Directory UMM :Data Elmu:jurnal:B:Biochemical Systematics and Ecology:Vol28.Issue7.Aug2000:
Biochemical Systematics and Ecology 28 (2000) 679}687
Geographical variation in the surface #avonoids
of Pulicaria dysenterica
Christine A. Williams*, Je!rey B. Harborne, Jenny Greenham
Department of Botany, School of Plant Sciences, The University of Reading, Whiteknights, P.O. Box 221,
Reading, RG6 6AS, UK
Received 11 May 1999; accepted 8 August 1999
Abstract
Four chemical races were detected in Pulicaria dysenterica, when sampled within Europe, on
the basis of the surface #avonoids present. One race uniquely contained quercetagetin
3,7-dimethyl ether and another 6-hydroxykaempferol 3,4@-dimethyl ether. A third race was
based on plants having 6-hydroxykaempferol 3,7-dimethyl ether together with quercetagetin
3,7,3@-trimethyl ether. The fourth race contained the above two compounds, as well as quercetagetin 3,7,3@,4@-tetramethyl ether and 6-hydroxykaempferol 3,7,4@-trimethyl ether. These
lipophilic constituents were variously present on the surfaces of leaf, ray #oret, disc #oret and
fruit. By contrast, the vacuolar #avonoid of all tissues and all races was uniformly quercetin
3-glucuronide. The kaempferol 3-glucoside previously reported in #owers was not detected. Of
the lipophilic #avonoids newly reported from this plant, one 6-hydroxykaempferol 3,7,4@trimethyl ether is new to nature. ( 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Compositae; Inuleae; Pulicaria dysenterica; Surface #avonoids; Methylated 6-hydroxy#avonols;
Geography
1. Introduction
Pulicaria dysenterica (L.) Bernh. Is a well-known medicinal plant of the past, which
was used for treating dysentery. However, the English name, #eabane, refers to its use
* Corresponding author. Tel.: 01189-318168; fax: 01189-753676.
E-mail address: [email protected] (C.A. Williams)
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 1 0 4 - 0
680
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
in mediaeval times when the plant itself or smoke from the burnt plant was used to
drive away #eas and other insects (Grieve, 1980).
Considering its long historic use, chemical information for this plant is limited.
Previous reports include two acetylenic compounds, thymol methyl ether, thymohydroquinone dimethyl ether and *8(9)-dehydrothymohydroquinone dimethyl ether from
the root tissue and ΒΈ-inositol from the leaf (Schulte et al., 1968). In a later study of the
aerial parts, Bohlmann and Zdero (1981) identi"ed nine new caryophyllene derivatives and a hydroxyisocomene in addition to two further thymol constituents. The
thymol derivatives are probably responsible for the insect repellent properties of the
plant since thymol itself has been shown to be insecticidal against house#ies (Dev and
Koul, 1997).
There have been two previous investigations of P. dysenterica #avonoids but these
di!er in some details. Thus, Schulte et al. (1968) working at the University of MuK nster,
Germany, reported a #avonol aglycone, quercetagetin 3,7,4@-trimethyl ether (oxyayanin B) and a #avonol glycoside, kaempferol 3-glucoside from #owers of this plant.
In a later investigation Pares et al. (1981) found that the leaf pro"le of a plant collected
in Istanbul (Turkey) was completely di!erent. They identi"ed: quercetagetin and
6-hydroxykaempferol 3,7-dimethyl ethers, 6-hydroxykaempferol 3,6,7-trimethyl ether
(penduletin), scutellarein, 6-hydroxykaemperol 3-methyl ether 6-glucoside and aesculetin from the leaves.
Since no attempt appears to have been made to explain the early medicinal use of
this plant, we decided to reinvestigate the lipophilic #avonoids to see whether they
had any useful activity. In fact, we have analysed a further seven leaf samples for both
lipophilic and polar #avonoid constituents from a number of di!erent geographical
localities to test for infraspeci"c variation. Furthermore, we have examined two of
these samples for #avonoid variation between the di!erent plant parts, i.e. leaf, disc
#oret, ray #oret and fruit.
2. Materials and methods
2.1. Plant material
Seven samples (1}7) of P. dysenterica were collected from various sources, both fresh
and herbarium, details of which are given in Table 1. Samples 1}3 were veri"ed by
Mr. R. Rutherford, School of Plant Sciences, The University of Reading and voucher
specimens have been deposited in RNG.
2.2. Flavonoid analysis
2.2.1. Lipophilic yavonoids
Lipophilic #avonoids were removed from leaves of samples 1}7 and additionally
from the disc #orets, ray #orets and fruits of samples 1 and 2, by brie#y dipping in
Me CO. The concentrated Me CO extracts were separated by multiple silica gel
2
2
TLC in toluene : HOAc,4 : 1, followed by puri"cation by PPC in 30% HOAc and
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
681
Table 1
Source data for Pullicaria dysenterica samples 1}7
Sample number
Source
1
Plants grown from seed provided by Conservatoire et Jardin Botaniques Gene`ve and
grown in the walled garden in the Harris garden at the School of Plant Sciences, The
University of Reading.
Fresh material collected from Bin"eld Heath nr. Reading, Berks, for which a Herbarium specimen has been lodged in RNG.
Fresh material collected from the top end of the Whiteknights Lake, The University of
Reading for which a herbarium specimen has been lodged in RNG.
Herbarium specimen (RNG) from Calabria, Italy. Optima Iter VIII, 1577; 15-6-97.
Herbarium specimen (RNG) from Caorle, prov. Venezia, Italy.; 11-9-93.
Herbarium specimen (RNG) from Gi!aumont-Champaubert, deH p. Marne, France.
No. 16592; 30-8-93.
Herbarium specimen (RNG) from nr. Santo Stafano, Italy. R.E.Longton 5106; 21-897.
2
3
4
5
6
7
H O. The following constituents: quercetagetin 3,7-dimethyl ether (1), 6-hy2
droxykaempferol 3,4@-dimethyl ether (2) and 3,7-dimethyl ether (3), quercetagetin
3,7,3@-trimethyl ether (4) and 3,7,3@,4@-tetramethyl ether (5) and 6-hydroxykaempferol
3,7,4@-trimethyl ether (6), were standardised as far as possible by UV spectral analysis,
MS and HPLC R and TLC R comparison with marker compounds available from
5
&
previous studies of feverfew and from our marker collection (Table 2).
2.2.2. Vacuolar yavonoids
The water soluble vacuolar #avonoids were extracted with hot 100% MeOH (fresh
material) or 80% MeOH (dried material), from the remaining tissues after removal of
the lipophilic #avonoids by dipping in Me CO. The concd MeOH extracts were
2
separated by PPC in BAW (n-BuOH : HOAc : H O,4 : 1 : 5, top layer) and the result2
ing #avonoid bands eluted and puri"ed by PPC in 15% HOAc, H O and/or BEW
2
(n-BuOH : EtOH : H O, 6 : 1 : 2.2). The identity of the only #avonoid constituent in
2
disc, leaf and ray tissues, quercetin 3-glucuronide, was con"rmed by UV spectral
analysis, by mobility during paper electrophoresis at pH 4.4 using an acetate bu!er, by
PC co-chromatography with an authentic sample in three solvents (BAW, BEW and
H O) and by acid hydrolysis to give quercetin and glucuronic acid.
2
3. Results
3.1. Leaf surface yavonoids
Seven accessions of P. dysenterica, three fresh and four herbarium specimens, from
di!erent localities in Europe (see Table 1), were analysed for their lipophilic leaf
#avonoid constituents. These surface #avonoids were removed by brie#y dipping the
leaves in acetone. Details of their puri"cation and identi"cation are given in the
Materials and Methods. The main constituents found in all seven accessions
682
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
Table 2
R and HPLC R data for the lipophilic #avonoids from Pulicaria dysenterica compared with the
&
5
corresponding isomers from feverfew
Flavonoid
TLCR ]100 in
&
HPLC RH
t
in minutes
Toluene : HOAc, 4 : 1
(silica gel)
30% HOAc (cellulose)
Quercetagetin
3,7-DiME
3,6-DiME
03
08
18
36
5.01
5.75
6-Hydroxykaempferol
3,4@-DiME
3,7-DiME
3,6-DiME
10
10
19
37
37
52
5.55
6.26
7.42
Quercetagetin
3,7,3@-TriME
3,6,3@-TriME
3,7,3@,4@-TetraME
16
30
28
30
49
22
6.48
7.84
9.6
6-Hydroxykaempferol
3,7,4@-TriME
3,6,4@-TriME
3,6,7-TriME
42
45
29
43
59
66
11.44
12.47
10.16
HC * phenyl reverse phase column at 253C using a linear gradient of 40% A:60% B over 20 min, #ow
18
rate 1 ml min ~1. A"2% HOAc, B" MeOH}HOAc}H O, 18 : 1 : 1. UV detection at 260 and 350 nm.
2
Abbreviations:HOAc"acetic acid;ME"methyl ether.
were:- 6-hydroxykaempferol 3,7-dimethyl ether (3) and quercetagetin 3,7,3@-trimethyl
ether (4). In a previous analysis of a Turkish leaf specimen Pares et al. (1981) also
identi"ed 6-hydroxykaemperol 3,7-dimethyl ether but the only quercetagetin derivative they recorded was the 3,7-dimethyl ether. However, Schulte et al. (1968) have
reported the 3,7,4@-isomer from #ower tissue of their German specimen. Mass spectral
data for 4:- [M] 360 and ions at 359 (M-1), 345 (M-15), 317 (M-43 characteristic of
#avonols), 274 (M-86, DiOH, monoOMe A ring) and 151 (monoOH, monoOMe
B ring) did not distinguish it from quercetagetin 3,7,4@-trimethyl ether. Therefore, to
con"rm the identity of 4 as the isomeric 3,7,3@-trimethyl ether we carried out a UV
spectral alkaline shift comparison with other 3@ and /or 4@-substituted #avonols.
A reduction in the alkaline shift and a depression in the absorbance of band II (long
wave) with alkali is a well known diagnostic feature of 4@-substituted #avonols. The
UV alkaline shift of 4 was compared with two other 3@-substituted #avonols: quercetagetin 3,6,3@-trimethyl ether and quercetin 3,3@-dimethyl ether and two 4@-substituted #avonols: quercetin 3,7,4@-trimethyl ether and 3,4@-dimethyl ether (Table 3).
Although, a marker of quercetagetin 3,7,4@-trimethyl ether was not available for direct
comparison it is clear from Table 3 that 4 does give a signi"cant shift in the wave
length of band II (#42 nm) and an increase in absorbance (#17%) with alkali
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
683
Table 3
UV spectral comparison of the alkaline shifts in #avonols with 3@ and/4@ substitution
Flavonoid
j
MeOH nm
.!9
Band II#alkali
Shift
Percent increase or
decrease in absorbance!
Quercetagetin 3,7,3@-trimethyl
ether
Quercetagetin 3,6,3@-trimethyl
ether
Quercetin 3,3@-dimethyl ether
Quercetin 3,7,4@-trimethyl
ether
Quercetin 3,4@-dimethyl ether
257, 281, 349
391
42
#17
256, 272, 352
411
59
#28
254, 268, 356
255, 267, 354
411
380
55
26
#46
!9
256, 266, 355
369
14
!36
!Approximate values based on adding 3 drops of 2N aqueous NaOH to compound in MeOH.
compared with the smaller shifts and decrease in absorbance shown by the two
4@-substituted #avonols.
In sample 7 from Santo Stephano in Italy and sample 6 from France, no
other lipophilic #avonoids were detected. However, the presence of additional minor
components in the other accessions suggested the existence of three further
chemotypes (Table 4). Thus, sample 1, which was grown from seed obtained from
Geneva, di!ered from all the other samples in producing quercetagetin 3,7-dimethyl
ether. Similarly, sample 2 from Bin"eld Heath, nr. Reading was distinguished
by the presence of a second 6-hydroxykaempferol dimethyl ether (2), whch had an
HPLC R of 5.55 compared with 6.26 for the 3,7-dimethyl ether. Compound 2
5
was identi"ed as the 3,4@-rather than the 7,4@-isomer on the basis of UV spectral
properties including the diagnostic low wavelength of band II, i.e. j
MeOH 280,
.!9
340 nm.
Two further 6-hydroxy#avonols : quercetagetin 3,7,3@,4@-tetramethyl ether (5) and
6-hydroxykaempferol 3,7,4@-trimethyl ether (6), a new compound, were identi"ed in
the Bin"eld Heath sample (No. 2) and also in the remaining accessions, i.e. No. 3 from
Whiteknights Lake, The University of Reading and samples 4 and 5 from two
di!erent localities in Italy (Table 1). The latter three samples (3}5) form the fourth
chemotype. 6-Hydroxykaempferol 3,7,4@-trimethyl ether (6) was characterised by
HPLC R and TLC comparison with the 3,6,4@- and 3,6,7-isomers and from the lack of
5
sodium acetate and boric acid shifts in the UV spectral analysis (UVj
MeOH : 280,
.!9
331;#NaOAc 282, 330 and #H BO 282,330. The band I at 280 also indicated that
3
3
the 6-hydroxyl was free. It is clearly di!erent from the previously reported (Schulte
et al., 1968) penduletin, 6-hydroxykaempferol 3,6,7-trimethyl ether which has
j
MeOH 271,343 and HPLC R 10.16 (see Table 2) compared with an R of 11.44
.!9
5
5
for the present compound and from the 3,6,4@-isomer, santin (j
MeOH 273,337 and
.!9
HPLC R 12.47). Compound 6 was provisionally identi"ed in earlier work on
5
feverfew, Tanacetum parthenium (L.) Schultz Bip. (Williams et al., 1995), but more
detailed studies including NMR showed that it was in fact the isomeric 3,6,4@-
684
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
Table 4
Leaf surface #avonoid variation amongst di!erent accessions of Pulicaria dysenterica
Sample number!
Quercet 3,7-diME
HPLC R
5
1
2
3
4
5
6
7
5.01
#
!
!
!
!
!
!
Leaf
surface
#avonoids
6-OHKm
3,4@-diME
5.55
!
#
!
!
!
!
!
6-OH Km
3,7,-diME
6.26
#
#
#
#
#
(#)
#
Quercet 3,7,3@triME 3
6.48
##
##
##
##
#
(#)
#
!For key to sample numbers Table 1. Abbreviations:Quercet"quercetagetin; 6-OHKm"6-hydroxykaempferol; ME"methyl ether.
trimethyl ether (santin) (Williams et al., 1999). This is therefore the "rst report of the
3,7,4@-trimethyl ether in nature.
3.2. Lipophilic yavonoid variation between tissues
The lipophilic #avonoids of the leaf, disc #orets, ray #orets and fruits were separately analysed in samples 1 and 2 (Table 1) and the results are summarized in Table 5. In
sample 1 each tissue can be distinguished either qualitatively or quantitatively by its
#avonoid pro"le. Thus, the leaf alone produced quercetagetin 3,7-dimethyl ether,
while the disc #orets were the richest source of 3 and 4. The ray tissue of this sample
gave only a trace of one compound, 4. This is in contrast to sample 2 where 4 is the
major constituent of ray and disc. Here disc, ray and fruit have the same qualitative
pro"le and only the leaf is distinguished by the presence of the two 6-hydroxykaempferol dimethyl ethers, 2 and 3.
3.3. Vacuolar yavonoids
The only vacuolar #avonoid constituent in all the leaf samples of #eabane and in
the fruits, ray and disc #orets of samples 1 and 2 was identi"ed as quercetin 3glucuronide. A previous report of kaempferol 3-glucoside from a German sample
(Schulte et al.) could not be substantiated. This represents the "rst record of quercetin
3-glucuronide in P. dysenterica.
4. Discussion
The present #avonoid results for P. dysenterica are remarkably at variance with the
data reported by previous workers (Schulte et al., 1968; Pares et al., 1981). With the
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
685
Table 5
The distribution of surface #avonoids in the di!erent tissues of two samples of Pulicaria dysenterica
Sample number!
Surface
Tissue type
Quercet 3,7- 6-OHKm
diME
diME
1
Leaf
Disc
Ray
Fruit
2
Leaf
Disc
Ray
Fruit
Flavonoids
(1)
(2)
6-OHKm 3,7- Quercet
Quercet
diME
3,7,3@-triME 3,7,3@,4@tetraME
(3)
(4)
(5)
#
!
!
!
!
!
!
!
#
##
!
#
#
###
(#)
#
!
!
!
!
!
!
!
!
!
!
!
#
!
!
!
#
!
!
!
##
###
###
##
#
(#)
#
#
(#)
(#)
(#)
(#)
6-OHKm
3,7,4@-triME
(6)
!See Table 1.
exception of kaempferol and quercetagetin 3,7-dimethyl ethers we found none of their
reported compounds (6-hydroxykaempferol 3,6,7-trimethyl ether, scutellarein, 6-hydroxykaempferol 3-methyl ether 6-glucoside, quercetagetin 3,7,4@-trimethyl ether and
kaempferol 3-glucoside) in our samples. Both previous research teams failed to look at
more than one plant sample or plant part and failed to separate the external from the
internal #avonoids. How they failed to "nd the major leaf and #ower constituent,
quercetagetin 3,7,3@-trimethyl ether, is surprising unless of course their plant material
was yet another chemical race. Similarly, how could they mistake kaempferol 3glucoside for the only vacuolar constituent, quercetin 3-glucuronide? This emphasizes
the need to look at as many veri"ed specimens from di!erent localities as possible and
to analyse separately the external and internal #avonoid constituents of the di!erent
plant parts.
It was not surprising to "nd quantitative di!erences in the lipophilic #avonoid
pro"les of the di!erent leaf samples as this has been observed in another temperate
medicinal composite, feverfew, Tanacetum parthenium (Williams et al., unpublished
results), where the amounts of the sesquiterpene lactone,parthenolide, also vary from
one accession to another. However, the latter species has been much cultivated and is
far more catholic in its choice of habitat than P. dysenterica, which grows only in
damp places. On the other hand, the qualitative di!erences in the lipophilic #avonoids
between samples and between plant parts were unexpected. There could be some
ecological explanation, e.g. a protective role against the damaging e!ects of UV light.
However, an increase in UV B irradiation has been shown to equally enhance both
external and epidermal cell vacuolar #avonoids (Cuadra et al., 1997).
Three other Pulicaria species have been analysed previously for their chemical
constituents. They all have very di!erent #avonoid pro"les from P. dysenterica. Thus,
686
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
from P. arabica the highly methylated #avonols: quercetagetin 3,6,7,3@,4@-penta,
3,5,6,7,3@-penta and 3,5,6,7,3@,4@-hexamethyl ethers were isolated from the aerial parts
(Melek et al., 1988). In P. incisa, kaempferol and quercetin 3-galactosides, kaempferol
and quercetin 3-methyl ethers, quercetin 3,7-dimethyl ether and dihydroquercetin
7-methyl ether were characterised (Mansour et al., 1990). In the third species,
P. undulata, kaempferol 3-methyl ether, quercetin 3,7-dimethyl ether and "ve dihydro#avonols: dihydrokaempferol and its 7-methyl ether, dihydroquercetin and its
7-mono and 7,3@-dimethyl ethers, were identi"ed from the aerial parts (Abdel-Mogib
et al., 1989). Pulicaria undulata is considered to be closely related to or synonymous
with P. incisa and this is re#ected in their #avonoid chemistry. In a previous study of
P. undulata (Metwally et al., 1986), a thymol derivative, 2-hydroxyacetylthymol, and
two #avanoids: dihydrokaempferol 7-methyl ether and the #avanone, eriodictyol
7-methyl ether were reported, again from the aerial parts. This is the only report of
a #avanone from a Pulicaria species. Thymol derivatives on the other hand appear to
be characteristic constituents of the genus. We would have analysed the only other
British species, P. vulgaris, to compare with P. dysenterica but it is now a very rare
plant and was not available to us. However, this species and all the other Pulicaria
species, which grow in Europe are morphologically distinct from and can not therefore be mistaken for P. dysenterica.
Having isolated the major lipophilic #avonoid constituents of P. dysenterica in
reasonable quantity, we now intend to test them for their anti-in#ammatory activity.
We have already tested some of the lipophilic #avonoids of feverfew and found that
6-hydroxykaempferol 3,6,4@-trimethyl ether (santin) inhibited both pathways of
arachidonate metabolism (cyclo-oxygenase and 5-lipoxygenase) with the same high
potency and that 6-hydroxykaempferol 3,6-dimethyl ether had an equivalent pro"le of
enzyme activity but was much less potent. On the other hand, quercetagetin 3,6,3@trimethyl ether showed preferential activity against cyclo-oxygenase (Williams et al.,
1999). Thus, it would be interesting to test the corresponding 7 methylated isomers
from #eabane, initially:6-hydroxykaempferol 3,7-dimethyl ether and quercetagetin
3,7,3@-trimethyl ether, the major #avonoid constituents.
References
Abdel-Mogib, M., Dawidar, A.-M., Metwally, M.A., Abou-Elzahab, M., 1989. Flavonols of Pulicaria
undulata. Pharmazie 44, 801.
Bohlmann, F., Zdero, C., 1981. Caryophyllene derivatives and a hydroxyisocomene from Pulicaria dysenterica. Phytochemistry 20, 2529}2534.
Cuadra, P., Harborne, J.B., Waterman, P., 1997. Increases in the surface #avonols and photosynthetic
pigments in Gnaphalium luteo-album in response to UV-B radiation. Phytochemistry 45, 1377}1383.
Dev, S., Koul, O., 1997. Insecticides of Natural Origin. Harwood Academic Publishers, Amsterdam,
pp. 118.
Grieve, M., 1980. A Modern Herbal. Penguin Books Ltd., Harmondsworth, England, pp. 32.
Mansour, R.M.A., Ahmed, A.A., Melek, F.R., Saleh, N.A.M., 1990. The #avonoids of Pulicaria incisa.
Fitoterapia 61, 186}187.
Melek, F.R., El-Ansari, M., Hassan, A., Regaila, A., Ahmed, A.A., Mabry, T.J., 1988. Methylated #avonoid
aglycones from Pulicaria arabica. Rev. Latinoam. Quim. 19, 119}120.
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
687
Metwally, R.M.A., Dawidar, A.A., Metwally, S., 1986. A new thymol derivative from Pulicaria undulata.
Chem. Pharm. Bull. 34, 378}379.
Pares, J.O., Oksuz, S., Ulubelen, A., Mabry, T.J., 1981. 6-Hydroxy#avonoids from Pulicaria dysenterica..
Phytochemistry 20, 2057.
Schulte, K.E., RuK cker, G., MuK ller, F., 1968. Einige Inhaltssto!e der BluK tenkoK pfchen von Pulicaria dysenterica. Arch. Pharm. 301, 115}119.
Williams, C.A., Harborne, J.B., Geiger, H., Hoult, J.R.S., 1999. The #avonoids of Tanacetum parthenium and
T. vulgare and their anti-in#ammatory properties. Phytochemistry 51, 417}423.
Williams, C.A., Hoult, J.R.S., Harborne, J.B., Greenham, J., Eagles, J., 1995. A biologically active lipophilic
#avonol from Tanacetum parthenium. Phytochemistry 38, 267}270.
Geographical variation in the surface #avonoids
of Pulicaria dysenterica
Christine A. Williams*, Je!rey B. Harborne, Jenny Greenham
Department of Botany, School of Plant Sciences, The University of Reading, Whiteknights, P.O. Box 221,
Reading, RG6 6AS, UK
Received 11 May 1999; accepted 8 August 1999
Abstract
Four chemical races were detected in Pulicaria dysenterica, when sampled within Europe, on
the basis of the surface #avonoids present. One race uniquely contained quercetagetin
3,7-dimethyl ether and another 6-hydroxykaempferol 3,4@-dimethyl ether. A third race was
based on plants having 6-hydroxykaempferol 3,7-dimethyl ether together with quercetagetin
3,7,3@-trimethyl ether. The fourth race contained the above two compounds, as well as quercetagetin 3,7,3@,4@-tetramethyl ether and 6-hydroxykaempferol 3,7,4@-trimethyl ether. These
lipophilic constituents were variously present on the surfaces of leaf, ray #oret, disc #oret and
fruit. By contrast, the vacuolar #avonoid of all tissues and all races was uniformly quercetin
3-glucuronide. The kaempferol 3-glucoside previously reported in #owers was not detected. Of
the lipophilic #avonoids newly reported from this plant, one 6-hydroxykaempferol 3,7,4@trimethyl ether is new to nature. ( 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Compositae; Inuleae; Pulicaria dysenterica; Surface #avonoids; Methylated 6-hydroxy#avonols;
Geography
1. Introduction
Pulicaria dysenterica (L.) Bernh. Is a well-known medicinal plant of the past, which
was used for treating dysentery. However, the English name, #eabane, refers to its use
* Corresponding author. Tel.: 01189-318168; fax: 01189-753676.
E-mail address: [email protected] (C.A. Williams)
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 1 0 4 - 0
680
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
in mediaeval times when the plant itself or smoke from the burnt plant was used to
drive away #eas and other insects (Grieve, 1980).
Considering its long historic use, chemical information for this plant is limited.
Previous reports include two acetylenic compounds, thymol methyl ether, thymohydroquinone dimethyl ether and *8(9)-dehydrothymohydroquinone dimethyl ether from
the root tissue and ΒΈ-inositol from the leaf (Schulte et al., 1968). In a later study of the
aerial parts, Bohlmann and Zdero (1981) identi"ed nine new caryophyllene derivatives and a hydroxyisocomene in addition to two further thymol constituents. The
thymol derivatives are probably responsible for the insect repellent properties of the
plant since thymol itself has been shown to be insecticidal against house#ies (Dev and
Koul, 1997).
There have been two previous investigations of P. dysenterica #avonoids but these
di!er in some details. Thus, Schulte et al. (1968) working at the University of MuK nster,
Germany, reported a #avonol aglycone, quercetagetin 3,7,4@-trimethyl ether (oxyayanin B) and a #avonol glycoside, kaempferol 3-glucoside from #owers of this plant.
In a later investigation Pares et al. (1981) found that the leaf pro"le of a plant collected
in Istanbul (Turkey) was completely di!erent. They identi"ed: quercetagetin and
6-hydroxykaempferol 3,7-dimethyl ethers, 6-hydroxykaempferol 3,6,7-trimethyl ether
(penduletin), scutellarein, 6-hydroxykaemperol 3-methyl ether 6-glucoside and aesculetin from the leaves.
Since no attempt appears to have been made to explain the early medicinal use of
this plant, we decided to reinvestigate the lipophilic #avonoids to see whether they
had any useful activity. In fact, we have analysed a further seven leaf samples for both
lipophilic and polar #avonoid constituents from a number of di!erent geographical
localities to test for infraspeci"c variation. Furthermore, we have examined two of
these samples for #avonoid variation between the di!erent plant parts, i.e. leaf, disc
#oret, ray #oret and fruit.
2. Materials and methods
2.1. Plant material
Seven samples (1}7) of P. dysenterica were collected from various sources, both fresh
and herbarium, details of which are given in Table 1. Samples 1}3 were veri"ed by
Mr. R. Rutherford, School of Plant Sciences, The University of Reading and voucher
specimens have been deposited in RNG.
2.2. Flavonoid analysis
2.2.1. Lipophilic yavonoids
Lipophilic #avonoids were removed from leaves of samples 1}7 and additionally
from the disc #orets, ray #orets and fruits of samples 1 and 2, by brie#y dipping in
Me CO. The concentrated Me CO extracts were separated by multiple silica gel
2
2
TLC in toluene : HOAc,4 : 1, followed by puri"cation by PPC in 30% HOAc and
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
681
Table 1
Source data for Pullicaria dysenterica samples 1}7
Sample number
Source
1
Plants grown from seed provided by Conservatoire et Jardin Botaniques Gene`ve and
grown in the walled garden in the Harris garden at the School of Plant Sciences, The
University of Reading.
Fresh material collected from Bin"eld Heath nr. Reading, Berks, for which a Herbarium specimen has been lodged in RNG.
Fresh material collected from the top end of the Whiteknights Lake, The University of
Reading for which a herbarium specimen has been lodged in RNG.
Herbarium specimen (RNG) from Calabria, Italy. Optima Iter VIII, 1577; 15-6-97.
Herbarium specimen (RNG) from Caorle, prov. Venezia, Italy.; 11-9-93.
Herbarium specimen (RNG) from Gi!aumont-Champaubert, deH p. Marne, France.
No. 16592; 30-8-93.
Herbarium specimen (RNG) from nr. Santo Stafano, Italy. R.E.Longton 5106; 21-897.
2
3
4
5
6
7
H O. The following constituents: quercetagetin 3,7-dimethyl ether (1), 6-hy2
droxykaempferol 3,4@-dimethyl ether (2) and 3,7-dimethyl ether (3), quercetagetin
3,7,3@-trimethyl ether (4) and 3,7,3@,4@-tetramethyl ether (5) and 6-hydroxykaempferol
3,7,4@-trimethyl ether (6), were standardised as far as possible by UV spectral analysis,
MS and HPLC R and TLC R comparison with marker compounds available from
5
&
previous studies of feverfew and from our marker collection (Table 2).
2.2.2. Vacuolar yavonoids
The water soluble vacuolar #avonoids were extracted with hot 100% MeOH (fresh
material) or 80% MeOH (dried material), from the remaining tissues after removal of
the lipophilic #avonoids by dipping in Me CO. The concd MeOH extracts were
2
separated by PPC in BAW (n-BuOH : HOAc : H O,4 : 1 : 5, top layer) and the result2
ing #avonoid bands eluted and puri"ed by PPC in 15% HOAc, H O and/or BEW
2
(n-BuOH : EtOH : H O, 6 : 1 : 2.2). The identity of the only #avonoid constituent in
2
disc, leaf and ray tissues, quercetin 3-glucuronide, was con"rmed by UV spectral
analysis, by mobility during paper electrophoresis at pH 4.4 using an acetate bu!er, by
PC co-chromatography with an authentic sample in three solvents (BAW, BEW and
H O) and by acid hydrolysis to give quercetin and glucuronic acid.
2
3. Results
3.1. Leaf surface yavonoids
Seven accessions of P. dysenterica, three fresh and four herbarium specimens, from
di!erent localities in Europe (see Table 1), were analysed for their lipophilic leaf
#avonoid constituents. These surface #avonoids were removed by brie#y dipping the
leaves in acetone. Details of their puri"cation and identi"cation are given in the
Materials and Methods. The main constituents found in all seven accessions
682
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
Table 2
R and HPLC R data for the lipophilic #avonoids from Pulicaria dysenterica compared with the
&
5
corresponding isomers from feverfew
Flavonoid
TLCR ]100 in
&
HPLC RH
t
in minutes
Toluene : HOAc, 4 : 1
(silica gel)
30% HOAc (cellulose)
Quercetagetin
3,7-DiME
3,6-DiME
03
08
18
36
5.01
5.75
6-Hydroxykaempferol
3,4@-DiME
3,7-DiME
3,6-DiME
10
10
19
37
37
52
5.55
6.26
7.42
Quercetagetin
3,7,3@-TriME
3,6,3@-TriME
3,7,3@,4@-TetraME
16
30
28
30
49
22
6.48
7.84
9.6
6-Hydroxykaempferol
3,7,4@-TriME
3,6,4@-TriME
3,6,7-TriME
42
45
29
43
59
66
11.44
12.47
10.16
HC * phenyl reverse phase column at 253C using a linear gradient of 40% A:60% B over 20 min, #ow
18
rate 1 ml min ~1. A"2% HOAc, B" MeOH}HOAc}H O, 18 : 1 : 1. UV detection at 260 and 350 nm.
2
Abbreviations:HOAc"acetic acid;ME"methyl ether.
were:- 6-hydroxykaempferol 3,7-dimethyl ether (3) and quercetagetin 3,7,3@-trimethyl
ether (4). In a previous analysis of a Turkish leaf specimen Pares et al. (1981) also
identi"ed 6-hydroxykaemperol 3,7-dimethyl ether but the only quercetagetin derivative they recorded was the 3,7-dimethyl ether. However, Schulte et al. (1968) have
reported the 3,7,4@-isomer from #ower tissue of their German specimen. Mass spectral
data for 4:- [M] 360 and ions at 359 (M-1), 345 (M-15), 317 (M-43 characteristic of
#avonols), 274 (M-86, DiOH, monoOMe A ring) and 151 (monoOH, monoOMe
B ring) did not distinguish it from quercetagetin 3,7,4@-trimethyl ether. Therefore, to
con"rm the identity of 4 as the isomeric 3,7,3@-trimethyl ether we carried out a UV
spectral alkaline shift comparison with other 3@ and /or 4@-substituted #avonols.
A reduction in the alkaline shift and a depression in the absorbance of band II (long
wave) with alkali is a well known diagnostic feature of 4@-substituted #avonols. The
UV alkaline shift of 4 was compared with two other 3@-substituted #avonols: quercetagetin 3,6,3@-trimethyl ether and quercetin 3,3@-dimethyl ether and two 4@-substituted #avonols: quercetin 3,7,4@-trimethyl ether and 3,4@-dimethyl ether (Table 3).
Although, a marker of quercetagetin 3,7,4@-trimethyl ether was not available for direct
comparison it is clear from Table 3 that 4 does give a signi"cant shift in the wave
length of band II (#42 nm) and an increase in absorbance (#17%) with alkali
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
683
Table 3
UV spectral comparison of the alkaline shifts in #avonols with 3@ and/4@ substitution
Flavonoid
j
MeOH nm
.!9
Band II#alkali
Shift
Percent increase or
decrease in absorbance!
Quercetagetin 3,7,3@-trimethyl
ether
Quercetagetin 3,6,3@-trimethyl
ether
Quercetin 3,3@-dimethyl ether
Quercetin 3,7,4@-trimethyl
ether
Quercetin 3,4@-dimethyl ether
257, 281, 349
391
42
#17
256, 272, 352
411
59
#28
254, 268, 356
255, 267, 354
411
380
55
26
#46
!9
256, 266, 355
369
14
!36
!Approximate values based on adding 3 drops of 2N aqueous NaOH to compound in MeOH.
compared with the smaller shifts and decrease in absorbance shown by the two
4@-substituted #avonols.
In sample 7 from Santo Stephano in Italy and sample 6 from France, no
other lipophilic #avonoids were detected. However, the presence of additional minor
components in the other accessions suggested the existence of three further
chemotypes (Table 4). Thus, sample 1, which was grown from seed obtained from
Geneva, di!ered from all the other samples in producing quercetagetin 3,7-dimethyl
ether. Similarly, sample 2 from Bin"eld Heath, nr. Reading was distinguished
by the presence of a second 6-hydroxykaempferol dimethyl ether (2), whch had an
HPLC R of 5.55 compared with 6.26 for the 3,7-dimethyl ether. Compound 2
5
was identi"ed as the 3,4@-rather than the 7,4@-isomer on the basis of UV spectral
properties including the diagnostic low wavelength of band II, i.e. j
MeOH 280,
.!9
340 nm.
Two further 6-hydroxy#avonols : quercetagetin 3,7,3@,4@-tetramethyl ether (5) and
6-hydroxykaempferol 3,7,4@-trimethyl ether (6), a new compound, were identi"ed in
the Bin"eld Heath sample (No. 2) and also in the remaining accessions, i.e. No. 3 from
Whiteknights Lake, The University of Reading and samples 4 and 5 from two
di!erent localities in Italy (Table 1). The latter three samples (3}5) form the fourth
chemotype. 6-Hydroxykaempferol 3,7,4@-trimethyl ether (6) was characterised by
HPLC R and TLC comparison with the 3,6,4@- and 3,6,7-isomers and from the lack of
5
sodium acetate and boric acid shifts in the UV spectral analysis (UVj
MeOH : 280,
.!9
331;#NaOAc 282, 330 and #H BO 282,330. The band I at 280 also indicated that
3
3
the 6-hydroxyl was free. It is clearly di!erent from the previously reported (Schulte
et al., 1968) penduletin, 6-hydroxykaempferol 3,6,7-trimethyl ether which has
j
MeOH 271,343 and HPLC R 10.16 (see Table 2) compared with an R of 11.44
.!9
5
5
for the present compound and from the 3,6,4@-isomer, santin (j
MeOH 273,337 and
.!9
HPLC R 12.47). Compound 6 was provisionally identi"ed in earlier work on
5
feverfew, Tanacetum parthenium (L.) Schultz Bip. (Williams et al., 1995), but more
detailed studies including NMR showed that it was in fact the isomeric 3,6,4@-
684
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
Table 4
Leaf surface #avonoid variation amongst di!erent accessions of Pulicaria dysenterica
Sample number!
Quercet 3,7-diME
HPLC R
5
1
2
3
4
5
6
7
5.01
#
!
!
!
!
!
!
Leaf
surface
#avonoids
6-OHKm
3,4@-diME
5.55
!
#
!
!
!
!
!
6-OH Km
3,7,-diME
6.26
#
#
#
#
#
(#)
#
Quercet 3,7,3@triME 3
6.48
##
##
##
##
#
(#)
#
!For key to sample numbers Table 1. Abbreviations:Quercet"quercetagetin; 6-OHKm"6-hydroxykaempferol; ME"methyl ether.
trimethyl ether (santin) (Williams et al., 1999). This is therefore the "rst report of the
3,7,4@-trimethyl ether in nature.
3.2. Lipophilic yavonoid variation between tissues
The lipophilic #avonoids of the leaf, disc #orets, ray #orets and fruits were separately analysed in samples 1 and 2 (Table 1) and the results are summarized in Table 5. In
sample 1 each tissue can be distinguished either qualitatively or quantitatively by its
#avonoid pro"le. Thus, the leaf alone produced quercetagetin 3,7-dimethyl ether,
while the disc #orets were the richest source of 3 and 4. The ray tissue of this sample
gave only a trace of one compound, 4. This is in contrast to sample 2 where 4 is the
major constituent of ray and disc. Here disc, ray and fruit have the same qualitative
pro"le and only the leaf is distinguished by the presence of the two 6-hydroxykaempferol dimethyl ethers, 2 and 3.
3.3. Vacuolar yavonoids
The only vacuolar #avonoid constituent in all the leaf samples of #eabane and in
the fruits, ray and disc #orets of samples 1 and 2 was identi"ed as quercetin 3glucuronide. A previous report of kaempferol 3-glucoside from a German sample
(Schulte et al.) could not be substantiated. This represents the "rst record of quercetin
3-glucuronide in P. dysenterica.
4. Discussion
The present #avonoid results for P. dysenterica are remarkably at variance with the
data reported by previous workers (Schulte et al., 1968; Pares et al., 1981). With the
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
685
Table 5
The distribution of surface #avonoids in the di!erent tissues of two samples of Pulicaria dysenterica
Sample number!
Surface
Tissue type
Quercet 3,7- 6-OHKm
diME
diME
1
Leaf
Disc
Ray
Fruit
2
Leaf
Disc
Ray
Fruit
Flavonoids
(1)
(2)
6-OHKm 3,7- Quercet
Quercet
diME
3,7,3@-triME 3,7,3@,4@tetraME
(3)
(4)
(5)
#
!
!
!
!
!
!
!
#
##
!
#
#
###
(#)
#
!
!
!
!
!
!
!
!
!
!
!
#
!
!
!
#
!
!
!
##
###
###
##
#
(#)
#
#
(#)
(#)
(#)
(#)
6-OHKm
3,7,4@-triME
(6)
!See Table 1.
exception of kaempferol and quercetagetin 3,7-dimethyl ethers we found none of their
reported compounds (6-hydroxykaempferol 3,6,7-trimethyl ether, scutellarein, 6-hydroxykaempferol 3-methyl ether 6-glucoside, quercetagetin 3,7,4@-trimethyl ether and
kaempferol 3-glucoside) in our samples. Both previous research teams failed to look at
more than one plant sample or plant part and failed to separate the external from the
internal #avonoids. How they failed to "nd the major leaf and #ower constituent,
quercetagetin 3,7,3@-trimethyl ether, is surprising unless of course their plant material
was yet another chemical race. Similarly, how could they mistake kaempferol 3glucoside for the only vacuolar constituent, quercetin 3-glucuronide? This emphasizes
the need to look at as many veri"ed specimens from di!erent localities as possible and
to analyse separately the external and internal #avonoid constituents of the di!erent
plant parts.
It was not surprising to "nd quantitative di!erences in the lipophilic #avonoid
pro"les of the di!erent leaf samples as this has been observed in another temperate
medicinal composite, feverfew, Tanacetum parthenium (Williams et al., unpublished
results), where the amounts of the sesquiterpene lactone,parthenolide, also vary from
one accession to another. However, the latter species has been much cultivated and is
far more catholic in its choice of habitat than P. dysenterica, which grows only in
damp places. On the other hand, the qualitative di!erences in the lipophilic #avonoids
between samples and between plant parts were unexpected. There could be some
ecological explanation, e.g. a protective role against the damaging e!ects of UV light.
However, an increase in UV B irradiation has been shown to equally enhance both
external and epidermal cell vacuolar #avonoids (Cuadra et al., 1997).
Three other Pulicaria species have been analysed previously for their chemical
constituents. They all have very di!erent #avonoid pro"les from P. dysenterica. Thus,
686
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
from P. arabica the highly methylated #avonols: quercetagetin 3,6,7,3@,4@-penta,
3,5,6,7,3@-penta and 3,5,6,7,3@,4@-hexamethyl ethers were isolated from the aerial parts
(Melek et al., 1988). In P. incisa, kaempferol and quercetin 3-galactosides, kaempferol
and quercetin 3-methyl ethers, quercetin 3,7-dimethyl ether and dihydroquercetin
7-methyl ether were characterised (Mansour et al., 1990). In the third species,
P. undulata, kaempferol 3-methyl ether, quercetin 3,7-dimethyl ether and "ve dihydro#avonols: dihydrokaempferol and its 7-methyl ether, dihydroquercetin and its
7-mono and 7,3@-dimethyl ethers, were identi"ed from the aerial parts (Abdel-Mogib
et al., 1989). Pulicaria undulata is considered to be closely related to or synonymous
with P. incisa and this is re#ected in their #avonoid chemistry. In a previous study of
P. undulata (Metwally et al., 1986), a thymol derivative, 2-hydroxyacetylthymol, and
two #avanoids: dihydrokaempferol 7-methyl ether and the #avanone, eriodictyol
7-methyl ether were reported, again from the aerial parts. This is the only report of
a #avanone from a Pulicaria species. Thymol derivatives on the other hand appear to
be characteristic constituents of the genus. We would have analysed the only other
British species, P. vulgaris, to compare with P. dysenterica but it is now a very rare
plant and was not available to us. However, this species and all the other Pulicaria
species, which grow in Europe are morphologically distinct from and can not therefore be mistaken for P. dysenterica.
Having isolated the major lipophilic #avonoid constituents of P. dysenterica in
reasonable quantity, we now intend to test them for their anti-in#ammatory activity.
We have already tested some of the lipophilic #avonoids of feverfew and found that
6-hydroxykaempferol 3,6,4@-trimethyl ether (santin) inhibited both pathways of
arachidonate metabolism (cyclo-oxygenase and 5-lipoxygenase) with the same high
potency and that 6-hydroxykaempferol 3,6-dimethyl ether had an equivalent pro"le of
enzyme activity but was much less potent. On the other hand, quercetagetin 3,6,3@trimethyl ether showed preferential activity against cyclo-oxygenase (Williams et al.,
1999). Thus, it would be interesting to test the corresponding 7 methylated isomers
from #eabane, initially:6-hydroxykaempferol 3,7-dimethyl ether and quercetagetin
3,7,3@-trimethyl ether, the major #avonoid constituents.
References
Abdel-Mogib, M., Dawidar, A.-M., Metwally, M.A., Abou-Elzahab, M., 1989. Flavonols of Pulicaria
undulata. Pharmazie 44, 801.
Bohlmann, F., Zdero, C., 1981. Caryophyllene derivatives and a hydroxyisocomene from Pulicaria dysenterica. Phytochemistry 20, 2529}2534.
Cuadra, P., Harborne, J.B., Waterman, P., 1997. Increases in the surface #avonols and photosynthetic
pigments in Gnaphalium luteo-album in response to UV-B radiation. Phytochemistry 45, 1377}1383.
Dev, S., Koul, O., 1997. Insecticides of Natural Origin. Harwood Academic Publishers, Amsterdam,
pp. 118.
Grieve, M., 1980. A Modern Herbal. Penguin Books Ltd., Harmondsworth, England, pp. 32.
Mansour, R.M.A., Ahmed, A.A., Melek, F.R., Saleh, N.A.M., 1990. The #avonoids of Pulicaria incisa.
Fitoterapia 61, 186}187.
Melek, F.R., El-Ansari, M., Hassan, A., Regaila, A., Ahmed, A.A., Mabry, T.J., 1988. Methylated #avonoid
aglycones from Pulicaria arabica. Rev. Latinoam. Quim. 19, 119}120.
C.A. Williams et al. / Biochemical Systematics and Ecology 28 (2000) 679}687
687
Metwally, R.M.A., Dawidar, A.A., Metwally, S., 1986. A new thymol derivative from Pulicaria undulata.
Chem. Pharm. Bull. 34, 378}379.
Pares, J.O., Oksuz, S., Ulubelen, A., Mabry, T.J., 1981. 6-Hydroxy#avonoids from Pulicaria dysenterica..
Phytochemistry 20, 2057.
Schulte, K.E., RuK cker, G., MuK ller, F., 1968. Einige Inhaltssto!e der BluK tenkoK pfchen von Pulicaria dysenterica. Arch. Pharm. 301, 115}119.
Williams, C.A., Harborne, J.B., Geiger, H., Hoult, J.R.S., 1999. The #avonoids of Tanacetum parthenium and
T. vulgare and their anti-in#ammatory properties. Phytochemistry 51, 417}423.
Williams, C.A., Hoult, J.R.S., Harborne, J.B., Greenham, J., Eagles, J., 1995. A biologically active lipophilic
#avonol from Tanacetum parthenium. Phytochemistry 38, 267}270.