Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

Analytica Chimica Acta 471 (2002) 61–75

High-performance liquid chromatographic analysis of Black


Cohosh (Cimicifuga racemosa) constituents with in-line
evaporative light scattering and photodiode array detection
Wenkui Li a,b,c , Shaonong Chen a,d , Daniel Fabricant a,b,d , Cindy K. Angerhofer e ,
Harry H.S. Fong a,b,c,d , Norman R. Farnsworth a,b,d , John F. Fitzloff a,b,c,∗
a Program for Collaborative Research in the Pharmaceutical Sciences (PCRPS, m/c 877), College of Pharmacy,
University of Illinois at Chicago, 833 South Wood Street, Chicago, IL 60612-7231, USA
b Department of Medicinal Chemistry and Pharmacognosy (m/c 781), College of Pharmacy, University of Illinois at Chicago,
833 South Wood Street, Chicago, IL 60612-7231, USA
c Functional Food for Health (FFH) Core Analytical Laboratory, College of Pharmacy, University of Illinois at Chicago,

833 South Wood Street, Chicago, IL 60612-7231, USA


d NIH/UIC Center for Botanical Dietary Supplements Research, College of Pharmacy, University of Illinois at Chicago,

833 South Wood Street, Chicago, IL 60612-7231, USA


e Tom’s of Maine Corporation, Lafayette Center, P.O. Box 710, Kennebunk, ME 04043-0710, USA

Received 3 April 2002; received in revised form 7 August 2002; accepted 7 August 2002

Abstract
A validated and reproducible high-performance liquid chromatography (HPLC) method with in-line evaporative light scat-
tering and photodiode array detection was developed for the analysis of major constituents in Black Cohosh (Cimicifuga
racemosa L.). The method is based on the baseline chromatographic separation of 18 compounds reported to be present in
Black Cohosh on a C-18 column (5 ␮m, 4.6 mm × 250 mm) with water (0.05% trifluoroacetic acid, TFA)–acetonitrile–water
as the mobile phase, and their in-line detection using photodiode array detector (PDA) and evaporative light scattering
detector (ELSD). Sixteen of these (caffeic acid, ferulic acid, isoferulic acid, cimicifugoside H-1, cimiracemoside A,
cimicifugoside H-2, (26R)-actein, 26-deoxycimicifugoside, (26S)-actein, 23-epi-26-deoxyactein, 23-OAc-shengmanol-3-O-
␤-d-xyloside, 26-deoxyactein, 25-OAc-cimigenol-3-O-␣-l-arabinoside, 25-OAc-cimigenol-3-O-␤-d-xyloside, cimigenol-
3-O-␣-l-arabinoside, cimigenol-3-O-␤-d-xyloside) were detected to be present and quantifiable. The reputed two flavonoid
constituents (kaempferol and formononetin) were not detectable.
The validation of the method included tests for sensitivity, linearity, reproducibility, recovery and stability. The detection
limits were found to be in the range of 26–55 ng (10 ␮l per injection) in ELSD for the 16 constituents of Black Cohosh. The
exponential linear calibration curves were observed for all the constituents tested with r2 being >0.99. The reproducibility
of the method was evaluated by analyzing three sets of controls on three consecutive days (n = 3) with R.S.D. (%) and
relative error (%) <8.14 and 11.27%, respectively. The observed recovery rates were better than 91.75%. The stability of
Black Cohosh methanolic sample solution was assessed by consecutively analyzing the methanolic solutions prepared from
liquid extract (Sample A), powered extract (Sample B), milled plant material (Sample C) and one of the commercial product
(Sample D1) for the content of 16 major constituents with the relative standard deviations (R.S.D., %) <6.39%, indicating
the sample matrix of Black Cohosh is stable.

∗ Corresponding author. Tel: +1-312-355-4660; fax: 312-996-7107.

E-mail address: fitzloff@uic.edu (J.F. Fitzloff).

0003-2670/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 3 - 2 6 7 0 ( 0 2 ) 0 0 7 7 6 - 6
62 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75

Contrary to literature reports, kaempferol and formononetin were not detected in either our field collected plant materials
or commercial products of Black Cohosh in the current assay.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Black Cohosh; Cimicifuga racemosa; Triterpene glycosides; Phenolic acids; Flavonoids; HPLC; PDA; ELSD

1. Introduction In general, direct reverse-phase high-performance


liquid chromatography (RP-HPLC) with UV detec-
Cimicifuga racemosa (L.) Nutt., or Actaea race- tion could be, and is used for the analysis of Black
mosa (Ranunculaceae) [1,2], commonly known as Cohosh samples. However, the detection of triterpene
Black Cohosh, is a perennial herb whose subterranean glycosides using UV is well known for its insensitiv-
portion consists of a thick knotted rhizome system and ity because of the weak chromophoric functionality of
aerial stem attains a height of 1–2.5 m. Black Cohosh triterpene glycosides in the 200–210 nm region. In ad-
is native to North America, with its distribution pri- dition, the complex matrix of Black Cohosh extracts
marily in the mountain regions of the Eastern United also interferes and challenges the sensitivity and se-
States (Blue Ridge and Great Smoky Mountains) and lectivity of the HPLC-UV methodology [12]. LC–MS
Southeastern Canada [3]. This plant has been used or LC–MS–MS methods may work well for the iden-
traditionally by Native Americans for gynecological tification and quantification of the wide range of triter-
conditions long before the arrival of European set- pene glycoside and phenolic compounds in the matrix
tlers in the New World [4]. Various preparations of [12]. However, such equipment is expensive and may
this plant have also been used for the treatment of not be readily available. Thus, the development of an
menopausal disorders in Germany for over 50 years alternative HPLC method with a reliable and repro-
[5], and are presently available in the USA as dietary ducible detection is desirable. The inexpensive evapo-
supplements. A number of clinical studies have been rative light scattering detector (ELSD) may be suitable
published in support of the use of Black Cohosh as al- for the routine analysis of Black Cohosh preparations.
ternative treatment of menopausal symptoms [6–10]. As a mass detection method, the ELSD is based
Chemically, Black Cohosh extracts contain at on the nebulization of the LC column effluent into
least two major natural product groups: cycloartane droplets by the nebulizing gas and the entrance of the
triterpene glycosides, such as actein, 26-deoxyactein resulting vapor into a temperature-controlled evapora-
[11–14], and phenolic compounds, such as ferulic tor tube, where the evaporation of mobile phase takes
acid and isoferulic acid [15]. The flavonoids, such as place. The resulting “clouds” of solid microparticles
formononetin, have also been reported as present in are then directed towards a narrow light beam. As a re-
Black Cohosh [16]. sult, light is scattered by microparticles and measured
In spite of extensive chemical, biological and clin- using a photomultiplier or photodiode. A plot of detec-
ical studies, neither the active constituents nor the tor response versus analyte concentration is sigmoidal,
mode of action of Black Cohosh have been deter- and the peak area (I) is related to the sample size and
mined. Currently, the standardization of Black Cohosh shape, but not the chemical identity of the residual
preparations is based on the content of triterpene gly- particles passing through the light beam, by the fol-
cosides, calculated as 26-deoxyactein [9]. Since the lowing relationship: I = amb , where b is the slope of
most common commercial preparation is a powdered the response line, m is the mass of the compound in-
root extract and since the biological active constituents jected, and a is the response factor. Plots of the peak
have not been identified, product standardization area versus the analyte concentration with logarithmic
should be considered from an aggregate of its major co-ordinates are linear [17]. ELSD has been applied
constituents, rather than on a single compound. The to a wide range of UV-transparent analytes includ-
availability of a robust method allowing the analysis ing lipids [18,19], peptides [20], carbohydrates [21],
of such a mixture of compounds in an extract is highly and botanical bioactive compounds [17,22]. Recently,
desirable. Ganzera et al. reported an HPLC–ELSD method for
W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 63

the determination of actein, 26-deoxyactein, and Company, St. Louis, MO 63178, USA; Reference
cimiracemoside A in Black Cohosh extact with LOD kaempferol and formononetin were obtained from
of 105, 156 and 316 ng of them on the column, re- Indofine Chemical Company, Somerville, NJ 08876,
spectively [23]. However, taking into consideration USA; Standard (26S)-actein, (26R)-actein, ferulic acid
of the fact that >18 triterpene glycosides and a num- [15], isoferulic acid [15], cimicifugoside H-1 [29],
ber of phenolic acids have been reported from Black cimicifugoside H-2 [29], 26-deoxycimicifugoside
Cohosh [11–14], it is necessary to expand on the ex- [30], 23-epi-26-deoxyactein [31], 23-OAc-sheng-
isting procedure and develop a more comprehensive manol-3-O-␤-d-xyloside [32], 26-deoxyactein [31],
method for routine analysis of the constituents in 25-OAc-cimigenol-3-O-␣-l-arabinoside [12], 25-
Black Cohosh and its commercial products. OAc-cimigenol-3-O-␤-d-xyloside [32], cimigenol-3-
In the development of a broad based analytical pro- O-␣-l-arabinoside [12], cimigenol-3-O-␤-d-xyloside
cedure, all chemical constituents in plant materials or [33] were isolated and purified from Cimicifuga race-
commercial products must be considered. Flavonoids, mosa roots. The identities and of purity of the iso-
such as formononetin, have been reported to be present lates were characterized by means of spectral (NMR
in Black Cohosh [16], but their presence have been and MS) analyses and/or X-ray crystallography. The
questioned because isoflavonoids (formononetin) are structures of these reference standards are shown in
known to be rich in the Legume family (Leguminosae Fig. 1.
or Fabaceae), but are relatively rare in other plant fam- All reference standards were dissolved in methanol
ilies [24–27]. Recent attempts to validate the presence at a range of concentrations as described under
of formononetin and kaempferol in isopropyl alcohol Section 3.2. For reproducibility determination, three
and ethanol extracts of Black Cohosh failed to either sets of control solutions containing (26S)-actein,
identify their presence or to show their being present in cimiracemoside A, ferulic acid, isoferulic acid, cimi-
any significant level [28]. Thus, further investigations cifugoside H-1, cimicifugoside H-2, 26-deoxyci-
to determine the presence or absence of flavonoids in micifugoside, 23-epi-26-deoxyactein, 26-deoxyactein,
Black Cohosh are warranted. 25-OAc-cimigenol-3-O-␣-l-arabinoside and 25-OAc-
In the present study, an HPLC–PDA–ELSD method cimigenol-3-O-␤-d-xyloside were prepared and stored
was developed for the quantitative and qualitative at −20 ◦ C and brought to room temperature before
analysis of 16 major constituents of Black Cohosh use.
along with two potential constituents (kaempferol, Black Cohosh samples: (A) liquid extracts dissolved
formononetin) in one single HPLC run. This paper in 15% EtOH and 40% glycerin was provided by
describes the details of the methodology. Tom’s of Maine Corporation (Kennebunk, ME 04043);
(B) dried powdered extract was provided by Tom’s
of Maine Corporation (Kennebunk, ME 04043); (C)
2. Experimental plant materials was collected by one of the authors and
milled after dried; and (D) Black Cohosh commercial
2.1. Reagents dietary supplement products were purchased from lo-
cal pharmacies in Chicago, IL, USA. To protect the
Methanol, acetonitrile, trifluoroacetic acid (TFA) manufacturers’ identity, the sample sources are labeled
were HPLC grade from Fisher Scientific Co. (Fair D1–7. Products (D1–7) were claimed to contain from
Lawn, NJ, USA). Deionized water was obtained with 9 to 40 mg of Black Cohosh extract (per serving), stan-
an in-house Nano-pure® water system (Barnstead, dardized to 2.5–8% triterpene glycosides, and contain
Newton, MA, USA). other herbal ingredients, including soy isoflavonoids,
red clover extract, licorice extract, kudzu extract, and
2.2. Preparation of standard solutions and samples others.
Sample preparation was carried out according to
Reference cimiracemoside A was purchased from the nature of samples. (A) Liquid extract: 1.0 ml of
ChromaDex Inc., Laguna Hills, CA 92653, USA; liquid formula of Black Cohosh was filtered through a
Reference caffeic acid was obtained from Sigma 0.22 ␮m of membrane filter into a 1 ml HPLC sample
64 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75

Fig. 1. Structures of reference standards.


W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 65

vial; (B) dried powdered extract: 30 mg of dried respectively, of reference isoferulic acid, cimicifugo-
powdered extract was dissolved in 1 ml of methanol side H-1, 26-deoxyactein, 25-OAc-cimigenol-3-O-␣-
using sonicator and the resulting solution was fil- l-arabinoside and cimigenol-3-O-␤-d-xyloside and
tered through a 0.22 ␮m of membrane filter into a thoroughly mixed. This material was then extracted
1 ml HPLC sample vial; (C) milled plant materials: with methanol and sonicated as previously described
0.5–1.0 g of samples was accurately weighed into a for the Sample C. Three replicate recovery samples
20 ml PTFE-capped sample vial, 15 ml of methanol were prepared for analysis. A blank recovery sam-
was added and intimately mixed followed by sonica- ple was prepared and analyzed for the chemically
tion at 25–30 ◦ C for 30 min. After cooling to room exhausted root powder for comparative analysis.
temperature, the resulting mixture was filtered through To assess the stability of the major constituents of
a Whatman #1 filter paper into a 250 ml round-bottom Black Cohosh in sample matrix. The sample solutions
flask. The residue was returned to the vial, a second prepared from liquid extract (Sample A), powered
15 ml of methanol was added, sonicated and filtered extract (Sample B), milled plant material (Sample C)
as described, This procedure was repeated a third time and one of the commercial product sample (D1)
as described. The combined methanol extracts were were stored at room temperature (20 ◦ C) in dark
evaporated to dryness, under vacuum, at 45–50 ◦ C. and analyzed on consecutive days (24, 48 and 72 h)
The residue was re-dissolved in HPLC grade methanol as described above. The relative standard deviation
(4 ml × 2 ml), transferred to a 10 ml volumetric flask, (R.S.D., %) for the content of each constituent was
and made up to the volume with methanol. The obtained by comparing the content measured on 24 h
sample solution was filtered through a 0.22 ␮m mem- (day 1), 48 h (day 2) and 72 h (day 3), respectively.
brane filter just prior to HPLC analysis; (D) Black To facilitate the clean-up and identification of the
Cohosh commercial dietary supplement products: one flavonoids in the analyte, 1 ml of Sample A solution
dosage of the product (the content of one capsule, was absorbed on polyamide CC-6 powder (1g, 0.05–
caplet or tablet) was accurately weighed into a 20 ml 0.16 mm, Macherey Nagel, Doren, Germany), dried
PTFE capped sample vial and extracted with vary- and loaded onto a column (1 cm × 45 cm) prepared
ing concentration of methanol (100% methanol for with 10 g of polyamide CC-6 in water and developed
capsule, 80% methanol for caplet, 50% methanol for with water and 20% methanol, consecutively (200 ml
tablet formulations) under sonication at 25–30 ◦ C for each). Finally, the column was eluted with 200 ml
60 min. After cooling, the resulting mixture was fil- of methanol (100%). Methanol fractions were evapo-
tered through a Whatman #1 filter paper into a 250 ml rated under reduced pressure at 40 ◦ C and the resulting
round-bottom flask, and the residue was returned residue was dissolved in 1 ml of methanol. Triplicate
to the vial and treated as described for milled plant samples were prepared and 10 ␮l of sample solution
materials with varying concentration of methanol, was subjected to HPLC–PDA–ELSD analysis.
dependent on the formulation as stipulated above.
The resulting sample solution was filtered through a 2.3. Chromatography
0.22 ␮m membrane filter just prior to HPLC analysis.
To evaluate the recovery rate, Black Cohosh root A Waters 2690 Alliance HPLC system (Waters
powder (Sample C, 3.0 g) was extracted in a 50 ml Corporation, Milford, MA, USA), equipped with a
tared flask with 25 ml of methanol by sonication for 996 photodiode array detector, an on-line degasser
30 min. After filtration, the residue was returned to the and an autosampler, and a Waters YMC ODS-AQTM
same flask and sonicated in 25 ml of methanol. The RP-18 column (5 ␮m, 120 Å, 4.6 mm × 250 mm, se-
process was repeated until the last filtrate was shown rial # 0425503853) was used as in the current study,
to be free of detectable substances by HPLC–ELSD along with a Waters Delta-Pak RP-18 guard column
as described above. The residue was thoroughly dried (4.6 mm × 5 mm). UV detection was achieved in the
and a portion (0.5 g) was accurately weighed into range of 200–400 nm on a photodiode array detector
a 20 ml PTFE capped sample vial. To this residue (PDA). After the PDA, the chromatographic column
in the vial, 1 ml of standard working solution con- effluent was subjected to the detection by an evap-
taining 0.749, 1.362, 0.54, 0.157 and 0.532 mg, orative light scattering detector (ELSD) (Sedex 75,
66 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75

Alfortville, France). Nebulization of the effluent in lowest limits of detection (S/N > 3) of the 18 refer-
the ELSD was provided by a stream of pressurized ence standards by ELSD detection were: (26S)-actein
air (3.4 bar) at room temperature, and the nebulized (43 ng), (26R)-actein (24 ng), caffeic acid (50 ng),
effluent was evaporated at 43 ◦ C. The detector was set kaempferol (50 ng), formononetin (53 ng), ferulic
at gain 11, with output interfaced, via a SATIN box acid (50 ng), isoferulic acid (52 ng), cimicifugoside
to the Waters Millennium 2000® chromatographic H-1 (49 ng), cimiracemoside A (55 ng), cimicifugo-
manager system (Waters) equipped with on a Compaq side H-2 (49 ng), 26-deoxycimicifugoside (26 ng),
6400X/10000/CDS computer (Houston, TX, USA) 23-epi-26-deoxyactein (45 ng), 23-OAc-shengmanol-
for data handling and chromatogram generation. 3-O-␤-d-xyloside (46 ng), 26-deoxyactein (50 ng),
The chromatographic elution was accomplished by 25-OAc-cimigenol-3-O-␣-l-arabinoside (46 ng), 25-
a gradient solvent system consisting of water contain- OAc-cimigenol-3-O-␤-d-xyloside (52 ng), cimigenol-
ing 0.05% TFA (A), acetonitrile (B) and water (C). The 3-O-␣-l-arabinoside (32 ng), and cimigenol-3-O-␤-d-
gradient conditions were: 0–8 min, 80% A, 20% B; xyloside (31 ng) [34]. On the other hand, it is not
8–15 min, 32% B, 68% C; 15–55 min, 64% B, 36% C; surprising that the UV detection at 254 nm showed
55–65 min, 95% B, 5% C, kept to 70 min; 70–85 min, a very good sensitivity for the detection of pheno-
back to 80% A and 20% B. The flow rate was kept at lic compounds with lowest measurable quantities
1.6 ml/min, and the injection volume was 10 ␮l (injec- of 1 ng of caffeic acid, ferulic acid, isoferulic acid,
tion loop 50 ␮l). The eluted constituents were identi- kaempferol and formononetin, each, on column. This
fied by comparison of the retention time in both ELSD sensitivity is acceptable for the detection of potential
and PDA chromatograms and by comparison of the kaempferol and formononetin in the samples tested.
UV spectra of the peaks in the samples with those of
reference standards. To simplify the computation of 3.2. Linearity
the chromatograms, peaks detected by the PDA were
integrated at the wavelength of 254 nm, which is nor- Data obtained for the analysis of each reference
mally used for the detection of aromatic compounds. standards in 6–11 concentrations individually or in
Prior to each run, the HPLC–PDA–ELSD system was a mixture were employed from at least nine experi-
allowed to warm up and the baseline was monitored ments. In the ELSD mode, a second-order polynomial
until stable before sample analysis. calibration (peak area against amount) was observed.
After log-transformation, the data provided a linear
function for all reference standards following the
3. Results and discussion equation: Y = a + bX with Y being the log value of
the peak area, X the log value of sample amount, a
3.1. Detection the intercept and b the slope (Table 1) [17].

Chromatograms of all 18 reference standards gen- 3.3. Reproducibility


erated by ELSD and PDA detection are presented in
Fig. 2. Baseline separation of all these reference stan- The precision and accuracy of the method were as-
dards was obtained by ELSD detection as shown in sessed by within and between run validations. The
Fig. 2A. UV detection at 203 and 254 nm in the PDA reproducibility was evaluated by injecting three sets
detector are presented in Fig. 2B and C. of controls on three separate days. By substituting
Due to their being poor chromophores, most of the peak-area into the calibration curve equation from
the Black Cohosh triterpene glycosides gave very the same run, the measured concentrations were ob-
weak UV absorption at 203 nm (Fig. 2B) and no tained. By comparing observed and theoretical con-
absorption at 254 nm (Fig. 2C), when compared to centrations, the relative errors (RE, %) were obtained.
same concentration of reference standards detected The relative standard deviation (R.S.D., %) was cal-
by the ELSD (Fig. 2A), which confirmed the useful- culated by comparing the observed concentrations. As
ness of the latter method for the detection of poor shown in the Table 2, the R.S.D. and the RE were
absorbing or non-absorbing compounds [17]. The found to be <8.14 and 11.27%, respectively.
W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 67
68 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75

Fig. 2. (Continued ).
W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 69

Table 1
Linearity of assays for 16 constituents of Black Cohosh and two flavonoids in HPLC–ELSD by regression analysis
# Name Slope Intercept r2 Rangea

1 Caffeic acid 0.6439 1.9709 0.9975 10–100


2 Ferulic acid 1.2014 1.7943 0.9988 10–100
3 Isoferulic acid 0.8347 1.8361 0.9980 10–100
4 Cimicifugoside H-2 1.1703 1.6881 0.9949 10–340
5 Kaempferol 1.1567 1.7082 0.9976 10–100
6 Cimiracemoside A 1.066 1.7323 0.9988 11–110
7 Formononetin 1.3553 1.7089 0.9984 10.5–105
8 Cimicifugoside H-1 1.0327 1.7005 0.9972 10–340
9 (26R)-Actein 1.3651 1.7708 0.9966 3.5–12
10 26-Deoxycimicifugoside 1.2818 1.7733 0.9974 5.3–53
11 (26S)-Actein 0.9752 1.6845 0.9972 8–80
12 23-epi-26-Deoxyactein 1.4261 1.6507 0.9953 10–320
13 23-OAc-shengmanol-3-O-␤-d-xyloside 1.1486 1.6465 0.9952 9–90
14 26-Dexoyactein 1.007 1.6080 0.9964 10–100
15 25-OAc-cimigenol-3-O-␣-l-arabinoside (24S) 1.1772 1.6154 0.9970 9–90
16 25-OAc-cimigenol-3-O-␤-d-xyloside (24S) 1.1850 1.6246 0.9971 10–100
17 Cimigenol-3-O-␣-l-arabinoside (24S) 0.7324 1.8148 0.9960 6.5–125
18 Cimigenol-3-O-␤-d-xyloside (24S) 0.9898 1.7995 0.9982 6.5–65
Y: log value of peak area; X: log value of concentration (␮g/ml).
a Lowest and highest concentration in ␮g/ml.

3.4. Recovery side H-2 (3.39%), 26-deoxycimicifugoside (4.72%),


23-epi-26-deoxyactein (3.85%), 23-OAc-shengmanol-
The three sets of recovery samples were analyzed as 3-O-␤-d-xyloside (3.77%), 26-deoxyactein (4.61%),
described in Section 2. The average recovery (%) and 25-OAc-cimigenol-3-O-␣-l-arabinoside (6.39%), 25-
R.S.D. (%) for the reference standards were: isoferulic OAc-cimigenol-3-O-␤-d-xyloside (4.64%), cimige-
acid (97.37, 1.26), cimicifugoside H-1 (98.50, 1.50), nol-3-O-␣-l-arabinoside (2.51%), and cimigenol-3-
26-deoxyactein (99.96, 1.34), 25-OAc-cimigenol-O- O-␤-d-xyloside (4.93%). However, the presence of
␣-l-arabinoside (95.38, 4.16) and cimigenol-3-O-␤-d- phenolic acids such as caffeic acid, ferulic acid,
xyloside (91.75, 5.54). Similar recoveries for closely isoferulic acid in the sample solutions might affect
related compounds would be expected. the stability of triterpene glycosides when the sam-
ple solutions were put at room temperature for long
3.5. Stability time. Thus, the long-term stability of Black Cohosh
preparation needs further investigation.
In the current study, the consecutive analyses of the
major constituents in the sample solutions prepared 3.6. Application
from liquid extract (Sample A), powered extract (Sam-
ple B), milled plant material (Sample C) and one of A commercial liquid extract (A) and a dried pow-
the commercial product (Sample D1) revealed that the dered extract (B) as well as a milled field collected
methanolic solutions of different Black Cohosh sam- plant material (C) of Black Cohosh were analyzed in
ple (Samples A, B, C and D1) are stable at room tem- triplicate. Most of the constituents in these samples
perature with the highest relative standard deviations were detectable and quantifiable. The content of the
(R.S.D., %) observed as follows: (26S)-actein (4.23%), major constituents in these materials is presented in
(26R)-actein (4.27%), caffeic acid (3.58%), ferulic Table 3. The typical HPLC chromatograms of liquid
acid (3.26%), isoferulic acid (3.91%), cimicifugoside extract (A) are shown in Fig. 3 ((A) HPLC–ELSD;
H-1 (3.07%), cimiracemoside A (4.41%), cimicifugo- (B) HPLC-UV at 203 nm; (C) HPLC-UV at 254 nm).
70 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75
W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 71

Fig. 3. Typical HPLC chromatograms of Black Cohosh liquid extract (Sample A) ((A) HPLC–ELSD; (B) HPLC-UV at 203 nm; (C)
HPLC-UV at 254 nm).
72 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75

Fig. 3. (Continued ).

Table 3
To determine the feasibility of the present method
Content (%) of major constituents in liquid extract (Sample A),
dried powdered root extract (Sample B) and milled field collected for post marketing analysis of Black Cohosh products,
samples (Sample C) of Black Cohosha seven commercial dietary supplement preparations
(Sample D) were analyzed. The results of the anal-
Constituent Liquid extract Powdered Field collected
(mg/ml) root extract crude material ysis (Table 4) showed triterpene glycosides content
(%, w/w) (%, w/w) in the range of 0.13–5.97 mg per serving. As previ-
1 0.07 0.05 0.01 ously indicated, Black Cohosh extract is generally
2 0.75 0.93 0.01 standardized to 2.5% triterpene glycosides, i.e. if the
3 0.36 4.94 0.07 product contains 40 mg of standardized extract per
4 0.11 1.31 0.01 serving (capsule, caplet or tablet), the assumed con-
6 0.40 0.09 0.05
8 0.73 5.01 0.01
tent of triterpene glycosides will be 1 mg per serving.
9 0.34 ND ND However, the current assay revealed that some Black
10 0.25 0.13 0.02 Cohosh preparations did not contain as much of the
11 2.80 1.08 0.17 triterpene glycosides as the manufacturers claimed.
12 1.25 1.24 0.09 On the other hand, the production of Black Cohosh
13 1.04 1.18 0.22
14 0.20 0.55 0.04
product, like any other botanical dietary supplement,
15 0.13 0.20 ND involves the acquisition of source plant materials by
16 0.10 0.63 ND cultivation or field collection, processing of the plant
17 ND ND 0.0001 materials into a standardized extract, and manufactur-
18 ND ND 0.0009 ing the standardized extract into suitable dosage forms
a ND: not detected. such as capsule, caplet, tablet, etc. Under the Dietary
W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 73
74 W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75

Table 4
Total triterpene glycosides content (per serving) in commercial Black Cohosh products
Source Amount/serving Label claim Total triterpene Triterpene glycosides detecteda
glycosides found
(mg) per serving
D1 500 mg of root powder NA 5.97 4, 6, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18
D2 9 mg of extract Triterpene glycosides 1.18 4, 6, 8, 11, 14, 15, 17, 18
in extract (8%)
D3 40 mg of extract Teriterpene glycoside 1.12 4, 6, 8, 11, 12, 13, 14, 15, 16, 17, 18
in extract (2.5%)
D4 40 mg of extract NA 1.33 4, 6, 8, 11, 12, 13, 14, 15, 16, 18
D5 40 mg of extract NA 0.87 4, 6, 8, 11, 14, 17, 18
D6 40 mg of extract Triterpene glycosides 3.23 4, 6, 8, 10, 11, 12, 13, 15, 16, 17, 18
in extract (2.5%)
D7 20 mg of extract NA 0.13 6, 11, 12
a For ID # of each constituent, please refer to Table 1.

Supplement Health and Education Act (DSHEA) of 4. Conclusion


1994, Black Cohosh products are marketed as dietary
supplements in the USA. Since DSHEA only places The present study describes a validated analytical
the responsibility of safety, but not the burden of method for the quantitation of major triterpene glyco-
proof for the product on the manufacturers, Black Co- sides and phenolic compounds in the Black Cohosh
hosh products have not been subjected to mandated samples. The limits of detection of those compounds
quality assurance (QA) standards. As a consequence, have been determined to be between 26 and 55 ng.
it is not uncommon to see that the product differ Kaempferol and formononetin were not detected in the
from brand to brand, even from lot to lot. In view of Black Cohosh extracts and in the commercial prod-
the variability of the major constituents detected in ucts, contrary to literature reports in the current assay.
the commercial samples (D1–7), a better regulation In the current study, some Black Cohosh commercial
and effective control (GAP, GLP and GMP) is highly products did not meet the label claim for triterpene
desirable. glycosides.
In contrast to their reputed presence in the lit-
erature [16], the flavonoids, kaempeferol and for-
mononetin were not detected in the current study, Acknowledgements
based on comparison of the retention times and UV
spectral library (200–400 nm) with the reference stan- This study was partially supported by funds from
dards. In order to eliminate interfering substances in NIH/NCCAM Grants no. 5950AT00155 and the
the materials tested, the samples were subjected to University of Illinois Functional Food for Health
a “clean-up” procedure as described in Section 2. Program.
Kaempferol and formononetin were still not detected
in the purified fraction. Although the HPLC–PDA References
chromatogram of a purified fraction of Black Cohosh
sample (A) showed several potential peaks with re- [1] J.A. Compton, A. Culham, J.G. Gibbings, S.L. Jury, Biochem.
tention times close to those of kaempferol (22.96 min) Syst. Ecol. 26 (1998) 185.
and formononetin (26.39 min) reference standards [2] J.A. Compton, A. Culham, S.L. Jury, Taxon. 47 (1998) 593.
[3] J.A. Duke, CRC Handbook of Medicinal Herbs, CRC Press,
(Fig. 4), their UV spectra (200–400 nm), however,
Boca Raton, FL, 1985, p. 677.
are very different. Confirmation of the presence of [4] D. McKenna, K. Jones, S. Humphrey, K. Hughes, Altern.
these flavonoids in Black Cohosh requires further Ther. Health Med. 7 (2001) 93.
investigation. [5] S. Foster, Herbal Gram. 45 (1999) 35.
W. Li et al. / Analytica Chimica Acta 471 (2002) 61–75 75

[6] M.L. Hardy, J. Am. Pharm. Assoc. 40 (2000) 234. [22] F.F. Camponovo, J.L. Wolfender, M.P. Maillard, O. Potterat,
[7] S. Lieberman, J. Women’s Health 7 (1998) 525. K. Hostettmann, Phytochem. Anal. 6 (1995) 141.
[8] B.L. McFarlin, M.H. Gibson, J. O’Rear, P. Harman, J. [23] M. Ganzera, E. Bedir, I.A. Khan, Chromatographia 52 (2000)
Nurse-Midwifery 44 (1999) 205. 301.
[9] J. Pepping, Am. J. Health Syst. Pharm. 56 (1999) 1400. [24] J.B. Harborne, T.J. Mabry, H. Mabry, The Flavonoids,
[10] E. Liske, Adv. Ther. 15 (1998) 45. Chapman and Hall, London, 1975.
[11] Y. Shao, A. Harris, M.F. Wang, H.J. Zhang, G.A. Cordell, [25] J.B. Harborne, T.J. Mabry, The Flavonoids: Advances in
M. Bowman, E. Lemmo, J. Nat. Prod. 63 (2000) 905. Research, Chapman and Hall, London, 1982.
[12] K. He, B.L. Zheng, C.H. Kim, L. Rogers, Q.Y. Zheng, Planta [26] J.B. Harborne, The Flavonoids: Advances in Research Since
Med. 66 (2000) 635. 1980, Chapman and Hall, New York, 1988.
[13] E. Bedir, I.A. Khan, Chem. Pharm. Bull. 48 (2000) 425. [27] J.B. Harborne, The Flavonoids: Advances in Research Since
[14] E. Bedir, I.A. Khan, Pharmazie 56 (2001) 268. 1986, CRC press, Boca Raton, FL, 1994.
[15] S.O. Kruse, A. Lohning, G.F. Pauli, H. Winterhoff, A. [28] D. Struck, M. Tegtmeier, G. Harnischfeger, Planta Med. 63
Nahrstedt, Planta Med. 65 (1999) 763. (1997) 289.
[16] H. Jarry, G. Harnischfeger, E. Duker, Planta Med. 51 (1985) [29] M. Koeda, Y. Aoki, N. Sakurai, M. Nagai, Chem. Pharm.
316. Bull. 43 (1995) 771.
[17] W.K. Li, J.F. Fitzloff, J. Pharm. Biomed. Anal. 25 (2001) [30] A. Kusano, M. Takahira, M. Shibano, T. Miyase, G. Kusano,
257. Chem. Pharm. Bull. 47 (1999) 511.
[18] T. Seppanen-Laakso, I. Laakso, H. Vanhanen, K. Kiviranta, [31] S.N. Chen, W.K. Li, D.S. Fabricant, B.D. Santarsiero, A.
T. Lehtimaki, R. Hiltunen, J. Chromatogr. B: Biomed. Appl. Mesecar, J.F. Fitzloff, H.H.S. Fong, N.R. Farnsworth, J. Nat.
754 (2001) 437. Prod. 65 (2002) 601.
[19] G.A. Picchioni, A.E. Watada, B.D. Whitaker, Lipids 31 (1996) [32] A. Kusano, M. Shibano, S. Kitagawa, G. Kusano, S. Nozoe,
217. S. Fushiya, Chem. Pharm. Bull. 42 (1994) 1940.
[20] J. Bongers, T.K. Chen, J. Liq. Chrom. Rel. Technol. 23 (2000) [33] C.J. Li, D.H. Chen, P.G. Xiao, Acta Pharm. Sin. 28 (1993)
925. 777.
[21] Y. Wei, M.Y. Ding, J. Chromatogr. A. 904 (2000) 113. [34] J. Vial, A. Jardy. Anal. Chem. 71 (1999) 2672.

You might also like