
visible in the gel). Lane 13, PCR products from the chemotype Au1 not digested by Taq I. Lane
14, purified PCR products from the chemotype Au2 digested using Taq I producing two
fragments of 197 and 130 bp.
Figure 5 - Aligments of the nucleotide sequences of 5S-rRNA gene spacer region of Artemisia
umbelliformis chemotypes Au1 and Au2. Universal primer sequences are indicated in squared
solid boxes. Artemisia umbelliformis forward primers are indicated in bold. Identical sequences
are indicated by (*). Gaps (-) are introduced for the best alignment. RsaI site is evidenced in the
squared large dot box, whereas the TaqI site is evidenced in the squared small dots box.
Forward and reverse specific primers of the Au2 chemotype are indicated.
Figure 6 - Position of the universal primers (5S-P1 and 5S-P2) flanking the spacer region of 5S-
rRNA gene and specific Artemisia umbelliformis forward (AuF) and A. umbelliformis
chemotype Au2 forward (Au2F1) and reverse (AuR1 and AuR2) specific primers used for PCR
amplification of the 5S-rRNA spacer region.
Table 1 - Components Characterising the Essential Oils of Artemisia umbelliformis Chemotypes Au1 and Au2.
α-Thujene α-Pinene Camphene Sabinene β-Pinene β-Myrcene α-Terpinene p-Cymene 1,8-Cineole γ-Terpinene cis-Sabinene hydrate α-Terpinolene trans-Sabinene hydrate α-Thujone β-Thujone Sabinol Camphor Borneol Terpinen-4-ol α-Terpineol Bornyl acetate α-Terpinyl acetate Sabinyl isobutirrate β-Caryophyllene Sabinyl isovalerianate Sabinyl valerianate Caryophyllene oxide Neryl isovalerianate
930 936 954 975 979 991 1017 1025 1031 1060 1070 1089 1098 1102 1114 1142 1143 1169 1177 1189 1289 1349 1416c 1419 1503c 1516c 1583 1584
929 936 951 975 978 993 1017 1026 1033 1061 1069 1089 1097 1108 1118 1141 1145 1168 1177 1190 1286 1351 1416 1418 1503 1516 1581 1585
1038 1031 953 1132 1117 1174 1189 1280 1213 1257 1268 1292 1488 1420 1446 1708 1516 1712 1619 1714 1378 1494 1494 1594 1577 1605 1965 1679
0.05 0.26 0.06 0.35 1.74 0.18 0.18 0.27 4.86 0.35 1.17 0.10 0.65 36.99 9.11 0.90 0.15 1.14 1.21 0.29 0.05 0.58 1.35 0.40 3.94 2.69 1.89 4.11
tr-0.10 tr-0.40 tr-0.30 0.09-1.40 0.10-2.30 tr-0.30 tr-0.30 0.10-0.40 0.30-9.50 0.10-0.50 0.30-1.50 tr-0.10 0.30-1.00 29.70-51.90 4.60-19.53 0.28-4.10 tr-1.54 tr-3.90 0.40-1.48 tr-0.50 tr-0.10 tr-1.10 1.00-3.83 tr-0.56 3.10-7.13 1.30-7.28 0.57-3.30 1.08-6.90
0.13 0.60 0.43 0.15 5.12 0.68 0.34 0.21 7.58 0.75 7.16 0.21 2.19 0.37 0.27 0.20 0.75 11.58 4.75 1.08 0.51 2.30 0.01 2.24 0.16 0.14 3.61 4.72
tr-0.57 0.33-1.20 tr-1.10 tr-0.36 1.73-11.50 0.10-2.90 0.10-0.55 0.10-0.70 4.05-14.00 0.43-1.00 2.70-20.08 0.10-0.30 1.30-3.35 tr-2.00 tr-0.65 0.10-0.50 tr-1.70 0.09-19.43 3.30-8.15 0.40-2.20 tr-1.97 0.70-4.19 tr-0.09 1.00-4.60 tr-0.50 tr-0.30 2.15-5.40 1.40-10.00
a: LRI from Adams library (18)y; b: experimental LRI c: LRI from standards synthesized in the authors’ laboratory (2) Mean and range values are expressed as percent areas normalized vs. nonane as internal standard.
Table 2 – Mean and Range Amount of α- + β- thujone Expressed as g/100g of Essential Oil in the Investigated Samples of Artemisia umbelliformis Chemotypes Au1 and Au2.
α-thujone (Y= 0.3105x + 3.311, R2 = 0.9995) β-thujone (Y = 0.2896x + 0.1380 ; R2 = 0.9994)
O 3 OH
R 5a OH 5b H
6a ∆ 6b ∆4(15) 3
6 O O
B B BB
p.c. 2 - var. sp. 0.13063
2 A 0
A A AA AAA A
AA A A A AaA a aa aa aa Aaa a a a a a a a a a
B b B
b B BB B b
b b b b b
b -6 -6
0 2 p.c. 1 - var. sp. 0.5273
650000 600000 550000 500000 450000 400000 350000 300000
Telekines 250000 200000
150000 100000 50000 0 2.5
600000 550000 500000 450000 400000 350000 300000 250000
200000 150000 100000 50000 0