Working with the natural complexity: selection and characterization of black cohosh root extract for use in toxicology testing

Suramya Waidyanatha; Tim Cristy; Jessica Pierfelice; Jon C Andre; Brian Burback; Esra Mutlu

doi:10.1016/j.fct.2021.112769

. Author manuscript; available in PMC: 2023 Feb 1.

Published in final edited form as: Food Chem Toxicol. 2021 Dec 18;160:112769. doi: 10.1016/j.fct.2021.112769

Working with the natural complexity: selection and characterization of black cohosh root extract for use in toxicology testing

Suramya Waidyanatha ^1,^*, Tim Cristy ², Jessica Pierfelice ², Jon C Andre ², Brian Burback ², Esra Mutlu ¹

PMCID: PMC9063431 NIHMSID: NIHMS1773688 PMID: 34929352

Abstract

Black cohosh (Actaea racemosa L.) is a botanical supplement marketed to women of all ages. Due to paucity of data to assess the safe use, the National Toxicology Program (NTP) is evaluating the toxicity of black cohosh. The use of an authentic, quality material is imperative to generate robust data. Because botanical materials are complex mixtures with variable composition, the selection of a material is challenging. We describe selection and phytochemical characterization of an unformulated black cohosh root extract (i.e., an extract that serves as source material for a formulated product) to be used in the NTP assessments. A material was selected using a combination of non-targeted and targeted chemical analyses, including confirmation of authenticity, absence of contaminants and adulterants, and similarity to a popular black cohosh product used by consumers. Thirty-nine constituents covering three major classes, triterpene glycosides, phenolic acids, and alkaloids were identified. Among constituents quantified, triterpene glycosides made up approximately 4.7% (w/w) with total constituents quantified making up 5.8% (w/w) of the extract. Non-targeted chemical analysis followed by chemometric analysis of various materials sold as black cohosh, and reference materials for black cohosh and other Actaea species further confirmed the suitability of the selected extract for use.

Keywords: Black cohosh, botanical dietary supplements, phytochemical composition, constituents

1. Introduction

Botanical dietary supplements are products made from plant materials and are intended to supplement the diet. Approximately 18% of the U.S population is taking these nonvitamin, nonmineral supplements for their purported therapeutic properties (Clarke et al., 2015). The industry continues to grow worldwide with the consumer spending in the U.S. alone over $11 billion in 2020, an increase of 17.3% from 2019 (Smith, 2021). The regulatory framework for botanical dietary supplements by the U.S. Food and Drug Administration (FDA) is via Dietary Supplement Health and Education Act (DSHEA) (https://ods.od.nih.gov/About/DSHEA_Wording.aspx). According to DSHEA, manufacturers are not required to notify the FDA prior to marketing products containing dietary ingredients or combination of ingredients in commerce prior to 1994. They are required to notify the FDA prior to marketing of only new dietary ingredients. In either case, pre-market safety data are not required, and the safety of these products is dependent upon the FDA enforcing current good manufacturing practices, history of use, and any reports of adverse events (https://www.fda.gov/food/dietarysupplements/default.htm).

Botanical dietary supplements are used by people of all ages and safe use of these products is of utmost importance. However, there is a paucity of safety data, likely stemming from the relatively lenient regulatory framework in the U.S. and the availability of large number of botanical products on the market. To fill this data gap, the National Toxicology Program (NTP) has an ongoing botanical testing program incorporating novel approaches and methods of chemistry and toxicology, including testing in rodent models (https://ntp.niehs.nih.gov/results/areas/botanical/index.html) (Collins et al., 2020; Hubbard et al., 2019; Rider et al., 2018; Ryan et al., 2019; Waidyanatha et al., 2020).

Because botanical dietary supplements are made from plant materials, they are complex mixtures with large number of constituents making up the phytochemical composition, and in many cases, with limited knowledge of the biologically active constituent(s). As highlighted in various reports, there are many sources of variation in the composition of a given botanical-from growing conditions to manufacturing processes, and to economically motivated adulteration (Ryan et al., 2019; Shipkowski et al., 2018; Waidyanatha et al., 2018). Additionally, there are large number of product types (e.g., extracts, tinctures, tablets) on the market for a given botanical material, either alone or in combination with other botanicals. For example, the National Institute of Health’s Dietary Supplement Label Database (DSLD) lists over 27,000 products in the botanical ingredient category (NIH, 2021c). Thus, there are numerous challenges in assessing the safety of these products.

We have previously used the concept of sufficient similarity to circumvent some these challenges with botanicas (Catlin et al., 2018; Collins et al., 2020; Ryan et al., 2019; Waidyanatha et al., 2020). The term sufficient similarity, or the read across approach for complex mixtures, refers to a determination, using a combination of analytical and statistical tools, of whether nominally related mixtures (i.e., black cohosh extracts) are similar enough in composition such that the toxicity data generated for one mixture can be used to determine the risk associated with the other similar materials. For example, prior to toxicity testing, this approach can be used to identify a material for testing that resembles either a known high-quality material or a reference material, as has been demonstrated with Echinacea purpurea (Waidyanatha et al., 2020). Post toxicity testing, a sufficient similarity assessment can be used to compare a toxicologically well-characterized product to other nominally related products on the market such that the findings of the tested product can be extrapolated to similar products on the market to determine their safety, as has been demonstrated with Ginkgo biloba extract (Collins et al., 2020, 2021; NTP, 2013).

Black cohosh (accepted Latin name Actaea racemosa L., previously known as Cimicifuga racemosa L.) is a perennial plant native to North America and belongs to the family Ranunculaceae (buttercup family). It is commonly known as snakeroot (Foster, 1999; Gardner, 2013). Black cohosh is a widely used botanical dietary supplement marketed to women to alleviate a variety of purported gynecological ailments (AHP, 2002; Blumenthal, 2003; Dugoua et al., 2006; Foster, 1999; NIH, 2021a) and was the 17^th top-selling botanical in mainstream outlets (e.g., grocery and drug stores) in 2020 (Smith, 2021). Although the consumer spending of black cohosh decreased approximately 12% from 2019, its sales surpassed $24 million in 2020 (Smith, 2021). Black cohosh preparations are made from its root and rhizomes and are available on the market in a variety of forms such as powdered materials, liquid extracts, and dried extracts in pill form (https://ods.od.nih.gov/factsheets/BlackCohosh-HealthProfessional). Over 100 black cohosh products are listed in the DSLD (NIH, 2021c). Commercial pharmaceutical formulations of black cohosh, such as Remifemin®, has been sold in Germany since 1956. Black cohosh products are also marketed in combination with other botanicals such as red clover, soy isoflavones, St. John’s wort, chasteberry, and dong quai, for a wide array of purported health benefits (Consumerlab, 2020; Healthline, 2020). Recommended daily dose of black cohosh vary from 20–120 mg, although products containing as high as 550 mg can be found on the market (NIH, 2021c). Additionally, consumers tend to use more than the daily recommended dose due to the perception that natural, plant-based products, are generally safe.

There are other North American species of Actaea such as A. americana, A. arizonica, A. cordifolia, A. elata, A. laciniata, and A. pachypoda, A. pordocarpa, and A. rubra and numerous Asian species including A. heracleifolia, A. dahurica (syn. Cimifuga dahurica or Chinese cohosh), and A. cimicifuga (syn. Cimifuga foetida), A. rubra (red cohosh), and A. podocarpa (yellow cohosh) used for their purported anti-inflammatory, analgesic and antipyretic effects, although none has been listed for treatment of gynecological symptoms (Ma et al., 2011). Hence, adulteration of black cohosh products on the market with other Actaea species either due to economical motivation or misidentification is common (Foster, 2013; Gafner, 2016; Harnly et al., 2016; Jiang et al., 2011; Verbitski et al., 2008). Some case reports of liver toxicity following consumption of products labelled as black cohosh have been suggested as due to potential adulteration with other Actaea species (Gafner, 2016; Mahady et al., 2008; NIH, 2021b; Painter, 2010).

Due to widespread use, potentially by women of all ages, and the paucity of data to assess the safety, the NTP selected black cohosh root extract for toxicity testing in in vitro assays and in vivo in rodent models (black cohosh testing status). The use of a representative and quality material, devoid of adulteration from other Actaea species is imperative to generate robust data that can be compared widely across studies and used to evaluate risk of black cohosh to humans. We describe selection and characterization of an unformulated material (i.e., a bulk extract that serves as source material for a formulated product) to be used in toxicology testing. Additionally, a combination of non-targeted and targeted analytical approaches combined with chemometric analysis, and the concept of sufficient similarity was used to compare and contrast the selected lot (i.e. the NTP test lot) with other cohosh materials procured over multiple years to further demonstrate the authenticity and quality of the NTP test lot. Th material procured included multiple commercially available unformulated materials, reference materials of various cohosh species, and formulated black cohosh products representing what humans are consuming (e.g., Remifemin®).

2. Materials and methods

2.1. Materials

Materials sold as black cohosh (A. racemosa) including unformulated materials (i.e., bulk extracts that serve as source material for formulated products), formulated products (i.e., commercially available black cohosh tablets, capsules etc.), and reference materials (e.g., black cohosh root extract reference material (XRM) were procured. Commercially available vouchered botanical reference materials (VBRM) for other cohoshes such as Cimicifuga dahurica (Chinese cohosh), A. rubra (red cohosh), and A. podocarpa (yellow cohosh) were also procured. The samples procured include a total of 17 unformulated materials from 7 suppliers (suppliers 2–8, Table 1), 13 finished products from 11 suppliers (suppliers 9–19, Table 1), and 4 commercially available reference materials, including those for other cohoshes from one supplier (supplier 1, Table 1). Potential black cohosh constituents were procured from several sources (Table 2). Samples were stored at either ambient temperature or −20 °C and standards were stored at −20 °C. Deionized water used was ASTM Type 1. All other chemicals and reagents were obtained from commercial sources.

Table 1.

Cohosh samples procured

Sample description^a	Sample ID^b	Type of Sample^c	Supplier ID
Black cohosh root XRM^d	1	Reference material	1
Chinese cohosh root VBRM^d	2	Reference material	1
Red cohosh root VBRM^d	3	Reference material	1
Yellow cohosh root VBRM^d	4	Reference material	1
Black cohosh powder extract	5	Unformulated material	2
Black cohosh powder extract	6	Unformulated material	2
Black cohosh extract 2.5%	7	Unformulated material	3
Black Cohosh Root Powder	8	Unformulated material	4
Black cohosh extract 2.5%	9	Unformulated material	5
Black cohosh extract 2.5%	10	Unformulated material	3
Black cohosh extract 2.5%	11	Unformulated material	3
Black cohosh extract 2.5%	12	Unformulated material	3
Black cohosh 2.5%	13	Unformulated material	6
Black cohosh 2.5%	14	Unformulated material	7
Black cohosh P.E. 2.5%	15	Unformulated material	8
Black cohosh P.E.	16	Unformulated material	6
Black cohosh extract 2.5%	17	Unformulated material	4
Black cohosh extract 2.5%	18	Unformulated material	4
Black cohosh extract 2.5%	19	Unformulated material	4
Black cohosh extract 2.5%	20	Unformulated material	3
Black cohosh extract 2.5%	21	Unformulated material	3
Black cohosh extract (Remifemin^®) (20 mg/tablet)	22	Formulated product	9
Black cohosh standardized extract (40 mg/capsule)	23	Formulated product	10
Black cohosh standardized root extract (40 mg/capsule)	24	Formulated product	11
Black cohosh root extract (400 mg/capsule)	25	Formulated product	12
Black cohosh root extract (20 mg/tablet)	26	Formulated product	13
Black cohosh root extract (40 mg/capsule)	27	Formulated product	14
Standardized black cohosh root extract (200 mg root extract + 200 mg raw root powder + 200 mg soy isoflavone concentrate/capsule)	28	Formulated product	15
Black cohosh root extract (100 mg root + 80 mg root extract/capsule)	29	Formulated product	16
Black cohosh standardized root extract (40 mg extract/softgel)	30	Formulated product	17
Black cohosh standardized root extract (40 mg standardized root extract/softgel)	31	Formulated product	18
Black cohosh One A Day Menopause Health (10 mg/capsule)	32	Formulated product	19
Black cohosh extract (Remifemin^®) (21 mg/tablet)	33	Formulated product	9
Black cohosh extract (Remifemin^®) (21 mg/tablet)	34	Formulated product	9

Open in a new tab

According to Certificate of Analysis or package labelling.

Sample identification code used in the manuscript.

Reference material indicates a standardized material; unfinished material indicates a bulk extract that serve as a source material for formulated product; formulated product indicates a commercially available product such as capsules, tablets, etc. that humans are consuming.

XRM, certified root extract reference material; VBRM, vouchered botanical reference material.

All reference materials were obtained from ChromaDex, Irvine, CA.

Table 2.

Cohosh constituents procured^a

Constituent	Sample ID	Lot Number	Vendor	Vendor Purity
Actein	A	01-01355-101	ChromaDex, Irvine, CA	99.9%
Actein	A	F0H264	Sigma-AldrichA1, St. Louis, MO	100.0%
27-Deoxyactein (23-epi-26-deoxyactein)	B	02-04130-103	ChromaDex, Irvine, CA	79.3%
27-Deoxyactein (23-epi-26-deoxyactein)	B	00004130-514	ChromaDex, Irvine, CA	95.3%
Cimicifugin	C	03649-104	ChromaDex, Irvine, CA	96.0%
Cimifugin	D	00003639-624	ChromaDex, Irvine, CA	96.4%
Ferulic Acid	E	357835/1	Fluka, Charlotte, NC	99.7%
Ferulic Acid	E	F0J193	Sigma-Aldrich, St. Louis, MO	100.0%
Isoferulic acid	F	01-09251-710	ChromaDex, Irvine, CA	99.4%
Isoferulic acid	F	00009251-001	ChromaDex, Irvine, CA	99.8%
Allocryptopine	G	-	Sigma-Aldrich, St. Louis, MO	NA
Caffeic acid	H	41K1147	Sigma-Aldrich, St. Louis, MO	99.4%
Caffeic acid	H	F0M198	Sigma-Aldrich, St. Louis, MO	100.0%
Cimicifugoside	I	01-03648-912	ChromaDex, Irvine, CA	95.6%
Cimicifugoside H-1	J	01-03646-101	ChromaDex, Irvine, CA	98.5%
Cimicifugoside H-1	J	PA VI-117.1	Planta Analytica, Danbury, CT	99.0 %
Cimiracemoside A	K	01-03647-101	ChromaDex, Irvine, CA	99.9*
Cimiracemoside C	L	00003644-202	ChromaDex, Irvine, CA	98.9%
26-Deoxycimicifugoside	M	FD656011550	Carbosynth, Compton, Berkshire, United Kingdom	99.97%
Formononetin	N	425615/1	Fluka, Charlotte, NC	100.0%
Formononetin	N	BCBS0873V	Sigma-Aldrich, St. Louis, MO	98.6%
Kaempferol	O	R015X0	Sigma-Aldrich, St. Louis, MO	100.0%
Magnoflorine	P	SLBM9622V	Sigma-Aldrich, St. Louis, MO	100.0%
Phellodendrine	Q	FP738501601	Carbosynth, Compton, Berkshire, United Kingdom	98.9%
Prim-O-glucosylcimifugin	R	00016221-511	ChromaDex, Irvine, CA	85.7%
Protocatechuic Acid	S	HWI01604	Sigma-Aldrich, St. Louis, MO	100.0%
Salsolinol hydrobromide	T	036M4730V	Sigma-Aldrich, St. Louis, MO	99.4%

Open in a new tab

For some standards multiple lots or procured from multiple sources were used over the duration of the study.

2.2. High performance liquid chromatography-evaporative light scattering detection (HPLC-ELSD)

Initially, 3 unformulated materials of black cohosh root extract (Table 1, samples 6, 8, and 9) from 3 suppliers, 3 finished products (Table 1, samples 32–34) in tablet form from 2 suppliers and potential black cohosh standards were procured and profiled using HPLC-ELSD. Unformulated materials were prepared at 100 mg/mL in acetonitrile:water (1:1, v/v). Finished products were prepared at approximately 7 (sample 32) and 30 (samples 33–34) mg/mL. Standard solutions were prepared in acetonitrile at 0.5 or 1 mg/mL. All samples and standards were analyzed using Waters (Milford, MA) high performance liquid chromatograph coupled to an evaporative light scattering detector. A Prodigy C18 column (250 × 4.6 mm, 5 μm, Phenomenex, Torrance, CA) was used with mobile phases A (70:30 1% aqueous formic acid:acetonitrile) and B (40:60 1% aqueous formic acid:acetonitrile). A flow rate of 1 mL/min was run with the following linear gradient (%B): 0–33 in 30 min; 33–100 in 30 min.

2.3. High performance thin layer chromatography (HPTLC) and DNA barcoding of sample 5

An unformulated material of black cohosh root extract (sample 5) was procured from a supplier identified following review of HPLC-ELSD data as a candidate bulk material. To determine the authenticity of sample 5, we identified three certified commercial laboratories to conduct HPTLC analysis and two certified commercial laboratories to conduct DNA barcoding. A third commercial certified laboratory for DNA barcoding could not be identified at the time. Although there is conflicting evidence on the utility of DNA barcoding in authentication of botanical extracts, we undertook the work to assess its applicability to black cohosh extract sample 5. HPTLC analysis was conducted by Alkemist Labs (Lab 1, Costa Mesa, CA), US Botanical Safety Laboratory (USBSL, Lab 2, Candler, NC), and Labs-Mart (Lab 3, Farmington Hills, MI). DNA barcoding was conducted by Molecular Epidemiology Inc. (Lab A, Lake Forest Park, WA) and NSF Health Sciences/AuthenTechnologies (Lab B, Richmond, CA). We selected multiple laboratories for each technique to build the confidence in the data generated. Sample preparation and analysis provided by each laboratory are presented in Supplemental Tables 1 and 2 for HPTLC analysis and DNA barcoding, respectively.

2.4. Composition and contaminant analysis of sample 5

The following analyses were conducted to determine the general composition of sample 5: Fourier-Transform infrared (FTIR) spectroscopy; elemental analysis by proton-induced X-ray emission (PIXE); weight loss by drying, moisture content, total inorganic content (determined as ash), nutritional analysis (fat, carbohydrate, protein content, total dietary fiber, sugars); unextractable and extractable fractions. In some cases, analysis was performed more than once and/or by multiple laboratories. Additionally, the presence/absence of contaminants such as heavy metals (antimony, arsenic, cadmium, lead, and mercury), pesticides (a panel of 310 compounds), mycotoxins (aflatoxins B1, B2, G1, and G2, ochratoxin, and zearalenone) were also determined. A summary of analyses and methods used is given in Supplemental Table 3.

2.5. Characterization of black cohosh extract sample 5 by mass spectrometry

To identify constituents and determine corresponding concentrations in sample 5, mass spectrometry methods were utilized as described below.

2.5.1. Identification of constituents in sample 5 by mass spectrometry

Phytochemical constituents in sample 5 were identified by liquid chromatography (LC)-mass spectrometry (MS) using either a quadrupole-time of flight-mass spectrometry (QTOF-MS) or quadrupole ion-trap-mass spectrometry (QTrap-MS).

For the QTrap-MS analysis, sample 5 was prepared in an extraction solution (methanol:water, 80:20, v/v) at approximately 40 mg/mL by brief vortex mixing followed by sonication for 30 min, followed by rotation overnight. Sample was centrifuged at 1000 x g for 5 min, and the supernatant was collected for Qtrap-MS analysis using an Agilent 1210 liquid chromatograph (Agilent, Palo Alto, CA) coupled to Sciex API-4000 QTrap mass spectrometer (Toronto, Canada). Chromatography was performed on an Aqua column (5 μm, 250 mm x 4.6 mm; Phenomenex, Torrance, CA). Mobile phase A (0.1% aqueous formic acid) and B (0.1% aqueous formic acid in acetonitrile) were run with a linear gradient from (%B) 5–15 in 15 min with 5 min hold, 15–30 in 15 min, 30–40 in 15 min, 40– 50 in 45 min, and 50–95 in 5 min at a flow rate of 1 mL/min. The electrospray ion source was operated in positive mode with a source temperature of 450 °C and spray voltage of 5500 V. The mass spectrometer was operated in Enhanced MS mode with dynamic ion trap fill time up to 1000 msec, 1000 Da/sec scan rate, a scan range of m/z 50 – 800, and Enhanced Product Ion scans of peaks with sufficient signal.

For QTOF-MS analysis, sample 5 was prepared in an extraction solution (methanol:water, 80:20, v/v) at approximately 40 mg/mL by brief vortex mixing followed by sonication for 30 min. Sample was centrifuged at 1100 x g for 5 min, and the supernatant was collected for QTOF-MS analysis using a Waters Acquity I-Class liquid chromatograph coupled to an AB Sciex Triple 5600 time of flight mass spectrometer. Chromatography was performed on a HyperClone C18 column (3 μm, 125 mm x 4 mm; Phenomenex, Torrance, CA). Mobile phases A (0.1% aqueous formic acid) and B (0.1% formic acid in acetonitrile) were run with a linear gradient from (%B) 5–50 in 40 min, 50–60 in 5 min, 60–80 in 10 min, 80–100 in 5 min, at a flow rate of 0.5 mL/min. The Turbo Spray ion source was operated in positive mode with a source temperature of 500 °C and spray voltage of 5500 V. The mass spectrometer was scanned from m/z 66–800 with an independent data acquisition range of m/z 40–800. Potential black cohosh constituent standards (Table 2) were prepared in same extraction solvent as that used to prepare samples and analyzed under same conditions as sample 5.

The primary method of identification of peaks for both analyses was based on methods commonly used for this field. Specifically, the following approach was used: 1) comparison of retention times and product ion (MS/MS) spectra between peaks in the sample and available constituent standards and/or 2) based on matching of exact mass of the peak in the MS1 spectrum to the calculated exact mass of the proposed constituent’s molecular formula within 0.006 Da and the fragments observed in the MS2 spectrum of the peak to the predicted fragmentation of the proposed constituent’s structure using Sciex Peakview software. When predictions were used, any proposed constituent where a peak was found in the MS1 data within 0.006 Da of its theoretical [M+H]⁺ and at least 80% of the intensity in the MS2 spectrum matched the predicted fragments for its structure was considered positively identified. Any proposed constituent where a peak was found in the MS1 data within 0.006 Da its theoretical [M+H]⁺ but less than 80% of the intensity in the MS2 spectrum matched the predicted fragments for its structure or if the MS1 peak was outside the set 0.006 Da limit but less than 0.05 Da was considered tentatively identified.

2.5.2. Quantitation of constituents in sample 5 by mass spectrometry

Based on the findings of mass spectrometry analysis and commercial availability of standards, 12 constituents were selected for quantitation by standard addition method (Meija et al., 2014) followed by LC with tandem mass spectrometry (MS/MS). Briefly, the method involved spiking of the black cohosh extract sample with standards of analytes to be quantified at approximately 0 (unspiked), 1x, 2x, and 4x of the expected concentration based on preliminary analysis. Using the analyte response data, a linear regression was generated for each analyte which was subsequently used to calculate the concentration of analytes in unspiked sample. Individual standards of actein, allocryptopine, caffeic acid, cimifugin, cimicifugoside H-1, cimiracemoside C, 27-deoxyactein, 26-deoxycimicifugoside, ferulic acid, isoferulic acid, magnoflorine, and prim-O-glucosylcimifugin were made in solvent (methanol:water:formic acid, 80:20:0.1) at 10000, 7500, 100, 1000, 3000, 100, 1000, 1940, 10000, 1000, 1500, and 100 μg/mL, respectively. For compounds needing a lower concentration for the assay, a combined standard solution was made in the same solvent by diluting the stocks made above. The combined solution consisted of cimifugin, ferulic acid, isoferulic acid, allocryptopine, caffeic acid, 26-deoxycimicifugoside, magnoflorine, and prim-o-glucosylcimifugin at final concentration of 2.5, 50, 450, 2.5, 75, 75, 225, and 2.5 g/mL, respectively.

Four sets of black cohosh extract (200 mg) samples were prepared in duplicate. The standard solutions prepared above (or solvent alone for unspiked sample, 0X) were added to each tube along with the extraction solvent (methanol:water:formic acid, 80:20:0.1) as given in Supplemental Table 4. Samples were vortex mixed for 5 min and then allowed to sit for 30 min. Following a brief mixing, samples were sonicated for 30 min with vortex mixing briefly each 10 min. The extraction was continued by end-over-end rotation overnight. Samples were centrifuged at 1900 x g for 5 min and the supernatants were transferred to 5-mL volumetric flasks. Pellets were extracted once with an additional 0.5 mL of solvent and supernatants were combined. The flasks were brought to volume with the solvent and mixed well. An aliquot from each sample collected for analysis.

All samples were analyzed by LC-MS/MS using a Shimadzu Prominence (Kyoto, Japan) liquid chromatograph coupled to a Sciex 4000 triple quadrupole mass spectrometer (Toronto, Ontario, Canada). A Phenomenex Aqua C18 column (250 × 4.6 mm, 5 μm, Torrance, CA) was used with mobile phases A (0.1% aqueous formic acid) and B (1% formic acid in acetonitrile) run at a flow rate of 1 mL/min. The following linear gradient was used (% B): 5–15 in 15 min and hold for 5 min, 15–30 in 15 min, 30–40 in 15 min, 40–50 in 45 min, 50–95 in 5 min and hold for 5 min. The Turboionspray ion source temperature was set at of 500 °C and was operated in positive ion mode for all except for caffeic acid, with a spray voltage of 5500 V. For caffeic acid, the ion source was operated in negative ion mode with a spray voltage of −4500 V. The transitions monitored and retention time for each analyte is shown in Supplemental Table 4.

A linear regression equation was used to relate the detector response (y) of each constituent to the nominal concentration spiked in sample (Supplemental Table 4). The concentration of each analyte in sample was calculated using its individual response, the regression equation, and any dilution factors, when applicable. The calculated concentration, dilution volume, and the initial weight of black cohosh extract was used to estimate the weight percent of each constituent in the black cohosh root extract sample 5.

2.6. Storage stability of bulk black cohosh extract sample 5

The stability of bulk black cohosh extract sample 5 overtime during storage and at various temperatures was assessed, including an accelerated stability study at 60 °C. Aliquots of black cohosh extract were stored in sealed amber glass vials at 60, 5, −20 °C, and at ambient temperature for up to 14 d. The appearance of sample was noted before and after storage. At the end of storage period, triplicate aliquots (100 mg) from each vial were weighed into 10 mL volumetric flasks, dissolved in, and diluted to, volume with extraction solvent (formic acid:water:acetonitrile 5:495:500 mL). A 1 mL aliquot of the supernatant was diluted with extraction solvent in 10 mL volumetric flasks. Samples were analyzed by HPLC with ultraviolet detection (UV) using a Waters Alliance HPLC (Milford, MA) with UV detection at 317 nm. A Hypersil Phenyl column (250 × 4.6 mm, 5 μm, Lake Forrest, CA) was used with mobile phases A (90:10 1% aqueous formic acid:acetonitrile) and B (40:60 1% aqueous formic acid:acetonitrile). A flow rate of 1 mL/min was run with the following linear gradient (%B): 0–60 in 30 min; 60–100 in 30 min.

Longer term stability was evaluated for black cohosh sample 5 by comparing chromatographic profiles of a sample taken shortly after receipt and stored at −20 °C to the bulk material that had been stored at ambient temperature for 13 years. Two different analytical methods were used, one optimized for the phenolic acid content and one optimized for the triterpene glycosides representing two major classes of constituents present in black cohosh. For the phenolic acid analysis, a 20 mg aliquot of each sample was extracted using sonication for 10 min with 5 mL of 70:30 methanol:water (v/v). After centrifugation, the supernatants were analyzed using an Agilent 1100 HPLC (Santa Clara, CA) with UV detection at 320 nm. An Aqua C18 column (250 × 4.6 mm, 5 μm, Phenomenex, Torrance, CA) was used with mobile phases A (10% aqueous formic acid) and B (acetonitrile). A flow rate of 1 mL/min was run with the following linear gradient (%B): 5–15 in 15 min and hold for 5 min; 15–50 in 30 min; 50–100 in 5 min. For the triterpene glycosides analysis, a 200 mg aliquot of each sample was extracted using sonication for 10 min with 5 mL of 80:20 methanol:water (v/v). After centrifugation, the supernatants were analyzed by HPLC coupled with charged aerosol detection (CAD) using an Agilent 1100 HPLC (Santa Clara, CA) and a Thermo Fisher Scientific Dionex Corona Plus charged aerosol detector (Chelmsford, MA). A Hyperclone C18 column (120 × 4 mm, 5 μm, Phenomenex, Torrance, CA) was used with mobile phases A (water) and B (acetonitrile). A flow rate of 1 mL/min was run with the following linear gradient (%B): 5–50 in 40 min; 50–60 in 15 min; 60–80 in 10 min; 80–100, 15 min.

2.7. Non-targeted chemical profiling of cohosh samples by high performance liquid chromatography-charged aerosol detection (HPLC-CAD)

Additional samples covering unformulated materials and formulated products sold as black cohosh, and commercially available cohosh reference materials were procured over multiple years as listed in Table 1. Samples 1–31 were included in the analysis, however, finished products 32–34 were not available to be included in this analysis. Samples were prepared at 40 mg/mL except for sample 26 where the final concentration was prepared at 20 mg/mL, inadvertently. For unformulated and reference materials, approximately 400 mg aliquots (exact weight recorded) were extracted with 10 mL of extraction solvent (methanol:water, 80:20, v/v). For finished products, the label claims were used determine the starting amount of material to achieve 400 mg black cohosh extract. Tablets (ground using a mortar and pestle) and the contents in capsules were extracted with 10 mL of extraction solvent. To soft gel supplements, 2 mL water was added and heated in a water bath at 40 °C to dissolve capsules, followed by addition of 8 mL of methanol. All samples were extracted by brief vortex mixing followed by sonication for 30 min, centrifuging at 1100 x g for ~5 min, and supernatants were collected for analysis. Individual standards of potential black cohosh constituents (Table 2) were prepared at a target concentration of 100 μg/mL in extraction solution. A combined standard solution of actein, 27-deoxyactein, cimifugin, ferulic acid, and isoferulic acid was prepared at a target concentration of 12.5 μg/mL in extraction solution.

Samples and standards were analyzed by HPLC-CAD using an Agilent 1100 (Santa Clara, CA) high performance liquid chromatograph coupled to a Thermo Fisher Scientific Dionex Corona Veo charged aerosol detector (Waltham, MA). A Phenomenex Aqua C18 column (250 × 4.6 mm, 5 μm, Torrance, CA) was used with mobile phases A (10% aqueous formic acid) and B (acetonitrile). A flow rate of 1 mL/min was run with the following linear gradient (%B): 5–15 in 15 min, hold for 5 min; 15–30 in 15 min, 30–40 in 15 min, 40–50 in 45 min, 50–95 in 10 min, hold for 5 min. Total run time was 120 min including the time to equilibrate back to the initial conditions. Standards were interspersed throughout the analytical runs to check for retention time variability and allow for peak alignments.

The HPLC-CAD chromatograms were imported as time and response/intensity pairs into SpecAlign (v2.4.1, University of Oxford, England) for preprocessing. Chromatograms were rescaled to positive to shift the negative baseline to zero, when needed. To remove the components eluting in the void volume and those during column flushing, the chromatograms were cropped from 5 to 97 min. The data were binned into 0.2-sec bins to reduce the effective data rate from 25 points per sec to 5 points per sec. Using all individual chromatograms, an average chromatogram was generated and was used as the reference to align the peaks of individual chromatograms. Standards of known constituents in black cohosh were used to constrain the movement of alignment software. The data from aligned chromatograms were exported as comma separated variable (CSV) files containing time and response pairs for statistical analyses. The evaluate data function within SpecAlign was used to generate a heatmap of correlations between the different samples.

The CSV files from SpecAlign were imported into the Eigenvector Research Solo chemometrics software (v8.5.1, Manson, WA for principal component analysis (PCA). In this chemometrics software, a dataset is transformed into a set of orthogonal variables that account for the greatest degree of variability in the data.

3. Results

3.1. Chemical profiling using HPLC-ELSD

Initially, three unformulated materials from three suppliers (samples 6, 8, and 9) were procured and screened along with formulated products including two different lots of Remifemin®, a popular black cohosh supplement (33 and 34), to identify a potential manufacturer to procure an unformulated bulk material from. The resulting chromatograms are shown in Figures 1A, 1B, and 1C for standards, unformulated materials, and formulated products, respectively. The key for sample and constituent identification is given in Tables 1 and 2, respectively. Of the formulated products, the two lots of Remifemin® (samples 33 and 34) had similar chemical profiles (Figure 1C), while the profile of the third formulated product (sample 32) lacked the general chromatographic pattern of Remifemin®. In general, the chromatographic profiles of three unformulated materials (samples 6, 8, 9) were similar to each other (Figure 1B). The pattern of peaks in unformulated materials (Figure 1B) were similar to Remifemin® (samples 33 and 34) based on a visual inspection, except that one of the unformulated materials (sample 6) showed higher intensity for multiple peaks. Because samples of formulated products were not prepared based on the label claims of black cohosh, rather the tablet weights, a direct comparison of intensity of peaks between unformulated materials and formulated products couldn’t be done. Using retention time matching with the standards (Figure 1A), several constituents were tentatively identified as marked in Figures 1B and 1C (Table 2, analytes A, B, C, E, F, H, K). Based on the similarity of the chemical profile of sample 6 to popular finished product Remifemin®, and the presence of numerous constituents known to be present in black cohosh extract at a higher intensity than the other two unformulated materials, subsequent to this analysis, a bulk unformulated black cohosh root extract material (sample 5) was obtained from the same manufacturer as sample 6, to be considered as a candidate bulk lot to be used in NTP research and testing activities. Analyses to confirm the quality and authenticity of sample 5 and resulting findings are given in subsequent sections.

Figure 1. — High performance liquid chromatography-evaporative light scattering detection (HPLC-ELSD) chromatograms: A) commercially available standards of potential black cohosh constituents B) unformulated materials procured as black cohosh root extract, and C) formulated black cohosh products. The key for sample and constituent identification can be found in Tables 1 and 2, respectively.

3.2. Authentication of sample 5 by HPTLC and DNA barcoding

HPTLC analysis and DNA barcoding were conducted to determine the authenticity of unformulated bulk black cohosh root extract material (sample 5). HPTLC analysis was conducted by three laboratories (Lab1, 2, and 3) and DNA barcoding was conducted by two laboratories (Lab A and B) to increase the confidence in the data generated.

In HPTLC analysis, the detection at ~365 nm suggested compounds with double bonds (e.g., flavonoids, polyphenols and phenolcarboxylic acids) which are common plant constituents. Plates were also treated with ethanolic sulfuric acid, in some cases, to enhance the detection under visible light. Data from Lab 1 is shown in Figure 2A without (top panel) and with (bottom panel) ethanolic sulfuric acid treatment. In the top panel, sample 5 (lanes 4 and 5) is similar to that of the testing laboratory’s black cohosh rhizome (lane 12) and root (lane 13) reference standards with respect to the number, retention factor (R_f)_, and intensity of analytes. The similarity of intensity of bands between the testing laboratory’s reference materials and the sample 5 suggests similar concentration of constituents between sample 5 and the reference materials. A major band in sample 5 at ~Rf 0.35 matches with the primary band in actein standard (lanes 1 and 17, Figure 2A, bottom panel). The standard of cimifugin (lane 18, Figure 2A, top panel) has two bands at ~Rf 0.22 and 0.5. The band at ~Rf 0.22 is specific for cohosh species A. cimicifuga, A. podocarpa (yellow cohosh), and A. heracleifolia as evident by corresponding standard reference materials in lanes 14, 15, and 16, respectively; the absence of strong band for cimifugin ~ Rf 0.2 in sample 5 (lanes 4 and 5) suggests absence of adulteration from other cohosh species (Figure 2A, top panel). In addition, there is a band at a ~Rf 0.04 in the treated plate under visible light (Figure 2A, bottom panel) that is characteristic of A. cimicifuga and A. heracleifolia (lanes 14 and 16) that is absent in sample 5 (lanes 4 and 5) further supports the absence of these two species in sample 5. The presence of actein and absence of or presence of low levels of cimifugin are characteristic of authentic black cohosh, A. racemosa. The testing laboratory concluded that the HPTLC profile of sample 5 was characteristic of black cohosh.

Figure 2B shows the data from Lab 2 without (top panel) and with (bottom panel) ethanolic sulfuric acid treatment. Based on Figure 2B, top and bottom panels, the pattern of the major bands in sample 5 (lanes 3 and 4) are similar to those in the testing laboratory’s black cohosh root reference materials (lanes 2, 5, and 6), although the intensities in the lanes of reference materials are lower than that of sample 5. However, the testing laboratory’s botanical reference material (lane 2) presents a band at ~ Rf 0.35 that is not apparent in either the sample 5 (lanes 3 and 4) or testing laboratory’s vouchered reference materials (lanes 5 and 6). Although an explanation was not provided by the testing laboratory for this observed difference in the pattern between the laboratory’s reference materials, the difference could be due to processing differences between materials. When treated with ethanolic sulfuric acid (Figure 2B, bottom panel) and viewed under white light, the faint band at ~Rf 0.4 likely corresponds to actein (lane 1) in both sample 5 and reference materials. Although a standard of cimifugin was not run, the testing laboratory concluded “based on information obtained from European Pharmacopoeia 8.1 monograph and absence of cimifugin in sample as well as reference materials, the sample 5 is characteristic of A. racemosa.”

Data from Lab 3 is shown in Figure 2C without ethanolic sulfuric acid treatment. The chromatographic profile of sample 5 (right lane) is the same as the testing laboratory’s A. racemosa root standard (left lane), matching all major and minor bands. The intensity of the bands is similar, and there were no apparent differences in the pattern of bands. The testing laboratory concluded that sample 5 is A. racemosa root extract. Taken collectively, the analyses by the three laboratories confirm that sample 5 is characteristic of black cohosh and unlikely adulterated with other Actaea species.

DNA barcoding was conducted by two laboratories (Supplemental Figure 1). The protocols used by the two laboratories differed substantially. Lab A provided little information on the methods used (Supplemental Table 2). Lab A extracted the DNA, screened biomarker genes for black cohosh, and concluded that sample 5 is greater than 96% A. racemosa. The analysis also identified that sample 5 contained material from Saccharomycetaceae (<3%, identified as yeast), Verbenaceae (<1%), and Balsaminaceae families (<0.5%) (Supplemental Figure 1A). Further details on the analyses or data were not available from the laboratory.

Lab B used specific primers to amplify specific gene regions using polymerase chain reaction (PCR). The DNA was sequenced using a Next Generation Sequencer and the results compared to a proprietary database of sequences. It is unclear whether the primers used were specific to A. racemosa or applicable to other members of the genus. Lab B concluded that sample 5 contained A. racemosa and no other Actaea species were identified (Supplemental Figure 1B). Lab B was able to distinguish against several Actaea species (A. cimicifuga, A. podocarpa, A. asiatica, A. brachycarpa, A. dahurica, A. frigida, A. heracleifolia, A. japonica, A. mairei, A. purpurea, A. simplex, A. vaginata, and A. yunnanensis).

3.3. Description, composition, and contaminant analysis of black cohosh root extract sample 5

Based on the above analyses, the unformulated black cohosh root extract sample 5 was confirmed as authentic black cohosh root extract and hence was selected as the candidate lot for NTP testing activities. The bulk material (CASRN 84776–26-1; lot 3012782), which has been identified as sample 5, was from PlusPharma Incorporated (Vista, California) and was manufactured by Frutarom Switzerland Ltd (Wädenswil, Switzerland). According to the manufacturer’s certificate of analysis (CoA), the extraction medium was 50% ethanol and the drug:extract ratio was 5.5:1. The following additional information was also provided in the manufacturer’s CoA: weight loss on drying, 4.0%; bulk density, 0.31 g/mL; triterpene glycosides by HPLC, 1.9% (w/w); total triterpene glycosides by HPLC using Institute for Nutraceutical Advancement method 7.8%; aerobic bacteria, <100 CFU/g; fungi, <10 CFU/g; E coli, Staphylococcus aureus and Salmonella were not detected. In addition, the CoA states that the material contains the followings: 13–21% (w/w) (syn povidone, polyvinylpyrrolidone, polyvidone) which is a binder used in pharmaceuticals. The bulk material was received in six containers with approximately 20 kg per container. Upon receipt, material was homogenized, and representative archival samples were taken and stored at −20 °C. Initially, the bulk material was stored at ambient temperature and was subsequently transferred to −20 °C storage. Visual inspection of an aliquot of the bulk material revealed that the unformulated black cohosh bulk material was a light brown powder and was consistent with the description provided in the manufacturer’s CoA as a brown powder.

Elemental analysis of sample 5 by PIXE showed that the material composed mainly of carbon (~ 69%), oxygen (~15%) and hydrogen (~ 12%); significant inorganic components detected were potassium (~ 2.5%), magnesium (~ 0.3%), and calcium (~0.3%). Weight loss on drying, moisture content, inorganic analysis (determined as ash), and nutrient analysis were conducted to estimate the composition of sample 5 and data are given in Table 3. The moisture content ranged from ~4 to 8%, depending on the laboratory conducting the analysis. The weight loss by drying was ~8% and was slightly higher than the value reported in the manufacturer’s CoA (4%); the similarity of the value to moisture content suggests absence of significant/volatiles in the sample. The total inorganic content, determined as ash, was ~6%, and was higher than that estimated by PIXE analysis (total ~3% based on 3 components as stated above) suggesting potential presence of other inorganic and/or non-combustible material in sample, potentially coming from excipients added during manufacturing of the dried bulk extract. The levels of fat, carbohydrate, protein, and total dietary fiber determined were <2%, ~22%, ~27%, and ~20%, respectively (Table 3); however, these values were not provided in the manufacturer’s CoA for direct comparison.

Table 3.

Composition and contaminant analysis of candidate black cohosh root extract sample 5.

Endpoint	Value
General composition
Moisture (%)^a	3.8, 5.6, 7.9
Weight loss by drying (%)	7.9
Ash (%)	6.12, 5.8
Fat (%)^a	1.8, 0.373
Carbohydrate (%)	22.4
Protein (%)^a	26.7, 25.8
Total dietary fiber (%)	20.2
Sugars (%)^b	6.2
Unextractable fraction (%)^c	26
Extractable fraction (%)^d	74
Contaminants
E. coli (CFU/g)	< 10
Staphylococcus enterotoxins (per 25 g)	Negative
Mycotoxins (ppb)
Aflatoxins (B1, B2, G1, G2)^e	-
Ochratoxin	< 5.00
Zearalenone	< 50.0
Heavy metals (ppb)
Antimony	< 10.0
Arsenic	276
Cadmium	16.3
Lead	164
Mercury	< 10.0
Pesticide screen (ppm) ^f
Chlorpropham	0.012
Phenylphenol, 2-	0.016

Open in a new tab

Analyses were done either more than one time and/or multiple laboratories.

Based on glucose, fructose, sucrose. Maltose and lactose were below limit of detection of 0.05.

Unextractable fraction was determined as the pellet weight following extraction of sample 5 with methanol:water (80:20), followed by hexane and drying to a constant weight. There was no difference in weight after methanol:water and after hexane extraction.

Extractable fraction was estimated as (100-unextractable fraction). This estimate likely includes moisture, volatiles, simple carbohydrates and some lipids.

Could not be determined due to matrix interferences.

Panel included 310 pesticides. Typical limits of detection (LOD) were between 0.01 to 0.05 ppm. Only the analytes above LOD are shown.

The unextractable fraction (using 80:20 methanol:water, Supplemental Table 3) was estimated at 26% based on the pellet weight, and by subtraction the extractable fraction was determined as ~74% (100–26%) (Table 3). This estimate of extractable fraction likely included moisture/volatiles, simple carbohydrates, and some lipids. (Table 3). Black cohosh root extract was also analyzed to determine presence/absence of contaminants. Of the heavy metals analyzed, only arsenic, cadmium, and lead were detected in parts per billion concentrations in the sample but were below the threshold limits for botanical dietary supplements of 5, 0.3, and 10 ppm, respectively (Table 3) (WHO, 2007; AHPA, 2009). Of >300 pesticides screened, only chlorpropham and 2-phenylphenol were detectable (0.012–0.016 ppm) (Table 3) above the limits of quantitation of assays, although, all were below the residue tolerance levels values of U.S. Environmental Protection Agency.

The FTIR spectrum of black cohosh root extract sample is shown in Figure 3. No literature reference spectrum was available to compare with the sample spectrum and hence the major bands in the spectrum were compared to those expected for black cohosh constituents such as phenolic acids, actein, caffeic and ferulic acids, sugars, terpenes, and glycosides. Absorption bands at 3345, 1288, and 1049 cm⁻¹ are representative of O-H stretch, C-O-H bend, and C-O-H stretch suggestive of presence of phenolic compounds. The data demonstrate that FTIR is useful in identifying functional groups in a complex mixture pointing to the presence of certain classes of compounds. Although the technique alone is not useful to characterize such mixtures, in combination with other techniques, FTIR data may be of value.

3.4. Identification of constituents in black cohosh root extract sample 5 by mass spectrometry

The material was analyzed by LC-QTOF- and -QTrap-MS and corresponding total ion chromatograms are shown in Figure 4 demonstrating presence of numerous peaks. A combination of commercially available standards, spectral data collected during analysis, and data base matching was used to guide the identification of as many constituents as possible. Compounds were identified as definitive when either a retention time and spectrum matched with an authentic standard or when MS1 data within 0.006 Da of its theoretical [M+H]+ value and >80% of total intensity of the MS2 spectrum matched to the software-predicted spectrum. Compounds were identified as tentative when MS1 data within 0.006 Da its theoretical [M+H]+ but <80% of the intensity in the MS2 spectrum matched the predicted fragments for its structure or if the MS1 peak was outside the set 0.006 Da limit.

Constituents definitively identified using authentic standards and/or spectral matching are given in Table 4. Twenty-five constituents belonging to the three major classes, triterpene glycosides (e.g., actein, 27-deoxyactein, cimicifugic acids, cimicifugosides), phenolic acids (e.g., caffeic acid, ferulic acid, and isoferrulic acids), and alkamides (e.g., dopargine, cimipronidine) were definitively identified. Fourteen other constituents were identified, including numerous isomers of cimiracemocides, and were classified as tentative as outlined in Table 5. Corresponding structures of compounds definitively or tentatively identified belonging to the three main classes, triterpene glycosides, phenolic acids, and alkaloids, are given in Figures 5A, 5B, and 5C, respectively, with (*) denoting constituents classified as definitively identified, and (**) denoting those constituents classified as tentatively identified.

Table 4.

Constituents identified in black cohosh root extract sample 5 by liquid chromatography coupled with quadrupole-time of flight- (QTOF) and quadrupole-ion trap (QTrap)-mass spectrometry.

Constituent	CAS	Molecular formula	Molecular weight	Monoisotopic mass (M+H)	MS	Found m/z	Mass Difference^a	Mode of identification^b,c
Actein	18642-44-9	C₃₇H₅₆O₁₁	676.8	677.390	QTrap	677.36	0.03	Standard
27-Deoxyactein (23-epi-26-deoxyactein)	264624-38-6	C₃₇H₅₆O₁₀	660.8	661.395	QTOF^f	661.391	0.004	Standard
26-Deoxycimicifugoside	214146-75-5	C₃₇H₅₄O₁₀	658.8	659.379	QTOF	659.380	0.001	Standard
Cimiracemoside C	256925-92-5	C₃₅H₅₆O₉	620.8	621.400	QTOF^f	621.395	0.005	Standard
Cimicifugoside H-1	163046-73-9	C₃₅H₅₂O₉	616.8	617.369	QTOF	617.365	0.004	Standard
25-O-Ethylcimigenol 3-O-β-D-xylopyranoside	914086-57-0	C₃₇H₆₀O₉	648.9	649.432	QTOF	649.427	0.005	Spectrum
Caffeic acid	331-39-5	C₉H₈O₄	180.2	181.050	QTOF^f	181.049	0.001	Standard
Ferulic acid	1135-24-6	C₁₀H₁₀O₄	194.2	195.066	QTOF	195.065	0.001	Standard
Isoferulic acid	537-73-5	C₁₀H₁₀O₄	194.2	195.066	QTOF^f	195.065	0.001	Standard
Cimicifugic acid D	219986-51-3	C₂₀H₁₈O₁₀	418.3	419.098	QTOF	419.096	0.002	Spectrum
Cimicifugic acid E^d	219986-67-1	C₂₁H₂₀O₁₀	432.4	433.113	QTOF	433.112	0.001	Spectrum
Cimicifugic acid F^d	220618-91-7	C₂₁H₂₀O₁₀	432.4	433.113	QTOF	433.112	0.001	Spectrum
Cimicifugic acid G	None	C₂₂H₂₂O₁₁	462.4	463.124	QTOF	463.122	0.002	Spectrum
Cimiracemate A^d	478294-16-5	C₁₉H₁₈O₇	358.3	359.113	QTOF	359.111	0.002	Spectrum
Cimiracemate B^d	478294-17-6	C₁₉H₁₈O₇	358.3	359.113	QTOF	359.111	0.002	Spectrum
Protocatechuic acid	99-50-3	C₇H₆O₄	154.1	155.034	QTOF	155.032	0.002	Standard
Dopargine	1111269-01-2	C₁₃H₂₀N₄O₂	264.3	265.166	QTOF	265.165	0.001	Spectrum
Cimipronidine	865266-71-3	C₇H₁₃N₃O₂	171.2	172.109	QTOF	172.108	0.001	Spectrum
Cimitrypazepine	1422514-69-9	C₁₂H₁₄N₂O	202.3	203.118	QTOF	203.118	0.000	Spectrum
Cyclocimipronidine	None	C₇H₁₁N₃O	153.2	154.098	QTOF	154.097	0.001	Spectrum
Laurolitsine	5890-18-6	C₁₈H₁₉NO₄	313.3	314.139	QTOF	314.138	0.001	Spectrum
Phellodendrine	6873-13-8	C₂₀H₂₄NO₄	342.4	342.171^e	QTOF	342.168	0.003	Standard
Salsolinol	27740-96-1	C₁₀H₁₃NO₂	179.2	180.102	QTOF	180.102	0.000	Standard
Cimifugin	37921-38-3	C₁₆H₁₈O₆	306.3	307.118	QTOF	307.116	0.002	Spectrum
Prim-O-glucosylcimifugin	80681-45-4	C₂₂H₂₈O₁₁	468.4	469.171	QTOF	469.167	0.004	Standard

Open in a new tab

Typical mass accuracy is 0.006 Da for the QTOF and 0.1 Da for the QTrap.

Standard=Based on retention time and spectrum comparison with an analytical standard.

Spectrum=based on automatic software assignment of fragments seen in the MS2 spectrum to predicted fragmentation of the structure (PeakView, V. 1.2.0.3). Those with >80% of total intensity assignable to the structure and M+H match within instrument accuracy were considered positively identified.

Without standards to confirm retention times, specific isomer(s) cannot be confirmed.

Since the parent is charged, M⁺ was detected and reported here rather than M+H.

Identity also confirmed with the QTrap, but only the higher resolution QTOF results shown here.

Table 5.

Constituents tentatively identified in black cohosh root extract sample 5 by liquid chromatography coupled with quadrupole-time of flight- (QTOF)-mass spectrometry.

Constituent	CAS	Molecular formula	Molecular weight	Monoisotopic mass (M+H)	Found m/z	Mass Difference^a	Mode of identification^b
Cimiracemoside A^d	264875-61-8	C₃₇H₅₆O₁₁	676.8	677.390	677.386	0.004	Spectrum. A single peak was detected at the M+H of these three isomers. 72% of total intensity in the MS2 spectrum was assignable to the structure of Cimiracemoside A, 73% to Cimiracemoside F, and 64% to Cimiracemoside G.
Cimiracemoside F^c	264875-61-8	C₃₇H₅₆O₁₁	676.8	677.390	677.386	0.004
Cimiracemoside G^c	289632-43-5	C₃₇H₅₆O₁₁	676.8	677.390	677.386	0.004
Cimiracemoside H^c	290821-41-9	C₃₇H₅₈O₁₁	678.8	679.406	679.402	0.004	Spectrum. There were 3 peaks detected at the M+H for these three isomers. The peaks were labeled in order of best match. The 21.0 min peak had 72% of total intensity assignable to Cimiracemoside H. The 35.1 min peak had 43% of total intensity assignable to Cimiracemoside D, and the 24.9-min peak had 32% of total intensity assignable to Cimiracemoside E.
Cimiracemoside D^c	290821-39-5	C₃₇H₅₈O₁₁	678.8	679.406	679.401	0.005
Cimiracemoside E^c	290821-40-8	C₃₇H₅₈O₁₁	678.8	679.406	679.402	0.004
Cimiracemoside B	290821-38-4	C₃₅H₅₆O₁₀	636.8	637.395	637.391	0.004	Spectrum. A peak was detected at the M+H and 66% of total intensity in the MS2 spectrum was assignable.
Divaricatic acid	491-62-3	C₂₁H₂₄O₇	388.4	389.160	389.161	0.001	Spectrum. A peak was detected at the M+H, but only 2% of total intensity was assignable.
Norsalsolinol	34827-33-3	C₉H₁₁NO₂	165.2	166.087	166.086	0.001	Spectrum. A peak was detected at the M+H and 77% of total intensity in the MS2 spectrum was assignable.
Allocryptopine	485-91-6	C₂₁H₂₃NO₅	369.4	370.165	370.164	0.001	Standard. A M+H peak was observed at the retention time of the standard, but intensity was too low to generate a good matching MS/MS spectrum.
Norcoclaurine	5843-65-2	C₁₆H₁₇NO₃	271.3	272.129	272.126	0.003	Spectrum. A peak was detected at the M+H and 50% of total intensity in the MS2 spectrum was assignable.
Laurotetanine	128-76-7	C₁₉H₂₁NO₄	327.4	328.155	328.111	0.044	Spectrum. Mass error outside of instrument accuracy. Only 35% of total intensity assignable.
N-Methyl cyclocimipronidine	None	C₈H₁₃N₃O	167.2	168.114	168.102	0.012	Spectrum. Mass error outside of instrument accuracy. Intensity too low to generate MS/MS spectrum.
3-Hydroxytyrosol-3-O-glucoside	142542-89-0	C₁₄H₂₀O₈	316.3	317.124	317.121	0.003	Spectrum. A peak was detected at the M+H and 73% of total intensity in the MS2 spectrum was assignable.

Open in a new tab

Typical mass accuracy of the QTOF is 0.006 Da.

Standard=based on comparison with an analytical standard. Spectrum=based on automatic software assignment of fragments seen in the MS2 spectrum to the predicted fragmentation of the structure (PeakView, V. 1.2.0.3).

Without standards to confirm retention times, specific isomer(s) present in sample cannot be confirmed.

graphic file with name nihms-1773688-f0006.jpg — Figure 5. Structures of constituents identified in black cohosh root extract sample 5: A) triterpene glycosides B) phenolic acids C) alkamides. *Denotes definitively identified constituents. ** Denotes constituents classified as tentatively identified.

graphic file with name nihms-1773688-f0007.jpg — Figure 5. Structures of constituents identified in black cohosh root extract sample 5: A) triterpene glycosides B) phenolic acids C) alkamides. *Denotes definitively identified constituents. ** Denotes constituents classified as tentatively identified.

3.5. Quantitation of constituents in black cohosh root extract sample 5 mass spectrometry

Constituents identified above, for which commercial standards were available, were quantified using standard addition followed by LC-MS/MS analysis. Twelve constituents were included as given in Supplemental Table 4. LC-MS/MS chromatograms of sample 5 and constituent standards are shown in Supplemental Figure 2. Linear regression equations for standards are given in Supplemental Table 4; coefficients of determination for all analytes were ≥ 0.96 demonstrating the suitability of the method for analyte quantitation. The amount of each analyte estimated using this method is presented as percent of analyte in black cohosh root extract per weight basis (% w/w) (Table 6). Of the 12 constituents included in the analysis, only 8 were above the limit of quantitation (LOQ) of 0.01% with 4 (allocryptonine, cimifugin, cimicifugoside H-1, and prim-O-glucosylcimifugin) below the LOQ. Of terpene glycosides, cimiracemoside C and 27-deoxyactein were the highest constituent present in the extract with values of 2.13 and 2.09%, respectively, with actein present at 0.47%. Of the phenolic acids, isoferrulic acid was the highest (0.70%), followed by caffeic acid (0.28%); the level of ferulic acid was low at 0.04%. The levels of the alkaloid, magnoflorine, was low at 0.01%. Based on this analysis, the extract contains 4.77% triterpene glycosides, 1.02%, phenolic acids, and 0.01% alkaloids bringing the total estimated constituents to 5.8% (w/w) (Table 6).

Table 6.

Concentration of constituents determined in black cohosh root extract sample 5 by standard addition followed by liquid chromatograph coupled with tandem mass spectrometry (LC-MS/MS) analysis^a

Constituent	Percent (w/w)
Actein	0.47
Allocryptonine	<0.01
Caffeic acid	0.28
Cimifugin	<0.01
Cimicifugoside H-1	<0.01
Cimiracemodise C	2.13
27-Deoxyactein (23-epi-26-deoxyactein)	2.09
26-Deoxycimicifugoside	0.08
Ferulic acid	0.04
Isoferrulic acid	0.70
Magnoflorine	0.01
Prim-O-glucosylcimifugin	<0.01

Open in a new tab

Values shown are average of 2 replicates. Percent of constituent in extract is estimated as: weight of constituent/weight of extract*100. The lower limit of quantitation for this method is approximately 0.01% (w/w).

3.6. Storage stability of black cohosh extract sample 5

Storage stability of the bulk was assessed in different ways. Aliquots of sample 5 were stored for 14 d under various temperatures. A visual observation showed that the appearance didn’t change except for the aliquots stored at 60 °C, which appeared to be slightly darker than those stored at other storage conditions (data not shown). This suggests potential oxidation of phenolic compounds under elevated temperatures. To assess any changes in the phytochemical profile, these stored aliquots of sample 5 were analyzed using HPLC-UV. Phytochemical profiles of samples stored at ambient temperature, 5 °C and 60 °C were similar to that stored at −20 °C, likely the most stable storage condition, suggesting that chemical profiles were unaffected due to storage conditions (Figure 6A). A similar analysis of samples stored for approximately 13 years at −20 °C and at ambient temperature, using HPLC-UV (Figure 6B) and HPLC-CAD (Figure 6C), showed the similarity of phytochemical profiles of bulk sample at ambient and frozen storage conditions. These data suggest that the bulk material may be stable at ambient, refrigerated, or frozen conditions over long-term storage.

Figure 6. — Phytochemical profile of black cohosh extract sample 5 stored at different conditions. A) High performance liquid chromatography (HPLC)-ultraviolet detection (UV) (317 nm) chromatograms of sample stored for 14 days at −20 °C, 5 °C, ambient temperature, and 60 °C. B) HPLC-UV (317 nm) chromatograms of sample stored for 13 years at −20 °C and ambient temperature C) HPLC-charge aerosol detection (CAD) chromatograms of sample stored for 13 years at −20 °C and ambient temperature.

3.7. Non-targeted chemical profiling using HPLC-CAD of cohosh materials procured overtime and chemometric analysis of data

Additional samples including unformulated black cohosh materials (17 samples), formulated black cohosh products (11 samples), and reference material of black cohosh root and those for other cohoshes (Chinese cohosh (A. dahurica), red cohosh (A. rubra), and yellow cohosh (A. podocarpa)) were procured over years and analyzed by HPLC-CAD to determine the similarity and/or differences of products on the market to that of sample 5, the material selected for the NTP toxicity testing. Chromatograms for commercially available standards of potential black cohosh constituents, cohosh reference materials, unformulated materials procured as black cohosh, and formulated products procured as black cohosh are shown in Figures 7A, 7B, 7C, and 7D, respectively. The key for sample and standard identification can be found in Tables 1 and 2, respectively.

The chromatographic profiles of authentic standards are shown in Figure 7A. Based on a visual examination, the chemical profile pattern of cohosh reference materials was different from each other (Figure 7B). Samples 1, 2, and 3 showed a distinct similar peak patterns at ~ 26 min, ~ 33–37 min, and were different from sample 4. Profiles of all four samples, in general, were different from each other after ~ 38 min. There were similarities and differences in chromatographic profiles of unformulated materials procured as black cohosh (samples 5–21) (Figure 7C). Based on a visual examination, the profiles of samples 5–8 were the most similar to each other with similar peaks or peak patterns spanning through the majority of the chromatogram from about 26 min to 84 min. These samples (samples 5–8) were different in that the distinct peaks at approximately 12, 22, 28, 46, and 55 min observed in samples 10–21 were either low or absent. Some materials had fewer peaks and/or low peak intensities (samples 15–18) while others (samples 10–18, 21) lack the distinct peak pattern between 32–35 min found in samples 5–8, which was also present in black cohosh root XRM (sample 1). Although the chromatographic profile of sample 9 was similar to those of samples 5–8 including a distinct peak pattern in the range 32–35 min, the profile has distinct peaks at approximately 22, 28, 46, and 55 min, which were absent in samples 5–8. As observed with unformulated materials, there were variability among the chromatographic profiles of finished products (samples 22–31) (Figure 7D). Chromatographic profiles of samples 22–25 and 29 were similar, while profiles of others (samples 26–28, 30, and 31) lacked the distinct peaks at approximately 26 and 33–36 min, although the peak pattern after approximately 60 min were similar between all samples. Taken collectively, the data demonstrated that chromatographic profiles of bulk extracts materials and finished products procured as black cohosh were highly variable.

HPLC-CAD profiles of samples were used to generate a heatmap of correlations between different samples using the evaluate data function within SpecAlign (Figure 8 and Supplemental Table 5). The correlation heatmap was sorted so that the black cohosh root extract XRM (sample 1, Table 1) was at the upper left and the rest of the samples were in order based on similarity with the XRM. The correlations were formatted as percentages (Supplemental Table 5) and colored based on degree of correlation (green, ≥80%; yellow, 60–79%; red, <60%). The popular black cohosh finished product, Remifemin® (sample 22) was highly correlated with the black cohosh XRM (sample 1) (97%). From the unformulated materials, Samples 5 (the NTP test lot) and 6, prepared by the same manufacturer, showed the best correlation with the XRM (97%) and with Remifemin® (sample 22) (96%). There were only two other unformulated materials with >90% correlation with the XRM (sample 7, 94%; sample 8, 90%). All the other unformulated materials (9–21) had <70% correlation with the black cohosh root XRM. Many of the unformulated materials that were poorly correlated with the black cohosh root XRM were highly correlated with each other (Figure 8, lower right). However, they were not well correlated with reference materials of the other cohoshes (samples 2,3,4), but likely contain the same unknown species or mixture of species. Of the formulated products sold as black cohosh, four (22–25) showed 90% percent correlation with the XRM, two (26,29) showed 60%. Others (27, 28, 30, and 31) were poorly correlated (<60%) with the XRM. In the chromatograms of two products which were sold as soft gels, high gelatin content obscured one region of the chromatogram which may have resulted in the poor match with the XRM (samples 30, 31).

Figure 8. — Heatmap of cohosh samples. SpecAlign software was used to generate a heatmap of correlations between the different samples. Black cohosh root extract reference material (XRM) (Sample 1, Table 1) is at the upper left and the rest of the samples are in order based on similarity with the XRM. The correlations were formatted as percentages and color represents the degree of correlation (green ≥80%; yellow 60–79%; red <60%). The key for sample numbers can be found in Tables 1.

HPLC-CAD data was also evaluated using chemometrics software to determine similarities and differences and the resulting PCA plots are shown in Figure 9. For PCA, two and three principal components (PC) captured ~66% and ~78% of the variance, respectively, and hence three PCs were used to report the data. Corresponding 2-dimensional scores plots of PC1 vs PC2 and PC1 vs PC3 are shown in Figures 9A and 9B, respectively, with the color of each sample indicating the year the sample was procured. Based on PC1 and PC2 in Figure 9A, cohosh reference standards procured in year 2014 clustered differently from one another demonstrating chemical composition differences amongst the cohoshes as expected; specifically, black cohosh root XRM (sample 1) grouped in a different quadrant from the Chinese, red, and yellow cohosh reference materials (samples 2, 3, and 4). The unformulated materials procured during 2002–2004 (5, 6, 8, 9) grouped in the same quadrant or closer to black cohosh XRM (sample 1). In contrast, most of the unformulated materials procured after 2010 (7, 10–21), except sample 7, clustered differently from those procured during 2002–2004 (5, 6, 8, 9) and from black cohosh root XRM (sample 1). Some materials (samples 12, 16–18) clustered in the same quadrant as Chinese, red, or yellow cohosh reference materials while others (samples 10, 11, 13–15, 20, 21) clustered in a different quadrant from all cohosh reference materials. Sample 7 appears to be black cohosh based on its proximity to black cohosh root XRM. The pattern of distribution and general conclusions based on PC1 and PC3 are similar to PC1 and PC2 (Figure 9B) with majority of the unformulated products procured after 2010 clustering differently from black cohosh root XRM. Of the unformulated products, sample 5, the selected lot for NTP research, is closet to black cohosh XRM (sample 1) (Figures 9A and 9B). Of the finished black cohosh products, samples 22–26 clustered in the same quadrant as the black cohosh root XRM (sample 1); while samples 30 and 31 clustered with reference materials of other cohoshes (samples 2–4), samples 27–29 clustered differently from all cohosh reference materials (Figure 9A). These data suggest that majority of the unformulated black cohosh material procured after 2010 are either not black cohosh or adulterated with other cohoshes. Of the finished products, some had characteristics of black cohosh based on comparison to XRM while others were either not black cohosh or adulterated black cohosh.

Figure 9. — Principal component (PC) analysis 2-dimensional score plots A) PC1 versus PC2 B) PC1 versus PC3 of data from high performance liquid chromatography-charged aerosol detection (HPLC-CAD) analysis of cohosh samples. Peaks were aligned using SpecAlign software (University of Oxford, England) and the plot was generated using Solo v8.5.1 software (Eigenvector Research, Manson, WA). Each point represents a single sample. Samples include cohosh reference materials (1–4), unformulated materials (5–21), and formulated products (22–31). For sample 22 and 26, half the amount of the product was used inadvertently. The color code represents the year the samples were procured as indicated in the color bar to the right of the graph. The key for sample numbers can be found in Table 1.

Figure 10 shows a direct comparison of chromatographic profiles of sample 5 (the NTP test lot), cohosh root extract reference materials, and popular black cohosh product, Remifemin®. As evident from visual examination of chromatograms, the chemical profile of sample 5 is very similar to that of black cohosh XRM (sample 1) and finished product Remifemin® (sample 22), and different from those of Chinese, red, and yellow cohosh (samples 2–4).

4. Discussion

Selection of a quality material has been a primary challenge in preclinical safety testing of botanicals and currently, there aren’t any established guidelines or strategies followed by researchers. This deficiency has been recognized by the research community and recommendations have been published by researchers of the Natural Product Drug Interaction Center (Kellogg et al., 2019) and by us recently, using Echinacea purpurea as a case study (Waidyanatha et al., 2020). Additionally, the National Center for Complementary and Integrative Health (NCCIH) recently put forth the Natural Product Integrity Policy for NCCIH-funded researchers on the characterization of botanical products to ensure reproducibility of botanical research (Kuszak et al., 2016). Overall, these publications highlight several key factors that one must take into consideration when using a material in preclinical safety testing. First, it should be an unformulated material suitable to be formulated in a vehicle to be administered in animals. Second, it must be authentic (i.e., whether the material belong to the correct plant genus and species), of good quality (free of contaminants and adulterants), and representative of products used by consumers. Third, its phytochemical composition should be well characterized, including the concentration of constituents. Fourth, because the botanical composition can vary due to various factors (Ryan et al., 2019), a quantity sufficient to support all planned studies should be obtained to eliminate use of multiple lots with divergent phytochemical compositions. Lastly, the storage stability of material should be confirmed to ensure that the phytochemical profile doesn’t change during planned use period. Here, we describe the selection and characterization of an unformulated black cohosh root extract lot for NTP toxicity testing activities. The selected lot was also compared to other unformulated materials and formulated products on the market, procured over time.

At the initiation of the project, we procured three unformulated materials from three suppliers with the intention of identifying a suitable supplier to obtain a larger quantity following initial screening. We compared the chemical profiles of the unformulated materials with popular formulated products available at the time on the market, representative of what consumers are using. By comparing the profiles, and similarity to a popular finished product on the market, and presence of constituents known to be present in black cohosh, we selected a supplier and procured a bulk unformulated material (sample 5), to meet the criteria identified based on the initial screening. We recommend, if available, including reference material from the botanical of interest as well as related plant species and potential adulterants in this assessment, as well as increasing the number of samples screened to increase the probability of identifying a quality material. This is especially important for black cohosh due to presence of a variety of other cohosh species and higher probability of adulteration. However, standard reference materials for black cohosh or any other cohoshes were not commercially available to be included in our initial comparison. Even today, reference materials are not available for many botanicals and those that are available for purchase are limited to a few sources such as those sold by the National Institute of Standards and Technology (https://www.nist.gov), the United States Pharmacopeia (USP) (https://www.usp.org), and Chromadex (https://chromadex.com). There is an ongoing effort by the NIH Office of Dietary Supplements funding research through the Analytical Methods and Reference Materials Program (https://ods.od.nih.gov/Research/AMRMProgramOverview.aspx) for the development of suitable materials to be used in research and testing.

It is critical to ensure that the candidate material is authentic (i.e., derived from black cohosh). The appropriateness of the authentication technique(s) depends on the type of product being considered for use. In our investigation with black cohosh extract, traditional authentication methods such as morphological and microscopical characteristics of the material compared to a vouchered specimen by a trained expert (Walker and Applequist, 2012) were precluded due to the highly processed nature of the material. DNA-based authentication methods have been shown to be effective and are becoming popular for authentication of botanicals, especially for unprocessed materials. Authentication of black cohosh and identifying adulterants using DNA barcoding have been reported in the literature (Baker et al., 2012; Harnly et al., 2015; Harnly et al., 2016). For example, one study used PCR amplification and bidirectional sequencing for selected genes to confirm the authenticity of samples. The authors identified all of the black cohosh samples as black cohosh while none of the other Actaea species in the set were falsely identified as black cohosh (Baker et al., 2012). Of the commercially available black cohosh supplements tested, 75% were identified as black cohosh while the remaining 25% had a sequence identical to three Asian species. In another study, DNA barcodes were determined at 2 loci (one nuclear ribosomal and one chloroplast DNA) to identify Actaea species. Authors reported that in general, the DNA sequencing confirmed the species identification with a couple of exceptions due to purity of the sample (e.g., mold growth). Of the commercial black cohosh products tested, four of the seven capsules contained DNA from A. racemosa while none of the liquid supplements were subjected to sequencing with the assumption that they would not contain DNA (Harnly et al., 2016).

In general, challenges in utilizing DNA-based techniques for authentication of botanical extracts have been reported due to DNA quantity and integrity in processed materials (Harnly et al., 2015; Harnly et al., 2016; Moraes et al., 2015; Parveen et al., 2016; Pawar et al., 2017; Raclariu et al., 2018; Ragupathy et al., 2019). Keeping these potential challenges in mind, we did an assessment of authenticity of sample 5 using DNA barcoding using two certified laboratories. Not surprisingly, the two laboratories used different protocols. Lab A provided limited information on the protocol used and when additional details were requested no further details were provided. The laboratory noted that DNA was extracted and screened for biomarker genes. It was not clear whether the laboratory amplified the DNA and used more specific primers or how the DNA sequencing was conducted. Lab B provided their Standard Operating Protocol. Lab B used specific primers and amplified gene regions using PCR, and DNA sequencing was done using the Next Generation Sequencer However, the laboratory didn’t specify whether the primers used were specific to A. racemosa or included all Actaea species. Consequently, the two laboratories provided somewhat divergent results. Although the presence of black cohosh was confirmed, Lab A did not confirm presence/absence of other Actaea species. While confirming the sample to be black cohosh, Lab B also distinguished against several other Actaea species. Although the DNA barcoding technique has been efficiently and successfully used to authenticate unprocessed plant materials, these data further highlight the need to advance the field such as validation and standardization of protocols, utilization of next generation sequencing (Lu et al., 2018), and expansion of databases containing DNA sequences for plant ingredients which will substantially increase our ability to successfully use this technique to confirm the authenticity of processed botanical materials (Moraes et al., 2015). Until such time, we recommend DNA barcoding of highly processed botanical materials be used in conjunction with other chemical analysis methods such as HPTLC, MS, and Nuclear magnetic resonance spectroscopy (NMR) (Collins et al., 2020).

HPTLC has been widely used in recent years for botanical authentication (Reich et al., 2008). We used multiple commercial laboratories to investigate the characteristics of candidate black cohosh extract by HPTLC using the laboratories’ reference materials, including reference materials for other common adulterants. All three laboratories came to the same conclusion regarding the authenticity of sample 5 as black cohosh root, although the thoroughness of the experimental design and the levels of details provided by the laboratories varied. Because HPTLC technique is well-established and widely used, analysis by a single laboratory may be adequate when authentication is needed, especially if other orthogonal techniques are used to arrive at the decision of authenticity, as has been utilized in the current work.

An added complexity in selecting a material for preclinical safety testing stems from adulteration of material. Adulteration of a material can arise from: unintentional adulteration due to misidentification of plant species; economically motivated adulteration, where a less expensive botanical is used in place of authentic more expensive material on the label; adulteration with pharmaceutical drugs or drug analogs to drive the efficacy. If an adulterant is responsible for the observed biology of a given botanical material, then using the data generated to compare across studies of the same botanical will be compromised. This is particularly important for black cohosh for two main reasons: 1) the genus Actaea comprise over 28 species and hence adulteration has been common either due to plant misidentification and/or economic motivation (Foster, 2013; Gafner, 2016; Harnly et al., 2016; Jiang et al., 2011; Verbitski et al., 2008), and 2) case reports of liver toxicity following consumption of products labelled as black cohosh have been linked to potential adulteration with other Actaea species (Gafner, 2016; Mahady et al., 2008; NIH, 2021b; Painter, 2010). The Botanical Adulteration Prevention Program was initiated by three leading nonprofit organizations to educate members of the dietary supplement industry about ingredient and product adulteration and currently is available on the American Botanical Council’s website (https://www.herbalgram.org/resources/botanical-adulterants-prevention-program/). FDA also maintains a database of botanicals adulterated with drugs and actions it has taken to remove those products from the market. Although some laboratories conducting the authentication did include other Actaea species during analysis, it was not comprehensive. Hence, additional analysis (e.g., MS) to confirm the presence of botanical-specific constituents and presence/absence of potential adulterants should be considered prior to selecting material for safety testing. It is also imperative that the candidate material is analyzed to identify potential contaminants, whether coming from the plant material due to growing, harvesting, or storage conditions and/or introduced during processing. In sample 5, these contaminants were either absent, or if present, were below the limits established for botanical materials (AHPA, 2009; WHO, 2007) further strengthening the argument for consideration of sample 5 as a candidate lot.

The extract was also analyzed to determine the general composition. The type of analyses to consider depend on the material (e.g., extract, tincture, powder). Here we determined the moisture content, total inorganic content (determined as ash), fiber, extractable material, and nutritional content (total protein, carbohydrate, and fat) to cover a variety of endpoints. The sample contains ~26% (w/w) unextractable fraction when methanol:water (80:20, v:v) was used for extraction and by subtraction the extractable fraction was estimated at ~74% (w/w). Since sample 5 is an aqueous ethanolic extract, the unextractable fraction likely contains the polymeric binder added when the final extract was produced per manufacturer’s CoA, which was 13–21%. This suggests that the majority of the other plant components including black cohosh constituents, fat, carbohydrates, and proteins are in the extractable fraction (74%). The fat (average ~1%), carbohydrate (~22%), protein (~25%), and moisture and other volatiles (~8%) makes up ~56% of the extract. Thus, all or some portion of the unaccounted extractable fraction of sample 5, which is ~18% may be black cohosh-specific constituents. It is noteworthy that some of these analyses may not have the level of precision required to determine the exact mass balance. Regardless, these analyses contribute to the understanding of general and constituent composition of the extract.

The identification of marker and/or bioactive constituents and establishing their concentrations are the key to ensuring that toxicity and other data across multiple studies can be compared. NMR spectroscopy and MS-based methods have been widely used to characterize marker constituents of botanical materials, although MS-based methods are more widely used due to versatility of available techniques, ability to detect low levels of analytes, and the availability of myriad of spectral libraries to identify constituents. For some botanicals, the known constituents make up a significant fraction of the botanical extract as in the case of Ginkgo biloba extract (>50%) (Collins et al., 2020), making it relatively easy to arrive at the decision on the authenticity and suitability of a given material based on targeted chemical analysis. In contrast, for botanical materials where the known constituents make up a very small fraction of the extract such as Echinacea purpurea (Waidyanatha et al., 2020) makes it difficult to determine the authenticity and suitability of the botanical material based on targeted chemical analysis alone, and hence multiple approaches such as those highlighted here are necessary. Information on marker constituents is available for some botanicals in the form of monographs such as those published by USP or the American Herbal Pharmacopoeia for black cohosh. It is important to note that these resources are not always intended be comprehensive or updated regularly (AHP, 2002; USP, 2015).

Investigations on the phytochemical composition of black cohosh have mostly focused on two major classes of constituents: triterpene glycosides (e.g., actein, 27-deoxyactein (also known as 23-epi-26-deoxyactein), and cimicifugosides) and phenolic acids (e.g., hydroxycinnamic acid derivatives and fukiic acid ester derivatives, fukinolic acids) (summarized in https://ods.od.nih.gov/factsheets/BlackCohosh-HealthProfessional/ (Kruse et al., 1999; Li and Yu, 2006)). In more recent years, a third class of constituents consisting of alkaloids and related compounds have been identified in black cohosh (Nikolic et al., 2015).

There are multiple ways to identify and classify compounds in complex matrices. Four levels of metabolite identification have been proposed in metabolomics research to facilitate replication and comparison of complex metabolomics data sets (Blazenovic et al., 2018; Sumner et al., 2007). In botanical research, a common approach has been to use the authentic standards and/or spectral matching with existing databases to categorize the identity as definitive (or matched) or tentative (Baker and Regg, 2018; Jin, 2018; Sica et al., 2018). We used MS methods with a combination of authentic standards and/or spectral matching with libraries to identify constituents in sample 5. We identified 39 constituents, 25 of which were identified with high confidence and hence were classified as definitive while the identity of the remaining of 14 constituents were classified as tentative due to lower confidence arising from numerous reasons including low level of analyte present to obtain a representative spectrum, the presence of multiple isomers with similar spectra and/or retention time as highlighted in Tables 4 and 5.

Of the constituents identified were compounds belonging to the three main classes known to be present in aqueous ethanolic extracts of black cohosh: triterpenoids (15), phenolic acids (12), and alkaloids (12). Among the members of the triterpenoid class identified in sample 5 were actein, 27-deoxyactein, cimifugin, and several isomers of cimiracemosides and cimiracemates (Figure 5). Due to close structural similarities of isomers and lack of authentic standards, the exact identity of some cimiracemoside isomers could not be discerned. More than 40 compounds belonging to the triterpenoid class have been reported in the literature for black cohosh (Jiang et al., 2011; Li and Yu, 2006; Qiu et al., 2012). Of the phenolic acids identified were the hydroxycinnamic acid derivatives (caffeic acid, ferulic acid, and isoferrulic acid) and their condensation products with glycoloyl phenylpropanoids, commonly known as cimifugic acids (Godecke et al., 2009; Kruse et al., 1999). Several cimifugic acid isomers were identified, although due to lack of authentic standards, the exact identity for some could not be determined. The alkaloid metabolome of black cohosh has been comprehensively investigated in more recent years and numerous alkaloids belonging to guanidino (e.g., cimipronidine), isoquinoline (e.g., dopargine), aporphlne (e.g., magnoflorine), protoberberberlne (e.g., allocryptopin), and indole classes have been isolated and identified when extractions were conducted using large quantities of starting material (summarized in (Nikolic et al., 2015)). In our investigation, the presence of some of these alkaloids were determined in sample 5, although the number identified was limited likely stemming from the low levels present in the extract and use of smaller amounts of starting material than those used in studies summarized in Nikolic et al. (Nikolic et al., 2015). It is noteworthy, that the scope of the current work is not an in-depth investigation of the phytochemical constituents of black cohosh, rather to utilize multiple approaches as described here to determine the quality and authenticity of the material.

For constituents where authentic standards were available, concentrations were quantified using LC-MS/MS, expressed as percent of analyte in extract per weight basis (% w/w). In our extract, only 3 (actein, 27-deoxyactein, and cimiracemoside C) of the 7 triterpene glycosides monitored were above the LOQ of the assay (0.01%) and the total triterpene glycoside content making up the extract, based on these constituents was 4.77%. This level is between the triterpene glycosides measured by HPLC (1.9%, w/w, specific analytes were not mentioned) and total triterpene glycosides measured by HPLC using Institute for Nutraceutical Advancement method (7.8%) provided in the manufacturer’s CoA. Actein and 27-deoxyactein are often used to standardize black cohosh preparations and hence are among the most common triterpenes quantified (USP, 2015). Black cohosh products are generally standardized to triterpene glycosides content and products containing 2.5% are common. Total phenolic acids present in extract (1.02% w/w) was lower than triterpene glycosides. Standards were available for only 2 alkaloids detected and both were at or below the LOQ. Although the alkaloid metabolome in black cohosh is suggested to be numerous, the levels present are known to be low in line with our observation here (Nikolic et al., 2015). Total constituents quantified, based on these selected analytes made up a small fraction of the total extract (5.8%, w/w) and was lower than that predicted from mass balance estimates (~18%). This may be due to the presence of other significant constituents that were not captured in the current analysis and/or imprecision in some of the composition endpoints used in the mass balance. Regardless, our findings on constituents present and concentrations present fall within the range expected for black cohosh extracts.

Data presented so far in this report supports that the bulk sample 5 is an authentic black cohosh material, devoid of contaminants and adulterants from other Actaea species and hence is suitable for use in toxicity testing. Because anticipated studies can span over multiple years, and botanicals are complex mixtures, it is important to use a single product, whenever possible, to ensure reproducibility of data and to facilitate comparison across different studies. Hence, ensuring the stability of the product during use period is of utmost importance. For complex mixtures like botanicals, although not ideal, the stability of the material has been established with regards to marker or biologically active constituent(s), if known (Waidyanatha et al., 2020). We investigated the phytochemical profiles of sample 5 stored at different storage conditions including an accelerated stability assessment at 60 °C. Phytochemical profiles, based on visual examination, showed that there was no difference due to storage conditions or duration, when compared to a sample stored at −20 °C, the best-case storage scenario suggesting that the bulk material may be stable at ambient, refrigerated, or frozen conditions over long-term storage. The bulk material was stored at −20 °C for long term storage. The stability of triterpene glycosides and phenolic acids in black cohosh products at ambient storage has also been demonstrated by other investigators (Jiang et al., 2008).

To generate additional data supporting the quality of the black cohosh material selected for NTP research and testing, and to monitor the variability of products on the market over time, we procured and analyzed additional samples, including unformulated black cohosh materials, formulated black cohosh products, and reference materials including those for other cohoshes. Chemometrics analysis clearly demonstrated that majority of the unformulated black cohosh material procured after 2010 were either not black cohosh or were adulterated with other Actaea species. Of the finished products, some have characteristics of black cohosh based on comparison to black cohosh root extract reference material while others are either not black cohosh or adulterated black cohosh. Adulteration of black cohosh with other Actea has been demonstrated by various investigators and been also implicated as the cause for observed hepatotoxicity following consumption of products labelled as black cohosh (Mahady et al., 2008; NIH, 2021b). The bulk extract sample 5 selected for NTP testing activities, stored under different conditions, and analyzed over time, consistently showed high correlation with the black cohosh reference material and the popular finished product Remifemin®, but lacked similarity to commercially available reference materials of other Actaea species, further supporting the authenticity, quality, and stability of the material. As has been demonstrated here, fit-for-purpose chemical analysis and data processing methods can be used to select material for pre-clinical safety assessment of complex botanical mixtures.

Conclusion

Selection of a quality authentic material is imperative to generate toxicity and other data that can be compared across studies to assess the safety of black cohosh. We used chemical fingerprinting techniques to identify a candidate material, the authenticity of which was subsequently confirmed using HPTLC and DNA barcoding techniques. Analysis of the candidate material showed that heavy metals, pesticides, and other contaminants are below the threshold values recommended for botanical materials. Utilizing MS methods, 39 constituents known to be present in black cohosh were identified. Among key constituents quantified, triterpene glycosides, cimiracemodise C, 27-deoxyactein and actein made up approximately 4.7% (w/w) of the extract with total constituents quantified making up 5.8% (w/w) of the black cohosh extract. These data demonstrate the suitability of the material for use in toxicity testing of black cohosh.

Supplementary Material

Supplemental materials

NIHMS1773688-supplement-Supplemental_materials.docx^{(3.8MB, docx)}

Acknowledgements

The authors are grateful to Mr. Brad Collins and Dr. Kristen Ryan for their review of this manuscript. This work was supported by the Intramural Research Program of the NIH, National Institute of Environmental Health Sciences, Intramural Research project ZIC ES103316-05, and performed for the National Toxicology Program, National Institute of Environmental Health Sciences, National Institutes of Health, U.S. Department of Health and Human Services, under contract HHSN273201400027C (Battelle, Columbus, OH).

Footnotes

Conflict of interest

Authors declare no competing interests

References

AHP, 2002. American Herbal Pharmacopeia. Black Cohosh Rhizom, Actaea racemosa L. syn. Cimicifuga racemosa (L.) Nutt.Standards of Analysis, Quality Control, and Therapeutics, in: Upton R (Ed.), AmericanHerbal Pharmacopoeia and Therapeutic Compendium, Santa Cruz, CA, USA. [Google Scholar]
AHPA, 2009. American Herbal Products Association. Heavy Metals: Analysis and Limits in Herbal Dietary Supplements. [Google Scholar]
Baker DA, Stevenson DW, Little DP, 2012. DNA barcode identification of black cohosh herbal dietary supplements. J AOAC Int 95, 1023–1034. [DOI] [PubMed] [Google Scholar]
Baker TR, Regg BT, 2018. A multi-detector chromatographic approach for characterization and quantitation of botanical constituents to enable in silico safety assessments. Anal Bioanal Chem 410, 5143–5154. [DOI] [PMC free article] [PubMed] [Google Scholar]
Blazenovic I, Kind T, Ji J, Fiehn O, 2018. Software Tools and Approaches for Compound Identification of LC-MS/MS Data in Metabolomics. Metabolites 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Blumenthal M, 2003. The ABC Clinical Guide to Health. American Botanical Council; Austin, TX Black Cohosh. 13–20. [Google Scholar]
Catlin NR, Collins BJ, Auerbach SS, Ferguson SS, Harnly JM, Gennings C, Waidyanatha S, Rice GE, Smith-Roe SL, Witt KL, Rider CV, 2018. How similar is similar enough? A sufficient similarity case study with Ginkgo biloba extract. Food Chem Toxicol 118, 328–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
Clarke TC, Black LI, Stussman BJ, Barnes PM, Nahin RL, 2015. Trends in the use of complementary health approaches among adults: United States, 2002–2012. Natl Health Stat Report, 1–16. [PMC free article] [PubMed] [Google Scholar]
Collins BJ, Kerns SP, Aillon K, Mueller G, Rider CV, DeRose EF, London RE, Harnly JM, Waidyanatha S, 2020. Comparison of phytochemical composition of Ginkgo biloba extracts using a combination of non-targeted and targeted analytical approaches. Anal Bioanal Chem 412, 6789–6809. [DOI] [PMC free article] [PubMed] [Google Scholar]
Collins BJ, Kerns SP, Aillon K, Mueller G, Rider CV, DeRose EF, London RE, Harnly JM, Waidyanatha S, 2021. Correction: Comparison of phytochemical composition of Ginkgo biloba extracts using a combination of non-targeted and targeted analytical approaches. http://link.springer.com/article/10.1007/s00216-021-03594-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Consumerlab, 2020. Menopause Supplements Review. Soy and Red Clover Isoflavones, Black Cohosh) and Progesterone Creams. https://www.consumerlab.com/reviews/soy-isoflavones-red-clover-black-cohosh-supplements/phytoestrogens/. Last accessed June 20, 2021.
Dugoua JJ, Seely D, Perri D, Koren G, Mills E, 2006. Safety and efficacy of black cohosh (Cimicifuga racemosa) during pregnancy and lactation. Can J Clin Pharmacol 13, e257–261. [PubMed] [Google Scholar]
Foster S, 1999. Black cohosh: A literature review. HerbalGram 45, 35–50. [Google Scholar]
Foster S, 2013. Exploring the peripatetic maze of black cohosh adulteration. HerbalGram 98, 32–51. [Google Scholar]
Gafner S, 2016. On adulteration of Actaea racemosa. Botanical Adulterants Bulletin.(accessed June 9th, 2021). HerbalGram; 111. [Google Scholar]
Gardner ZM, M., 2013. American Herbal Products Association’s botanical safety handbook., 2nd ed. CRC Press, Boca Raton, FL. [Google Scholar]
Godecke T, Nikolic D, Lankin DC, Chen SN, Powell SL, Dietz B, Bolton JL, van Breemen RB, Farnsworth NR, Pauli GF, 2009. Phytochemistry of cimicifugic acids and associated bases in Cimicifuga racemosa root extracts. Phytochem Anal 20, 120–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
Harnly J, Chen P, Colson K, McCoy J-A, 2015. MS, NMR, and DNA barcoding, complementary methods for identification and authentication of Black Cohosh (Actaea racemosa L.). Planta Medica 81. [DOI] [PMC free article] [PubMed] [Google Scholar]
Harnly J, Chen P, Sun J, Huang H, Colson KL, Yuk J, McCoy JA, Reynaud DT, Harrington PB, Fletcher EJ, 2016. Comparison of Flow Injection MS, NMR, and DNA Sequencing: Methods for Identification and Authentication of Black Cohosh (Actaea racemosa). Planta Med 82, 250–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
Healthline, 2020. Black Cohosh: Benefits, Dosage, Side Effects, and More. https://www.healthline.com/health/food-nutrition/black-cohosh. Last accessed June 20, 2021.
Hubbard TD, Hsieh JH, Rider CV, Sipes NS, Sedykh A, Collins BJ, Auerbach SS, Xia M, Huang R, Walker NJ, DeVito MJ, 2019. Using Tox21 High-Throughput Screening Assays for the Evaluation of Botanical and Dietary Supplements. Appl In Vitro Toxicol 5, 10–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
Jiang B, Lyles JT, Reynertson KA, Kronenberg F, Kennelly EJ, 2008. Stability evaluation of selected polyphenols and triterpene glycosides in black cohosh. J Agric Food Chem 56, 9510–9519. [DOI] [PMC free article] [PubMed] [Google Scholar]
Jiang B, Ma C, Motley T, Kronenberg F, Kennelly EJ, 2011. Phytochemical fingerprinting to thwart black cohosh adulteration: a 15 Actaea species analysis. Phytochem Anal 22, 339–351. [DOI] [PubMed] [Google Scholar]
Jin YM, Y; Xie W; Hou L; Xu H; Zhang K; Zhang L; Du Y, 2018. UHPLC-Q-TOF-MS/MS-oriented characteristic components dataset and multivariate statistical techniques for the holistic quality control of Usnea. RSC Adv. 8, 15487–15500. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kellogg JJ, Paine MF, McCune JS, Oberlies NH, Cech NB, 2019. Selection and characterization of botanical natural products for research studies: a NaPDI center recommended approach. Nat Prod Rep 36, 1196–1221. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kruse SO, Lohning A, Pauli GF, Winterhoff H, Nahrstedt A, 1999. Fukiic and piscidic acid esters from the rhizome of Cimicifuga racemosa and the in vitro estrogenic activity of fukinolic acid. Planta Med 65, 763–764. [DOI] [PubMed] [Google Scholar]
Kuszak AJ, Hopp DC, Williamson JS, Betz JM, Sorkin BC, 2016. Approaches by the US National Institutes of Health to support rigorous scientific research on dietary supplements and natural products. Drug Test Anal 8, 413–417. [DOI] [PMC free article] [PubMed] [Google Scholar]
Li JX, Yu ZY, 2006. Cimicifugae rhizoma: from origins, bioactive constituents to clinical outcomes. Curr Med Chem 13, 2927–2951. [DOI] [PubMed] [Google Scholar]
Lu Z, Rubinsky M, Babajanian S, Zhang Y, Chang P, Swanson G, 2018. Visualization of DNA in highly processed botanical materials. Food Chem 245, 1042–1051. [DOI] [PubMed] [Google Scholar]
Ma C, Kavalier AR, Jiang B, Kennelly EJ, 2011. Metabolic profiling of Actaea species extracts using high performance liquid chromatography coupled with electrospray ionization time-of-flight mass spectrometry. J Chromatogr A 1218, 1461–1476. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mahady GB, Low Dog T, Barrett ML, Chavez ML, Gardiner P, Ko R, Marles RJ, Pellicore LS, Giancaspro GI, Sarma DN, 2008. United States Pharmacopeia review of the black cohosh case reports of hepatotoxicity. Menopause 15, 628–638. [DOI] [PubMed] [Google Scholar]
Moraes DFC, Still DW, Lum MR, Hirsch AM, 2015. DNA-Based Authentication of Botanicals and Plant-Derived Dietary Supplements: Where Have We Been and Where Are We Going? Planta Medica 81, 687–695. [DOI] [PubMed] [Google Scholar]
NIH, 2021a. Black cohosh. https://ods.od.nih.gov/factsheets/BlackCohosh-HealthProfessional/. Last accessed July 22, 2021.
NIH, 2021b. Black cohosh. https://www.ncbi.nlm.nih.gov/books/NBK547990/pdf/Bookshelf_NBK547990.pdf. Last accessed June 20, 2021.
NIH, 2021c. Dietary Supplement Label Database. https://ods.od.nih.gov/Research/Dietary_Supplement_Label_Database.aspx. Last accessed June 9, 2021.
Nikolic D, Lankin DC, Cisowska T, Chen SN, Pauli GF, van Breemen RB, 2015. Nitrogen-Containing Constituents of Black Cohosh: Chemistry, Structure Elucidation, and Biological Activities. Recent Adv Phytochem 45, 31–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
NTP, 2013. NTP Technical Report on the Toxicology and Carcinogenesis Studies of Ginkgo biloba extract (CAS No. 90045–36-6) in F344/N rats and B6C3F1/N mice (gavage studies) NTP TR 578 https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr578_508.pdf. [PubMed]
Painter DP, S.; Murty M, 2010. Black cohosh products and liver toxicity: update. Canadian Adverse Reaction Newsletter 20, 12–15. [Google Scholar]
Parveen I, Gafner S, Techen N, Murch SJ, Khan IA, 2016. DNA Barcoding for the Identification of Botanicals in Herbal Medicine and Dietary Supplements: Strengths and Limitations. Planta Med 82, 1225–1235. [DOI] [PubMed] [Google Scholar]
Pawar RS, Handy SM, Cheng R, Shyong N, Grundel E, 2017. Assessment of the Authenticity of Herbal Dietary Supplements: Comparison of Chemical and DNA Barcoding Methods. Planta Med 83, 921–936. [DOI] [PubMed] [Google Scholar]
Qiu F, Imai A, McAlpine JB, Lankin DC, Burton I, Karakach T, Farnsworth NR, Chen SN, Pauli GF, 2012. Dereplication, residual complexity, and rational naming: the case of the Actaea triterpenes. J Nat Prod 75, 432–443. [DOI] [PMC free article] [PubMed] [Google Scholar]
Raclariu AC, Heinrich M, Ichim MC, de Boer H, 2018. Benefits and Limitations of DNA Barcoding and Metabarcoding in Herbal Product Authentication. Phytochem Anal 29, 123–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ragupathy S, Faller AC, Shanmughanandhan D, Kesanakurti P, Shaanker RU, Ravikanth G, Sathishkumar R, Mathivanan N, Song J, Han J, Newmaster S, 2019. Exploring DNA quantity and quality from raw materials to botanical extracts. Heliyon 5, e01935. [DOI] [PMC free article] [PubMed] [Google Scholar]
Reich E, Schibli A, DeBatt A, 2008. Validation of high-performance thin-layer chromatographic methods for the identification of botanicals in a cGMP environment. J AOAC Int 91, 13–20. [PMC free article] [PubMed] [Google Scholar]
Rider CV, Walker NJ, Waidyanatha S, 2018. Getting to the Root of the Matter: Challenges and Recommendations for Assessing the Safety of Botanical Dietary Supplements. Clin Pharmacol Ther 104, 429–431. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ryan KR, Huang MC, Ferguson SS, Waidyanatha S, Ramaiahgari S, Rice JR, Dunlap PE, Auerbach SS, Mutlu E, Cristy T, Peirfelice J, DeVito MJ, Smith-Roe SL, Rider CV, 2019. Evaluating Sufficient Similarity of Botanical Dietary Supplements: Combining Chemical and In Vitro Biological Data. Toxicol Sci 172, 316–329. [DOI] [PMC free article] [PubMed] [Google Scholar]
Shipkowski KA, Betz JM, Birnbaum LS, Bucher JR, Coates PM, Hopp DC, MacKay D, Oketch-Rabah H, Walker NJ, Welch C, Rider CV, 2018. Naturally complex: Perspectives and challenges associated with Botanical Dietary Supplement Safety assessment. Food Chem Toxicol 118, 963–971. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sica VP, Mahony C, Baker TR, 2018. Multi-Detector Characterization of Grape Seed Extract to Enable in silico Safety Assessment. Front Chem 6, 334. [DOI] [PMC free article] [PubMed] [Google Scholar]
Smith TM, F.; Eckl V; Reynolds CM,, 2021. Herbal Supplement Sales in US Increase by record-breaking 17.3% in 2020. HerbalGram 131, 52–65. [Google Scholar]
Sumner LW, Amberg A, Barrett D, Beale MH, Beger R, Daykin CA, Fan TW, Fiehn O, Goodacre R, Griffin JL, Hankemeier T, Hardy N, Harnly J, Higashi R, Kopka J, Lane AN, Lindon JC, Marriott P, Nicholls AW, Reily MD, Thaden JJ, Viant MR, 2007. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3, 211–221. [DOI] [PMC free article] [PubMed] [Google Scholar]
USP, 2015. United States Pharmacopea Dietary supplements compendium. The United States Pharmacopeal Convention, Twinbrook Parkway, Rockville, MD. [Google Scholar]
Verbitski SM, Gourdin GT, Ikenouye LM, McChesney JD, Hildreth J, 2008. Detection of Actaea racemosa adulteration by thin-layer chromatography and combined thin-layer chromatography-bioluminescence. J AOAC Int 91, 268–275. [PMC free article] [PubMed] [Google Scholar]
Waidyanatha S, Pierfelice J, Cristy T, Mutlu E, Burback B, Rider CV, Ryan K, 2020. A strategy for test article selection and phytochemical characterization of Echinacea purpurea extract for safety testing. Food Chem Toxicol 137, 111125. [DOI] [PMC free article] [PubMed] [Google Scholar]
Waidyanatha S, Ryan K, Roe AL, Jia W, Paine MF, Ferguson S, Gurley BJ, Welch K, Chow MSS, Devito M, Rider C, 2018. Follow that botanical: Challenges and recommendations for assessing absorption, distribution, metabolism and excretion of botanical dietary supplements. Food Chem Toxicol 121, 194–202. [DOI] [PubMed] [Google Scholar]
Walker KM, Applequist WL, 2012. Adulteration of Selected Unprocessed Botanicals in the U.S. Retail Herbal Trade. Economic Botany 66, 321–327. [Google Scholar]
WHO, 2007. WHO guidelines for assessing quality of herbal medicines with reference to contaminants and residues.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental materials

NIHMS1773688-supplement-Supplemental_materials.docx^{(3.8MB, docx)}

[R1] AHP, 2002. American Herbal Pharmacopeia. Black Cohosh Rhizom, Actaea racemosa L. syn. Cimicifuga racemosa (L.) Nutt.Standards of Analysis, Quality Control, and Therapeutics, in: Upton R (Ed.), AmericanHerbal Pharmacopoeia and Therapeutic Compendium, Santa Cruz, CA, USA. [Google Scholar]

[R2] AHPA, 2009. American Herbal Products Association. Heavy Metals: Analysis and Limits in Herbal Dietary Supplements. [Google Scholar]

[R3] Baker DA, Stevenson DW, Little DP, 2012. DNA barcode identification of black cohosh herbal dietary supplements. J AOAC Int 95, 1023–1034. [DOI] [PubMed] [Google Scholar]

[R4] Baker TR, Regg BT, 2018. A multi-detector chromatographic approach for characterization and quantitation of botanical constituents to enable in silico safety assessments. Anal Bioanal Chem 410, 5143–5154. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] Blazenovic I, Kind T, Ji J, Fiehn O, 2018. Software Tools and Approaches for Compound Identification of LC-MS/MS Data in Metabolomics. Metabolites 8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] Blumenthal M, 2003. The ABC Clinical Guide to Health. American Botanical Council; Austin, TX Black Cohosh. 13–20. [Google Scholar]

[R8] Catlin NR, Collins BJ, Auerbach SS, Ferguson SS, Harnly JM, Gennings C, Waidyanatha S, Rice GE, Smith-Roe SL, Witt KL, Rider CV, 2018. How similar is similar enough? A sufficient similarity case study with Ginkgo biloba extract. Food Chem Toxicol 118, 328–339. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] Clarke TC, Black LI, Stussman BJ, Barnes PM, Nahin RL, 2015. Trends in the use of complementary health approaches among adults: United States, 2002–2012. Natl Health Stat Report, 1–16. [PMC free article] [PubMed] [Google Scholar]

[R10] Collins BJ, Kerns SP, Aillon K, Mueller G, Rider CV, DeRose EF, London RE, Harnly JM, Waidyanatha S, 2020. Comparison of phytochemical composition of Ginkgo biloba extracts using a combination of non-targeted and targeted analytical approaches. Anal Bioanal Chem 412, 6789–6809. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] Collins BJ, Kerns SP, Aillon K, Mueller G, Rider CV, DeRose EF, London RE, Harnly JM, Waidyanatha S, 2021. Correction: Comparison of phytochemical composition of Ginkgo biloba extracts using a combination of non-targeted and targeted analytical approaches. http://link.springer.com/article/10.1007/s00216-021-03594-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] Consumerlab, 2020. Menopause Supplements Review. Soy and Red Clover Isoflavones, Black Cohosh) and Progesterone Creams. https://www.consumerlab.com/reviews/soy-isoflavones-red-clover-black-cohosh-supplements/phytoestrogens/. Last accessed June 20, 2021.

[R13] Dugoua JJ, Seely D, Perri D, Koren G, Mills E, 2006. Safety and efficacy of black cohosh (Cimicifuga racemosa) during pregnancy and lactation. Can J Clin Pharmacol 13, e257–261. [PubMed] [Google Scholar]

[R14] Foster S, 1999. Black cohosh: A literature review. HerbalGram 45, 35–50. [Google Scholar]

[R15] Foster S, 2013. Exploring the peripatetic maze of black cohosh adulteration. HerbalGram 98, 32–51. [Google Scholar]

[R16] Gafner S, 2016. On adulteration of Actaea racemosa. Botanical Adulterants Bulletin.(accessed June 9th, 2021). HerbalGram; 111. [Google Scholar]

[R17] Gardner ZM, M., 2013. American Herbal Products Association’s botanical safety handbook., 2nd ed. CRC Press, Boca Raton, FL. [Google Scholar]

[R18] Godecke T, Nikolic D, Lankin DC, Chen SN, Powell SL, Dietz B, Bolton JL, van Breemen RB, Farnsworth NR, Pauli GF, 2009. Phytochemistry of cimicifugic acids and associated bases in Cimicifuga racemosa root extracts. Phytochem Anal 20, 120–133. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] Harnly J, Chen P, Colson K, McCoy J-A, 2015. MS, NMR, and DNA barcoding, complementary methods for identification and authentication of Black Cohosh (Actaea racemosa L.). Planta Medica 81. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] Harnly J, Chen P, Sun J, Huang H, Colson KL, Yuk J, McCoy JA, Reynaud DT, Harrington PB, Fletcher EJ, 2016. Comparison of Flow Injection MS, NMR, and DNA Sequencing: Methods for Identification and Authentication of Black Cohosh (Actaea racemosa). Planta Med 82, 250–262. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] Healthline, 2020. Black Cohosh: Benefits, Dosage, Side Effects, and More. https://www.healthline.com/health/food-nutrition/black-cohosh. Last accessed June 20, 2021.

[R22] Hubbard TD, Hsieh JH, Rider CV, Sipes NS, Sedykh A, Collins BJ, Auerbach SS, Xia M, Huang R, Walker NJ, DeVito MJ, 2019. Using Tox21 High-Throughput Screening Assays for the Evaluation of Botanical and Dietary Supplements. Appl In Vitro Toxicol 5, 10–25. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] Jiang B, Lyles JT, Reynertson KA, Kronenberg F, Kennelly EJ, 2008. Stability evaluation of selected polyphenols and triterpene glycosides in black cohosh. J Agric Food Chem 56, 9510–9519. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] Jiang B, Ma C, Motley T, Kronenberg F, Kennelly EJ, 2011. Phytochemical fingerprinting to thwart black cohosh adulteration: a 15 Actaea species analysis. Phytochem Anal 22, 339–351. [DOI] [PubMed] [Google Scholar]

[R25] Jin YM, Y; Xie W; Hou L; Xu H; Zhang K; Zhang L; Du Y, 2018. UHPLC-Q-TOF-MS/MS-oriented characteristic components dataset and multivariate statistical techniques for the holistic quality control of Usnea. RSC Adv. 8, 15487–15500. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] Kellogg JJ, Paine MF, McCune JS, Oberlies NH, Cech NB, 2019. Selection and characterization of botanical natural products for research studies: a NaPDI center recommended approach. Nat Prod Rep 36, 1196–1221. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] Kruse SO, Lohning A, Pauli GF, Winterhoff H, Nahrstedt A, 1999. Fukiic and piscidic acid esters from the rhizome of Cimicifuga racemosa and the in vitro estrogenic activity of fukinolic acid. Planta Med 65, 763–764. [DOI] [PubMed] [Google Scholar]

[R28] Kuszak AJ, Hopp DC, Williamson JS, Betz JM, Sorkin BC, 2016. Approaches by the US National Institutes of Health to support rigorous scientific research on dietary supplements and natural products. Drug Test Anal 8, 413–417. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] Li JX, Yu ZY, 2006. Cimicifugae rhizoma: from origins, bioactive constituents to clinical outcomes. Curr Med Chem 13, 2927–2951. [DOI] [PubMed] [Google Scholar]

[R30] Lu Z, Rubinsky M, Babajanian S, Zhang Y, Chang P, Swanson G, 2018. Visualization of DNA in highly processed botanical materials. Food Chem 245, 1042–1051. [DOI] [PubMed] [Google Scholar]

[R31] Ma C, Kavalier AR, Jiang B, Kennelly EJ, 2011. Metabolic profiling of Actaea species extracts using high performance liquid chromatography coupled with electrospray ionization time-of-flight mass spectrometry. J Chromatogr A 1218, 1461–1476. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] Mahady GB, Low Dog T, Barrett ML, Chavez ML, Gardiner P, Ko R, Marles RJ, Pellicore LS, Giancaspro GI, Sarma DN, 2008. United States Pharmacopeia review of the black cohosh case reports of hepatotoxicity. Menopause 15, 628–638. [DOI] [PubMed] [Google Scholar]

[R33] Moraes DFC, Still DW, Lum MR, Hirsch AM, 2015. DNA-Based Authentication of Botanicals and Plant-Derived Dietary Supplements: Where Have We Been and Where Are We Going? Planta Medica 81, 687–695. [DOI] [PubMed] [Google Scholar]

[R34] NIH, 2021a. Black cohosh. https://ods.od.nih.gov/factsheets/BlackCohosh-HealthProfessional/. Last accessed July 22, 2021.

[R35] NIH, 2021b. Black cohosh. https://www.ncbi.nlm.nih.gov/books/NBK547990/pdf/Bookshelf_NBK547990.pdf. Last accessed June 20, 2021.

[R36] NIH, 2021c. Dietary Supplement Label Database. https://ods.od.nih.gov/Research/Dietary_Supplement_Label_Database.aspx. Last accessed June 9, 2021.

[R37] Nikolic D, Lankin DC, Cisowska T, Chen SN, Pauli GF, van Breemen RB, 2015. Nitrogen-Containing Constituents of Black Cohosh: Chemistry, Structure Elucidation, and Biological Activities. Recent Adv Phytochem 45, 31–75. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] NTP, 2013. NTP Technical Report on the Toxicology and Carcinogenesis Studies of Ginkgo biloba extract (CAS No. 90045–36-6) in F344/N rats and B6C3F1/N mice (gavage studies) NTP TR 578 https://ntp.niehs.nih.gov/ntp/htdocs/lt_rpts/tr578_508.pdf. [PubMed]

[R39] Painter DP, S.; Murty M, 2010. Black cohosh products and liver toxicity: update. Canadian Adverse Reaction Newsletter 20, 12–15. [Google Scholar]

[R40] Parveen I, Gafner S, Techen N, Murch SJ, Khan IA, 2016. DNA Barcoding for the Identification of Botanicals in Herbal Medicine and Dietary Supplements: Strengths and Limitations. Planta Med 82, 1225–1235. [DOI] [PubMed] [Google Scholar]

[R41] Pawar RS, Handy SM, Cheng R, Shyong N, Grundel E, 2017. Assessment of the Authenticity of Herbal Dietary Supplements: Comparison of Chemical and DNA Barcoding Methods. Planta Med 83, 921–936. [DOI] [PubMed] [Google Scholar]

[R42] Qiu F, Imai A, McAlpine JB, Lankin DC, Burton I, Karakach T, Farnsworth NR, Chen SN, Pauli GF, 2012. Dereplication, residual complexity, and rational naming: the case of the Actaea triterpenes. J Nat Prod 75, 432–443. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] Raclariu AC, Heinrich M, Ichim MC, de Boer H, 2018. Benefits and Limitations of DNA Barcoding and Metabarcoding in Herbal Product Authentication. Phytochem Anal 29, 123–128. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] Ragupathy S, Faller AC, Shanmughanandhan D, Kesanakurti P, Shaanker RU, Ravikanth G, Sathishkumar R, Mathivanan N, Song J, Han J, Newmaster S, 2019. Exploring DNA quantity and quality from raw materials to botanical extracts. Heliyon 5, e01935. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] Reich E, Schibli A, DeBatt A, 2008. Validation of high-performance thin-layer chromatographic methods for the identification of botanicals in a cGMP environment. J AOAC Int 91, 13–20. [PMC free article] [PubMed] [Google Scholar]

[R46] Rider CV, Walker NJ, Waidyanatha S, 2018. Getting to the Root of the Matter: Challenges and Recommendations for Assessing the Safety of Botanical Dietary Supplements. Clin Pharmacol Ther 104, 429–431. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] Ryan KR, Huang MC, Ferguson SS, Waidyanatha S, Ramaiahgari S, Rice JR, Dunlap PE, Auerbach SS, Mutlu E, Cristy T, Peirfelice J, DeVito MJ, Smith-Roe SL, Rider CV, 2019. Evaluating Sufficient Similarity of Botanical Dietary Supplements: Combining Chemical and In Vitro Biological Data. Toxicol Sci 172, 316–329. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] Shipkowski KA, Betz JM, Birnbaum LS, Bucher JR, Coates PM, Hopp DC, MacKay D, Oketch-Rabah H, Walker NJ, Welch C, Rider CV, 2018. Naturally complex: Perspectives and challenges associated with Botanical Dietary Supplement Safety assessment. Food Chem Toxicol 118, 963–971. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] Sica VP, Mahony C, Baker TR, 2018. Multi-Detector Characterization of Grape Seed Extract to Enable in silico Safety Assessment. Front Chem 6, 334. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] Smith TM, F.; Eckl V; Reynolds CM,, 2021. Herbal Supplement Sales in US Increase by record-breaking 17.3% in 2020. HerbalGram 131, 52–65. [Google Scholar]

[R51] Sumner LW, Amberg A, Barrett D, Beale MH, Beger R, Daykin CA, Fan TW, Fiehn O, Goodacre R, Griffin JL, Hankemeier T, Hardy N, Harnly J, Higashi R, Kopka J, Lane AN, Lindon JC, Marriott P, Nicholls AW, Reily MD, Thaden JJ, Viant MR, 2007. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3, 211–221. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] USP, 2015. United States Pharmacopea Dietary supplements compendium. The United States Pharmacopeal Convention, Twinbrook Parkway, Rockville, MD. [Google Scholar]

[R53] Verbitski SM, Gourdin GT, Ikenouye LM, McChesney JD, Hildreth J, 2008. Detection of Actaea racemosa adulteration by thin-layer chromatography and combined thin-layer chromatography-bioluminescence. J AOAC Int 91, 268–275. [PMC free article] [PubMed] [Google Scholar]

[R54] Waidyanatha S, Pierfelice J, Cristy T, Mutlu E, Burback B, Rider CV, Ryan K, 2020. A strategy for test article selection and phytochemical characterization of Echinacea purpurea extract for safety testing. Food Chem Toxicol 137, 111125. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] Waidyanatha S, Ryan K, Roe AL, Jia W, Paine MF, Ferguson S, Gurley BJ, Welch K, Chow MSS, Devito M, Rider C, 2018. Follow that botanical: Challenges and recommendations for assessing absorption, distribution, metabolism and excretion of botanical dietary supplements. Food Chem Toxicol 121, 194–202. [DOI] [PubMed] [Google Scholar]

[R56] Walker KM, Applequist WL, 2012. Adulteration of Selected Unprocessed Botanicals in the U.S. Retail Herbal Trade. Economic Botany 66, 321–327. [Google Scholar]

[R57] WHO, 2007. WHO guidelines for assessing quality of herbal medicines with reference to contaminants and residues.

PERMALINK

Working with the natural complexity: selection and characterization of black cohosh root extract for use in toxicology testing

Suramya Waidyanatha

Tim Cristy

Jessica Pierfelice

Jon C Andre

Brian Burback

Esra Mutlu

Abstract

1. Introduction

2. Materials and methods

2.1. Materials

Table 1.

Table 2.

2.2. High performance liquid chromatography-evaporative light scattering detection (HPLC-ELSD)

2.3. High performance thin layer chromatography (HPTLC) and DNA barcoding of sample 5

2.4. Composition and contaminant analysis of sample 5

2.5. Characterization of black cohosh extract sample 5 by mass spectrometry

2.5.1. Identification of constituents in sample 5 by mass spectrometry

2.5.2. Quantitation of constituents in sample 5 by mass spectrometry

2.6. Storage stability of bulk black cohosh extract sample 5

2.7. Non-targeted chemical profiling of cohosh samples by high performance liquid chromatography-charged aerosol detection (HPLC-CAD)

3. Results

3.1. Chemical profiling using HPLC-ELSD

Figure 1.

3.2. Authentication of sample 5 by HPTLC and DNA barcoding

Figure 2.

3.3. Description, composition, and contaminant analysis of black cohosh root extract sample 5

Table 3.

Figure 3.

3.4. Identification of constituents in black cohosh root extract sample 5 by mass spectrometry

Figure 4.

Table 4.

Table 5.

3.5. Quantitation of constituents in black cohosh root extract sample 5 mass spectrometry

Table 6.

3.6. Storage stability of black cohosh extract sample 5

Figure 6.

3.7. Non-targeted chemical profiling using HPLC-CAD of cohosh materials procured overtime and chemometric analysis of data

Figure 7.

Figure 8.

Figure 9.

Figure 10.

4. Discussion

Conclusion

Supplementary Material

Acknowledgements

Footnotes

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases