Article pubs.acs.org/JAFC
Mid-Infrared Spectral Characteristics of Lipid Molecular Structures in Brassica carinata Seeds: Relationship to Oil Content, Fatty Acid and Glucosinolate Profiles, Polyphenols, and Condensed Tannins Hangshu Xin,† Nazir A. Khan,† Kevin C. Falk,§ and Peiqiang Yu*,†,# †
Department of Animal and Poultry Science, College of Agricultural and Bioresources, University of Saskatchewan, 51 Campus Drive, Saskatoon, Saskatchewan S7N 5A8, Canada § Agriculture and Agri-Food Canada, 107 Science Place, Saskatoon, Saskatchewan S7N 0X2, Canada # Tianjin Agricultural University, Xiqing District, Tianjin 300384, China ABSTRACT: The objectives of this study were to quantify lipid-related inherent molecular structures using a Fourier transform infrared spectroscopy (FT-IR) technique and determine their relationship to oil content, fatty acid and glucosinolate profile, total polyphenols, and condensed tannins in seeds from newly developed yellow-seeded and brown-seeded Brassica carinata lines. Canola seeds were used as a reference. The lipid-related molecular spectral band intensities were strongly correlated to the contents of oil, fatty acids, glucosinolates, and polyphenols. The regression equations gave relatively high predictive power for the estimation of oil (R2 = 0.99); all measured fatty acids (R2 > 0.80), except C14:0, C20:3n-3, C22:2n-9, and C22:2n-6; 3-butenyl, 2-OH-3-butenyl, 4-OH-3-CH3-indolyl, and total glucosinolates (R2 > 0.686); and total polyphenols (R2 = 0.935). However, further study is required to obtain predictive equations based on large numbers of samples from diverse sources to illustrate the general applicability of these regression equations. KEYWORDS: carinata seed, fatty acids, chemical compounds, lipid molecular structures
■
INTRODUCTION Ethiopian mustard (Brassica carinata), locally known as gomenzer, has many superior agronomic traits such as high potential seed yield, disease and stress tolerance, and adequate growth under the hot, semiarid growing conditions in western Canada and the Northern Plains of the United States.1−3 Moreover, compared with conventional canola seed, B. carinata has a larger seed size, resulting in a higher protein/fiber ratio in the defatted meals and cakes that are used for feeding animals. As a result, the development of B. carinata has been strongly supported by the Saskatchewan Ministry of Agriculture. Recently, interest in the development of specific lines of B. carinata for human and industrial consumption has been increased. The oil, free of erucic acid,4,5 can be used as a source of vegetable oil for human consumption, whereas the oil rich in erucic acid may be suitable for the production of bioethanol,6 biodiesel,7,8 biolubricants,9 or specialty fatty acids such as conjugated linoleic acid.10 Therefore, development of carinata lines with low erucic acid or high erucic acid content has become a breeding objective of increasing interest. Another remarkable characteristic of Brassica species is the presence of glucosinolate compounds, predominantly allylglucosinolate, butenyl-3-enylglucosinolate, and 2-OH-butenyl-3-enylglucosinolate.11 Several types of glucosinolates in the oilseeds limit its utilization in both humans and animals, owing to antinutritional and toxic effects. In the livestock industry, oilseed meals with less glucosinolate are more competitive and often receive premium payment in the commercial market. To the best of our knowledge, many B. carinata breeding programs for developing specific lines (for agricultural or industrial application) have been primarily focused on oil content and © 2014 American Chemical Society
functional fatty acid composition, and less attention has been given to the other compounds.12−14 It is possible that such selection may have altered the contents of glucosinolates, polyphenols, and condensed tannins. However, comparison research on the fatty acid profile and contents of antinutritional compounds between the yellow-seeded and brown-seeded B. carinata lines is still lacking. More recently, numerous research studies from our group have demonstrated that attenuated total reflectance−Fourier transform infrared (ATR-FT/IR) molecular spectroscopy can be used to determine the inherent differences in molecular structures of various biopolymers such as protein,15 carbohydrates,16 and lipids17 among different types of feeds. Moreover, these studies have shown that the differences in digestive behavior and metabolic characteristics of protein and carbohydrates among different feeds can be partly explained by the differences in their molecular structural makeup. However, limited attention has been paid to the intrinsic molecular structural makeup of lipids. We hypothesize that the lipid-related molecular spectral characteristics in B. carinata seeds might be related to the oil content, fatty acid and glucosinolate profiles, polyphenols, and/or condensed tannins. The present study was designed to evaluate and compare (1) oil content and fatty acid profile; (2) glucosinolate profile, total polyphenols, and condensed tannins; and (3) lipid-related inherent molecular structures in terms of CH stretching bands, Received: Revised: Accepted: Published: 7977
January 7, 2014 July 15, 2014 July 21, 2014 July 21, 2014 dx.doi.org/10.1021/jf502209x | J. Agric. Food Chem. 2014, 62, 7977−7988
Journal of Agricultural and Food Chemistry
Article
Figure 1. Typical ATR-FT/IR molecular spectrum of B. carinata seeds and canola seed (as a reference control) in the mid-IR region ca. 4000−800 cm−1. Lipid Molecular Structure by ATR-FTIR Spectroscopy. The lipid-related molecular structural band intensities (Figure 1) of each sample were generated in the mid-IR region (ca. 4000−800 cm−1) with 128 co-added scans at a spectral resolution of 4 cm−1 in a transmission mode, using a JASCO FT/IR 4200 with ATR (JASCO Corp., Tokyo, Japan). Each sample was run six times. The spectral data were collected with JASCO Spectra Manager II software, and the lipidrelated spectral band intensities were quantified using OMNIC 7.2 software (Madison, WI, USA). The lipid-related functional bands including asymmetric CH3 (CH3-as, ca. 2959 cm−1), asymmetric CH2 (CH2-as, ca. 2921 cm−1), symmetric CH3 (CH3-s, ca. 2871 cm−1), symmetric CH2 (CH2-s, ca. 2853 cm−1), ULB (ca. 3008 cm−1), and LECC (ca. 1744 cm−1) were identified according to published methodologies.15,25,26 The asymmetric and symmetric CH3 and CH2 stretching bands were identified in the CH region (ca. 2999−2800 cm−1) of the original spectra, using the second-derivative and Fourier self-deconvolution functions of the OMNIC 7.2 software. The spectral peak height, area, and their ratios were calculated from the original spectra. Multivariate Spectral Analysis. Agglomerative hierarchical cluster analysis (AHCA) and principal component analysis (PCA), as described in detail earlier,27−30 were performed on the overall data of lipid-related spectral bands, to visualize and classify the differences in the inherent molecular structures of lipids among the three types of oilseeds. The overall spectral data of CH3 and CH2 asymmetric and symmetric stretchings (ca. 2999−2800 cm−1), ULB (ca. 3035−2991 cm−1), and LECC (ca. 1788−1707 cm−1) regions were used for multivariate analyses. The AHCA results are presented as dendrograms using the linkage distance method and Ward’s algorithm cluster method without prior parametrization. The PCA results are presented as a scatter plot between the first principal factor and the second principal factor, based on loading scores. These analyses were carried out using Statistica software (version 8, StatSoft Inc., Tulsa, OK, USA). Statistical Analysis. Data on the oil content, fatty acid and glucosinolate profiles, polyphenols, and condensed tannins and univariate lipid spectral data were statistically analyzed using the Mixed Model procedure of the Statistical Analysis System (SAS version 9.2, SAS Institute Inc., Cary, NC, USA). The following model was used:
unsaturated lipid bands (ULB), and lipid ester CO carbonyl (LECC) using ATR-FT/IR spectroscopy, between the yellowseeded and brown-seeded B. carinata. Canola seeds were used as a reference. The second objective was to determine the relationship between the molecular spectral data and the contents of oil, fatty acid and glucosinolate profiles, total polyphenols, and condensed tannins in the oilseed.
■
MATERIALS AND METHODS
Seed Sampling and Processing. Seeds from newly developed yellow-seeded (AAC A100) and brown-seeded (110915EM) lines of B. carinata were evaluated in the present study. These two carinata lines were bred for the industrial oil market using traditional pedigree breeding methods. Canola seed with a brown seed coat was used as a reference. Samples (n = 4) from all seed types were obtained from plots at the Saskatoon Research Centre experimental farm of Agriculture and Agri-Food Canada (AAFC) and from contra-season research plots in the central region of Chile during the winter of 2010/ 2011. Each seed type was obtained from two replicate plots in each location. For wet chemical and spectral analysis, seeds were ground in a coffee grinder (PC770, Loblaws Inc., Toronto, Canada) for 10 s, then chilled, and reground for another 10 s. Oil Content and Fatty Acid Profile Determination. The samples (∼0.500 g) were analyzed for oil content according to the standard procedure of AOAC (method 920.39).18 Fat from the ground oilseeds (∼0.375 g) was extracted with chloroform/methanol (2:1 v/ v) according to the method of Folch et al.19 with slight modification as described by Khan et al.20 Tridecanoic acid (C13:0) was used as an internal standard. The fatty acids were separated and quantified using gas chromatography as described by Khan et al.21 All samples were analyzed in duplicate and repeated if error exceeded the acceptable level of each analysis. Determination of Glucosinolates, Polyphenols, and Condensed Tannins. All seed samples (∼0.100 g) were analyzed for glucosinolate profile according to the official method of the Canadian Grain Commission.22 The method of Slinkard and Singleton23 was used to analyze the contents of total polyphenols in samples (∼1.00 g). The HCl−butanol procedure as described by Reed24 was used to analyze the content of condensed tannins in the seeds (∼0.050 g). All samples were analyzed in duplicate and repeated if error exceeded the acceptable level of each analysis.
Yijk = μ + Fi + S(F )j + eijk 7978
dx.doi.org/10.1021/jf502209x | J. Agric. Food Chem. 2014, 62, 7977−7988
Journal of Agricultural and Food Chemistry
Article
Table 1. Oil Content and Fatty Acid Profile in the Yellow- and Brown-Seeded B. carinata with Comparison to Canola Seeds contrast, P value
carinata oil, % DM total fatty acids, mg/g oil fatty acids profile, % total fatty acids C14:0, myristic acid C16:0, palmitic acid C16:1n-7, palmitoleic acid C17:0, margaric acid C18:0, stearic acid C18:1n-9, oleic acid C18:1, octadecenoic acid C18:2n-6, linoleic acid C18:3n-3, α-linolenic acid C20:0, arachidic acid C20:1, eicosenoic acid C20:2n-6, eicosadienoic acid C20:3n-3, eicosatrienoic acid C22:0, behenic acid C22:1n-9, erucic acid C22:2n-9, docosadienoic acid C22:2n-6, docosadienoic acid C24:0, lignoceric acid C24:1n-9, nervonic acid total SFAd total MUFAe total PUFAf total n-3 fatty acids total n-6 fatty acids total n-9 fatty acids a
yellowa
browna
canola browna
SEMb
P value
carinata vs canola
carinata (yellow vs brown)
41.36b 913.21
39.45b 914.34
46.27a 912.33
0.36 1.98
0.002 0.780
0.001 0.590
0.03 0.710
0.07 3.07 0.19b 0.12b 1.07b 8.67b 0.77b 15.0b 12.7a 0.91a 8.73a 0.98a 0.17 0.79b 41.3b 0.63b 1.42b 0.58b 2.04a 6.60ab 61.60b 30.91a 12.92a 17.42b 61.32
0.05 2.73 0.18b 0.10c 0.94b 7.44b 0.65b 14.7b 12.1a 0.91a 7.54b 0.91b 0.15 0.87a 45.0a 0.79a 1.64a 0.64a 1.94b 6.25b 62.72b 30.32a 12.20a 17.30b 62.73
0.06 3.79 0.33a 0.13a 1.91a 61.73a 2.63a 18.4a 7.88b 0.63b 1.37c 0.08c NDc 0.34c 0.12c ND ND 0.14c 0.13c 6.99a 66.33a 26.42b 7.88b 18.52a 63.33
0.01 0.09 0.01 0.00 0.03 0.66 0.04 0.13 0.32 0.01 0.03 0.00 0.01 0.01 0.07 0.01 0.01 0.01 0.01 0.07 0.54 0.43 0.32 0.14 0.59
0.36 0.01 0.001
Mid-Infrared Spectral Characteristics of Lipid Molecular Structures in Brassica carinata Seeds: Relationship to Oil Content, Fatty Acid and Glucosinolate Profiles, Polyphenols, and Condensed Tannins Hangshu Xin,† Nazir A. Khan,† Kevin C. Falk,§ and Peiqiang Yu*,†,# †
Department of Animal and Poultry Science, College of Agricultural and Bioresources, University of Saskatchewan, 51 Campus Drive, Saskatoon, Saskatchewan S7N 5A8, Canada § Agriculture and Agri-Food Canada, 107 Science Place, Saskatoon, Saskatchewan S7N 0X2, Canada # Tianjin Agricultural University, Xiqing District, Tianjin 300384, China ABSTRACT: The objectives of this study were to quantify lipid-related inherent molecular structures using a Fourier transform infrared spectroscopy (FT-IR) technique and determine their relationship to oil content, fatty acid and glucosinolate profile, total polyphenols, and condensed tannins in seeds from newly developed yellow-seeded and brown-seeded Brassica carinata lines. Canola seeds were used as a reference. The lipid-related molecular spectral band intensities were strongly correlated to the contents of oil, fatty acids, glucosinolates, and polyphenols. The regression equations gave relatively high predictive power for the estimation of oil (R2 = 0.99); all measured fatty acids (R2 > 0.80), except C14:0, C20:3n-3, C22:2n-9, and C22:2n-6; 3-butenyl, 2-OH-3-butenyl, 4-OH-3-CH3-indolyl, and total glucosinolates (R2 > 0.686); and total polyphenols (R2 = 0.935). However, further study is required to obtain predictive equations based on large numbers of samples from diverse sources to illustrate the general applicability of these regression equations. KEYWORDS: carinata seed, fatty acids, chemical compounds, lipid molecular structures
■
INTRODUCTION Ethiopian mustard (Brassica carinata), locally known as gomenzer, has many superior agronomic traits such as high potential seed yield, disease and stress tolerance, and adequate growth under the hot, semiarid growing conditions in western Canada and the Northern Plains of the United States.1−3 Moreover, compared with conventional canola seed, B. carinata has a larger seed size, resulting in a higher protein/fiber ratio in the defatted meals and cakes that are used for feeding animals. As a result, the development of B. carinata has been strongly supported by the Saskatchewan Ministry of Agriculture. Recently, interest in the development of specific lines of B. carinata for human and industrial consumption has been increased. The oil, free of erucic acid,4,5 can be used as a source of vegetable oil for human consumption, whereas the oil rich in erucic acid may be suitable for the production of bioethanol,6 biodiesel,7,8 biolubricants,9 or specialty fatty acids such as conjugated linoleic acid.10 Therefore, development of carinata lines with low erucic acid or high erucic acid content has become a breeding objective of increasing interest. Another remarkable characteristic of Brassica species is the presence of glucosinolate compounds, predominantly allylglucosinolate, butenyl-3-enylglucosinolate, and 2-OH-butenyl-3-enylglucosinolate.11 Several types of glucosinolates in the oilseeds limit its utilization in both humans and animals, owing to antinutritional and toxic effects. In the livestock industry, oilseed meals with less glucosinolate are more competitive and often receive premium payment in the commercial market. To the best of our knowledge, many B. carinata breeding programs for developing specific lines (for agricultural or industrial application) have been primarily focused on oil content and © 2014 American Chemical Society
functional fatty acid composition, and less attention has been given to the other compounds.12−14 It is possible that such selection may have altered the contents of glucosinolates, polyphenols, and condensed tannins. However, comparison research on the fatty acid profile and contents of antinutritional compounds between the yellow-seeded and brown-seeded B. carinata lines is still lacking. More recently, numerous research studies from our group have demonstrated that attenuated total reflectance−Fourier transform infrared (ATR-FT/IR) molecular spectroscopy can be used to determine the inherent differences in molecular structures of various biopolymers such as protein,15 carbohydrates,16 and lipids17 among different types of feeds. Moreover, these studies have shown that the differences in digestive behavior and metabolic characteristics of protein and carbohydrates among different feeds can be partly explained by the differences in their molecular structural makeup. However, limited attention has been paid to the intrinsic molecular structural makeup of lipids. We hypothesize that the lipid-related molecular spectral characteristics in B. carinata seeds might be related to the oil content, fatty acid and glucosinolate profiles, polyphenols, and/or condensed tannins. The present study was designed to evaluate and compare (1) oil content and fatty acid profile; (2) glucosinolate profile, total polyphenols, and condensed tannins; and (3) lipid-related inherent molecular structures in terms of CH stretching bands, Received: Revised: Accepted: Published: 7977
January 7, 2014 July 15, 2014 July 21, 2014 July 21, 2014 dx.doi.org/10.1021/jf502209x | J. Agric. Food Chem. 2014, 62, 7977−7988
Journal of Agricultural and Food Chemistry
Article
Figure 1. Typical ATR-FT/IR molecular spectrum of B. carinata seeds and canola seed (as a reference control) in the mid-IR region ca. 4000−800 cm−1. Lipid Molecular Structure by ATR-FTIR Spectroscopy. The lipid-related molecular structural band intensities (Figure 1) of each sample were generated in the mid-IR region (ca. 4000−800 cm−1) with 128 co-added scans at a spectral resolution of 4 cm−1 in a transmission mode, using a JASCO FT/IR 4200 with ATR (JASCO Corp., Tokyo, Japan). Each sample was run six times. The spectral data were collected with JASCO Spectra Manager II software, and the lipidrelated spectral band intensities were quantified using OMNIC 7.2 software (Madison, WI, USA). The lipid-related functional bands including asymmetric CH3 (CH3-as, ca. 2959 cm−1), asymmetric CH2 (CH2-as, ca. 2921 cm−1), symmetric CH3 (CH3-s, ca. 2871 cm−1), symmetric CH2 (CH2-s, ca. 2853 cm−1), ULB (ca. 3008 cm−1), and LECC (ca. 1744 cm−1) were identified according to published methodologies.15,25,26 The asymmetric and symmetric CH3 and CH2 stretching bands were identified in the CH region (ca. 2999−2800 cm−1) of the original spectra, using the second-derivative and Fourier self-deconvolution functions of the OMNIC 7.2 software. The spectral peak height, area, and their ratios were calculated from the original spectra. Multivariate Spectral Analysis. Agglomerative hierarchical cluster analysis (AHCA) and principal component analysis (PCA), as described in detail earlier,27−30 were performed on the overall data of lipid-related spectral bands, to visualize and classify the differences in the inherent molecular structures of lipids among the three types of oilseeds. The overall spectral data of CH3 and CH2 asymmetric and symmetric stretchings (ca. 2999−2800 cm−1), ULB (ca. 3035−2991 cm−1), and LECC (ca. 1788−1707 cm−1) regions were used for multivariate analyses. The AHCA results are presented as dendrograms using the linkage distance method and Ward’s algorithm cluster method without prior parametrization. The PCA results are presented as a scatter plot between the first principal factor and the second principal factor, based on loading scores. These analyses were carried out using Statistica software (version 8, StatSoft Inc., Tulsa, OK, USA). Statistical Analysis. Data on the oil content, fatty acid and glucosinolate profiles, polyphenols, and condensed tannins and univariate lipid spectral data were statistically analyzed using the Mixed Model procedure of the Statistical Analysis System (SAS version 9.2, SAS Institute Inc., Cary, NC, USA). The following model was used:
unsaturated lipid bands (ULB), and lipid ester CO carbonyl (LECC) using ATR-FT/IR spectroscopy, between the yellowseeded and brown-seeded B. carinata. Canola seeds were used as a reference. The second objective was to determine the relationship between the molecular spectral data and the contents of oil, fatty acid and glucosinolate profiles, total polyphenols, and condensed tannins in the oilseed.
■
MATERIALS AND METHODS
Seed Sampling and Processing. Seeds from newly developed yellow-seeded (AAC A100) and brown-seeded (110915EM) lines of B. carinata were evaluated in the present study. These two carinata lines were bred for the industrial oil market using traditional pedigree breeding methods. Canola seed with a brown seed coat was used as a reference. Samples (n = 4) from all seed types were obtained from plots at the Saskatoon Research Centre experimental farm of Agriculture and Agri-Food Canada (AAFC) and from contra-season research plots in the central region of Chile during the winter of 2010/ 2011. Each seed type was obtained from two replicate plots in each location. For wet chemical and spectral analysis, seeds were ground in a coffee grinder (PC770, Loblaws Inc., Toronto, Canada) for 10 s, then chilled, and reground for another 10 s. Oil Content and Fatty Acid Profile Determination. The samples (∼0.500 g) were analyzed for oil content according to the standard procedure of AOAC (method 920.39).18 Fat from the ground oilseeds (∼0.375 g) was extracted with chloroform/methanol (2:1 v/ v) according to the method of Folch et al.19 with slight modification as described by Khan et al.20 Tridecanoic acid (C13:0) was used as an internal standard. The fatty acids were separated and quantified using gas chromatography as described by Khan et al.21 All samples were analyzed in duplicate and repeated if error exceeded the acceptable level of each analysis. Determination of Glucosinolates, Polyphenols, and Condensed Tannins. All seed samples (∼0.100 g) were analyzed for glucosinolate profile according to the official method of the Canadian Grain Commission.22 The method of Slinkard and Singleton23 was used to analyze the contents of total polyphenols in samples (∼1.00 g). The HCl−butanol procedure as described by Reed24 was used to analyze the content of condensed tannins in the seeds (∼0.050 g). All samples were analyzed in duplicate and repeated if error exceeded the acceptable level of each analysis.
Yijk = μ + Fi + S(F )j + eijk 7978
dx.doi.org/10.1021/jf502209x | J. Agric. Food Chem. 2014, 62, 7977−7988
Journal of Agricultural and Food Chemistry
Article
Table 1. Oil Content and Fatty Acid Profile in the Yellow- and Brown-Seeded B. carinata with Comparison to Canola Seeds contrast, P value
carinata oil, % DM total fatty acids, mg/g oil fatty acids profile, % total fatty acids C14:0, myristic acid C16:0, palmitic acid C16:1n-7, palmitoleic acid C17:0, margaric acid C18:0, stearic acid C18:1n-9, oleic acid C18:1, octadecenoic acid C18:2n-6, linoleic acid C18:3n-3, α-linolenic acid C20:0, arachidic acid C20:1, eicosenoic acid C20:2n-6, eicosadienoic acid C20:3n-3, eicosatrienoic acid C22:0, behenic acid C22:1n-9, erucic acid C22:2n-9, docosadienoic acid C22:2n-6, docosadienoic acid C24:0, lignoceric acid C24:1n-9, nervonic acid total SFAd total MUFAe total PUFAf total n-3 fatty acids total n-6 fatty acids total n-9 fatty acids a
yellowa
browna
canola browna
SEMb
P value
carinata vs canola
carinata (yellow vs brown)
41.36b 913.21
39.45b 914.34
46.27a 912.33
0.36 1.98
0.002 0.780
0.001 0.590
0.03 0.710
0.07 3.07 0.19b 0.12b 1.07b 8.67b 0.77b 15.0b 12.7a 0.91a 8.73a 0.98a 0.17 0.79b 41.3b 0.63b 1.42b 0.58b 2.04a 6.60ab 61.60b 30.91a 12.92a 17.42b 61.32
0.05 2.73 0.18b 0.10c 0.94b 7.44b 0.65b 14.7b 12.1a 0.91a 7.54b 0.91b 0.15 0.87a 45.0a 0.79a 1.64a 0.64a 1.94b 6.25b 62.72b 30.32a 12.20a 17.30b 62.73
0.06 3.79 0.33a 0.13a 1.91a 61.73a 2.63a 18.4a 7.88b 0.63b 1.37c 0.08c NDc 0.34c 0.12c ND ND 0.14c 0.13c 6.99a 66.33a 26.42b 7.88b 18.52a 63.33
0.01 0.09 0.01 0.00 0.03 0.66 0.04 0.13 0.32 0.01 0.03 0.00 0.01 0.01 0.07 0.01 0.01 0.01 0.01 0.07 0.54 0.43 0.32 0.14 0.59
0.36 0.01 0.001
