GC–MS Profiling of Triterpenoid Saponins from 28 Quinoa Varieties

Aug 15, 2016 - Woldemichael , G. M.; Wink , M. Identification and biological activities of triterpenoid saponins from Chenopodium quinoa J. Agric. Foo...
9 downloads 10 Views 1MB Size
Article pubs.acs.org/JAFC

GC−MS Profiling of Triterpenoid Saponins from 28 Quinoa Varieties (Chenopodium quinoa Willd.) Grown in Washington State Ilce G. Medina-Meza,† Nicole A. Aluwi,† Steven R. Saunders,‡ and Girish M. Ganjyal*,† †

School of Food Science, Washington State University, Pullman 99164, Washington, United States The Gene and Linda Voiland School of Chemical Engineering and Bioengineering, Washington State University, Pullman 99164, Washington, United States



ABSTRACT: Quinoa (Chenopodium quinoa Willd) contains 2 to 5% saponins in the form of oleanane-type triterpenoid glycosides or sapogenins found in the external layers of the seeds. These saponins confer an undesirable bitter flavor. This study maps the content and profile of glycoside-free sapogenins from 22 quinoa varieties and 6 original breeding lines grown in North America under similar agronomical conditions. Saponins were recovered using a novel extraction protocol and quantified by GC−MS. Oleanolic acid (OA), hederagenin (HD), serjanic acid (SA), and phytolaccagenic acid (PA) were identified by their mass spectra. Total saponin content ranged from 3.81 to 27.1 mg/g among the varieties studied. The most predominant sapogenin was phytolaccagenic acid with 16.72 mg/g followed by hederagenin at 4.22 mg/g representing the ∼70% and 30% of the total sapogenin content. Phytolaccagenic acid and the total sapogenin content had a positive correlation of r2 = 0.88 (p < 0.05). Results showed that none of the varieties we studied can be classified as “sweet”. Nine varieties were classified as “lowsapogenin”. We recommend six of the varieties be subjected to saponin removal process before consumption. A multivariate analysis was conducted to evaluate and cluster the different genotypes according their sapogenin profile as a way of predicting the possible utility of separate quinoa in food products. The multivariate analysis showed no correlations between origin of seeds and saponin profile and/or content. KEYWORDS: triterpenoid sapogenins, quinoa, total saponin content, GC−MS, multivariate analysis



INTRODUCTION Saponins are secondary metabolites of plants and are divided into two major classes, triterpenoid and steroid glycosides. Saponins vary by the numbers and position of sugar units linked to the hydrophobic aglycon, mainly arabinose, glucose, galactose, glucuronic acid, xylose, and rhamnose.1 The biological activities of triterpene saponins have sparked interest in pharmacology due to clinical applications,2 in agriculture for their antifungal properties,3,4 and in the food industry as additives, flavor modifiers, and preservatives.5 Saponins are prominent in the Amaranthaceae and Chenopodiaceae families and were merged into a single family (Amaranthaceae) in the past decade due to many shared relevant features of morphology, anatomy, and phytochemistry.6 The saponins in quinoa (Chenopodium quinoa Willd.) are glycosylated triterpenoids, consisting of a pentacyclic C30 skeleton (sapogenin). Saponins are present in all parts of the quinoa plant (including leaves, flowers, fruits, seeds, and seed coats) and their concentration ranges from 0.01% to 5% on a dry weight basis.7,8 Quinoa varieties have been classified according to their saponin content as “sweet” (free or less than 0.11 g/100 g DW) and “bitter” varieties (more than 0.11 g/100 g DW).9 Because of the associated bitterness and toxicity, quinoa is treated to reduce saponin levels by washing, dehulling, or thermal processing.10−12 Currently, there is a major effort underway to develop new quinoa varieties for cultivation in North America. The quinoa breeding program at the Washington State University (WSU) is testing several quinoa varieties in the state of Washington, which are being extensively studied in their functionality and © 2016 American Chemical Society

processability to develop a better understanding of the cultivar characteristics before they are released for broader cultivation. This will aid in the building of a successful quinoa industry. For rapid detection of saponins in quinoa flours, the plant material is dispersed in distilled water and vigorously shaken for 2 min, a method known as the afrosimetric method.13 The formation of a stable and persistent foam at the liquid/air interface (15 min) indicates the presence of saponins. Since this method is qualitative, or semiquantitative at best, a significant overestimation of saponin content can often occur.11 This procedure is effective since saponin analysis is difficult due to the high polarity and structural complexity of these compounds that complicates isolation, identification, and quantification of saponins analytically. The conventional analytical methods for saponin quantification involve extraction with a large quantity of solvent (ethanol, butanol) at high temperatures (110 °C).14 The extraction protocols are generally time-consuming with maceration steps, a Soxhlet or reflux extraction, and followed by a second n-butanol extraction. The n-butanol separation step is problematic since saponins with higher polarity may remain in the aqueous phase and are not recovered.15 To overcome this problem, solid phase extraction (SPE) and/or thin layer chromatography (TLC) have been used as preliminary separation techniques.1 More advanced chromatographic techniques including gas chromatography,12,16 low pressure Received: Revised: Accepted: Published: 8583

May 11, 2016 July 19, 2016 August 15, 2016 August 15, 2016 DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry Table 1. Names, Source, and Origin of the Quinoa Varieties Used in the Study sample ID

name

Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Q19 Q20e Q21 Q22 Q23 Q24 Q25 Q26 Q27 Q28 Q29

Titicacaa Kaslaea QQ74 Black Blanca KU2 Isluga Linares Punoa Cahuil QuF9P1-20 Red Headc NL-6 Cherry Vanillac Oro de Vallec CO 407 WMF CO 407 Dave French Vanillac Temuco Black WWA BBR (Brightest Brilliant Rainbow)c UDEC1 QuF9P40-29 QuF9P39-51 QuF9P39-72 Japanese Strain QuF9P39-73 QuF9P39-64 UBCd

type variety variety variety variety variety variety variety variety variety variety breeding variety variety variety variety variety variety variety variety variety variety variety breeding breeding breeding variety breeding breeding variety

line

line line line line line

seed source

origin

Sven-Erik Jacobsen (University of Copenhagen) USDA NPGS, Ames 13745 (New Mexico, USA) USDA-NPGS, PI 614886 White Mountain Farm (Colorado, USA) White Mountain Farm (Colorado, USA) BYU-WSUb (Cambridge, UK) USDA NPGS, Ames 13743 Backyard beans and grains Sven-Erik Jacobsen (University of Copenhagen) White Mountain Farm (Colorado, USA) BYU-WSUb Wild Garden Seeds Wageningen University (Netherlands) Wild Garden Seeds Wild Garden Seeds White Mountain Farm, (Colorado, USA) USDA-NPGS, PI 596293 (Colorado, USA) Wild Garden Seeds Bountiful Gardens (California, USA) White Mountain Farm (Colorado, USA) Wild Garden Seeds USDA-NPGS, PI 634923 BYU-WSUb BYU-WSUb BYU-WSUb USDA-NPGS, PI 677100 (Mink Gaylord) BYU-WSUb BYU-WSUb Lundberg Family Farms

Denmark Bolivia Maule, Chile Bolivia Peru Chile Chile Chile Denmark Chile Washington, USA Oregon, USA Chile Oregon, USA Oregon, USA Chile Chile Oregon, USA Chile USA Oregon, USA Chile Washington, USA Washington, USA Washington, USA Washington, USA Washington, USA Washington, USA Bolivia

a

Titicaca and Puno are based on crosses between Chilean and Peruvian material, followed by selection for long day length in North Europe.31,32 BYU-WSU: Brigham Young University−Washington State University breeding program. cBBR, Cherry Vanilla, French Vanilla, Oro de Valle, and Red Head come from the lines CO 406, CO407 and CO 409 (Frank Morton, personal communication). dUBC: Unscarified black commercial, unknown agronomical conditions. eBlack WWA originated from the heterogeneous variety Black from White Mountain Farm, and was subsequently subjected to two years of selection in western Washington, USA. b

Quinoa Samples. Twenty-eight quinoa (Chenopodium quinoa Willd.) varieties were obtained from the quinoa breeding program between Brigham Young University (BYU) and Washington State University (WSU), as well as other sources. The main characteristics and the geographic origin of the varieties are given in Table 1. Six varieties were breeding lines (Q11, Q23, Q24, Q25, Q27, and Q28) from BYU and WSU.12,17All varieties were grown in field plots (6′ by 20′) at a seeding rate of 5.4 kg/ha under organic conditions near Chimacum, WA (48.0167° N, 122.7667° W). Soil conditions included 26 μg/g of P, 52 μg/g of NO2, 50 μg/g of NH3, and 460 μg/g of K with 9% organic matter and a soil pH of 5.8. The quinoa was harvested during September 2014 and stored in a dry room until milling. An unscarified black commercial variety (UBC, abbreviated Q29 in the rest of this work) was obtained from Lundberg Family Farms and used for comparison due to its high saponin content. Sapogenin Extraction. The flowchart of the extraction protocol is shown in Figure 1. Quinoa seeds were ground into a fine powder using a cyclone mill (model #3010-030, UDY Corporation, Fort Collins, CO) equipped with a 2 mm sized screen. The flour was then defatted according to the AACCI method (30-25.01) and stored in polyethylene zip closure bags at room temperature for further analysis. To assess the saponin content, ca. 5.00 g of defatted quinoa was extracted using a Soxhlet reflux extractor with methanol/water (4:1 v/ v) for 3 h. The crude extract was then concentrated under reduced pressure (68 kPa) at 60 °C. A 5 mL aliquot of the crude concentrate was hydrolyzed with HCl 6 N at 110 °C for 2 h. The hydrolysate was cooled, neutralized with a NH4OH, and centrifuged at 3000g for 5 min. Sapogenins were then extracted by liquid−liquid partition with

liquid (LPLC), medium pressure liquid (MPLC), and high pressure liquid (HPLC), or normal and reverse phase or ion exchange chromatography have been used for the analytical separation of saponins.2 Due to the complexity of the extraction protocols, there is a need for a faster and more sensitive method for the analysis of sapogenins in crops such as quinoa. Plant breeders require rapid and reliable methods to quantify the total content of these compounds such as UV−vis spectroscopy; food chemists require identification and quantification of individual compounds in addition to changes in the total content that may occur during processing. Having a more effective and reliable analytical method is critical to assess functionality and processability of quinoa. Thus, the aim of this research was to evaluate the total saponin contents and the sapogenin profiles of 28 quinoa varieties grown in the state of Washington (USA) with a new protocol for extraction, separation, and quantification using UV−vis spectroscopy and GC mass spectrometry. Along with these 28 varieties one commercial variety was also included in the study.



MATERIALS AND METHODS

Chemicals. All reagents, oleanolic acid, and solvents unless noted were obtained from Sigma-Aldrich (St. Louis, MO). HPLC grade methanol, ethyl acetate, and n-hexane were purchased from J.T. Baker (Center Valley, PA). 8584

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry

afterward the mixture was cooled in ice water. The absorbance at 527 nm was measured (UV−vis spectrometer: UV-3100PC, VWR International LLC., Visalia, CA). Glacial acetic acid was used as blank. The total saponin contents were obtained by comparing with a standard curve of oleanolic acid (100 μg/mL to 1000 μg/mL) with results expressed as mg/mL of oleanolic acid equivalent. The analysis was done in triplicate. Triterpenoid Sapogenin Quantification by Gas Chromatography−Mass Spectrometry (GC−MS). Quinoa triterpenoid profiles were determined by taking an aliquot of the final extract dried under a N2 stream. Then to derivatize the extracted saponins, 100 μL of anhydrous pyridine, 100 μL of bis(trimethylsilyl) trifluoroacetamide (BSTFA), and 10 μL of cholesteryl decanoate (internal standard, IS) were added to the dried extract and heated at 70 °C for 1 h. The derivatized extracts (2 μL) were injected into a Trace 1310 gas chromatograph (ThermoScientific, Waltham, MA) fitted with a Zebron ZB-5HT (Phenomenex, Torrance, CA) capillary column (30 m × 0.25 mm × 0.25 μm). Injector and detector temperatures were both set at 350 °C, whereas the oven temperature was programmed from 160 to 220 °C at 15 °C/min, from 220 to 290 °C at 10 °C/min, for 7 min, from 290 to 330 °C at 8 °C/min hold for 15 min. Splitless mode and carrier pressure of 100 kPa were used. Identification and structural confirmation of saponins was done by an ISQ single quadrupole mass spectrometer (Thermo Scientific, Waltham, MA) in the electron impact mode (70 eV). The MS data were recorded in full scan mode (mass range of m/z 50−900 amu). MS transfer line and ion source temperature were set at 350 °C. The analyses were performed in triplicate for each variety of quinoa and sapogenin content quantified as peak area under the curve relative to that of the internal standard. Statistical Analyses. Results were expressed as the mean value ± standard deviation (SD) of the independent extractions (n = 3). Oneway analysis of variance was used to compare the means by Tukey’s test. A comparison of assay was made by correlation and linear regression analysis. Principal component analysis (PCA), hierarchical cluster analysis, and k-means test were performed. The PCA was performed on the correlation matrix using the normalized response variables (y′ij) defined as follows:

Figure 1. Flowchart showing the extraction protocol used for the quantification of sapogenins. ethyl acetate (3 × 5 mL). The fractions were then combined and filtered in a bed of anhydrous Na2SO4. The final extract was stored at −20 °C in a tube until further analysis. All extractions were performed in triplicate. Total Saponin Content by UV−Vis Spectrophotometry. Total saponin content was measured according to Penafiel et al.14 with minor modifications. A saponin extract (250 μL) was placed in a tube with 1000 μL of the reagent mix (glacial acetic acid/sulfuric acid 1:1 v/ v) for color development, vortexed vigorously (30 s), and then heated at 60 °C for 30 min, during which time a purple color was developed;

yij′ =

yij − y·j̅ y·j̅

where yij is the response j obtained in the experiment i and y.̅ j is the arithmetic average of the response j on the whole of the experiments. The principal components were chosen by visualization of the scree plot and considering the cumulative variance.18 The selected components were used to perform a hierarchical clustering analysis

Table 2. Summary of Reported Extraction Methods and Quantification Conditions of Saponins from Chenopodium Willd. plant material

extraction method

seeds seeds, flour, bran, perisperm, and embryo flour seed hulls flour seeds and callus flour seeds seeds

maceration Soxhleta

seeds

maceration for 3 days (two times) macerationa stirring for 1 h

husks seeds a

Soxhlet for 16 and 30 h stirring for 3.5 h reflux for 3 h Soxhlet three times Soxhlet for 24 h Soxhlet for 72 h macerationa

solvent methanol 80% Eethanol/H2O v/v and butanol chloroform, methanol 50% methanol/H2O v/v methanol/HCL (0.2 N) 1-butanol/ethanol/water methanol methanol 90% Methanol/H2O v/v 95% methanol/H2O v/v NaOH/H2O v/v 40% ethanol/H2O v/v

separation LPLC TLC/LPLC GC LPLC GC GC HPLC LPLC LPLC (Sephadex LH-20 column) LPLC RP-HPLC LPLC (Diaion HP-20 column)

quantification

reference

HPLC gravimetric method MS HPLC MS MS HPLC HPLC HPLC

7 8 12 21 20 16 25 19 22

HPLC

33

LC/MS UPLC/Q-TOFMS

34 35

Information about the extraction time is not provided by the authors. 8585

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry to group the varieties according to a particular variety’s sapogenin profile. The distances among clusters were computed with a k-means test on the calculated centers. Differences were considered significant at p ≤ 0.01. All statistical analyses were performed with OriginPro version 9.4 (OriginLab, Northampton, MA, USA).

protocols for the extraction of saponins from quinoa in terms of solvent, extraction method, quantification method, and extraction times. In this study, we used a shorter reflux extraction (3 h) in comparison with previous studies.12,19 A solvent mixture of 80% methanol in water was used to achieve a complete extraction of both polar and nonpolar components from the sample instead of solvent−acid mixtures (methanol/ HCl 0.2 N)20 or the most used n-butanol neat19,21 or mixed with H2O.22 The reflux extraction used in this study was capable of removing most of the pigments present resulting in a light yellow extract for the final quantification. A longer reflux time did not significantly improve the extraction yield (data not shown). The hydrolysis step was included to remove glycosides from the triterpenoid before its characterization by chromatography. Finally, the liquid−liquid partition reduced the extraction time avoiding tedious isolation and purification steps such as TLC or SPE, as well as the volume of solvent used. The extraction method developed in this study may be applied for total saponin quantification either by UV−vis or by chromatographic methods. The quantification of total saponins by UV−vis spectroscopy is shown in Figure 2. This method leverages the reaction of oxidized triterpenoid saponins with acetic acid. Sulfuric acid was used as an oxidant, and the reaction mixture develops a purple color. Significant differences in the total saponin contents were detected in the 28 varieties used in the study, and as a first approximation we were able to distinguish three groups (p < 0.01). The total saponin content ranged from 0.66



RESULTS AND DISCUSSION Extraction and Total Saponin Quantification by UV− Vis Spectrophotometry. Table 2 summarizes reported

Figure 2. Total saponin content (g/100 g DW) in the quinoa varieties.

Figure 3. An example of GC/MS chromatogram of trimethylsilylated of triterpenoid saponins from quinoa. 8586

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry

Figure 4. GC−MS spectrum and fragmentation assignment of serjanic acid from quinoa.

Quantification of Sapogenins by GC−MS. Gas chromatography methods are useful for the aglycon (sapogenin) separation and quantification. Calibration curves of the internal standard and authentic triterpenoids standards were carried out. Calibration curves were prepared over concentration ranges similar to those expected to be present in the extracts over a linear response range (0.01−1 mg/mL) and with correlation coefficients (r2) greater than 0.98. Five silylated aglycons were simultaneously separated in our experiments. The chromatographic profile of quinoa sapogenins obtained is shown in Figure 3. Trimethylsilylated oleanolic acid (OA), hederagenin (HD), serjanic acid (SA), and phytolaccagenic acid (PA) were confirmed by comparison of retention times (26.3, 27.0, 30.5, 32.7 min, respectively) and mass spectra to the pure standards. This profile is in agreement with previous reports.21,24 A high resolution separation was achieved within 35 min. The major fragmentation reactions of the silylated aglycons consisted in the elimination of TMSi, formic acid, and the formaldehyde resulting in a characteristic fragmentation pattern for each aglycon. Minors fragmentations included rupture of the polycyclic ring (ABC*) and loss of moiety at the lateral chain (C28). OA fragmentation ions are found at m/z 600 (molecular ion, M+), 585 (M+ − CH3), 482 (M+ − TMSiOOH), and 393 (M+ − TMSiOH − TMSiOOCH). HD fragmentation was found at m/z 688 (molecular ion, M+), 598 (M+ − TMSiOH), 570 (M+ − TMSiOOCH), 320 (M+ − ABC* rings), and 203 (M+ − ABC* rings − C28 moiety). SA fragmentation was at m/z 717 (molecular ion, M+), 644 (M+ −

to 3.09 g/100 g DW for Q21 and Q12 respectively. The total saponin content for Q12 was higher in comparison with the commercial variety (Q29) with a value of 2.7 g/100 g DW. These results agreed with the concentration range of several quinoa varieties previously reported.9 The quantification of total saponins by UV−vis spectroscopy applied to quinoa seed extract was simple, fast, and inexpensive to perform. Several spectrophotometry studies have been reported for the quantification of plant saponins, such as for soymilk, rice bran, and Ipomoea batatas.1 Unfortunately, due to differences in the spectrophotometric protocols, any comparison with our method in terms of extraction and quantification yields is hard. It is important to note that since the total saponin method is based on the reaction of saponins with oxidizing agents (perchloric acid, acetic acid, or sulfuric acid), the selection of the saponin standard used for quantification is critical. In fact, the standard should belong to the classification group of the most representative sapogenin in the considered food matrix (e.g., oleanolic acid for triterpenoids and diosgenin, hecogenin, or smilagenin for steroidal saponins).23 Furthermore, other factors including the selection of reagents, wavelength of light absorbed, and purity of extract should be considered before performing any spectroscopic quantitation, since an error of about 10% may occur in comparison with chromatographic quantification.14 A full screening of wavelengths should be performed to determine the wavelength of maximum absorption (λmax), avoiding reading errors during the development of purple color. 8587

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry

Figure 5. Percentage of individual sapogenins vs total saponin content for quinoa varieties.

TMSi), 629 (M+ − TMSiOH), 526 (M+ − TMSiOOCH), and 364 (M+ − CH3COOH − TMSiOH − CH3) (Figure 4). PA fragmentation signal was at m/z 732 (molecular ion, M+), 642 (M+ − TMSiOH), 614 (M+ − TMSiCOOH), 364 (M+ − ABC* rings), and 247 (M+ − ABC* rings − C28 moiety). The overall fragmentation patterns were similar to those in the literature.12,16,20 In the present study, PA was found to be the major triterpenoid in all varieties ranging from 44.55% to 87.97% (Figure 5). A positive relationship (r2 = 0.882, p < 0.01) between total saponin content and PA was found. These results are in agreement with other literature studies including two Chilean ecotypes grown in the U.K.12 and a Latinreco-40057 variety from Ecuador.25 Other authors20,24,26 showed that oleanoic acid and serjanic acid are the most prominent triterpenoids in quinoa seeds. Differences observed in the sapogenin contents could be due to the cross-fertilization that quinoa exhibits (almost 10%).24 It is important to note that the varieties we studied were all grown in the same agronomical conditions. A fifth sapogenin was separated at 34 min (compound 5), and the observed fragmentation was 732 (molecular ion, M+), 717 (M+ − CH3), 688 (M+ − CO2), 642 (M+ − TMSiOH), 614 (M+ − TMSiOOCH), 512 (unknown fragmentation), 435 (loss of benzyl ring from fragment 512 m/z), and 278 (unknown fragmentation) m/z. Up to seven sapogenins have been identified in quinoa.7 We initially speculated that the compound 5 could be an enantiomer of HD (possibly echinocystic acid or queretaroic acid); however, the molecular weight and fragmentation pattern did not correspond to those compounds. Further analysis should be performed to determine the structure of the fifth sapogenin.

It has been shown that, in quinoa seed, the amount of individual saponins varies according to the cellular location, with the bran fraction containing more than 80% of the total content.8 Industrial washing of the seeds increases the content of OA from 30 to 36% and decreases the content of PA from 43 to 36% after washing.12 Bitterness may be related to the PA content, while sweet varieties have no detectable amount of this class of sapogenin.27 This suggests that it will be useful to explore the functional properties of this triterpenoid and its glycosides. Literature regarding single sapogenin content is hard to compare with our results for some of the varieties studied. Some of the quinoa varieties (Q11, Q23, Q24, Q25, Q27, and Q28) are original breeding lines from the WSU breeding program (Table 1). Q21 contained the lowest content of OA while Q2 had the highest value with 0.09 mg/g and 2.80 mg/g, respectively; in comparison with the commercial variety Q29 with 5.18 mg/g. Q21 presents the lowest value of HD while Q12 contains the highest amount with 0.44 mg/g and 4.22 mg/ g, respectively. The commercial variety Q29 contained 5.86 mg/g. Q15 and Q2 were found with the lowest (0.06 mg/g) and highest (1.66 mg/g) amounts respectively of SA whereas Q29 had 1.54 mg/g. Remarkably, both OA and SA were not detected in the WSU breeding line ecotype Q11. Q13 was found with the lowest value of PA with 1.70 mg/g, whereas Q26 was found with the highest content (16.72 mg/g). Total saponin content of variety Q1 (Titicaca) was 16.75 mg/g. This result is in agreement with previous studies20,28 which used the same variety (Titicaca). Further, those works indicated that different irrigation regimes can quantitatively affect the saponin response in quinoa cultivars. Q10 (Cahuil) 8588

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry Table 3. Sapogenin Profiles in Quinoa Varieties (mg/g DW)a variety Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Q19 Q20 Q21 Q22 Q23 Q24 Q25 Q26 Q27 Q28 Q29 a

oleanolic acid 1.83 2.80 0.63 0.84 0.56 0.60 0.69 0.45 0.15 2.44 ND 1.61 0.48 0.74 0.15 0.38 0.21 0.62 0.39 0.50 0.09 0.16 0.51 1.21 0.79 0.50 1.04 0.42 5.18

± ± ± ± ± ± ± ± ± ±

0.45 0.16 0.28 0.05 0.14 0.29 0.17 0.13 0.07 0.82

ab ab b b b b b b b ab

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.69 0.32 0.58 0.04 0.15 0.04 0.32 0.19 0.20 0.01 0.05 0.37 0.29 0.19 0.08 0.12 0.13 1.09

b b b b b b b b b b b b b b b b b a

hederagenin 2.59 3.92 2.28 1.07 0.75 2.40 3.26 1.57 0.99 1.91 2.00 4.22 1.15 1.75 0.55 0.77 0.70 1.52 1.61 0.79 0.44 0.64 1.46 2.79 1.39 2.56 1.08 1.36 5.86

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

1.42 1.27 1.48 0.53 0.27 1.25 1.04 0.28 0.34 0.84 0.80 0.69 0.18 0.77 0.39 0.85 0.87 0.11 0.25 0.15 0.92 0.31 0.59 0.68 0.45 0.45 0.64 0.13 0.74

abc abc abc bc bc abc abc bc bc bc bc ab bc bc c bc bc bc bc bc c bc bc abc bc abc bc bc a

sejanic acid 0.54 1.66 0.29 0.44 0.46 0.44 0.40 0.66 0.10 0.68 ND 0.27 0.45 0.75 0.06 0.38 0.25 0.34 0.22 0.45 0.07 0.32 0.41 1.14 0.69 0.51 0.11 0.46 1.54

± ± ± ± ± ± ± ± ± ±

0.32 0.50 0.06 0.04 0.05 0.06 0.13 0.03 0.05 0.20

abc a bc abc abc abc abc abc c abc

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.08 0.13 0.18 0.03 0.09 0.04 0.04 0.03 0.02 0.05 0.05 0.09 0.36 0.12 0.13 0.04 0.05 0.52

bc abc abc c abc bc abc bc abc c abc abc abc abc abc c abc ab

phytolaccagenic acid 10.89 ± 2.38 abc 11.47 ± 2.66 abc 7.08 ± 3.00 abc 4.51 ± 1.31 bc 3.42 ± 1.11 bc 10.86 ± 2.35 abc 13.80 ± 3.57 ab 9.58 ± 2.82 abc 6.78 ± 1.29 abc 5.91 ± 0.08 abc 8.46 ± 0.20 abc 10.08 ± 0.29 abc 1.70 ± 0.10 c 2.50 ± 0.97 bc 4.72 ± 1.19 bc 4.82 ± 1.34 bc 4.94 ± 1.78 abc 7.07 ± 2.17 abc 6.29 ± 1.82 abc 4.51 ± 1.14 bc 4.33 ± 1.65 bc 5.14 ± 1.87 abc 10.03 ± 2.92 abc 13.36 ± 3.59 abc 8.25 ± 2.20 abc 16.72 ± 2.68 a 6.63 ± 1.64 abc 8.86 ± 1.97 abc 14.02 ± 1.37 ab

peak 5 RT 34.0

total

± ± ± ± ± ± ±

16.75 ± 3.95 ab 20.28 ± 3.82 ab 10.34 ± 1.10 ab 7.05 ± 3.31 b 5.29 ± 2.73 b 14.91 ± 3.95 ab 18.44 ± 4.06 ab 12.25 ± 3.41 ab 8.05 ± 3.04 b 10.95 ± 2.21 ab 10.46 ± 2.10 ab 16.17 ± 2.02 ab 3.81 ± 1.45 b 5.78 ± 1.00 b 5.61 ± 2.23 b 6.35 ± 3.95 b 6.10 ± 3.26 b 9.54 ± 3.09 ab 9.05 ± 3.23 b 6.27 ± 2.45 b 4.93 ± 1.85 b 6.33 ± 2.02 b 12.73 ± 3.91 ab 18.89 ± 4.07 ab 11.12 ± 2.79 ab 21.08 ± 3.05 ab 9.53 ± 3.01 ab 11.10 ± 3.19 ab 27.11 ± 3.90 a

0.91 0.30 0.07 0.15 0.10 0.46 0.20 ND ND ND ND ND ND ND 0.07 ND ND ND 0.40 ND ND 0.07 0.30 0.38 ND 0.40 0.29 ND 0.52

0.32 0.18 0.06 0.01 0.00 0.00 0.12

± 0.01

± 0.05

± 0.00 ± 0.00 ± 0.00 ± 0.00 ± 0.00 ± 0.00

P < 0.01. ND: Not detected.

Figure 6. PCA and k-means cluster analysis (A); hierarchical tree (B) for triterpenoid saponins.

higher (18.44 mg/g) in comparison with the range reported by Ward (1.12 to 6.62 mg/g).11 It is important to point out that, in our study, both Isluga and CO-407 varieties are from Chile (according to the USDA-NPGS database), whereas Ward11 used seeds from Bolivia (Isluga) and Chile (Cahuil and CO407). Although the seeds used in the our study come from a South American country, most of them have been adapted and cultivated for several years in the United States.17 A genetic study could help to understand the interaction between the

contained 10.95 mg/g of saponin, a higher amount than those reported by Ward11 and Miranda29 studies, with 4.65 mg/g and 0.39 mg/g, respectively. Different climatic and soil conditions are the principal factors for this variation. Varieties Q16 (CO 407-WMF) and Q17 (CO 407-Dave) contained saponin levels of 6.35 and 6.10 mg/g, respectively. These results were slightly lower than those reported by Ward,11 with saponin content ranging from 2.10 to 8.95 mg/g for Q16. This difference can be attributed to the fact that quantification was carried out with the crude standardized protocol involving foam formation.13 Contrarily, the total sapogenin content for Q7 (Isluga) was 8589

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry

be derived from the fragmentation patterns. 2D-NMR and FTIR analyses are ongoing for complete characterization of the sapogenins in the tested varieties. Overall, phytolaccagenic acid was the major sapogenin found, which is of great interest since its functional properties are largely unknown. More research is necessary in the area of isolation and purification of sapogenins from different seed fractions to evaluate their functional properties. Multivariate analysis showed no correlations between origin of seeds and saponin profile and/or content, probably because all were grown in similar agronomical conditions. Thus, a complete genetic mapping of the new hybrids grown in the US is suggested in order to identify specific molecular markers associated with both sapogenin content and profile.

genetic composition of the genotype and the growing environment. To the best of our knowledge, this is the first study that profiles sapogenins for the Red Head, Cherry Vanilla, French Vanilla, Oro de Valle, BBR, Japanese Strain variety types, as well as for the six breeding lines (QuF9P1-20, QuF9P40-29, QuF9P39-51, QuF9P39-72, QuF9P39-73, and QuF9P39-64) (Table 3). Our results show that none of varieties we studied can be classified as “sweet”. Varieties Q13, Q21, Q5, Q14, Q15, Q16, Q17, Q20, and Q22 can be classified as “low-saponin” with less than 1% of total saponins content, and the remaining varieties can be classified as “bitter”. The varieties Q1, Q2, Q7, Q12, Q24, and Q 26 need to be subjected to saponin reduction prior to human consumption. The saponins can be removed by methods such as washing with tap water, scarification, polishing,25 or pearling.10 As alternatives, thermal processes such as steaming, roasting, or extrusion can also be used.30 Principal Component and Clustering Analyses. We performed a principal component analysis (PCA) to determine potential variables that can explain the variance in the data set (Figure 6). Only two components were retained (PC1 and PC2), according to the cumulative percentage of total variation (89.97%), corresponding to the first “elbow” in the scree plot (data not shown).18 The combinations of linear functions with their respective loading factors describing the principal components are the following:



AUTHOR INFORMATION

Corresponding Author

*FSHN 110, Pullman, WA 99164, USA. Phone: +001-509-3355613. Fax: +001-509-335-4815. E-mail: [email protected]. Funding

Thanks to the USDA-FSMIP program for providing funding in part for this research work. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We are very grateful to Dr. Kevin Murphy at the Department of Crop and Soil Sciences at the Washington State University for providing the quinoa seeds used in this study. We also want to thank Dr. Jeff Maughan and Dr. Rick Jeller from Brigham Young University for providing us the genetic information on the quinoa breeding lines. We are grateful to Frank Morton from Wild Gardens Seeds for his valuable information and support. We also want to thank the Lundberg Family Farms for providing us the commercial quinoa variety.

PC1 = 0.3841 × OA + 0. 4401 × HD + 0.3590 × SA + 0.3942 × PA + 0. 4560 × Tot + 0. 4082 × TotUV PC2 = 0. 5315 × OA + 0.0411 × HD + 0. 5572 × SA − 0. 5254 × PA − 0.2108 × Tot − 0.2913 × TotUV



According to the loadings, PC1 explains the total saponin contents and the HD, whereas PC2 better explains the content of OA and SA, having a contrast for PA. The scores for the two components were used as new variables to perform a hierarchical cluster analysis (Figure 6). Four clusters were identified, and the distance between clusters was computed with a k-means test on the cluster centers obtained from the hierarchical test (Figure 6A). Cluster 1 grouped the varieties with high content of PA and total content of saponins being classified as “bitter”. The main cluster (cluster 2) grouped the low-saponin content varieties, in which both principal components had an overall low score. Cluster 3 grouped those varieties whose characteristics were significantly different from the rest, specifically in the amount of PA. Finally, cluster 4 grouped the commercial variety Q29 and Q2, which are characterized for their higher content in SA and OA. In addition, these two varieties Q29 and Q2 are from the same origin (Bolivia). In conclusion, the GC−MS profiling of saponins in 28 varieties of quinoa grown under same agronomical conditions was completed. Nine varieties were classified as “low saponin” content; however, the rest of the varieties were bitter varieties with high saponin content. The varieties with high saponin content would need to be subjected to a saponin removal process before consumption. Structural information on the different aglycons was extracted from the mass spectra; however, both stereochemistry and exact structure could not

REFERENCES

(1) Cheok, C. Y.; Salman, H. A. K.; Sulaiman, R. Extraction and quantification of saponins: A review. Food Res. Int. 2014, 59, 16−40. (2) Mroczek, A. Phytochemistry and bioactivity of triterpene saponins from Amaranthaceae family. Phytochem. Rev. 2015, 14, 577−605. (3) Martin, R. S.; Ndjoko, K.; Hostettmann, K. Novel molluscicide against Pomacea canaliculata based on quinoa (Chenopodium quinoa) saponins. Crop Prot. 2008, 27, 310−319. (4) Stuardo, M.; Martin, R. S. Antifungal properties of quinoa (Chenopodium quinoa Willd) alkali treated saponnins against Botrytis cinerea. Ind. Crops Prod. 2008, 27, 296−302. (5) Bazile, D.; Bertero, H.; Nieto, C. State of the Art Report on Quinoa around the World in 2013; Food and Agriculture Organization of the United Nations (FAO) & CIRAD (Centre de coopér ation internationale en recherche agronomique pour le développement): Rome 2015. (6) Judd, W. S.; Campbell, C. S.; Kellogg, E. A.; Stevens, P. F. Plant systematics: a phylogenetic approach; Sinauer Associates: 1999; Vol. 1. (7) Kuljanabhagavad, T.; Thongphasuk, P.; Chamulitrat, W.; Wink, M. Triterpene saponins from Chenopodium quinoa Willd. Phytochemistry 2008, 69, 1919−1926. (8) Ando, H.; Chen, Y.-c.; Tang, H.; Shimizu, M.; Watanabe, K.; Mitsunaga, T. Food Components in Fractions of Quinoa Seed. Food Sci. Technol. Res. 2002, 8, 80−84. (9) Vega-Gálvez, A.; Miranda, M.; Vergara, J.; Uribe, E.; Puente, L.; Martínez, E. A. Nutrition facts and functional potential of quinoa (Chenopodium quinoa willd.), an ancient Andean grain: a review. J. Sci. Food Agric. 2010, 90, 2541−2547. 8590

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591

Article

Journal of Agricultural and Food Chemistry (10) Gómez-Caravaca, A. M.; Iafelice, G.; Verardo, V.; Marconi, E.; Caboni, M. F. Influence of pearling process on phenolic and saponin content in quinoa (Chenopodium quinoa Willd). Food Chem. 2014, 157, 174−178. (11) Ward, S. M. Response to selection for reduced grain saponin content in quinoa (Chenopodium quinoa Willd.). Field Crops Research 2000, 68, 157−163. (12) Ridout, C. L.; Price, K. R.; Dupont, M. S.; Parker, M. L.; Fenwick, G. R. Quinoa saponinsanalysis and preliminary investigations into the effects of reduction by processing. J. Sci. Food Agric. 1991, 54, 165−176. (13) Koziol, M. J. Afrosimetric estimation of threshold saponin concentration for bitterness in quinoa (Chenopodium quinoa Willd). J. Sci. Food Agric. 1991, 54, 211−219. (14) Elías Peñ a fiel, C. C.; Díaz Villar, L. Determinación espectrofotométrica de acido oleanólico y saponinas de quinua (Chenopodium quinoa Willd, variedad Kancolla). Arch. Latinoam. Nutr. 1988, 38, 113−31. (15) Oleszek, W.; Bialy, Z. Chromatographic determination of plant saponinsan update (2002−2005). Journal of chromatography A 2006, 1112, 78−91. (16) Burnouf-Radosevich, M.; Delfel, N. E.; England, R. Gas chromatography-mass spectrometry of oleanane-and ursane-type triterpenesapplication to Chenopodium quinoa triterpenes. Phytochemistry 1985, 24, 2063−2066. (17) Maughan, P. J.; Smith, S. M.; Rojas-Beltran, J. A.; Elzinga, D.; Raney, J. A.; Jellen, E. N.; Bonifacio, A.; Udall, J. A.; Fairbanks, D. J. Single Nucleotide Polymorphism Identification, Characterization, and Linkage Mapping in Quinoa. Plant Genome 2012, 5, 114−125. (18) Cadima, J.; Jolliffe, I. T. Loading and correlations in the interpretation of principle compenents. Journal of Applied Statistics 1995, 22, 203−214. (19) Woldemichael, G. M.; Wink, M. Identification and biological activities of triterpenoid saponins from Chenopodium quinoa. J. Agric. Food Chem. 2001, 49, 2327−2332. (20) Gomez-Caravaca, A. M.; Iafelice, G.; Lavini, A.; Pulvento, C.; Caboni, M. F.; Marconi, E. Phenolic Compounds and Saponins in Quinoa Samples (Chenopodium quinoa Willd.) Grown under Different Saline and Nonsaline Irrigation Regimens. J. Agric. Food Chem. 2012, 60, 4620−4627. (21) Mad, T.; Sterk, H.; Mittelbach, M.; Rechberger, G. N. Tandem mass spectrometric analysis of a complex triterpene saponin mixture of Chenopodium quinoa. J. Am. Soc. Mass Spectrom. 2006, 17, 795−806. (22) Dini, I.; Schettino, O.; Simioli, T.; Dini, A. Studies on the constituents of Chenopodium quinoa seeds: Isolation and characterization of new triterpene saponins. J. Agric. Food Chem. 2001, 49, 741− 746. (23) Baccou, J.; Lambert, F.; Sauvaire, Y. Spectrophotometric method for the determination of total steroidal sapogenin. Analyst 1977, 102, 458−465. (24) Mastebroek, H. D.; Limburg, H.; Gilles, T.; Marvin, H. J. P. Occurrence of sapogenins in leaves and seeds of quinoa (Chenopodium quinoa Willd). J. Sci. Food Agric. 2000, 80, 152−156. (25) Ruales, J.; Nair, B. M. Saponins, phytic acid, tannins and protease inhibitors in quinoa (Chenopodium quinoa, Willd) seeds. Food Chem. 1993, 48, 137−143. (26) Cuadrado, C.; Ayet, G.; Burbano, C.; Muzquiz, M.; Camacho, L.; Cavieres, E.; Lovon, M.; Osagie, A.; Price, K. Occurrence of saponins and sapogenols in Andean crops. J. Sci. Food Agric. 1995, 67, 169−172. (27) Ng, K. G.; Price, K. R.; Fenwick, G. R. A TLC method for the analysis of quinoa (Chenopodium quinoa) saponins. Food Chem. 1994, 49, 311−315. (28) Pulvento, C.; Riccardi, M.; Lavini, A.; Iafelice, G.; Marconi, E.; d’Andria, R. Yield and Quality Characteristics of Quinoa Grown in Open Field Under Different Saline and Non-Saline Irrigation Regimes. J. Agron. Crop Sci. 2012, 198, 254−263. (29) Miranda, M.; Vega-Galvez, A.; Quispe-Fuentes, I.; Rodriguez, M. J.; Maureira, H.; Martinez, E. A. Nutritional Aspects of Six Quinoa

(Chenopodium quinoa WILLD.) Ecotypes from Three Geographical Areas of Chile. Chilean J. Agric. Res. 2012, 72, 175−181. (30) Brady, K.; Ho, C.-T.; Rosen, R. T.; Sang, S.; Karwe, M. V. Effects of processing on the nutraceutical profile of quinoa. Food Chem. 2007, 100, 1209−1216. (31) Stikic, R.; Glamoclija, D.; Demin, M.; Vucelic-Radovic, B.; Jovanovic, Z.; Milojkovic-Opsenica, D.; Jacobsen, S.-E.; Milovanovic, M. Agronomical and nutritional evaluation of quinoa seeds (Chenopodium quinoa Willd.) as an ingredient in bread formulations. J. Cereal Sci. 2012, 55, 132−138. (32) Sun, Y.; Liu, F.; Bendevis, M.; Shabala, S.; Jacobsen, S. E. Sensitivity of Two Quinoa (ChenopodiumquinoaWilld.) Varieties to Progressive Drought Stress. J. Agron. Crop Sci. 2014, 200, 12−23. (33) Zhu, N.; Sheng, S.; Sang, S.; Jhoo, J.-W.; Bai, N.; Karwe, M. V.; Rosen, R. T.; Ho, C.-T. Triterpene saponins from debittered quinoa (Chenopodium quinoa) seeds. J. Agric. Food Chem. 2002, 50, 865− 867. (34) San Martín, R.; Ndjoko, K.; Hostettmann, K. Novel molluscicide against Pomacea canaliculata based on quinoa (Chenopodium quinoa) saponins. Crop Prot. 2008, 27, 310−319. (35) Verza, S. G.; Silveira, F.; Cibulski, S.; Kaiser, S.; Ferreira, F.; Gosmann, G.; Roehe, P. M.; Ortega, G. G. Immunoadjuvant activity, toxicity assays, and determination by UPLC/Q-TOF-MS of triterpenic saponins from Chenopodium quinoa seeds. J. Agric. Food Chem. 2012, 60, 3113−3118.

8591

DOI: 10.1021/acs.jafc.6b02156 J. Agric. Food Chem. 2016, 64, 8583−8591