Talanta 194 (2019) 795–802

Contents lists available at ScienceDirect

Talanta

journal homepage: www.elsevier.com/locate/talanta

Determination of bioactive nonvolatile constituents in dietary T supplements by a rapid and economic HPLC method: Analytical method development and single-laboratory validation ⁎ Hong Youa, , Bailey Irelanda, Michael Moeszingera, Haoshu Zhanga, Laura Snowb, Scott Krepichb, Vivian Takagawac a Eurofins Scientific, Inc., 1365 Redwood Way, Petaluma, CA 94954, United States b Phenomenex, Inc., 411 Madrid Avenue, Torrance, CA 90501, United States c ChromaDex, Inc., 10005 Muirlands Blvd Suite #G, Irvine, CA 92618, United States

ARTICLE INFO ABSTRACT

Keywords: Most of the validated methods for ginger-containing dietary supplements have long run time and low sensitivity Nonvolatile ginger constituents and only analyze and . 6- and become popular in modern dietary supple- AOAC validation ment industry as bioactive ginger constituents. Therefore, we developed an efficient HPLC-UV/Vis method to Gingerols analyze all above major constituents. Compared to 282/280 nm used by the current compendial United States Shogaols Pharmacopeia (USP) monograph method and International Organization for Standardization (ISO) 13685-1997 6-paradol method, detection wavelength was optimized to 230 nm which showed a higher sensitivity (signal-to-noise ratio) Zingerone and better peak resolution. For measuring the ginger constituents in AOAC required matrices, the method was demonstrated to be selective, linear (R2 > 0.999), specific, accurate (91.1–103.2% spike recovery rate) and

precise (RSDr < 5%, RSDR < 8%). Among 10 commercial ginger-containing samples that we screened using this method, the results were 80–123% of the products’ labeling value. The HPLC running time was successfully shortened from 29 min (USP method) and 40 min (ISO method) to 12 min without the need of using an expensive Mass Spectrometer for analyte separation. The method is the first method that meets all AOAC SMPR 2017.12 requirements and therefore has the potential to be adopted as a consensus industrial reference method for meeting FDA's cGMP Compliance for the manufacture and quality control of dietary supplements and in- gredients.

1. Introduction and oleoresin contain the non-volatile components including gingerols, , shogaols and zingerone. Previous methods and publications Ginger (Zingiber officinale) rhizome is used in several systems of focused on analytical methods that only quantify gingerols and sho- medicine, including traditional Chinese medicine, Ayurveda and gaols [4–7]. However, compared to other ginger non-volatile con- Western herbal medicine. It has been used to treat a great variety of stituents, 6-paradol has shown better anti-hyperglycemic assay activity disorders including colds and flu, dyspepsia, flatulence and colic, mi- in stimulating glucose utilization of 3T3-L1 adipocytes and C2C12 graine, nausea, rheumatic disorders, and vomiting [1]. Gingerols are myotubes [3]. In a separate study, 6-paradol was suggested as a ther- the major bioactive nonvolatile pungent constituents of fresh ginger apeutic agent to effectively protect the brain after cerebral ischemia, rhizome [2]. Gingerols are mainly converted to zingerone and shogaols likely by attenuating neuroinflammation in microglia [8]. As an alka- during drying, prolonged storage, processing, and cooking [3]. The loid with strong antioxidantive activity, zingerone has been proven to shogaols can then be partially transformed to paradols upon cooking or have an ONOO(-) peroxynitrite scavenging ability, which indicates its metabolized to paradols in the animals’ body after being consumed and cellular defensive potential against related-symptoms in several human absorbed by digestive system [3]. Major ingredients of ginger-con- diseases such as stroke, Alzheimer's disease, and atherosclerosis [9]. By taining dietary supplements such as ginger rhizome, powder extract, increasing systemic superoxide dismutase activity, zingerone was also

⁎ Correspondence to: Eurofins Supplement Analysis Center, Eurofins Scientific, Inc., 1365 Redwood Way, Petaluma, CA, 94954, UnitedStates. E-mail addresses: [email protected] (H. You), [email protected] (B. Ireland), [email protected] (M. Moeszinger), [email protected] (H. Zhang), [email protected] (L. Snow), [email protected] (S. Krepich), [email protected] (V. Takagawa). https://doi.org/10.1016/j.talanta.2018.10.075 Received 25 September 2018; Received in revised form 22 October 2018; Accepted 22 October 2018 Available online 25 October 2018 0039-9140/ © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). .Yue al. et You H.

Table 1 Test samples used in method validation and optimization studies.

ID code Matrices Sample form Major Component(s) Manufacturer Product name Label claim Estimated level (w/w purity)

251 Rhizome powder Raw material Ginger USP Powdered Ginger 100% Ginger rhizome powder 1% Total nonvolatile ginger constituentsa 099 Rhizome dry extract Raw material Ginger Suan Fharm Ginger extract 100% Ginger rhizome dry extract 5% Total nonvolatile powder ginger constituents 063 Tablet Finished Ginger The Ginger People Ginger Rescue, Chewable Ginger 40 mg ginger powder/1.3 g 0.03% Total nonvolatile product Tablets for Kids (665.78 mg/tablet, 3% ginger ginger constituentsa powder) 580 Capsule Finished Ginger Now Foods Ginger Root extract, 250 mg 12.5 mg gingerols/capsule 2.51% Total nonvolatile 796 product (498.02 mg/capsule) ginger constituents 101 Liquid Capsule Finished Bacopa monnieri, Caffeine, p-Synephrine, Nutrex Research Labs Lipo6 Black 10 mg zingerone/3 capsules 0.58% Zingerone product Vinpocetine, Zingerone, Yohimbine (572.61 mg/capsule) 986 Rhizome oleoresin - Raw material Ginger Shaanxi Guanjie Ginger supercritical CO2 extract 100% Ginger rhizome dry extract 50% Total nonvolatile supercritical CO2 extract Technology powder ginger constituents 648 Lean muscle recovery Finished Glycyl-Alanyl-Lysine-L-Leucine, -Isoleucine, -Valine, BPI sports Best BCAA Shredded Lean Muscle 1.25 mg 6-Paradol/11 g 0.01% 6-Paradol supplement powder product and -Citrulline, L-Carnitine, Paradoxine Grains of Recovery Formula, Watermelon Paradise extract (6-Paradol) Ice 889 Softgel capsules Finished Tumeric, ginger Life Extension Advanced Bio-Curcumin, with 60 mg Gingerols/softgel 5.10% total nonvolatile product Ginger & Turmerones (1176.75 mg/softgel) Ginger constituents 740 Rhizome powder Raw material Ginger NIST SRM 3398 (not published) 100% Ginger powder 1% total nonvolatile ginger constituentsa 692 Rhizome dry extract Raw material Ginger NIST SRM 3399 (not published) 100% Ginger extract Unknown

a Estimated assuming ginger rhizome powder contains 1% total nonvolatile ginger constituents. Talanta 194(2019)795–802 H. You et al. Talanta 194 (2019) 795–802 suggested to have possible value in the treatment of Parkinson's disease water diluent. The rest of the sample preparation was the same as [10]. The dietary supplement industry has recognized the market of 6- Control. paradol and zingerone and has made them the major bioactive com- • Published enzymatic method [12]: Used 0.5% v/w cellulase solution ponents in commercially available dietary supplements (Table 1; for each gram of ginger powder and the reaction occurs in pH = 5 Sample 101 and 648). The AOAC SMPRs (Standard Method Perfor- water. Incubated sample for 1 h in 50 °C waterbath before adding mance Requirement) are developed by working groups that are com- methanol for further extraction. Final methanol/water ratio was posed of stakeholders from academia, industry, and non-profit organi- 80%/20%. The rest of the sample preparation was the same as zations and designed to serve as guidance for the development of a control. consensus industrial reference method for meeting FDA's cGMP com- • Optimal enzymatic method: Reaction conditions were designed pliances of analyzing dietary supplements and ingredients. Targeting to based on the enzyme manufacturer's recommendations (email meet the AOAC SMPR 2017.12 (Quantitation of Select Nonvolatile communication with Sigma-Aldridge). 0.5% v/w cellulase solution Ginger Constituents) [11] requirements for ginger-containing dietary was used for each gram of ginger powder and the reaction occurs in ingredient and supplements, the method in development is designed to pH = 8 water. Incubated sample for 1 h in 60 °C waterbath before identify and quantitate 6-paradol and zingerone, in addition to the adding methanol for further extraction. Final methanol/water ratio traditional target nonvolatile ginger pungent constituents including 6-, was 80%/20%. Rest of sample preparations was the same as control. 8- and 10-gingerols and 6-, 8- and 10-shogaols. The details of SMPR 2017.12 are discussed later. To further evaluate the impacts of changing 7 factors in the analy- tical procedures, a Youden Ruggedness Trial [13] was conducted. 2. Material and methods Challenging factors were listed in Table 1S (Supplementary data), and 8 aliquots of the ginger rhizome powder sample were extracted and tested 2.1. Reagents and materials using different factor combinations as shown in Table 2S (Supplementary data). Standard 6-, 8-gingerol, 10-gingerol, 6-, 8-shogaol and 10-shogaol were provided by ChromaDex (Irvine, CA, USA). 2.4. Sample preparation Zingerone, cellulase (cellulase from Aspergillus sp.; 1200 IU/mL; Enzyme Commission (EC) Number 3.2.1.4; Sigma Product # C2605) In dietary supplements and ingredients sample, estimated levels of and citric acid (reagent grade) were obtained from Sigma-Aldrich (St. total nonvolatile ginger constituents may vary from 0.03% w/w to 50% Louis, MO, USA). 6-Paradol was donated by Dalton Research (Toronto, w/w. They were determined from the dietary supplement formulation Canada). HPLC grade methanol and acetonitrile were purchased from or dietary ingredient supplier's certificate of analysis. Because of the Fisher Scientific (Pittsburgh, PA, USA). large range and importance of sample/diluent ratio (shown later in the Ten samples (Table 1.) in the forms of both dietary supplements 3.2.3 Optimization of extraction studies), the amount to weigh for each (tablet, capsule, softgel, etc.) and dietary ingredients (rhizome powder, sample (Table 2) was adjusted based on the expected level and the mid- dry extract, and oleoresin) were assigned by random ID codes and used level of the calibration curve. To assure peak area lied within the curve, for validation studies. Sample 251, 099, 986, 740, and 692 were kind the sample's weight was calculated to target levels 2–4 of the Calibra- donations from their manufactures. Other samples were purchased from tion Solution's concentration. The maximum sample weight was www.iHerb.com. A ginger rhizome powder sample (Botanical Re- 1000 mg for all the matrices to avoid the oversaturation of diluent. ference Materials 30290-5) was obtained from ChromaDex (Irvine, CA, USA). Samples were stored in room temperature prior to testing. (a) Rhizome powder and rhizome dry extract – e.g. for ginger rhizome powder that has an expected level of 1% total nonvolatile ginger 2.2. Preparation of standard solution constituents, 60 ± 6 mg (Table 2) sample was accurately weighed and transferred into a 20-mL amber VOA vial. Before weighing, Standard stock solution was prepared by accurately weighing and mixed the sample thoroughly until homogenized. Samples were transferring 2.5 ± 0.3 mg of 6-gingerol, 0.75 ± 0.08 mg of 8-gingerol, grinded if necessary. 1.0 ± 0.1 mg of 10-gingerol, 1.0 ± 0.1 mg of 6-shogaol, (b) Tablet and capsule – composited 10 tablets or capsules (remove 0.75 ± 0.08 mg of 8-shogaol, 1.0 ± 0.1 mg of 10-shogaol, shells), grinded the samples until homogenized. E.g. for a tablet that 0.5 ± 0.1 mg of 6-paradol and 1.0 ± 0.1 mg of zingerone to a 20-mL has 3% ginger powder and an expected level of 0.03% total non- amber VOA vial and then diluting with 20.0 mL of methanol. Stock volatile ginger constituents, 1000 ± 100 mg (Table 2) was solution was stored in freezer until use. This solution may be stored in - weighed into a 20-mL amber VOA vial. 20 ± 2 °C freezer for up to 50 days (stability data not shown). 6-point (c) Liquid capsule, softgel capsule, and oleoresin– a Pasteur pipette, calibration curve was prepared by diluting the standard stock solution. positive displacement pipette, or other appropriate tool was used to transfer sample (for softgel liquid capsule, mix and composite the 2.3. Optimization of extraction

To optimize the extraction conditions, the ginger rhizome powder Table 2 sample was used as study material. Extraction solvents (methanol; Sample preparation instructions for samples with different matrices and ex- methanol:water (80%:20% v/v); methanol:water (50%:50% v/v); pected levels. ethanol, ethyl acetate; dichloromethane-methanol (50%:50% v/v)) Example Rhizome Tablet, capsule Liquid capsule, were tested for the extraction efficiency optimization using 1 h (soni- estimated level, powder, softgel capsule, and cate 30 min, vortex 5 s, sonicate another 30 min) sonication. % w/w rhizome dry oleoresin extract After determining the extraction solvent, acidic environment and Example sample's weight, mg enzyme application in extraction solvent were also studied using the same ginger rhizome powder sample. The treatments included: 0.03 N/A 1000 ± 100 1000 ± 25 0.3 N/A 1000 ± 100 250 ± 25 1 60 ± 6 300 ± 30 75 ± 8 Control: 80%methanol/20%water was used as extraction solvent. • 5 12 ± 3 60 ± 6 15 ± 2 • Acidic extraction: Adjusted water pH to 5 using citric acid solution 50 1.2 ± 0.3 6 ± 0.6 1.5 ± 0.2 before mixing with methanol for making the 80%methanol/20%

797 H. You et al. Talanta 194 (2019) 795–802

content of 5 capsules before sample transfer). E.g. for ginger calibration curve. Each calibration curve was made up of 6 data points, oleoresin that has an expected level as 50% total nonvolatile ginger and the resulting R2 coefficients for all curves must exceed the re- constituents, 1.5 ± 0.2 mg (Table 2) was weighed into a 20-mL quirement of NLT 0.999. amber VOA vial using a microbalance. 2.6.4. LOD and LOQ After weighing, each sample was diluted with 10.0 mL of 80% me- The International Union of Pure and Applied Chemistry (IUPAC) thanol: 20% water (the water was acidified with citric acid, pH =5), method (under the recommendation by AOAC INERNATIONSIONAL) vortexed for 1 ± 0.1 min, sonicated for 30 ± 3 min in cold water (if was used to determine the LOD (limit of detection) and LOQ (limit of the sonicator has no temperature control function, use icebags to keep quantification) of select nonvolatile ginger constituents. The dataset the temperature low, ≤ 30 °C), vortexed for 5 s, sonicated for another from seven replications of blank matrix (prepared separately) from 30 ± 3 min in cold water and filtered through a 0.45-µm PTFE filter galangal root was used. LOD is defined as the sum of mean response and into an amber autosampler vial for future analysis. three times the standard deviation. LOQ is defined as the sum of mean response and ten times the standard deviation. The instrument's limits 2.5. HPLC analysis (mg/mL) are divided by the sample concentration to convert to a % w/ w basis. HPLC (Agilent; Santa Clara, CA, USA) equipped with a binary pump and a diode-array detector (190–400 nm) was employed for the ana- lyses. A Kinetex C18 column (5 µm, 150 ×3 mm) with a SecurityGuard 2.6.5. Accuracy Ultra Cartridge (Phenomenex, Torrance, CA, USA) was used for the To demonstrate method accuracy, galangal root (considered as the chromatographic separation. The optimal instrument conditions are: “Placebo” of ginger rhizome) and the excipient blend (considered as the injection volume of 5.0 μL, column temperature at 30 °C, flow rate at “Placebo” of tablet and capsule dietary supplements) were spiked with 1.1 mL/minute and detection wavelength at 230 nm. The gradient nonvolatile ginger constituent reference material solutions at 2 dif- program with water as mobile phase A and acetonitrile as mobile phase ferent levels. The details of both “Placebo” samples have been described B is listed below: 0–1.5 min, hold 35% B; 1.5–1.8 min, from 35% to 60% in Section 2.6.1. The spike recovery experiment was performed by B; 1.8–5 min, hold 60% B; 5–6.5 min, from 60% to 100% B; 6.5–9 min, spiking the nonvolatile ginger constituent reference material solution hold 100% B; 9–9.1 min, 100–35% B; 9.1–12 min, hold 35% B. The total (known concentration) immediately after weighing the samples. Three run time was 12.0 min. separate sample preparations were conducted (n = 3). The diluent was added afterwards to make the total extraction solution volume 10 mL. 2.6. Single-laboratory validation parameters 2.6.6. Precision 2.6.1. Selectivity Eight dietary ingredients and supplements (first eight samples in For major target matrices (ginger rhizome dietary ingredient, and Table 1) in different matrices were run to demonstrate the repeatability tablet and capsule dietary supplements), method selectivity was de- and reproducibility of the method. Two NIST SRM samples (sample 740 monstrated by running “Placebo” samples expected to be free of non- and 692) were tested to further evaluate the accuracy and repeatability volatile ginger constituents. Galangal (Kaempferia galangal, Aromatic of the method. Ginger) root was used as a placebo of ginger rhizome matrix because of All of ten dietary ingredients and supplements examined for re- their morphological similarities. Galangal root is also considered as an peatability were tested in quadruplicate (n = 4) in day 1 by analyst #1 adulterant of ginger rhizome [14]. An excipient blend placebo was on instrument #1. The within-day results, repeatability (RSDr), and formulated (50% maltodextrin, 42% hydroxypropyl methyl cellulose, HorRatr values for each nonvolatile ginger constituent were then cal- 5% stearic acid, 2% magnesium stearate, and 1% silicon dioxide) and culated for all the materials investigated. tested to evaluate potential chromatographic interferences from the For evaluating method intermediate precision, eight dietary in- filler ingredients used in common tablet and capsule dietary supple- gredients (ginger rhizome, dry extract, oleoresin (super critical fluid ments. CO2 soft extract)) and dietary supplements (tablets, capsule, liquid softgel capsule) were screened as target matrices. Three runs were 2.6.2. System suitability conducted. Two analysts prepared 4 replicates of each sample, on 3 To demonstrate the overall chromatographic system suitability, five separate days, on 3 different instruments (Run 1: Day#1, Analyst#1, replicate injections of the reference material solution were performed Instrument#1; Run 2: Day#2, Analyst#2, Instrument#2; Run 3: for all 8 nonvolatile ginger constituents. Peak retention time, peak area, Day#3, Analyst#2, Instrument#3). A total of 12 replicate preparations and peak shape were analyzed. %RSD (relative standard deviation) of (n = 12) were determined for each type of sample for precision de- peak area, %RSD on retention time, USP tailing factor, and the relative termination. The between-day results, reproducibility (RSDR), and retention time of each analyte to 6-gingerol were calculated. As the HorRatR values were documented. quality control criteria for system suitability, reference material peak area %RSD must be ≤ 2.5. Reference material retention time %RSD 2.7. Data analysis must be ≤ 2.5. USP tailing factor of the reference material peak must be less than 2.0 for all analyte peaks. Agilent ChemStation was used to automatically determine in- dividual nonvolatile ginger constituents as follows: 2.6.3. Linearity Six calibration solutions were injected at the beginning of each in- Area(sample)–Calibration equation intercept % Analyte in sample = jection sequence. The calibration curves of the method cover the range Calibration equation slope× [Sample, mg/mL] approximately 0.5–50 µg/mL for zingerone; 1.2–120 µg/mL for 6-gin- × 100% gerol; 0.35–35 µg/mL for 8-gingerol; 0.5–50 µg/mL for 6-shogaol; 0.25–25 µg/mL for 6-paradol; 0.5–50 µg/mL for 10-gingerol; Results were analyzed by using one-way analysis of variance 0.35–35 µg/mL for 8-shogaol; 0.5–50 µg/mL for 10-shogaol. The cor- (ANOVA) followed by independent t-tests with Tukey's adjustment relation coefficients, calibration equation slopes, and y-intercepts were when there were significant differences among groups. All data were automatically generated by the HPLC data processing software (e.g. analyzed using Statistix software (version 10.0; Analytical Software, Agilent ChemStation). The blank was not considered as a part of the Tallahassee, U.S.).

798 H. You et al. Talanta 194 (2019) 795–802

Fig. 1. An overlaid chromatogram of nonvolatile ginger constituents reference material mixture (red line), galangal root powder sample (green line), and excipient blend placebo (blue line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

3. Results and discussion 3.2.2. Evaluating acidic environment and enzyme application in extraction solvent 3.1. HPLC detection wavelength selection Sample pretreatment was further tested using the ginger rhizome powder as study material. Acidic environment (low pH) has been shown UV wavelengths 230 nm, 282 nm, and 200 nm were evaluated in the to facilitate shogaol production [19]. Therefore, acidic extraction was early stages of method development. A greater amount of matrix in- tested on the basis of using methanol:water (80%:20% v/v) as extrac- terference was found when using 200 nm for screening samples that tion solvent (pH adjustment was performed on aqueous solution before have complicate matrices (data not shown). Wavelength 282 nm was mixing with methanol). Moreover, multiple publications showed the adapted by the USP Ginger Monograph method and wavelength 280 nm benefits of using enzyme to pretreat ginger rhizome samples for better was adapted by the ISO (International Organization for extraction efficiency. Amylase, cellulase, pectinase, protease, and vis- Standardization) 13685:1997 compendial method. Compared to 282/ cozyme were tested. Cellulase yielded the highest gingerol-rich oleor- 280 nm, 230 nm provided a higher signal-to-noise ratio (Table 3S, esin followed by either alcohol (ethanol) extraction [12] or three phase Supplementary data) and better peak resolution (Figure. 1S, partitioning (TPP) separation [20]. Thus, cellulase pretreatment was Supplementary data) for all 8 nonvolatile ginger constituents, especially tested in this study with multiple reaction conditions. for 6-, 8-, and 10-shogaols which have 230 nm as their maximum ab- The experiment results did not show significant difference between sorbance wavelength (λmax). This feature is of great importance for each treatment groups (Table 4S, Supplementary data). Although there method accuracy and precision when analyzing samples that have low is no statistical difference between different treatments, acidic condi- level shogaols. tion showed the best extraction efficiency, and therefore was chosen as the extraction diluent (methanol:water rather than pure methanol was 3.2. Experimental design and optimization used because most standard laboratory pH electrodes are not designed to be used in organic solvents) of the final method. 3.2.1. Extraction solvent optimization The extraction procedure has been thoroughly evaluated for raw 3.2.3. Evaluation of other extraction conditions-Youden Ruggedness Trial ginger powder and dried aqueous ginger extract by other research The Youden Ruggedness Trial [13] test results were shown in groups [15]. Compared to methanol: water (50%:50% v/v), water, Table 5S (Supplementary data) and Fig. 3S (Supplementary data). Total ethanol, and hexane, methanol was found to be most efficient when 6-, nonvolatile ginger constituent values were evaluated as the outcomes of 8-, 10-gingerol, and 6-shogaol were considered as analytes of interest. A interest. Among all factors that were being tested, sonication time single extraction with a 1 h sonication time (sonicate 30 min, vortex 5 s, (Factor A-a) and sample/diluent ratio (Factor G-g) were found to be sonicate another 30 min) was able to recover as high as ≥98.5% of the most critical. However, the difference between using method factor and analytes in the raw ginger powder [15]. However, additional solvents challenging factor were relatively small (0.042% w/w purity difference were evaluated in our validation study because analytes 8-, 10-shogaol, for Factor A-a = 3.4% difference from average; 0.052% w/w purity for 6-paradol, and zingerone were also screened for this project. Angeli Factor G-g = 4.3% difference from average). The method was therefore et al. [16] demonstrated that methanol:water (80%:20% v/v) has a found to be robust. better extraction efficiency than methanol for fresh ginger root. Ethyl acetate [4,6] and dichloromethane:methanol [17] were also used for 3.3. Method validation gingerol extraction in some published methods. Ethanol extraction yielded almost 100% higher gingerol-rich oleoresin comparing to A reference material mixture solution (Fig. 1) and eight test samples extraction [12]. Yudthavorasit et al. [18] confirmed that 1h (Figure 4S, Supplementary data) are shown as representative chroma- (sonicate 30 min, vortex 5 s, sonicate another 30 min) sonication is tograms. The result chromatograms showed that the method is specific optimal for the ginger non-volatile constituents extraction. for the target nonvolatile ginger constituents in the ginger rhizome and The extraction solvent comparison test results were shown in Figure tablet and capsule containing ginger, because there is no or negligible 2S (Supplementary data). Methanol: water (80%:20% v/v) has sig- chromatographic interference in the region where nonvolatile ginger nificantly higher extraction efficiency (p < 0.05) than other extraction constituents would elute. All required system suitability criteria were solvent (except pure methanol) when total nonvolatile ginger con- met as shown in Table 6S and 7S (Supplementary data). Calibration stituent value was evaluated as the outcome of interest (n = 3). curves were shown in Fig. 5S (Supplementary data). In the approximate

799 H. You et al. Talanta 194 (2019) 795–802

Table 3 AOAC SMPR 2017.012 requirements and HPLC method results.a

Parameter Requirement HPLC Method Results

6-Gingerol 8-Gingerol 6-Shogaol 10-Gingerol

Analytical rangeb 0.05–50% w/w 0.02–50% w/w LOQ ≤0.05% w/w 0.0103% w/w 0.0045% w/w 0.0022% w/w 0.0015% w/w Recovery 90–107% 95.9–99.6% 98.3–103.3% 98.7–101.1% 93.7–101.0%

RSDr ≤5% ≤2% ≤3% ≤2% ≤2%

RSDR ≤8% ≤6% ≤7% ≤4% ≤7%

a Table only summarizes the results of AOAC SMPR 2017.12 required analytes (6-, 8-, 10-gingerols and 6-shoganol) in required matrices (rhizome powder, rhizome dry extract, tablet and capsules). b Analytical range was calculated from total select nonvolatile ginger constituents contents. range of 0.25–50 μg/mL, the method was found to be linear for different from 0.03% w/w purity for a tablet dietary supplement (Sample 063) to analytes (R2 > 0.999). 40.2% w/w purity for a super critical fluid CO2 soft extract dietary As a part of method specificity evaluation, peak purity was checked ingredient (Sample 986). Zingerone's reproducibility RSDR failed AOAC for each analyte in each sample that was determined in precision and SMPR 2017.12 on sample 580 and 986. However, zingerone showed accuracy tests. A UV–Vis spectrum library (Fig. 6S, Supplementary acceptable reproducibility RSDR (3.66%) on Sample 101 that is the only data) was established in the HPLC Agilent ChemStation according to the target dietary supplement that has a label claim for zingerone (label

UV–Vis spectra of corresponding bioactive nonvolatile ginger con- claim recovery=86%). 6-Paradol's reproducibility RSDR failed AOAC stituent reference materials. A peak purity factor (spectrum matching SMPR 2017.12 on Sample 251. However, 6-paradol showed acceptable score) of 950 was set as the failing threshold in this validation study and reproducibility RSDR (3.66%) on Sample 648 that is the only target any peak that has a peak purity factor < 950 was excluded from the dietary supplement that has a label claim for 6-paradol (label claim integration. The LOD and LOQ for the sample with lowest estimated recovery=100%). Sample 740 and 692 are National Institute of Stan- level of nonvolatile ginger constituents are calculated (Table 8S, dards and Technology (NIST) Standard Reference Materials (SRM) 3398 Supplementary data) as the method limitation. LOQs meet the SMPR (ginger rhizome) and 3399 (ginger dry extract), respectively. The 2017.12's requirements for LOQ < 0.05% for all analytes (Table 3). analytical results of these 2 samples have been sent to NIST, and the Spike recovery results were summarized in Table 4. The high level accuracy of this method was confirmed by NIST through an email and low level of analytical ranges for ginger containing dietary in- communication. gredient (ginger rhizome) and dietary supplements (tablets and cap- All the required analytes in required matrices have passed the SMPR sules) were tested. The reference material spiked levels bracketed the 2017.12 requirements for accuracy and precision (Table 3). ranges as: 0.02–2% w/w for zingerone, 0.06–6% w/w for 6-gingerol, 0.02–2% w/w for 8-gingerol, 0.02–2% w/w for 6-shogaol, 0.015–1.5% 3.4. Application to sample w/w for 6-paradol, 0.02–2% w/w for 10-gingerol, 0.017–1.7% w/w for 8-shogaol, and 0.02–2% w/w for 10-shogaol. The mean recovery value A variety of dietary ingredients and supplements were tested in the (detected level vs. theoretical level) obtained among different spiked repeatability study using the proposed method. The results were com- levels of nonvolatile ginger constituent ranged from 91.1% (zingerone) parable to their label/manufacture's claims (Table 9S, Supplementary to 103.2% (8-gingerol) for the galangal root sample, and from 97.2% data). Because the claims may have been developed using different (zingerone) to 101.5% (8-gingerol) for the excipient blend sample. All analytical methods and terminologies, the recovery results were docu- of the results fulfilled the SMPR 2017.12's requirements for recovery mented as reference information instead of failing criteria for accuracy (90–107%). The repeatability precision (RSDr) for all analytes among test. Among 10 commercial ginger-containing samples that we screened the spike recovery test were ≤5%, which is acceptable by the AOAC using this method, the results were 80–123% of the products’ labeling SMPR 2017.12. value. Repeatability results (Table 5) showed that all of the total non- volatile ginger constituents, as well as target analytes with levels above 4. Conclusion LOQ, met the AOAC SMPR 2017.12 requirement for repeatability RSDr (≤5%). Intermediate precision test results (Table 5) showed that all 8 The validation results showed that the proposed method is selective, samples have their total nonvolatile ginger constituent reproducibility linear, accurate, precise, and robust. As shown in Table 3, all SMPR RSDR acceptable by AOAC SMPR 2017.12 (≤8%). The results ranged 2017.12 requirements were met for required matrices (rhizome

Table 4 Spike recovery results for individual nonvolatile ginger constituent in 2 spiking levels.

Matrix Spike level Results Zingerone 6-Gingerol 8-Gingerol 6-Shogaol 6-Paradol 10-Gingerol 8-Shogaol 10-Shogaol Total constitutes

Galangal root Low Mean, % (w/w) 0.022 0.06 0.021 0.022 0.014 0.021 0.016 0.022 0.198 RSD, % 1.76% 5.11% 5.48% 4.93% 3.61% 1.70% 4.62% 3.65% 3.36% Mean Recovery 91.11% 95.90% 103.23% 101.03% 95.81% 93.77% 96.15% 99.76% 96.82% High Mean, % (w/w) 2.33 6.1 1.98 2.18 1.45 2.15 1.68 2.25 20.1 RSD, % 2.17% 1.94% 1.83% 2.06% 2.25% 1.85% 2.29% 2.36% 2.04% Mean Recovery 97.99% 98.14% 98.39% 98.70% 99.01% 98.60% 99.26% 101.63% 98.79% Excipient blend Low Mean, % (w/w) 0.024 0.062 0.021 0.022 0.015 0.022 0.017 0.023 0.206 RSD, % 1.19% 0.58% 0.64% 0.47% 1.37% 1.39% 0.59% 0.72% 0.56% Mean Recovery 97.20% 98.70% 101.50% 99.80% 98.70% 100.70% 101.20% 100.30% 99.50% High Mean, % (w/w) 2.36 6.16 2.01 2.18 1.46 2.2 1.68 2.21 20.3 RSD, % 2.22% 2.22% 2.10% 2.36% 2.12% 2.53% 2.23% 2.27% 2.25% Mean Recovery 99.60% 99.60% 100.30% 99.40% 100.60% 101.00% 99.90% 100.20% 99.90%

800 H. You et al. Talanta 194 (2019) 795–802

Table 5 Table 5 (continued) Precision results summary of the test materials. Sample Analyte Mean, Repeatability Intermediate Precision Sample Analyte Mean, Repeatability Intermediate Precision % (w/

% (w/ w) RSDr HorRatr RSDR HorRatR w) RSDr HorRatr RSDR HorRatR 740b Zingerone < LOQ N/A N/A N/T a 251 Zingerone < LOQ N/A N/A N/A N/A 6-Gingerol 0.212 0.85% 0.168 6-Gingerol 0.554 1.01% 0.231 1.39% 0.320 8-Gingerol 0.036 2.52% 0.383 8-Gingerol 0.140 1.23% 0.230 6.80% 1.268 6-Shogaol 0.230 0.59% 0.119 6-Shogaol 0.116 1.59% 0.290 2.98% 0.540 6-Paradol 0.012 1.46% 0.187 6-Paradol 0.015 1.51% 0.204 11.11% 1.476 10-Gingerol 0.096 1.62% 0.286 10-Gingerol 0.176 1.14% 0.222 3.89% 0.751 8-Shogaol 0.064 0.63% 0.105 8-Shogaol 0.025 4.33% 0.631 7.45% 1.074 10-Shogaol 0.146 0.38% 0.071 10-Shogaol 0.048 0.96% 0.154 4.03% 0.641 Total 0.796 0.48% 0.116 Total 1.074 1.15% 0.292 2.52% 0.638 692b Zingerone 0.030 2.87% 0.850 N/T a 099 Zingerone 0.063 1.64% 0.274 7.14% 1.182 6-Gingerol 1.935 1.75% 0.971 6-Gingerol 2.969 0.92% 0.272 5.82% 1.718 8-Gingerol 0.266 1.81% 0.744 8-Gingerol 0.523 3.30% 0.752 4.41% 1.002 6-Shogaol 0.497 1.09% 0.494 6-Shogaol 0.505 0.82% 0.186 3.98% 0.900 6-Paradol 0.073 3.63% 1.229 6-Paradol 0.075 1.48% 0.254 5.36% 0.911 10-Gingerol 0.543 1.73% 0.793 10-Gingerol 0.736 1.22% 0.296 7.37% 1.764 8-Shogaol 0.098 1.18% 0.419 8-Shogaol 0.095 0.54% 0.096 8.04% 1.415 10-Shogaol 0.202 1.37% 0.541 10-Shogaol 0.210 2.12% 0.422 7.96% 1.579 Total 3.643 1.60% 0.972 Total 5.177 1.18% 0.381 4.90% 1.571 a 063 Zingerone < LOQ N/A N/A N/A N/A N/A = Not applicable; N/T = Not tested. 6-Gingerol 0.014 0.86% 0.115 3.16% 0.419 a The mean of each sample (except 740 and 692) was calculated from 12 8-Gingerol < LOQ N/A N/A N/A N/A replicates in the Intermediate Percision test. 6-Shogaol 0.005 0.53% 0.060 3.00% 0.340 b 6-Paradol < LOQ N/A N/A N/A N/A the mean values of 740 or 692 were calculated from 4 replicates in the 10-Gingerol 0.003 0.65% 0.069 2.68% 0.285 Repeatability test. 8-Shogaol 0.001 2.40% 0.211 5.15% 0.450 10-Shogaol 0.002 0.30% 0.029 5.98% 0.580 powder, dry extract, tablets and capsule dietary supplements). This Total 0.026 0.64% 0.092 2.38% 0.345 method avoided extensive sample preparation, only requires one 580a Zingerone 0.011 4.01% 0.517 8.84% 1.129 6-Gingerol 1.702 1.25% 0.341 2.48% 0.672 column, had relatively high sensitivity, shortened the run time to 8-Gingerol 0.309 1.02% 0.214 6.13% 1.288 12 min, and was able to distinguish eight nonvolatile ginger con- 6-Shogaol 0.338 0.35% 0.075 3.51% 0.748 stituents. 6-Paradol 0.041 1.19% 0.186 5.45% 0.846 10-Gingerol 0.456 1.73% 0.386 2.35% 0.523 Acknowledgement 8-Shogaol 0.068 0.48% 0.081 4.16% 0.696 10-Shogaol 0.149 0.52% 0.098 5.19% 0.977 Total 3.075 0.90% 0.269 1.87% 0.555 Authors would like to thank Dr. Michael Chan from the British a 101 Zingerone 0.498 0.97% 0.218 3.66% 0.826 Columbia Institute of Technology for providing valuable assistance in 6-Gingerol < LOQ N/A N/A N/A N/A data processing. The funding was provided through an inter-company 8-Gingerol < LOQ N/A N/A N/A N/A 6-Shogaol < LOQ N/A N/A N/A N/A Research & Development Grant. 6-Paradol < LOQ N/A N/A N/A N/A 10-Gingerol < LOQ N/A N/A N/A N/A Appendix A. Supporting information 8-Shogaol < LOQ N/A N/A N/A N/A 10-Shogaol < LOQ N/A N/A N/A N/A Supplementary data associated with this article can be found in the Total 0.498 0.97% 0.218 3.66% 0.826 986a Zingerone 0.111 2.81% 0.517 12.46% 2.246 online version at doi:10.1016/j.talanta.2018.10.075. 6-Gingerol 24.419 0.31% 0.124 1.92% 0.776 8-Gingerol 4.212 4.11% 1.277 5.18% 1.610 References 6-Shogaol 2.862 0.45% 0.132 3.14% 0.922 6-Paradol 0.492 1.03% 0.232 6.59% 1.485 [1] H. Wohlmuth, Phytochemistry and pharmacology of plants from the ginger family, 10-Gingerol 6.460 0.60% 0.198 3.86% 1.279 Zingiberaceae (2008). 8-Shogaol 0.538 0.29% 0.068 6.70% 1.530 [2] M.S. Butt, M.T. Sultan, Ginger and its health claims: molecular aspects, Crit. Rev. 10-Shogaol 1.081 0.56% 0.141 4.74% 1.201 Food Sci. Nutr. 51 (5) (2011) 383–393. Total 40.175 0.12% 0.051 1.19% 0.517 a [3] C.-K. Wei, Y.-H. Tsai, M. Korinek, P.-H. Hung, M. El-Shazly, Y.-B. Cheng, Y.-C. Wu, 648 Zingerone < LOQ N/A N/A N/A N/A T.-J. Hsieh, F.-R. Chang, 6-Paradol and 6-shogaol, the pungent compounds of 6-Gingerol < LOQ N/A N/A N/A N/A ginger, promote glucose utilization in adipocytes and myotubes, and 6-paradol 8-Gingerol < LOQ N/A N/A N/A N/A reduces blood glucose in high-fat diet-fed mice, Int. J. Mol. Sci. 18 (1) (2017) 168. 6-Shogaol < LOQ N/A N/A N/A N/A [4] W. Marx, E.A. Isenring, A.E. Lohning, Determination of the concentration of major 6-Paradol 0.011 1.16% 0.147 3.07% 0.391 active anti-emetic constituents within commercial ginger food products and dietary 10-Gingerol < LOQ N/A N/A N/A N/A supplements, Eur. J. Integr. Med. 10 (2017) 19–24. 8-Shogaol < LOQ N/A N/A N/A N/A [5] Y. Tao, W. Li, W. Liang, R.B. Van Breemen, Identification and quantification of 10-Shogaol < LOQ N/A N/A N/A N/A gingerols and related compounds in ginger dietary supplements using high-per- Total 0.011 1.16% 0.147 3.07% 0.391 formance liquid chromatography–tandem mass spectrometry, J. Agric. Food Chem. a 57 (21) (2009) 10014–10021. 889 Zingerone < LOQ N/A N/A N/A N/A [6] H.A. Schwertner, D.C. Rios, High-performance liquid chromatographic analysis of 6-Gingerol 2.592 1.54% 0.449 8.18% 2.364 6-gingerol, 8-gingerol, 10-gingerol, and 6-shogaol in ginger-containing dietary 8-Gingerol 0.434 0.87% 0.191 7.98% 1.764 supplements, spices, teas, and beverages, J. Chromatogr. B 856 (1–2) (2007) 41–47. 6-Shogaol 1.064 0.64% 0.161 7.74% 1.957 [7] H.A. Schwertner, D.C. Rios, J.E. Pascoe, Variation in concentration and labeling of 6-Paradol < LOQ N/A N/A N/A N/A ginger root dietary supplements, Obstet. Gynecol. 107 (6) (2006) 1337–1343. 10-Gingerol 0.671 0.56% 0.133 6.16% 1.454 [8] B.P. Gaire, O.W. Kwon, S.H. Park, K.-H. Chun, S.Y. Kim, D.Y. Shin, J.W. Choi, 8-Shogaol 0.274 0.55% 0.113 8.14% 1.680 Neuroprotective effect of 6-paradol in focal cerebral ischemia involves theat- 10-Shogaol 0.434 0.42% 0.094 8.49% 1.878 tenuation of neuroinflammatory responses in activated microglia, PLoS One 10(3) Total 5.469 1.05% 0.341 4.14% 1.338 (2015) e0120203. [9] S.-G. Shin, J.Y. Kim, H.Y. Chung, J.-C. Jeong, Zingerone as an antioxidant against

801 H. You et al. Talanta 194 (2019) 795–802

peroxynitrite, J. Agric. Food Chem. 53 (19) (2005) 7617–7622. the raw herb and dried aqueous extract, J. AOAC Int. 90 (5) (2007) 1219–1226. [10] H. Kabuto, M. Nishizawa, M. Tada, C. Higashio, T. Shishibori, M. Kohno, Zingerone [16] G.G. Angeli, et al., Extraction and Analysis of Fresh Ginger Root and Ginger Dietary [4-(4-hydroxy-3-methoxyphenyl)–2-butanone] prevents 6-hydroxydopamine-in- Supplement. 〈https://www.academia.edu/10167567/Extraction_and_Analysis_of_ duced dopamine depression in mouse striatum and increases superoxide scavenging Fresh_Ginger_Root_and_Ginger_Dietary_Supplement〉. activity in serum, Neurochem. Res. 30 (3) (2005) 325–332. [17] R. Lantz, G. Chen, M. Sarihan, A. Solyom, S. Jolad, B. Timmermann, The effect of [11] A. Bzhelyansky, G. Atkinson, L. Brown, P. Burns, N.E. Es-Safi, C. Farthing, extracts from ginger rhizome on inflammatory mediator production, Phytomedicine A. Horkey, H. Johnson, A. Kuszak, A. Lippert, AOAC SMPR® 2017.012, J. AOAC Int. 14 (2–3) (2007) 123–128. 100 (4) (2017) 1192–1196. [18] S. Yudthavorasit, K. Wongravee, N. Leepipatpiboon, Characteristic fingerprint [12] D. Manasa, P. Srinivas, H. Sowbhagya, Enzyme-assisted extraction of bioactive based on gingerol derivative analysis for discrimination of ginger (Zingiber offici- compounds from ginger (Zingiber officinale Roscoe), Food Chem. 139 (1–4) (2013) nale) according to geographical origin using HPLC-DAD combined with chemo- 509–514. metrics, Food Chem. 158 (2014) 101–111. [13] W.J. Youden, E. Steiner, Ruggedness test for procedures, Stat. Man. AOAC (1975) [19] S. Ok, W.-S. Jeong, Optimization of extraction conditions for the 6-shogaol-rich 33–36. extract from ginger (Zingiber officinale Roscoe), Prev. Nutr. Food Sci. 17 (2) (2012) [14] A.H.P. Association, Zingiber officinale (rhizome), 2018. 〈http://www. 166. botanicalauthentication.org/index.php/Zingiber_officinale_(rhizome)〉 (Accessed [20] S. Varakumar, K.V. Umesh, R.S. Singhal, Enhanced extraction of oleoresin from 19 August 2018). ginger (Zingiber officinale) rhizome powder using enzyme-assisted three phase [15] S. Lee, C. Khoo, C.W. Halstead, T. Huynh, A. Bensoussan, Liquid chromatographic partitioning, Food Chem. 216 (2017) 27–36. determination of 6-, 8-, 10-gingerol, and 6-shogaol in ginger (Zingiber officinale) as

802