PHYTOCHEMICAL ANALYSIS

86 Phytochem. Anal. 16, 86 –89 (2005) S. LIU ET AL.

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002.pca.816

Isolation and Identification of Arctiin and

Arctigenin in Leaves of Burdock (Arctium lappa
L.) by Polyamide Column Chromatography in
Combination with HPLC-ESI/MS

Shiming Liu,1,3* Kaoshan Chen1, Willibald Schliemann2 and Dieter Strack2

1
Institute of Biochemistry and Molecular Biology, School of Life Science, Shandong University, Jinan 250100, China
2
Institute of Plant Biochemistry, Weinberg 3, 06120 Halle, Germany
3
Low Temperature Physiology Lab, Department of Low Temperature Sciences, National Agricultural Research Center for Hokkaido Region,
1 Hitsujigaoka, Toyohira-ku, Sapporo, Hokkaido 062-8555, Japan

A simple method involving polyamide column chromatography in combination with HPLC-PAD and HPLC-
ESI/MS for isolating and identifying two kinds of lignans, arctiin and arctigenin, in the leaves of burdock
(Arctium lappa L.) has been established. After extraction of burdock leaves with 80% methanol, the aqueous
phase of crude extracts was partitioned between water and chloroform and the aqueous phase was fractionated
on a polyamide glass column. The fraction, eluting with 100% methanol, was concentrated and gave a white
precipitate at 4°C from which two main compounds were purified by semi-preparative HPLC. In comparison
with the UV and ESI-MS spectra and the HPLC retention time of authentic standards, the compounds were
determined to be arctiin and arctigenin. The extraction/separation technique was validated using an internal
standard method. Copyright © 2005 John Wiley & Sons, Ltd.
Keywords: polyamide column chromatography; HPLC-ESI/MS; arctiin; arctigenin; burdock leaves; Arctium lappa L.

INTRODUCTION

Lignans (neolignans) are a large group of natural prod-ucts

characterised by the coupling of two C6C3 units. They
possess many kinds of bioactivities (Macrae and Towers,
1984) and a number of important pharmacologi-cal
properties including demutagenic (Morita et al., 1985;
Shinohara et al., 1988), cytotoxic, anti-proliferative (Ryu et
al., 1995; Moritani et al., 1996), platelet activating factor
(PAF) antagonist (Iwakami et al., 1992), calcium
antagonist (Ichikawa et al., 1986) and anti-carcinogenesis
(Hirose et al., 2000) activities. Furthermore, (−)-arctigenin
(2) inhibits HIV-1 in tissue culture systems and is also a
potent inhibitor of DNA topoisomerase II and of HIV-1
integrase (Eich et al., 1990, 1996; Pfeifer et al., 1992;
Umehara et al., 1993, 1996; Vlietinck et al., 1998).
Seeds, fruits, stalks and roots of plants are good sources
of lignans, but the isolation of lignans from these organs is
typically achieved through several complex pro-cedures of
partition and silica column chromatography or precipitating
with lead salts and then displacement of lead after
extraction.
A variety of lignans, including arctigenin (2; C21H24O6;
molecular weight 372) and its glucoside arctiin (1;
C27H34O11; molecular weight 534), have been isolated
from the seeds of the perennial herb burdock (Arctium
* Correspondence to: S. Liu, Low Temperature Physiology Laboratory, Depart-ment lappa L.; Compositae) (Ichihara et al., 1976, 1977, 1978;
of Low Temperature Sciences, National Agricultural Research Center for Hokkaido Ichikawa et al., 1986; Wang and Yang, 1993; Han et el.,
Region, 1 Hitsujigaoka, Toyohira-ku, Sapporo, Hokkaido 062-8555, Japan. 1994; Umehara et al., 1996; Vlietinck et al., 1998). How-
Email: smliuhn@yahoo.com
ever, there are no reports on the use of leaves of burdock to
Contract/grant sponsor: National High New Technology Development Programs isolate these valuable lignans. During our analytical studies
of China, (‘863’ Programs); Contract/grant number: 2002AA244031. on the chemical constituents of burdock leaves

Copyright © 2005 John Wiley & Sons, Ltd. Phytochem. AnalReceived.16:8623– 89July(200035)
Copyright © 2005 John Wiley & Sons, Ltd. Revised 22 December 2003
Accepted 3 March 2004
 DETERMINATION OF ARCTIIN AND ARCTIGENIN IN BURDOCK
using polyamide column chromatography with methanol
and water as the mobile phases, we observed that the
fraction that eluted with 100% methanol precipitated out
after concentration and cooling. This precipitate con-sisted
of a mixture of arctiin (1) and arctigenin (2) as determined
by HPLC-ESI/MS analysis. The isolation method is simple
and effective, and 1 and 2 could be separated out directly
through just one step of polyamide chromatography after
extraction.
EXPERIMENTAL
Chemicals and plant materials. Standards of arcti-genin
and arctiin were provided by Dr. Jutta Kalbitz (Bioservice
Halle GmbH, Halle, Germany). All solvents used for HPLC
and HPLC-ESI/MS were of HPLC grade. Leaves of
burdock (Arctium lappa L.) were col-lected from the
Production Base for Burdock, Gaomi, China. A voucher
specimen was deposited in the her-barium of the School of
Life Science, Shandong Univer-sity, China.
Extraction of samples. Whole leaves of burdock were
lyophilised after harvest and ground to an homogeneous
powder. The lyophilised leaves were extracted for 30 min
with 80% methanol (1:10, w/v) using a homogeniser cooled
on ice. The crude extracts were filtered and the filtrate
evaporated using a rotary evaporator under re-duced
pressure. The residue obtained was re-dissolved in water
and partitioned against chloroform in order to eliminate the
chlorophylls. The aqueous phase obtained was concentrated
to about one-fiftieth of the extraction volume and was used
for chromatographic separation.
Chromatographic analyses. Column chromatography was
performed using a glass column (70 × 5 cm i.d.) loaded
with polyamide SC 6 (Macherey-Nagel, Düren, Germany).
An aliquot (20 mL) of the sample was loaded onto the top
of the polyamide and eluted sequentially with 500 mL each
of water, 30 and 60% aqueous metha-nol, and methanol at a
flow rate of 10 mL/min.
The HPLC system consisted of a Waters (Milford, MA,
USA) model 2690 separation module and a model 996
photodiode array detector (PAD), and was equipped with a
Nucleosil C18 (Macherey-Nagel, Düren, Germany) column
(250 × 4 mm i.d.; 5 µm particle size). The output from the
analogue detector was acquired by a computer interface,
digitised and then processed using Waters Millennium
software. For analytical HPLC, a linear sol-vent gradient of
acetonitrile: 1.5% aqueous phosphoric acid from 0:100 to
50:50 was applied over 60 min at a flow rate of 1 mL/min,
and UV absorption at 280 nm was measured. The
conditions for the semi-preparative puri-fication of the
lignans by HPLC were the same as those for analytical
HPLC except that the 1.5% aqueous phos-phoric acid was
replaced by 1.0% aqueous acetic acid.
For HPLC-MS, a Micro-Tech (San Jose, CA, USA)
Ultra-Plus Micro LC system equipped with an Ultra-Sep
(Restek: San Jose, CA, USA) C18 column (100 × 1 mm
i.d.; 4 µm) was coupled to a Finnigan (San Jose, CA, USA)
MAT TSQ 7000 instrument. The mobile phase consisted of
0.2% aqueous acetic acid (solvent A) and 0.2% aqueous
acetic acid in methanol (solvent B) and the gradient system
started from 90:10 (A:B) chang-
Copyright © 2005 John Wiley & Sons, Ltd.
87
ing to 50:50 within 10 min, followed by isocratic elution
with 50:50 (A:B) for a further 10 min at a flow rate of 70
µL/min. Positive and negative electrospray MS were
obtained under the following conditions: electrospray
voltage 4.5 kV (positive ions), 3.5 kV (negative ions);
capillary temperature 220°C; sheath gas nitrogen; the
injection volume was 2 µL.
Isolation, identification and quantification of 1 and 2.
After the polyamide column chromatography, the frac-tion
that eluted with 100% methanol was concentrated and
stored at 4°C for 24 h, after which time a white pre-cipitate
separated out. This precipitate was dissolved in methanol,
centrifuged and analysed by HPLC. Two main peaks were
observed in the chromatogram indicating the presence in
the precipitate of at least two major compo-nents that were
subsequently isolated and purified by semi-preparative
HPLC and then analysed by HPLC-ESI/MS. The HPLC
profile of the two main compounds from the precipitate
were compared with those of au-thentic standards of arctiin
(1) and arctigenin (2), and co-injection studies were carried
out by analytical HPLC so as to compare their retention
times and UV absorptions. On the bases of these results, the
two compounds in the precipitate were identified.
Quantification of these two lignans was conducted using
the internal standard method. Since the standards
themselves were not of ultra-high purity, the actual
concentrations of the stand-ards were calculated from their
peak area ratios in the HPLC profiles together with the
concentrations of the solutions injected.
RESULTS AND DISCUSSION
Following the separation of the aqueous phase of crude
extracts of burdock leaves by polyamide column chro-
matography, the fraction that eluted with 100% metha-nol
was concentrated and stored at 4°C for 24 h to yield a white
precipitate, the HPLC profile of which is depicted in Fig.
1(a). The two main peaks were at retention times of 37.11
and 47.17 min, and their UV ab-sorption maxima (λ max)
were at 226.2/277.4 and 228.6/ 280.2 nm, respectively.
However, the HPLC profile of the aqueous phase of the
crude extracts of burdock leaves did not reveal two peaks at
these retention times (data not shown), indicating that the
contents of these two compounds in burdock leaves is very
low. Under the same HPLC conditions, the retention times
and UV absorption maxima of the two compounds in the
pre-cipitate were identical with those of standard arctiin (1)
and arctigenin (2), respectively [Fig. 1(a–c)]. Further
evidence of the identities of the components of the pre-
cipitate was obtained by co-injection with authentic
standards of 1 and 2 [Fig. 1(d)].
The ESI-MS spectra of the two compounds purified from
the precipitate are shown in Fig. 2(a– d). The m/z values of
the [M+Na]+ and [M−H]− peaks of the com-pound with
retention time 37.11 min were 557 and 533, respectively
[Fig. 3(a, b)], indicating a molecular weight of 534, which
is identical to that of 1. The m/z values of the [M+Na]+ and
[M−H]− peaks of the compound with retention time 47.17
min were 395 and 371, respectively [Fig. 3(c, d)],
indicating a molecular weight of 372, which is identical to
that of 2.
Phytochem. Anal. 16: 86 –89 (2005)
88 S. LIU ET AL.

Figure 1. HPLC profiles of: (a) the precipitate obtained from the fraction of a burdock leaf
extract that eluted with 100% methanol from a polyamide column chromatography;
(b) authentic standard of arctiin (1); (c) authentic standard of arctigenin (2); and (d) co-
injection of the precipitate and standards of 1 and 2. (For extraction, separation and chro-
matographic protocols see Experimental section: the peak labelled 3 is a system peak.)

From the results above (HPLC retention times, UV (2). In order to validate the method of extraction and
absorption maxima and ESI-MS spectra) it was con-cluded separation of 1 and 2 from burdock leaves, different
that the compounds at retention times 37.11 and 47.17 min weights (50, 100, 250 and 500 g) of leaf material were
were, respectively, arctiin (1) and arctigenin extracted and separated using the described procedures.

Figure 2. ESI-MS spectra of the two compounds purified from the precipitate obtained from
the fraction of a burdock leaf extract that eluted with 100% methanol from a polyamide
column chromatography showing the positive ion ([M+Na] +) and negative ion ([M−H]−)
modes of the compound with retention time 37.11 min (a and b, respect-ively), and of the
compound with retention time 47.17 min (c and d, respectively).

Copyright © 2005 John Wiley & Sons, Ltd. Phytochem. Anal. 16: 86 – 89 (2005)
 DETERMINATION OF ARCTIIN AND ARCTIGENIN IN BURDOCK
Figure 3. Yields (mg) of precipitate ( ), arctiin ( ), and arctigenin
( ) vs. weight (g) of burdock leaves extracted and separated
using the described methods. The calibration curves were
linear (y = 0.06145x + 0.09898, r 2 = 0.9997; y = 0.03408x −
0.11837, r 2 = 0.9998; and y = 0.01876x + 0.10510, r 2 =
0.9991, respectively) over the range of burdock leaf samples
extracted from 50 to 500 g.
Calibration plots showing the yield of the precipitate and of
1, 2 and precipitate vs. the weight of burdock leaves
employed (Fig. 3) were linear over the leaf sample weight
range (r2 values  0.999). The precipitate obtained, which
consisted of ca. 54.5% of 1 and 31.8% of 2, represented
only 0.0062% of the burdock leaves extracted. The method
described is clearly suitable to isolate and iden-tify arctiin
and arctigenin in burdock leaves. Because the amounts of
standards were limited, and 1 and 2 could not be detected in
the crude extracts, the recovery of the method was not
evaluated.
REFERENCES
Eich E, Schulz J, Trumm S, Sarin PS, Maidhef A, Merz H,
Schröder HC, Müller MEG. 1990. Lignanolides: novel in
vitro anti-HIV active agents. Plant Med 56: 506.
Eich E, Pertz H, Kaloga M, Schulz J, Fesen MR, Mazumder A,
Pommier Y. 1996. (−)-Arctigenin as a lead structure for
inhibitors of human immunodeficiency virus type-1
integrase. J Med Chem 39: 86–95.
Han BH, Kang YH, Yang HO, Park MK. 1994. A butyrolactone
lignan dimer from Arctium lappa L. Phytochemistry 37:
1161–1613.
Hirose M, Yamaguchi T, Lin C, Kimoto N, Futakuchi M, Kono T,
Nishibe S, Shirai T. 2000. Effects of arctiin on PhlP-induced
mammary, colon and pancreatic carcino-genesis in female
Sprague-Dawley rats and MelQx-induced
hepatocarcinogenesis in male F344 rats. Cancer Lett 155:
79–88.
Ichihara A, Oda K, Numata Y, Sakamura S. 1976. Lappaol A
and B, novel lignans from Arctium lappa L. Tetrahedron Lett
17: 3961–3964.
Ichihara A, Numata Y, Kanai S, Sakamura S. 1977. New
sesquilignans from Arctium lappa L. the structure of lappaol
C, D and E. Agric Biol Chem 41: 1813–1814.
Ichihara A, Kanai S, Nakamura Y, Sakamura S. 1978. Struc-
tures of lappaol F and H, dilignans from Arctium lappa L.
Tetrahedron Lett 19: 3035–3038.
Ichikawa K, Kinoshita T, Nishibe S, Sankawa U. 1986. The
calcium antagonist activity of lignans. Chem Pharm Bull 34:
1514–1517.
Iwakami S, Wu JB, Ebizuka Y, Sankawa U. 1992. Platelet acti-
vating factor (PAF) antagonists contained in medicinal
plants: lignans and sesquiterpenes. Chem Pharm Bull 40:
1196–1198.
Copyright © 2005 John Wiley & Sons, Ltd.
89
The seed extracts of burdock contain high con-
centrations of lipophilic substances, especially oils, and
various other constituents that interfere with the isolation of
the lignans. The isolation of 1 and 2 from burdock seed
normally has to be performed using a number of complex
procedures of partition and silica column chromatography
with several kinds of toxic organic solvents or by
precipitating with lead salts and then displacement of lead
after extraction. In con-trast, crude extracts of burdock
leaves contain few lipophilic substances except for
chlorophylls. In order to avoid the influence of chlorophylls
in the described method, the crude extracts were partitioned
with chlo-roform before column chromatography.
Following poly-amide column chromatography, the
fraction that eluted with 100% methanol could be
concentrated and pre-cipitated to yield a mixture of arctiin
and arctigenin directly. The method has been shown to be
effective for the isolation of 1 and 2 using burdock leaves
as source material and avoids the problems involved in
using large amounts of toxic organic solvents and heavy
metals as are required for isolating 1 and 2 from seed
material.
Acknowledgements
The authors wish to thank Dr. Jutta Kalbitz (Bioservice Halle GmbH) for
providing the standards of arctiin and arctigenin and Mrs. Barbara Kolbe
(Institute of Plant Biochemistry, Halle, Germany) for skilful technical
assistance. This research was partly supported by a grant from National
High New Technology and Development Programs of China (‘863’
Programs, no. 2002AA244031).
Macrae WD, Towers GH. 1984. Biological activities of lignans.
Phytochemistry 23: 1207–1220.
Morita K, Nishijima Y, Kada T. 1985. Chemical nature of a
demutagenic factor from burdock (Arctium lappa Linne).
Agric Biol Chem 49: 925–932.
Moritani S, Nomura M, Takeda Y, Kada T. 1996. Cytotoxic
components of bardanae fructus (goboshi). Biol Pharm Bull
19: 1515–1517.
Pfeifer K, Merz H, Steffen R, Müller WEG, Trumm S, Schulz J,
Eich E, Schröder HC. 1992. In-vitro activity of lignans—
differential inhibition of HIV-1 integrase reaction, topoiso-
merase activity and cellular microtubules. J Pharm Med 2:
75–97.
Ryu SY, Ahn JW, Kang YH, Han BH. 1995. Antiproliferative
effect of arctigenin and arctiin. Arch Pharm Res 18: 462–
463.
Shinohara K, Kuroki S, Miwa M, Kong LZ, Hosoda H. 1988.
Antimutagenicity of dialysates of vegetables and fruits.
Agric Biol Chem 52: 1369–1375.
Umehara K, Sugawa A, Kuroyanagi M, Ueno A, Taki T. 1993.
Studies on differentiation-inducers from Arctium fructus.
Chem Pharm Bull 41: 1774–1779.
Umehara K, Nakamura M, Miyase T, Kuroyanagi M, Ueno A.
1996. Studies on differentiation inducers VI. Lignan de-
rivatives from Arctium fructus (2). Chem Pharm Bull 44:
2300–2304.
Vlietinck AJ, De Bruyne T, Apers S, Pieters LA. 1998. Plant-
derived leading compounds for chemotherapy of human
immunodeficiency virus (HIV) infections. Planta Med 64:
97–109.
Wang HY, Yang JS. 1993. Research on constituents of
Arctium lappa L. seeds. Acta Pharm Sin 28: 911–917.
Phytochem. Anal. 16: 86 –89 (2005)