Detection of Actaea racemosa Adulteration by Thin-Layer Chromatography and Combined Thin-Layer Chromatography-Bioluminescence

Abstract

Actaea racemosa L. (black cohosh; syn. Cimicifuga racemosa L. Nutt.) is a native North American perennial whose root and rhizome preparations are commercially available as phytomedicines and dietary supplements, primarily for management of menopausal symptoms. Despite its wide use, methods that accurately identify processed A. racemosa are not well established; product adulteration remains a concern. Because of its similar appearance and growing locales, A. racemosa has been unintentionally mixed with other species of the genus, such as Actaea pachypoda Ell. (white cohosh) and more commonly Actaea podocarpa DC. (yellow cohosh). The genus Actaea also has 23 temperate species with numerous common names, which can also contribute to the misidentification of plant material. Consequently, a variety of Actaea spp. are common adulterants of commercially available black cohosh preparations. Thin-layer chromatography (TLC) and combined TLC-bioluminescence (Bioluminex™) are efficient, economical, and effective techniques which provide characteristic patterns and toxicity profiles for each plant species. These data indicate that common black cohosh adulterants, such as yellow cohosh, can be differentiated from black cohosh by TLC and TLC-bioluminescence. This study also showed that unknown contaminants that were not detected using standard A. racemosa identity techniques were readily detected by TLC and TLC-bioluminescence.

It is estimated that more than 50% of adults in the United States use some form of dietary supplement (1). This prevalent use generated more than $4.4 billion in annual herb sales for 2005 (2). Actaea racemosa L. (black cohosh; syn. Cimicifuga racemosa L. Nutt.) was ranked as the third fastest growing herbal product in the mass market in 1998 and is currently one of the 10 top-selling herbals in the United States (2, 3). The medicinal use of A. racemosa dates back to Native Americans and has been prescribed for the treatment of a variety of ailments, such as inflammation and gynecological disorders (4). Today, the roots and rhizomes of A. racemosa are used in the United States as a dietary supplement, primarily to manage menopausal symptoms. In recent years, its consumption has increased as a result of its perceived activity as a nonestrogenic alternative to hormone replacement therapy, which has come under scrutiny as a possible cause of increased risk of breast cancer and thrombosis (5–12).

This surge in black cohosh consumption places enormous harvest pressure on the species (13) and increases the potential for product adulteration. It is estimated that over 1.4 million pounds of black cohosh were harvested between 1997 and 2001 (13, 14). Because of insufficient cultivation sources for A. racemosa, an estimated 96.7% comes from wild-crafting sources (15), which increases the potential for collection errors such as collecting the wrong plant and/or the wrong plant part.

Adulteration and/or plant misidentification of black cohosh is a serious health concern. Several species of Actaea that are listed as potential adulterants of black cohosh are also described as poisonous (14, 16–18). For example, Actaea podocarpa DC. (yellow cohosh; formerly Cimicifuga americana) grows in the same habitat and has a similar above-ground appearance as A. racemosa and is consequently reported as commonly being mixed with A. racemosa (14). After drying, the rhizomes of A. racemosa and A. podocarpa are extremely difficult to distinguish, making post-harvest identification problematic. Actaea pachypoda Ell. (white cohosh), which also has underground portions that are difficult to distinguish from A. racemosa, is reported as an occasional adulterant of black cohosh (14). Adding to the concern, there are little to no data on the toxicological properties of A. podocarpa, and all plant parts of A. pachypoda are reported as poisonous (19, 20). This is of grave concern since it is speculated that these 2 Actaea species, including their aerial parts, are common adulterants of commercially available black cohosh preparations and biomass.

Unfortunately, the low UV activity of A. racemosa constituents and high number of triterpene glycosides make identification and characterization of processed A. racemosa a difficult task. Current mainstream methods of A. racemosa standardization rely on triterpene glycoside determination by high-performance liquid chromatography–evaporative light-scattering detection (HPLC-ELSD; 21). This technique uses the standard marker compound 23-epi-26-deoxyactein to quantitate total triterpene glycosides but does not distinguish between the triterpene glycosides or identify Actaea species or common adulterants. Several methods have been developed in an attempt to establish an effective method that addresses these issues. One promising validated method using HPLC-photodiode array (PDA)/ELSD has been developed for the analysis of the major constituents of A. racemosa but has not been tested to determine if it can detect other Actaea species (22). A variety of analytical methods using mass spectrometry (MS) techniques, such as HPLC-MS (23, 24) and HPLC-PDA/MS/ELSD (25) have also been used to identify chemical constituents and determine biomass identity. Even molecular biology methods that use amplified DNA analyses to create DNA fingerprints have been used to identify Actaea species (26, 27). Although each of these methods is promising, the complicated data fingerprints, expensive equipment, and lack of trained personnel make routine analysis inefficient and quite expensive.

An inexpensive and effective methodology to evaluate black cohosh biomass for identity and adulteration would be advantageous. Presented in this study are 2 such techniques, a method confirmation of high-performance thin-layer chromatography (HPTLC; 14) and method development using TLC-bioluminescence (Bioluminex™) for detection of Actaea spp. adulteration in A. racemosa. HPTLC methods can separate up to 20 (20 × 10 cm plate) complex mixtures simultaneously and provide multiple fingerprints in one analysis by using various UV and staining reagent detection measures. Adding bioluminescence detection to the HPTLC data simultaneously provides a unique chemical and biological profile for each analyzed mixture, which can be used to help support material identity and detect potentially toxic compounds. This approach provides multiple measures by which reproducible identification, quality, and efficacy can be established.

Experimental

Plant Material

Authenticated rhizome/root and leaf samples of dried A. racemosa L. (Ranunculaceae), A. pachypoda Ell. (Ranunculaceae), and A. podocarpa DC. (Ranunculaceae) were supplied by Trish Flaster (Botanical Liaisons, LLC, Boulder, CO) and are referred to as Reference 1. Authenticated rhizome/root samples of dried A. racemosa L. (Ranunculaceae), A. pachypoda Ell. (Ranunculaceae), and A. podocarpa DC. (Ranunculaceae) were supplied by Edward J. Fletcher (Strategic Sourcing Inc., Reading, PA) and are referred to as Reference 2. Voucher specimens of all plant material supplied by Trish Flaster are archived at Botanical Liaisons, LLC with reference No. DB 71001 (A. racemosa), No. DB 727 (A. pachypoda), and No. DB 710303 (A. podocarpa). Voucher specimens of all plant material supplied by Edward Fletcher are stored at ChromaDex Analytics (Boulder, CO) with reference No. ACRA5/16”14-2”7.9.05-W (A. racemosa), No. ACPA10/2“2-8”-.10.05-C (A. pachypoda), and No. ACPO10/6”6-14”-8.13.05C (A. podocarpa). All dried plant materials were milled to a fine powder in a Cyclone Sample Mill (UDY Corp., Ft. Collins, CO). One commercial sample and one additional vouchered sample of A. racemosa rhizome/root were obtained (Botanical Liaisons) and subjected to the same procedure as the reference materials.

Biomass Extraction

A sample (0.5 g) of each powdered plant material was suspended in 10 mL CH₃OH in a 25 mL scintillation vial with Teflon cap. Samples were sonicated at 40°C for 15 min, and then extracted in an oven at 60°C for 1 h without shaking. Supernatants (10 mL) were filtered through a 0.45 µm PTFE syringe filter and brought to a 10 mL volume in a 10 mL volumetric flask.

Total Extracted Solids

Total extracted solids (TES) for each sample extract were obtained in duplicate and then averaged. A predried aluminum weigh pan tare weight was recorded. A 1.00 mL sample of each extract was transferred to the aluminum pan (no cover). The samples were then heated at 100°C for 1 h. After cooling, the pan was again weighed to obtain a final weight. Resulting percent TES for each sample extract was calculated as follows:

• (Final weight – tare weight)/1.00 mL × 100

Standards

Solutions of actein (1.0 mg/mL in 60% C₂H₅OH), 23-epi-26-deoxyactein (0.5 mg/mL in CH₃OH), cimiracemoside (0.5 mg/mL in CH₃OH), cimifugoside (0.5 mg/mL in CH₃OH), caffeine (1.0 mg/mL in CH₃OH), and tyrosine (0.5 mg/mL in H₂O) were supplied by ChromaDex (Irvine, CA).

HPTLC and Bioluminex Analysis

Chromatographic analyses were performed on prewashed (CH₃OH elution to the top of the plate followed by 100°C) Bioluminex silica gel 60 F₂₅₄ HPTLC plates (20 × 10 cm; 0.20 mm sorbent thickness; ChromaDex). Samples were applied at y = 8 mm as 6 mm-wide bands, using the band spray application mode of a CAMAG (Muttenz, Switzerland) Automatic TLC Sampler 4. The plates were then developed to 70 mm in a presaturated (30 min with saturation pad) 20 × 10 cm ridged-bottom chamber (Alltech, Nicholasville, KY) with toluene–ethyl formate–formic acid (5 + 3 + 2, v/v/v) under laboratory conditions (14). Developed plates were dried in a mechanical oven at 40°C for 2 h before detection. All images were captured using a Fluorchem® 8900 system, and retention factors (R_f) were calculated using AlphaEase FC software (Alpha Innotech, San Leandro, CA).

HPTLC Detection

Components were visualized by UV irradiation at 254 and 365 nm. After initial UV images were recorded, plates were sprayed with a 5% H₂SO₄–anisaldehyde reagent, then heated (125°C, 5 min), and a white light image was recorded (14). The 5% H₂SO₄–anisaldehyde reagent was prepared by adding 5 mL concentrated sulfuric acid to 95 mL anisaldehyde reagent (40 mL acetic acid, 2 mL anisaldehyde, 340 mL CH₃OH).

Bioluminex Detection

Lyophilized Vibrio fischeri cultures were obtained from ChromaDex and stored at 4°C. Bacteria cells were grown for the assay overnight (30 h) in a 200 mL batch culture in complex Bioluminex medium at 120 rpm (Forma Scientific Orbital Shaker Model No. 4518; Marietta, OH) and 28°C under normal aerobic conditions. Directly before assay, 15.62 g Bioluminex Buffer B was added to fully luminescent bacteria and dissolved at 120 rpm and 28°C under atmospheric conditions. The developed HPTLC plate was coated with buffered luminescent V. fischeri using an automatic immersion device (200 mL; CAMAG). Excess bacteria were removed from the plate using a squeegee device, and images were recorded after an 8 min incubation using an exposure time of 2 min with a cooled (−30°C absolute temperature) charge-coupled device (CCD) camera (Alpha Innotech Fluorchem 8900) and dark box. Compounds that interfere with the respiratory system or adenosine triphosphate pool of the bacteria proportionally inhibit bioluminescence. These compounds are detected as dark bands (inhibited bacteria) on a luminescent background (viable bacteria).

Results and Discussion

For this study, rhizome/root and leaf extracts of black (A. racemosa), yellow (A. podocarpa), and white (A. pachypoda) cohosh were examined by TLC and TLC-bioluminescence using the standardized triterpene glycoside analysis method for black cohosh (14). A 10 mg amount of each biomass extract, 4 cohosh standards (actein, 23-epi-26-deoxyactein, cimifugoside, cimiracemoside) and a Bioluminex positive and negative control (caffeine and tyrosine, respectively) were separated by TLC using toluene–ethyl formate–formic acid (5 + 3 + 2, v/v/v; Table 1). After development, UV detection at 254 and 365 nm were recorded for each plate (Figure 1, chromatograms 1 and 2, respectively). The TLC-only plate was then stained with anisaldehyde reagent and the resulting chromatogram was recorded under white light (Figure 1, chromatogram 3). The Bioluminex plate was coated with bioluminescent V. fischeri and the resulting image was recorded (Figure 1, chromatogram 4). All R_f values were calculated using AlphaEaseFC (Alpha Innotech) software analysis tools and are presented in Table 2.

Table 1.

Sample identification, application amounts, and lane assignments for Figure 1

Lane	Amount, µg	Sample	Lane	Amount, µg	Sample
1	0.6	Actein	11	10.0	A. pachypoda Reference 2
2	10.0	A. racemosa Vouchered	12	0.5	Cimiracemoside
3	10.0	A. racemosa Reference 2	13	4.0	Caffeine
4	10.0	A. racemosa Reference 1	14	4.0	Tyrosine
5	10.0	A. racemosa Commericial	15	NA	Blank
6	0.5	23-epi-26-deoxyactein	16	NA	Blank
7	10.0	A. podocarpa Reference 1	17	10.0	A. racemosa Reference 1 (leaf)
8	10.0	A. podocarpa Reference 2	18	10.0	A. podocarpa Reference 1 (leaf)
9	0.5	Cimifugoside	19	10.0	A. pachypoda Reference 1 (leaf)
10	10.0	A. pachypoda Reference 1

Figure 1.

Portions (10 µg) of rhizome/root (tracks 2–5, 7, 8, 10, 11) and leaf (tracks 17–19) extract from black, yellow, and white cohosh were developed with toluene–ethyl formate–formic acid (5 + 3 + 2, v/v/v) and detected under 254 nm (chromatogram 1), 365 nm (chromatogram 2), white light (chromatogram 3) after coating with anisaldehyde reagent, and via the Bioluminex assay (chromatogram 4).

Table 2.

Calculated compound R_f values by detection at UV 254 and 365 nm, anisaldehyde staining, and Bioluminex analysis

	Rf
Track	254 nm	365 nm	Anisaldehyde	Bioluminex
1	NDa	ND	0.36	ND
2	0.10, 0.14, 0.23, 0.27, 0.36, 0.45	0.07, 0.10, 0.14, 0.23, 0.27	0.05, 0.32, 0.36, 0.40–0.41, 0.45, 0.72, 0.83	0.07, 0.14, 0.23, 0.27, 0.40, 0.45, 0.50, 0.59, 0.72, 0.83
3	0.10, 0.14, 0.36, 0.45	0.10, 0.14, 0.23, 0.27	0.05, 0.32, 0.36, 0.40–0.41, 0.45, 0.72, 0.83	0.14, 0.23, 0.27, 0.40, 0.45, 0.50, 0.59, 0.72, 0.83
4	0.10, 0.14, 0.36, 0.45	0.10, 0.14	0.05, 0.32, 0.36, 0.40, 0.45, 0.72, 0.83	0.14, 0.40, 0.45, 0.59, 0.72, 0.83
5	0.10, 0.14, 0.36, 0.45	0.10, 0.14	0.05, 0.32, 0.36, 0.40–0.41, 0.45, 0.72, 0.83	0.14, 0.40, 0.45, 0.50, 0.59, 0.72, 0.83
6	ND	ND	0.32	ND
7	0.07, 0.15, 0.21, 0.27, 0.31, 0.36, 0.45	0.15, 0.21	0.05, 0.21, 0.27, 0.31, 0.35, 0.44, 0.72, 0.83	0.14, 0.21, 0.27, 0.35, 0.39, 0.44, 0.48, 0.59, 0.72, 0.83
8	0.15, 0.21, 0.27, 0.31, 0.36, 0.45	0.15, 0.21	0.05, 0.21, 0.27, 0.31, 0.35, 0.44, 0.72, 0.83	0.14, 0.21, 0.27, 0.35, 0.39, 0.44, 0.48, 0.59, 0.72, 0.83
9	ND	ND	0.40	ND
10	0.10, 0.14, 0.36b, 0.45b	0.10, 0.14	0.05, 0.28, 0.33, 0.35, 0.39, 0.43, 0.72b, 0.83b	0.13, 0.39, 0.43, 0.56, 0.72, 0.83
11	0.10, 0.36b, 0.45b	0.10	0.05, 0.28, 0.34, 0.39, 0.43, 0.72b, 0.83b	0.34, 0.39, 0.43, 0.72, 0.83
12	ND	ND	0.33	ND
13	0.08	ND	0.08	Slight
14	0.32	ND	ND	0.32
15	—	—	—	—
16	—	—	—	—
17	0.10, 0.15, 0.46, 0.72	0.10, 0.15, 0.46, 0.53, 0.63, 0.72	0.05, 0.13, 0.21, 0.25, 0.28, 0.33, 0.50, 0.72, 0.83	0.43, 0.46, 0.50, 0.59, 0.72
18	0.46, 0.72	0.14, 0.46, 0.53, 0.63, 0.72	0.05, 0.13, 0.23, 0.27, 0.50, 0.72, 0.83	0.43, 0.46, 0.50, 0.59, 0.72
19	0.10, 0.46, 0.72	0.07, 0.10, 0.46, 0.72	0.05, 0.13, 0.31, 0.36, 0.50, 0.72, 0.83	0.43, 0.46, 0.50, 0.59, 0.72, 0.83

^aND = Not detected.

^bPeaks detectable at amounts >25 g (data not shown).

Bands that are characteristic to A. racemosa rhizome identification were detected by TLC in all analyzed A. racemosa extracts at R_f ≅ 0.10 and 0.14 under 254 and 365 nm and at R_f ≅ 0.32 (23–epi-26-deoxyactein), 0.36 (actein), and 0.40 after anisaldehyde staining and white light illumination. Bands with similar but not identical R_f values to those characteristic of A. racemosa are also detected in A. podocarpa and in A. pachypoda rhizome extracts (see Table 2 for a full list of bands). These similarities might contribute to the misidentification or lack of adulterant identification of these products.

The most concern for distinction by TLC and TLC-bioluminescence is between the pachypoda and racemosa spp. The only clearly distinguishable peak detected in both A. pachypoda samples that is not detected in the A. racemosa extracts is at R_f ≅ 0.28 after anisaldehyde derivatization and white light detection. Alternatively, several bands of distinction between A. podocarpa and A. racemosa are detected in A. podocarpa, which is more commonly found mixed with A. racemosa (14). Two notable bands that could be used to discern A. podocarpa from A. racemosa are detected at R_f ≅ 0.21 and 0.27 under 254 nm, white light after anisaldehyde derivatization, and after bioluminescence application. Additionally, the peak at R_f ≅ 0.21 is detected under 365 nm. To determine if A. podocarpa rhizome/root material can be detected in A. racemosa rhizome/root extract by TLC and/or TLC-bioluminescence, rhizome/root extracts of A. racemosa were spiked with 17% (5.7 µg black and 1.2 µg yellow cohosh), 39% (3.8 µg black and 2.4 µg yellow cohosh), and 65% (1.9 µg black and 3.6 µg yellow cohosh) A. podocarpa rhizome/root extract. Each extract, including the 100% A. racemosa and A. podocarpa extracts, was chromatographed as previously described using a toluene–ethyl formate–formic acid (5 + 3 + 2, v/v/v) solvent system, and resulting chromatograms were recorded under 254 and 365 nm, white light after anisaldehyde staining, and after Bioluminex analysis (Figure 2, chromatograms 5–8, respectively, and Table 3).

Figure 2.

Chromatographed rhizome/root extracts of black cohosh adulterated with various amounts of yellow cohosh rhizome/root were developed with toluene–ethyl formate–formic acid (5 + 3 + 2, v/v/v) and detected under 254 nm (chromatogram 5), 365 nm (chromatogram 6), white light (chromatogram 7) after coating with anisaldehyde reagent, and via the Bioluminex assay (chromatogram 8).

Table 3.

Sample identification, application amounts, and lane assignments for Figure 2

Lane	Amount, µg	Sample	Lane	Amount, µg	Sample
1	0.6	Actein	6	8.4	A. podocarpa
2	13.3	A. racemosa	7	0.5	23-epi-26-deoxyactein
3	5.7/1.2	A. racemosa/A. podocarpa	8	4	Tyrosine
4	3.8/2.44	A. racemosa/A. podocarpa	9	4	Caffeine
5	1.9/3.6	A. racemosa/A. podocarpa

Figure 2 shows the TLC and Bioluminex fingerprints of each spiked cohosh extract tested. Peaks at R_f ≅ 0.21 and 0.27, which are detected in A. podocarpa but not in A. racemosa, can be used to distinguish between the 2 species. Under 254 nm detection, the band at R_f ≅ 0.27 is easily detected when ≥17% A. podocarpa is present. Under 365 nm and after anisaldehyde staining, the band at R_f ≅ 0.21 is readily detected when ≥17% A. podocarpa is present. After Bioluminex analysis, both bands are detected, but not until ≥39% A. podocarpa. These characteristic bands could be used to identify adulteration of A. racemosa with A. podocarpa and determine approximate contaminant concentrations of A. podocarpa when the concentration is >17%.

Examination of the leaf extracts of black, yellow, and white cohosh demonstrated that these plant parts can be easily distinguished from root and rhizome extracts primarily by using TLC analysis and, to some degree, TLC-bioluminescence. The leaf bands that are the easiest to distinguish are detected under 365 nm. Compounds are revealed at R_f ≅ 0.50, 0.72, and 0.83 as bright red bands that correlate to the leaf pigment components. These red bands are not observed in the rhizome/root extracts. Interestingly, the band at R_f ≅ 0.72 in the A. podocarpa leaf and rhizome/root extract display significant bioluminescence inhibition that is not as prevalent in the other extracts.

One of most interesting observations revealed in this study are the 2 peaks found at R_f ≅ 0.23 and 0.27 in the A. racemosa vouchered and the A. racemosa Reference 2 material, lanes 2 and 3, respectively, in Figure 1. These peaks are readily detected under 365 nm as bright orange bands and, after Bioluminex detection, as bands that significantly inhibit bioluminescence. The vouchered material in lane 2 of Figure 1 was collected in July 2003 in Tennessee; Reference 2 material was collected from a cultivated field in August 2005 in North Carolina. Since both suppliers confirmed that these materials were not treated with a pesticide, it is our conjecture that these materials were infected with a pest such as a fungus or larvae, and its presence is detected by the observation of unique compounds detected at R_f ≅ 0.23 and 0.27. Of serious concern, both contaminants were readily detected with TLC-bioluminescence, indicating that they may be biologically active.

The results of this study show that A. racemosa root/rhizome and A. podocarpa root/rhizome display unique profiles by TLC and TLC-bioluminescence. Additionally, a potential marker compound that could be used to distinquish A. pachypoda from A. racemosa root/rhizome was detected in both analyzed A. pachypoda samples. These characteristic fingerprints or bands could be used to differentiate between the root/rhizome extracts of the 3 species, thereby indicating that TLC and TLC-bioluminescence are effective methods to support A. racemosa root/rhizome identity.

These methods are also effective at discovering adulteration of A. racemosa root/rhizome not only with A. podocarpa, A. pachypoda, and various leaf parts, but also with other contaminants. The unknown adulterants detected in this study were discovered by using 2 additional detection methods: UV 365 nm and TLC-bioluminescence that are not described in the standard TLC analysis of triterpene glycosides in A. racemosa (14). Additionally, the 2 A. racemosa samples containing the unknown adulterants were examined under blind conditions using a standard HPLC-ELSD method for triterpene glycosides determination, but the contaminants were not detected. These results emphasize the effectiveness of using TLC with a variety of detection methods. It also suggests that the addition of bioluminescence as a detection medium could be an effective tool in screening complex mixtures such as biomass for biologically active contaminants. The Bioluminex assay measures biological activity that may be indicative of toxicity, whereas a chemical test, which may be more sensitive, does not provide information about the potential liability of an adulterant.