Metabolomics Study by Amino and Organic Acid Profiling Analyses in Pre- and Post-Milling Barley Using Gas Chromatography-Tandem Mass Spectrometry

Abstract

Barley (Hordeum vulgare) is a major cereal grain grown in temperate climates globally and provides various nutrients in a peeled form after milling. However, milling causes changes in nutritional composition, including metabolites. Thus, a metabolomics study was conducted to monitor the changes in nutritional composition before and after the milling of Hordeum vulgare seeds (Saechalssal, Hinchalssal, and Yeongbaekchal) focusing on the development and application of new analytical methods for organic acids (OA) and amino acids (AA). Profiling analyses of OAs and AAs were performed using GC-MS/MS. This analytical method showed good linearity (r ≥ 0.995) with limit of detection (0.1 ng, 21.2 ng) and limit of quantitation (0.3 ng, 63.6 ng), respectively. Repeatability varied from 0.1 to 12.4 (% RSD) and accuracy varied from –12.3 to 14.8 (% RE), respectively. Altered levels of 36 metabolites (16 OAs, 20 AAs) were monitored post-milling and compared with pre-milling in the three Hordeum vulgare cultivars. Radar plots of OAs and AAs to corresponding mean levels of each pre-milling group in the three Hordeum vulgare cultivars were easily distinguished from those in each post-milling group. The pre-and post-milling groups of the three Hordeum vulgare cultivars were completely separated by partial least square discriminant analysis, and the lysine, cysteine, glutamic acid, asparagine, 4-hydroxyphenylacetic acid, and citric acid were significantly different. Therefore, this study will be useful for monitoring altered metabolites following milling and discrimination of varieties.

1. Introduction

Barley (Hordeum vulgare) is the fourth major grain crop worldwide in terms of cultivation area, along with corn, wheat, and rice, and contains nutrients such as starch, protein, lipids, and minerals, β-glucan reduces blood cholesterol content [1,2,3]. It is effective in reducing cardiovascular health and diseases such as diabetes [4,5]. Barley is mainly used in the form of malt or a variety of processed foods such as bread, cookies, noodles, and cooked foods [6]. Grains such as barley are made into processed foods through a milling process that removes the pericarp, seed coat, endosperm, and aleurone layer, and the nutritional characteristics of the grain differ depending on the milling process [7]. For example, milled white rice helps digestion and quickly raises blood sugar levels, whereas brown rice, which is rich in nutrients such as dietary fiber, minerals, γ-aminobutyric acid (GABA), γ-inositol, and oryzanol, has functional ingredients destroyed when polished [8,9]. In addition, GABA contained in unrefined grains improves the brain’s memory ability and has an improving and protective effect against diseases such as depression, cardiovascular disease, and cancer [10,11]. Although milling affects the quality and characteristics of processed products, including their nutritional components, there is currently a lack of research on suitable varieties for barley processing and the changes in the components owing to milling. Organic acids (OAs) play an important role in maintaining the nutritional value and organoleptic quality of foods and are widely used as preservatives, acidity regulators, and antioxidants [12]. Amino acids (AAs) are essential components of proteins. Nutritionally speaking, essential AAs that cannot be produced in the body are consumed through the diet [13]. The taste of vegetables can be attributed to various factors such as sweetness and AA content [14]. Therefore, in this study, we optimized and validated profiling methods for OAs as methoxime (MO)/tert-butyldimethylsilyl (TBDMS) derivatives and AAs as ethoxylcarbonyl (EOC)-TBDMS derivatives using gas chromatography-tandem mass spectrometry (GC-MS/MS). Profiling analyses of OAs and AAs were performed on three varieties of barley (Saechalssal, Hinchalssal, and Yeongbaekchal) to monitor changes in nutritional components according to variety and milling for barley quality control.

2. Materials and Methods

2.1. Chemicals and Reagents

Standards of AAs and OAs, internal standards (IS; norvaline and 3,4-dimethoxybenzoic acid), triethylamine (TEA), ethyl chloroformate (ECF), methoxyamine hydrochloride were obtained from Sigma-Aldrich (St. Louis, MO, USA). Dichloromethane (DCM), toluene, diethyl ether (DEE), and ethyl acetate (EA) were purchased from Kanto Chemical (Chuoku, Tokyo, Japan), and distilled water (DW) sodium chloride, sulfuric acid, and sodium hydroxide were purchased from Deajung (Siheungsi, Republic of Korea). N-Methyl-N-tert-butyldimethylsilyl trifluoroacetamide (MTBSTFA) was purchased from Pierce Biotechnology (Rockford, IL, USA). All chemicals were of analytical grade.

2.2. Barley Seeds

The barley cultivars as Hordeum vulgare used in this study, Saechalssal, Hinchalssal, and Yeongbaekchalssal are hulless barley cultivars developed in the Republic of Korea. The three cultivars were grown in 2018 at the National Institute of Crop Science, Rural Development Administration, Republic of Korea. All of these barley cultivars are 6-row waxy barley with low amylose content. Additionally, Yeongbaekchalssal was developed in 2013 with a reduced proanthocyanidin content to inhibit browning [15]. The age of the barley seeds is approximately 5 months. All barley cultivars seeds (200 g) were pearled 23% of their original weight using the Sadake Test Mill (M05, Satake, Tokyo, Japan). And then, pearled grains were ground by a Retsch centrifugal mill (Zm 100, I. Kurt Rotech GmbH, Haan, Germany) with 0.2 mm sieve.

For metabolite extraction, each individual barley cultivar seeds (50 mg) were ground using a blender, then 1 mL of DW was added to adjust the concentration of 50 mg/mL. The mixture was then subjected to hot water extraction for 1 h. The mixture was then filtered using a 0.45 µm filter. This process was repeated five times for each individual variety to obtain a total of five samples per variety. All barley solutions were stored at −80 °C until analysis.

2.3. Preparation of Standard Solutions

For standard and IS stock solutions, 20 AAs and norvaline (IS) were prepared at a concentration of 10 mg/mL in 0.1 M HCl. 18 OAs and 3,4-dimethoxybenzoic acid (IS) were dissolved in methanol at 10 mg/mL. All standard stock solutions were stored at −20 °C. For GC-MS/MS analysis, 19 OAs standard working solution mixtures were prepared at 10 µg/mL and 100 µg/mL in methanol, and 3,4-dimethoxybenzoic acid was prepared at 10 µg/mL in methanol. The 32 AAs standard working solution mixtures were prepared at 10 µg/mL and 100 µg/mL in 0.1 M HCl, and norvaline was prepared at 20 µg/mL in 0.1 M HCl.

2.4. GC-MS/MS

Analyses of the standards and samples were performed using a GCMS-TQ8040 (Shimadzu, Kyoto, Japan) interfaced with a triple quadrupole mass spectrometer (70 eV, electron impact ionization mode) in the selected reaction monitoring mode for quantitative analysis. The injector, interface, and ion source temperatures were maintained at 260, 300, and 230 °C, respectively. An Ultra-2 (5% phenyl–95% methylpolysiloxane bonded phase; 25 m × 0.20 mm i.d; 0.11 μm film thickness) cross-linked capillary column (Agilent Technologies, Palo Alto, CA, USA) was used for analysis. Helium was used as the carrier gas at a flow rate of 0.5 mL/min in constant flow mode. Samples (1.0 μL) were introduced using AOC-20i auto-injector and AOC-20s auto-sampler in the split-injection mode (10:1). The oven temperature was programmed as follows: AAs, 140 °C was maintained for 3 min, increased to 300 °C at a rate of 8 °C/min and maintained for 5 min; OAs, 100 °C was maintained for 2 min, increased to 300 °C at a rate of 10 °C/min and maintained for 8 min.

2.5. Method Validation of Profiling Analysis of AAs and OAs in Barley Using GC-MS/MS

Pooled barley, including Saechalssal, Hinchalssal, and Yeongbaekchalssal, was used for matrix validation of AAs and OAs. For method validation of simultaneous AA and OA profiling analyses as EOC/MO/TBDMS derivatives, mixed standard solutions containing 20 AAs and 18 OAs of 10–5000 ng, and ISs [norvaline (200 ng) as the IS of AAs and 3,4-dimethoxybenzoic acid (100 ng) as the IS of OAs] were used in this study [16,17]. These solutions were then spiked into 2 mg of the pooled barley solution. Method validation was performed according to the following protocol.

For simultaneous validation of the AA and OA methods, the supernatant was transferred to a vial containing ECF (40 μL) and DCM (2 mL), and then DW (960 μL) was added. The aqueous phase was then adjusted to a pH > 12 using 5.0 M NaOH and subjected to a sequential EOC reaction. Then, methoxyamine hydrochloride (1 mg) was added to the aqueous phase, adjusted to pH ≥ 12 using 5 M NaOH, and reacted at 60 °C for 60 min to form the MO derivative. The aqueous phase was adjusted to pH ≤ 2 with 10% H₂SO₄, saturated with NaCl, and sequentially extracted using DEE (3 mL) and EA (2 mL). The extract was evaporated to dryness under a gentle stream of nitrogen at 40 °C. Toluene (15 μL), MTBSTFA (20 μL), and TEA (5 μL) were added to the residue and then reacted at 60 °C for 60 min to obtain the TBDMS derivative. Finally, 1.0 µL was injected into the GC-MS/MS with selected reaction monitoring (SRM) mode. These methods were performed in triplicate under optimal conditions and validated for analytical parameters including linearity, repeatability, accuracy, limits of detection (LOD), and limits of quantitation (LOQ). The slope, intercept, and correlation coefficient values were determined by a linearity test using least squares regression analysis on a calibration curve constructed based on the relative peak area ratios to the IS. The LOD and LOQ values for each metabolite were calculated as three- and 10-times the standard deviation of the blank divided by the slope of the calibration curve, respectively. Repeatability as relative standard deviation (% RSD) and accuracy as relative error (% RE) were determined in triplicate within the range that included the metabolite levels detected.

2.6. Sample Preparation for Profiling Analysis of AAs and OAs in Barley Using GC-MS/MS

Profiling analyses of AAs and OAs were performed using GC-MS/MS as EOC/MO/TBDMS. Briefly, each barley sample was homogenized in DW and centrifuged at 13,500 rpm for 3 min. For AA and OA profiling analysis, an aliquot of the homogenate equivalent to the weight of each barley (2 mg), 3,4-dimethoxybenzoic acid (100 ng), and norvaline (200 ng) were added to a vial containing ECF (40 μL) and DCM (2 mL), and then DW (1 mL) was added. The aqueous phase was then adjusted to a pH > 12 using 5.0 M NaOH and subjected to a sequential EOC reaction. Methoxyamine hydrochloride (1 mg) was then added to the aqueous phase, adjusted to pH ≥ 12 using 5 M NaOH, and reacted at 60 °C for 60 min to form the MO derivative. The aqueous phase was adjusted to pH ≤ 2 with 10% H₂SO₄, saturated with NaCl, and sequentially extracted using DEE (3 mL) and EA (2 mL). The extract was evaporated to dryness under a gentle stream of nitrogen at 40 °C. Toluene (15 μL), MTBSTFA (20 μL), and TEA (5 μL) were added to the residue, and then reacted at 60 °C for 60 min to obtain the TBDMS derivative. Finally, 1.0 µL was injected into the GC-MS/MS with SRM mode.

2.7. Radar Plot and Statistical Analyses

The levels of 20 AAs and 15 OAs in the barley samples from the pre- and post-milling groups were determined using calibration curves and converted into percentage compositions (%). The percentage composition was normalized to the corresponding mean values of the pre-milling group. For drawing a radar plot, the mean levels of AA and OA in the pre-milling group were normalized to the mean level in the post-milling group, and Microsoft Excel Office 365 was used to draw a radar plot [18,19,20,21].

Univariate statistical analysis was performed to evaluate the normality of the data using the Shapiro-Wilk test. Some variables did not follow a normal distribution; therefore, a nonparametric test was used for comparisons between the study groups. Comparisons between the pre- and post-milling groups were performed using the Wilcoxon rank-sum test, p-values were adjusted with the false discovery rate (FDR), and metabolites with Q-value < 0.05 were considered significant. In addition, the non-parametric Kruskal-Wallis test was used to detect statistical differences between the three Hordeum vulgare.

Multivariate analysis was conducted using principal component analysis (PCA) and partial least square discriminant analysis (PLS-DA). PCA as unsupervised learning, was used for trend and pattern analyses of the data. PLS-DA, as a supervised learning method, was used for the classification of Hordeum vulgare and for discriminating milling. The validity of the PLS-DA model was verified using the correlation coefficients (R²) and cross-validation correlation coefficients (Q²). The variable importance in projection (VIP) score was used to measure the importance of variables in the PLS-DA model; variables with a VIP score > 1.0 were considered significant in the PLS-DA model. The following analyses were performed with log10-transformed and auto-scaled data using MetaboAnalyst (version 6.0) based on the R project (http://www.metaboanalyst.ca, accessed on 30 September 2024) [18,19,20,21].

3. Results

3.1. Method Validation of Profiling Analyses of AAs and OAs in Barley Samples

3.1.1. Optimization of AA and OA Profiling Analyses

In this study, the precursor ions of AA and OA generated by a 70 eV electron impact (EI) in the ion source were detected and selected in the first quadrupole (Q1). Precursor ions of AAs and OAs were fragmented by collision energy (CE) in the range of −5 to −45 V using argon gas, a collision-induced dissociation gas, in a collision cell (Q2) to generate product ions, which were collected in the third quadrupole (Q3). Three product ions were selected for the identification of AA and OA, and one product ion with high sensitivity and selectivity was selected as the quantitative ion for AA and OA without the matrix effect of the barley sample. The SRM mode conditions for the AAs and OAs are listed in Supplementary Table S1.

3.1.2. Method Validation of Profiling Analysis of 20 AAs and 16 OAs

The profiling method for 20 AAs and 16 OAs was validated under optimal conditions. The calibration curves of 20 AAs and 16 OAs ranging from 10 to 2000 ng examined under optimal conditions were linearity (r) better than 0.9950 with good LODs (0.1–21.2 ng) and LOQs (0.3–63.6 ng). The repeatability and accuracy of the analysis method were measured in concentration range and varied from 0.1 to 12.4 (% RSD) and −12.3 to 14.8 (% RE), respectively. The repeatability and accuracy of the overall procedure, measured at three different concentrations, are listed in Supplementary Table S2. The results of the validated assays indicated that it was suitable for the quantitative analysis of 20 AAs and 16 OAs in pooled barley samples.

3.2. AA and OA Profiles in Three Hordeum vulgare Cultivars (Saechalssal, Hinchalssal, and Yeongbaekchalssal) of Pre- and Post-Milling Barley

3.2.1. AA and OA Profiles by Milling

The metabolites in each barley sample were analyzed using the Wilcoxon rank-sum test by comparing the pre- and post-milling groups.

In the Saechalssal, the levels of 36 metabolites were determined in the pre- (n = 5) and post-milling (n = 5) groups. Glutamic acid was the most abundant amino acid in the 20 AA profiles, followed by glutamine. Among the 16 OAs profiles, lactic acid was most abundant in the pre- and post-milling groups, followed by malic acid (

Table 1. Metabolite composition, Wilcoxon rank-sum test, PCA, and PLS-DA of seeds of Saechalssal, Hordeum vulgare.

No.	Metabolite	Concentration (ng/mg)		Composition (%)		Normalized Value *	Wilcoxon Rank-Sum Test		PCA Loading Scores		VIP Score(PLS-DA)
		Pre-Milling	Post-Milling	Pre-Milling	Post-Milling		p-Value	Q-Value	PC1	PC2
	Amino acid
1	Alanine	71.8 ± 13.9	65.1 ± 15.2	5.33 ± 1.02	4.94 ± 1.03	0.93	0.690	1.000	0.191	0.230	0.61
2	Glycine	13.2 ± 2.9	10.7 ± 2.4	0.97 ± 0.14	0.79 ± 0.09	0.82	0.151	0.905	0.033	0.127	1.77
3	α-Aminobutyric acid	0.8 ± 0.2	0.7 ± 0.1	0.06 ± 0.02	0.05 ± 0.01	0.94	0.841	1.000	0.243	0.121	0.31
4	Valine	34.5 ± 7.2	34.2 ± 3.0	2.65 ± 0.91	2.72 ± 0.89	1.02	0.841	1.000	0.246	0.107	0.15
5	Leucine	18.4 ± 1.7	16.7 ± 1.7	1.41 ± 0.37	1.31 ± 0.34	0.93	0.548	0.986	0.229	0.119	0.46
6	Isoleucine	22.3 ± 3.5	20.1 ± 1.8	1.74 ± 0.63	1.59 ± 0.52	0.91	1.000	1.000	0.247	0.095	0.35
7	Proline	32.6 ± 7.1	26.0 ± 2.1	2.47 ± 0.71	2.07 ± 0.71	0.84	0.310	0.946	0.253	0.050	0.90
8	Pipecolic acid	2.8 ± 0.7	1.4 ± 0.3	0.23 ± 0.12	0.11 ± 0.05	0.50	0.056	0.905	0.169	0.156	1.78
9	Pyroglutamic acid	31.3 ± 9.2	38.2 ± 4.7	2.38 ± 0.84	3.00 ± 0.97	1.26	0.421	0.946	0.223	0.027	1.12
10	Serine	26.9 ± 8.5	29.5 ± 6.6	1.97 ± 0.43	2.18 ± 0.20	1.11	0.421	0.946	0.171	−0.041	0.96
11	Threonine	28.9 ± 16.8	31.6 ± 8.0	2.38 ± 1.66	2.54 ± 1.08	1.07	1.000	1.000	0.234	0.026	0.61
12	γ-Aminobutyric acid	59.7 ± 10.0	53.8 ± 11.7	4.77 ± 2.11	4.40 ± 2.06	0.92	0.841	1.000	0.209	0.179	0.34
13	Phenylalanine	25.3 ± 5.8	22.4 ± 6.3	1.84 ± 0.22	1.64 ± 0.18	0.89	0.222	0.946	−0.016	−0.086	1.35
14	Cysteine	7.5 ± 1.8	6.7 ± 1.1	0.55 ± 0.11	0.51 ± 0.08	0.92	0.690	1.000	0.032	0.059	0.56
15	Aspartic acid	51.1 ± 15.0	49.6 ± 14.3	3.71 ± 0.86	3.59 ± 0.28	0.97	0.095	0.905	−0.140	−0.062	1.88
16	Glutamic acid	369.4 ± 129.2	374.6 ± 126.8	25.97 ± 2.96	26.72 ± 4.90	1.03	0.548	0.986	−0.169	−0.063	0.14
17	Asparagine	169.6 ± 141.0	59.9 ± 28.8	11.20 ± 8.28	4.15 ± 1.48	0.37	0.841	1.000	−0.178	−0.025	0.09
18	Glutamine	191.4 ± 119.0	286.2 ± 131.1	12.62 ± 4.72	19.99 ± 5.33	1.58	0.151	0.905	−0.093	−0.127	1.84
19	Lysine	139.9 ± 44.9	127.2 ± 31.2	9.95 ± 0.97	9.35 ± 0.86	0.94	0.310	0.946	0.024	−0.311	0.94
20	Tyrosine	102.3 ± 28.2	107.6 ± 18.9	7.81 ± 2.61	8.34 ± 2.36	1.07	1.000	1.000	0.218	−0.012	0.44
	Organic acid
21	Pyruvic acid	16.8 ± 2.2	13.5 ± 2.5	0.66 ± 0.11	0.63 ± 0.21	0.95	1.000	1.000	0.154	−0.247	0.53
22	Lactic acid	947.4 ± 82.5	879.0 ± 78.2	37.70 ± 7.38	39.82 ± 8.86	1.06	0.548	0.986	−0.017	−0.306	0.32
23	Glycolic acid	369.4 ± 195.1	342.4 ± 150.8	13.93 ± 5.83	14.27 ± 3.42	1.02	0.548	0.986	−0.230	0.109	0.28
24	2-Hydroxybutyric acid	2.3 ± 0.46	2.2 ± 0.50	0.09 ± 0.03	0.10 ± 0.01	1.03	0.841	1.000	−0.150	−0.157	0.37
25	3-Hydroxypropionic acid	14.9 ± 4.3	12.1 ± 4.1	0.58 ± 0.15	0.52 ± 0.12	0.89	0.421	0.946	−0.111	0.042	0.62
26	Malonic acid	36.9 ± 14.0	15.5 ± 2.6	1.48 ± 0.57	0.68 ± 0.09	0.46	0.151	0.905	0.143	−0.143	1.82
27	Succinic acid	60.6 ± 6.3	62.9 ± 10.8	2.39 ± 0.35	2.76 ± 0.26	1.15	0.310	0.946	−0.037	−0.194	1.59
28	Fumaric acid	54.4 ± 8.6	47.7 ± 5.5	2.13 ± 0.29	2.12 ± 0.35	1.00	1.000	1.000	0.199	−0.214	0.07
29	Oxaloacetic acid	345.3 ± 200.0	157.9 ± 71.6	13.02 ± 6.51	7.45 ± 4.19	0.57	0.222	0.946	0.204	−0.096	1.38
30	α-Ketoglutaric acid	83.1 ± 16.2	55.7 ± 9.9	3.24 ± 0.46	2.46 ± 0.41	0.76	0.056	0.905	0.080	−0.213	1.99
31	4-Hydroxyphenylacetic acid	3.7 ± 0.81	3.1 ± 0.11	0.15 ± 0.04	0.14 ± 0.03	0.93	0.421	0.946	0.197	−0.218	0.27
32	Malic acid	531.2 ± 155.7	593.6 ± 352.6	20.04 ± 3.75	23.87 ± 8.44	1.19	0.690	1.000	0.019	0.290	0.62
33	2-Hydroxyglutaric acid	38.5 ± 9.5	35.7 ± 8.4	1.48 ± 0.22	1.58 ± 0.38	1.07	1.000	1.000	0.128	−0.120	0.37
34	cis-Aconitic acid	8.9 ± 2.0	9.9 ± 1.5	0.36 ± 0.11	0.44 ± 0.08	1.23	0.421	0.946	0.165	−0.184	1.16
35	Citric acid	59.6 ± 21.1	69.9 ± 58.0	2.30 ± 0.67	2.68 ± 1.51	1.16	1.000	1.000	−0.070	0.292	0.27
36	Isocitric acid	11.5 ± 2.1	12.1 ± 5.9	0.45 ± 0.03	0.50 ± 0.11	1.11	0.841	1.000	−0.110	0.269	0.76

* Values normalized to the corresponding mean value of each metabolite composition in the pre-milling groups.

).

For Hinchalssal, the levels of 36 metabolites were determined in the pre- (n = 5) and post-milling (n = 5) groups. Among the 20 AAs profiles, the pre- and post-milling groups exhibited the highest levels of asparagine, followed by glutamic acid. Among the 16 OAs profiles, lactic acid was most abundant in the pre- and post-milling groups, followed by malic acid (

Table 2. Metabolite composition, Wilcoxon rank-sum test, PCA, and PLS-DA of seeds of Hinchalssal, Hordeum vulgare.

No.	Metabolite	Concentration (ng/mg)		Composition (%)		Normalized Value *	Wilcoxon Rank-Sum Test		PCA Loading Scores		VIP Score(PLS-DA)
		Pre-Milling	Post-Milling	Pre-Milling	Post-Milling		p-Value	Q-Value	PC1	PC2
	Amino acid
1	Alanine	32.1 ± 4.8	12.6 ± 3.0	5.10 ± 1.11	3.95 ± 1.2	0.78	0.421	0.658	0.183	0.227	0.90
2	Glycine	11.8 ± 1.0	6.2 ± 0.8	1.85 ± 0.13	1.91 ± 0.2	1.03	0.690	0.857	0.180	0.029	0.39
3	α-Aminobutyric acid	0.2 ± 0.03	0.1 ± 0.03	0.03 ± 0.01	0.03 ± 0.01	1.10	0.841	0.918	0.248	0.075	0.23
4	Valine	11.4 ± 1.5	7.8 ± 2.5	1.82 ± 0.37	2.52 ± 1.0	1.38	0.421	0.658	0.263	−0.019	0.63
5	Leucine	7.5 ± 0.5	4.7 ± 1.0	1.19 ± 0.18	1.47 ± 0.5	1.23	0.548	0.758	0.251	−0.058	0.68
6	Isoleucine	7.4 ± 1.1	4.5 ± 1.6	1.18 ± 0.28	1.46 ± 0.7	1.24	0.690	0.857	0.261	0.028	0.32
7	Proline	18.3 ± 2.1	10.2 ± 2.9	2.91 ± 0.54	3.28 ± 1.3	1.13	1.000	1.000	0.222	−0.004	0.20
8	Pipecolic acid	1.8 ± 0.4	0.7 ± 0.2	0.29 ± 0.09	0.23 ± 0.1	0.82	0.421	0.658	0.208	0.141	0.62
9	Pyroglutamic acid	14.1 ± 1.7	12.5 ± 2.2	2.26 ± 0.50	3.97 ± 1.0	1.76	0.056	0.222	0.173	−0.195	1.37
10	Serine	8.8 ± 0.9	5.8 ± 1.0	1.40 ± 0.29	1.84 ± 0.5	1.32	0.222	0.471	0.272	−0.056	0.89
11	Threonine	4.9 ± 3.2	5.5 ± 3.3	0.81 ± 0.63	1.83 ± 1.2	2.25	0.310	0.619	0.267	−0.050	0.85
12	γ-Aminobutyric acid	17.7 ± 3.4	8.1 ± 2.1	2.81 ± 0.72	2.57 ± 0.9	0.91	1.000	1.000	0.235	0.140	0.38
13	Phenylalanine	11.1 ± 0.7	6.6 ± 1.1	1.76 ± 0.25	2.02 ± 0.2	1.14	0.151	0.362	0.192	−0.119	1.03
14	Cysteine	6.1 ± 0.4	5.5 ± 0.5	0.95 ± 0.06	1.69 ± 0.1	1.77	0.008	0.071	0.109	−0.271	1.88
15	Aspartic acid	48.2 ± 7.0	37.1 ± 7.8	7.50 ± 0.40	11.29 ± 1.4	1.51	0.222	0.471	0.014	−0.252	0.71
16	Glutamic acid	71.0 ± 14.7	35.3 ± 16.9	11.09 ± 2.21	10.78 ± 5.2	0.97	0.548	0.758	0.084	0.150	0.29
17	Asparagine	219.4 ± 105.3	77.4 ± 53.3	32.83 ± 13.45	22.09 ± 12.6	0.67	0.008	0.071	−0.270	0.036	1.74
18	Glutamine	42.2 ± 8.6	32.4 ± 7.0	6.62 ± 1.33	10.17 ± 2.8	1.54	0.151	0.362	0.087	−0.202	1.21
19	Lysine	37.5 ± 9.8	30.4 ± 5.0	5.85 ± 1.26	9.50 ± 2.3	1.62	0.016	0.095	0.157	−0.166	1.40
20	Tyrosine	70.1 ± 49.9	23.8 ± 6.7	11.75 ± 9.25	7.39 ± 2.3	0.63	1.000	1.000	0.152	0.042	0.19
	Organic acid
21	Pyruvic acid	15.8 ± 3.7	14.8 ± 2.8	0.51 ± 0.17	0.48 ± 0.2	0.95	0.841	0.918	0.187	0.098	0.17
22	Lactic acid	1213.3 ± 418.5	800.9 ± 133.9	35.81 ± 2.94	24.69 ± 6.2	0.69	0.016	0.095	0.021	0.290	1.42
23	Glycolic acid	501.5 ± 244.0	455.6 ± 178.4	14.29 ± 2.66	12.86 ± 1.4	0.90	0.421	0.658	−0.082	0.092	0.58
24	2-Hydroxybutyric acid	3.0 ± 1.9	2.2 ± 0.5	0.08 ± 0.03	0.07 ± 0.01	0.82	0.690	0.857	−0.105	0.111	0.46
25	3-Hydroxypropionic acid	20.7 ± 9.3	11.0 ± 3.0	0.59 ± 0.11	0.32 ± 0.04	0.54	0.008	0.071	−0.119	0.238	1.67
26	Malonic acid	36.7 ± 20.4	8.5 ± 6.0	1.09 ± 0.52	0.22 ± 0.1	0.20	0.008	0.071	−0.090	0.209	1.65
27	Succinic acid	81.7 ± 27.2	99.2 ± 35.0	2.48 ± 0.86	2.89 ± 0.5	1.16	0.548	0.758	0.050	−0.101	0.63
28	Fumaric acid	53.5 ± 17.7	40.4 ± 13.1	1.63 ± 0.43	1.17 ± 0.2	0.72	0.095	0.286	−0.105	0.166	1.10
29	Oxaloacetic acid	245.6 ± 70.4	138.1 ± 27.2	8.10 ± 3.97	4.41 ± 1.5	0.54	0.095	0.286	0.073	0.208	1.13
30	α-Ketoglutaric acid	177.3 ± 27.4	196.4 ± 115.2	5.62 ± 1.49	5.67 ± 2.5	1.01	0.841	0.918	0.197	0.099	0.23
31	4-Hydroxyphenylacetic acid	3.1 ± 0.5	2.0 ± 0.2	0.10 ± 0.02	0.06 ± 0.02	0.65	0.095	0.286	0.076	0.269	1.32
32	Malic acid	689.9 ± 160.6	1233.4 ± 546.6	20.96 ± 2.32	33.66 ± 6.5	1.61	0.032	0.143	−0.012	−0.300	1.55
33	2-Hydroxyglutaric acid	56.4 ± 13.0	59.2 ± 21.7	1.71 ± 0.13	1.69 ± 0.2	0.99	0.841	0.918	0.108	0.057	0.13
34	cis-Aconitic acid	19.0 ± 3.3	18.4 ± 3.5	0.60 ± 0.14	0.55 ± 0.1	0.92	0.421	0.658	0.051	0.067	0.25
35	Citric acid	201.5 ± 100.4	384.2 ± 218.1	5.73 ± 1.41	10.26 ± 3.2	1.79	0.032	0.143	−0.110	−0.266	1.35
36	Isocitric acid	22.8 ± 5.7	37.8 ± 21.1	0.70 ± 0.11	1.00 ± 0.3	1.43	0.151	0.362	−0.021	−0.249	1.12

* Values normalized to the corresponding mean value of each metabolite composition in the pre-milling groups.

). In particular, the levels of two AAs (cysteine and lysine) and two OAs (malic acid and citric acid) increased (p < 0.05, Table 2), whereas those of asparagine and three OAs (lactic acid, 3-hydroxypropionic acid, and malonic acid) decreased (p < 0.05, Table 2).

In Yeongbaekchalssal, the levels of 36 metabolites were determined in the pre-milling (n = 5) and post-milling (n = 5) groups. Glutamic acid was the most abundant amino acid in the 20 AA profiles, followed by asparagine. Regarding 16 OAs profiles, the pre- and post-milling groups, lactic acid was the most abundant, followed by glycolic acid (

Table 3. Metabolite composition, Wilcoxon rank-sum test, PCA, and PLS-DA of seeds of Yeongbaekchalssal, Hordeum vulgare.

No.	Metabolite	Concentration (ng/mg)		Composition (%)		Normalized Value *	Wilcoxon Rank-Sum Test		PCA Loading Scores		VIP Score(PLS-DA)
		Pre-Milling	Post-Milling	Pre-Milling	Post-Milling		p-Value	Q-Value	PC1	PC2
	Amino acid
1	Alanine	67.3 ± 9.6	68.8 ± 14.2	6.17 ± 1.45	4.67 ± 0.82	0.76	0.056	0.125	0.221	0.144	0.92
2	Glycine	15.7 ± 4.0	14.4 ± 2.7	1.37 ± 0.06	0.97 ± 0.15	0.71	0.008	0.029	0.222	0.006	1.35
3	α-Aminobutyric acid	0.5 ± 0.05	0.6 ± 0.1	0.04 ± 0.02	0.04 ± 0.01	0.95	1.000	1.000	0.111	0.275	0.06
4	Valine	24.8 ± 4.1	26.7 ± 1.9	2.32 ± 0.74	1.97 ± 0.84	0.85	0.690	0.777	0.172	0.249	0.45
5	Leucine	21.1 ± 8.0	19.6 ± 2.1	1.86 ± 0.47	1.40 ± 0.48	0.76	0.151	0.259	0.207	0.150	0.76
6	Isoleucine	19.8 ± 5.9	22.3 ± 1.7	1.79 ± 0.49	1.64 ± 0.68	0.92	0.841	0.891	0.161	0.241	0.30
7	Proline	24.3 ± 4.4	22.2 ± 2.4	2.25 ± 0.64	1.60 ± 0.59	0.71	0.151	0.259	0.208	0.182	0.85
8	Pipecolic acid	3.2 ± 0.7	1.1 ± 0.2	0.30 ± 0.10	0.08 ± 0.02	0.25	0.008	0.029	0.244	−0.054	1.54
9	Pyroglutamic acid	17.8 ± 2.2	29.2 ± 3.1	1.70 ± 0.61	2.11 ± 0.77	1.24	0.548	0.657	0.009	0.320	0.49
10	Serine	21.9 ± 3.4	25.3 ± 6.9	1.98 ± 0.32	1.68 ± 0.22	0.85	0.222	0.333	0.192	0.152	0.79
11	Threonine	20.8 ± 8.2	18.1 ± 6.1	2.06 ± 1.16	1.38 ± 0.91	0.67	0.421	0.582	0.170	0.180	0.50
12	γ-Aminobutyric acid	49.9 ± 10.5	36.4 ± 4.1	4.83 ± 2.27	2.55 ± 0.66	0.53	0.032	0.076	0.227	0.111	1.08
13	Phenylalanine	24.5 ± 9.1	24.6 ± 4.1	2.09 ± 0.33	1.69 ± 0.34	0.81	0.222	0.333	0.175	0.057	0.84
14	Cysteine	7.1 ± 1.7	8.0 ± 1.3	0.63 ± 0.10	0.55 ± 0.12	0.87	0.421	0.582	0.112	0.150	0.54
15	Aspartic acid	51.5 ± 15.4	71.9 ± 20.7	4.44 ± 0.47	4.75 ± 0.96	1.07	0.548	0.657	−0.093	0.025	0.39
16	Glutamic acid	207.7 ± 69.3	410.6 ± 140.3	17.67 ± 2.35	26.29 ± 1.51	1.49	0.008	0.029	−0.235	0.030	1.46
17	Asparagine	189.5 ± 124.4	320.2 ± 157.8	15.37 ± 8.27	19.16 ± 6.94	1.25	0.841	0.891	−0.169	−0.203	0.29
18	Glutamine	101.1 ± 49.3	266.3 ± 130.8	8.69 ± 3.45	16.47 ± 2.78	1.89	0.008	0.029	−0.202	0.069	1.24
19	Lysine	155.2 ± 59.8	99.8 ± 34.7	13.04 ± 1.97	6.56 ± 1.29	0.50	0.008	0.029	0.209	−0.134	1.46
20	Tyrosine	126.5 ± 66.5	59.7 ± 8.1	11.39 ± 5.02	4.41 ± 1.96	0.39	0.032	0.076	0.242	0.052	1.20
	Organic acid
21	Pyruvic acid	26.1 ± 2.8	20.3 ± 3.2	0.61 ± 0.06	0.59 ± 0.14	0.97	1.000	1.000	−0.009	−0.197	0.25
22	Lactic acid	1808.8 ± 274.8	1046.4 ± 125.6	42.15 ± 3.99	30.16 ± 4.62	0.72	0.008	0.029	0.157	−0.241	1.31
23	Glycolic acid	1112.0 ± 265.7	630.8 ± 165.2	25.57 ± 3.85	17.52 ± 2.29	0.69	0.032	0.076	0.129	−0.238	1.31
24	2-Hydroxybutyric acid	5.2 ± 1.4	3.0 ± 0.5	0.12 ± 0.02	0.09 ± 0.01	0.71	0.032	0.076	0.120	−0.213	1.11
25	3-Hydroxypropionic acid	44.4 ± 15.6	53.8 ± 31.6	1.02 ± 0.29	1.37 ± 0.55	1.34	0.690	0.777	−0.056	0.125	0.52
26	Malonic acid	17.1 ± 6.3	6.4 ± 5.0	0.41 ± 0.17	0.21 ± 0.21	0.52	0.222	0.333	0.152	−0.008	0.87
27	Succinic acid	66.3 ± 8.8	95.1 ± 21.3	1.55 ± 0.15	2.75 ± 0.76	1.77	0.008	0.029	−0.141	0.152	1.30
28	Fumaric acid	60.6 ± 5.5	41.7 ± 8.1	1.44 ± 0.28	1.23 ± 0.39	0.85	0.548	0.657	0.060	−0.067	0.58
29	Oxaloacetic acid	263.5 ± 108.3	190.7 ± 126.6	6.11 ± 2.51	5.66 ± 4.09	0.93	0.548	0.657	0.052	−0.146	0.33
30	α-Ketoglutaric acid	92.6 ± 16.7	191.6 ± 19.4	2.16 ± 0.33	5.56 ± 0.99	2.57	0.008	0.029	−0.192	0.122	1.56
31	4-Hydroxyphenylacetic acid	9.2 ± 1.2	2.9 ± 0.3	0.21 ± 0.01	0.09 ± 0.03	0.40	0.008	0.029	0.177	−0.221	1.49
32	Malic acid	592.4 ± 177.3	1084.1 ± 518.6	14.14 ± 4.69	28.70 ± 7.51	2.03	0.008	0.029	−0.147	0.250	1.20
33	2-Hydroxyglutaric acid	79.9 ± 13.8	46.5 ± 10.6	1.87 ± 0.31	1.30 ± 0.14	0.69	0.016	0.052	0.168	−0.108	1.29
34	cis-Aconitic acid	8.3 ± 2.4	12.6 ± 1.5	0.20 ± 0.07	0.38 ± 0.12	1.90	0.095	0.190	−0.128	0.096	1.14
35	Citric acid	84.7 ± 47.3	147.8 ± 83.2	2.01 ± 1.10	3.84 ± 1.35	1.91	0.095	0.190	−0.156	0.201	1.03
36	Isocitric acid	18.0 ± 9.6	21.0 ± 8.4	0.42 ± 0.21	0.57 ± 0.11	1.35	0.151	0.259	−0.161	0.101	0.80

* Values normalized to the corresponding mean value of each metabolite composition in the pre-milling groups.

). In particular, two AAs (glutamic acid and glutamine) and three OAs (succinic acid, α-ketoglutaric acid and malic acid) were increased (p-value < 0.05, Table 3), whereas four AAs (glycine, pipecolic acid, γ-Aminobutyric acid, and lysine) and five OAs (lactic acid, glycolic acid, 2-hydroxybutyric acid, 4-hydroxyphenylacetic acid, and 2-hydroxyglutaric acid) were decreased (p-value < 0.05, Table 2). After FDR correction, two AAs (glutamic acid and glutamine) and three OAs (succinic acid, α-ketoglutaric acid and malic acid) were increased in the post-milling groups, whereas three AAs (glycine, pipecolic acid, and lysine) and two OAs (lactic acid and 4-hydroxyphenylacetic acid) were increased in the post-milling groups (Q-value < 0.05, Table 3).

3.2.2. Univariate Analysis Among Three Hordeum vulgare Cultivars in Pre- and Post-Milling Groups

The compositions of 20 AA and 16 OA were compared among the three Hordeum vulgare cultivars (Saechalssal, Hinchalssal, and Yeongbaekchalssal) pre- and post-milling.

As a result of comparing the three Hordeum vulgare cultivars (Saechalssal, Hinchalssal, and Yeongbaekchalssal) pre-milling, five AA (glycine, cysteine, glutamic acid, asparagine, and lysine) and five OAs (3-hydroxypropionic acid, α-ketoglutaric acid, 4-hydroxyphenylacetic acid, cis-aconitic acid, citric acid) were significantly altered (Q-value < 0.05,

Table 4. Metabolite composition, Kruskal-Wallis test, and PLS-DA of the three Hordeum vulgare cultivars.

No.	Metabolite	Composition (%)						Kruskal- Wallis Test		VIP Score (PLS-DA)
		Pre-Milling			Post-Milling			Pre- Milling	Post- Milling	Pre- Milling	Post- Milling
		Saechalssal	Hinchalssal	Yeongbaekchal	Saechalssal	Hinchalssal	Yeongbaekchal
	Amino acid
1	Alanine	5.33 ± 1.02	5.10 ± 1.11	6.17 ± 1.45	4.94 ± 1.03	3.95 ± 1.2	4.67 ± 0.82	0.382	0.486	0.58	0.75
2	Glycine	0.97 ± 0.14	1.85 ± 0.13	1.37 ± 0.06	0.79 ± 0.09	1.91 ± 0.2	0.97 ± 0.15	0.028	0.042	1.21	1.73
3	α-Aminobutyric acid	0.06 ± 0.02	0.03 ± 0.01	0.04 ± 0.02	0.05 ± 0.01	0.03 ± 0.01	0.04 ± 0.01	0.114	0.132	0.85	0.98
4	Valine	2.65 ± 0.91	1.82 ± 0.37	2.32 ± 0.74	2.72 ± 0.89	2.52 ± 1.0	1.97 ± 0.84	0.426	0.415	0.67	0.03
5	Leucine	1.41 ± 0.37	1.19 ± 0.18	1.86 ± 0.47	1.31 ± 0.34	1.47 ± 0.5	1.40 ± 0.48	0.171	0.801	1.03	0.26
6	Isoleucine	1.74 ± 0.63	1.18 ± 0.28	1.79 ± 0.49	1.59 ± 0.52	1.46 ± 0.7	1.64 ± 0.68	0.141	0.932	0.98	0.32
7	Proline	2.47 ± 0.71	2.91 ± 0.54	2.25 ± 0.64	2.07 ± 0.71	3.28 ± 1.3	1.60 ± 0.59	0.444	0.078	0.77	0.97
8	Pipecolic acid	0.23 ± 0.12	0.29 ± 0.09	0.30 ± 0.10	0.11 ± 0.05	0.23 ± 0.1	0.08 ± 0.02	0.270	0.042	0.14	1.24
9	Pyroglutamic acid	2.38 ± 0.84	2.26 ± 0.50	1.70 ± 0.61	3.00 ± 0.97	3.97 ± 1.0	2.11 ± 0.77	0.382	0.111	0.53	0.82
10	Serine	1.97 ± 0.43	1.40 ± 0.29	1.98 ± 0.32	2.18 ± 0.20	1.84 ± 0.5	1.68 ± 0.22	0.141	0.189	1.13	0.47
11	Threonine	2.38 ± 1.66	0.81 ± 0.63	2.06 ± 1.16	2.54 ± 1.08	1.83 ± 1.2	1.38 ± 0.91	0.276	0.385	0.96	0.38
12	γ-Aminobutyric acid	4.77 ± 2.11	2.81 ± 0.72	4.83 ± 2.27	4.40 ± 2.06	2.57 ± 0.9	2.55 ± 0.66	0.141	0.194	0.98	0.70
13	Phenylalanine	1.84 ± 0.22	1.76 ± 0.25	2.09 ± 0.33	1.64 ± 0.18	2.02 ± 0.2	1.69 ± 0.34	0.276	0.132	0.74	0.98
14	Cysteine	0.55 ± 0.11	0.95 ± 0.06	0.63 ± 0.10	0.51 ± 0.08	1.69 ± 0.1	0.55 ± 0.12	0.031	0.042	1.36	1.73
15	Aspartic acid	3.71 ± 0.86	7.50 ± 0.40	4.44 ± 0.47	3.59 ± 0.28	11.29 ± 1.4	4.75 ± 0.96	0.111	0.042	1.02	1.06
16	Glutamic acid	25.97 ± 2.96	11.09 ± 2.21	17.67 ± 2.35	26.72 ± 4.90	10.78 ± 5.2	26.29 ± 1.51	0.028	0.042	1.30	1.54
17	Asparagine	11.20 ± 8.28	32.83 ± 13.45	15.37 ± 8.27	4.15 ± 1.48	22.09 ± 12.6	19.16 ± 6.94	0.031	0.042	1.48	1.76
18	Glutamine	12.62 ± 4.72	6.62 ± 1.33	8.69 ± 3.45	19.99 ± 5.33	10.17 ± 2.8	16.47 ± 2.78	0.141	0.042	0.66	1.40
19	Lysine	9.95 ± 0.97	5.85 ± 1.26	13.04 ± 1.97	9.35 ± 0.86	9.50 ± 2.3	6.56 ± 1.29	0.028	0.067	1.68	0.35
20	Tyrosine	7.81 ± 2.61	11.75 ± 9.25	11.39 ± 5.02	8.34 ± 2.36	7.39 ± 2.3	4.41 ± 1.96	0.733	0.132	0.19	0.15
	Organic acid
21	Pyruvic acid	0.66 ± 0.11	0.51 ± 0.17	0.61 ± 0.06	0.63 ± 0.21	0.48 ± 0.2	0.59 ± 0.14	0.426	0.739	0.80	0.61
22	Lactic acid	37.70 ± 7.38	35.81 ± 2.94	42.15 ± 3.99	39.82 ± 8.86	24.69 ± 6.2	30.16 ± 4.62	0.232	0.063	0.73	1.15
23	Glycolic acid	13.93 ± 5.83	14.29 ± 2.66	25.57 ± 3.85	14.27 ± 3.42	12.86 ± 1.4	17.52 ± 2.29	0.051	0.069	0.88	0.54
24	2-Hydroxybutyric acid	0.09 ± 0.03	0.08 ± 0.03	0.12 ± 0.02	0.10 ± 0.01	0.07 ± 0.01	0.09 ± 0.01	0.218	0.042	0.90	1.37
25	3-Hydroxypropionic acid	0.58 ± 0.15	0.59 ± 0.11	1.02 ± 0.29	0.52 ± 0.12	0.32 ± 0.04	1.37 ± 0.55	0.048	0.042	0.88	0.98
26	Malonic acid	1.48 ± 0.57	1.09 ± 0.52	0.41 ± 0.17	0.68 ± 0.09	0.22 ± 0.1	0.21 ± 0.21	0.053	0.042	0.74	0.49
27	Succinic acid	2.39 ± 0.35	2.48 ± 0.86	1.55 ± 0.15	2.76 ± 0.26	2.89 ± 0.5	2.75 ± 0.76	0.078	0.800	0.81	0.20
28	Fumaric acid	2.13 ± 0.29	1.63 ± 0.43	1.44 ± 0.28	2.12 ± 0.35	1.17 ± 0.2	1.23 ± 0.39	0.123	0.042	0.03	0.99
29	Oxaloacetic acid	13.02 ± 6.51	8.10 ± 3.97	6.11 ± 2.51	7.45 ± 4.19	4.41 ± 1.5	5.66 ± 4.09	0.222	0.408	0.10	0.53
30	α-Ketoglutaric acid	3.24 ± 0.46	5.62 ± 1.49	2.16 ± 0.33	2.46 ± 0.41	5.67 ± 2.5	5.56 ± 0.99	0.028	0.094	1.64	0.68
31	4-Hydroxyphenylacetic acid	0.15 ± 0.04	0.10 ± 0.02	0.21 ± 0.01	0.14 ± 0.03	0.06 ± 0.02	0.09 ± 0.03	0.031	0.045	1.53	1.29
32	Malic acid	20.04 ± 3.75	20.96 ± 2.32	14.14 ± 4.69	23.87 ± 8.44	33.66 ± 6.5	28.70 ± 7.51	0.070	0.305	0.86	0.82
33	2-Hydroxyglutaric acid	1.48 ± 0.22	1.71 ± 0.13	1.87 ± 0.31	1.58 ± 0.38	1.69 ± 0.2	1.30 ± 0.14	0.141	0.175	0.06	0.63
34	cis-Aconitic acid	0.36 ± 0.11	0.60 ± 0.14	0.20 ± 0.07	0.44 ± 0.08	0.55 ± 0.1	0.38 ± 0.12	0.035	0.161	1.44	0.83
35	Citric acid	2.30 ± 0.67	5.73 ± 1.41	2.01 ± 1.10	2.68 ± 1.51	10.26 ± 3.2	3.84 ± 1.35	0.048	0.042	1.46	1.53
36	Isocitric acid	0.45 ± 0.03	0.70 ± 0.11	0.42 ± 0.21	0.50 ± 0.11	1.00 ± 0.3	0.57 ± 0.11	0.070	0.048	1.24	1.44

). Six AA (glycine, cysteine, aspartic acid, glutamic acid, asparagine, and glutamine) and seven OAs (2-hydroxybutyric acid, 3-hydroxypropionic acid, malonic acid, fumaric acid, 4-hydroxyphenylacetic acid, citric acid, and isocitric acid) were significantly altered (Q-value < 0.05; Table 4).

3.3. AAs and OAs Radar Plots in Three Hordeum vulgare Cultivars (Saechalssal, Hinchalssal, and Yeongbaekchal) by Milling

3.3.1. Radar Plots of AAs Among Three Hordeum vulgare Cultivars in Pre- and Post-Milling Groups

The AA levels in the post-milling group were normalized to the corresponding mean levels in the pre-milling group. Radar plots of the normalized AAs in each barley were shown in Figure 1.

In Saechalssal, the normalized values of the 20 AAs ranged from 0.37 to 1.58 in the post-milling group (Table 1). In particular, three AAs (pyroglutamic acid, serine, and glutamine) increased by 11–58%, whereas five AAs (glycine, proline, pipecolic acid, phenylalanine, and asparagine) decreased by 11–63% in the post-milling group (Figure 1a).

In Hinchalssal, normalized values of the 20 AAs ranged from 0.63 to 2.25 in the post-milling group (Table 2). Especially, 13 AAs (α-Aminobutyric acid, valine, leucine, isoleucine, proline, pyroglutamic acid, serine, threonine, phenylalanine, cysteine, aspartic acid, glutamine, and lysine) were increased by 10–125%, whereas four AAs (alanine, pipecolic acid, asparagine, and tyrosine) were decreased by 18–37% in the post-milling group (Figure 1b).

In Yeongbaekchal, normalized values of the 20 AAs ranged from 0.25 to 1.89 in the post-milling group (Table 3). Especially, four AAs (pyroglutamic acid, glutamic acid, asparagine, and glutamine) were increased by 24–89%, whereas 13 AAs (alanine, glycine, valine, leucine, proline, pipecolic acid, serine, threonine, γ-aminobutyric acid, phenylalanine, cysteine, lysine, and tyrosine) were decreased by 15–75% in the post-milling group (Figure 1c).

3.3.2. Radar Plots of OAs Among Three Hordeum vulgare Cultivars in Pre- and Post-Milling Groups

The OA levels in the post-milling group were normalized to the corresponding mean levels in the pre-milling group. Radar plots of the normalized OAs in each barley were shown in Figure 2.

In Saechalssal, the normalized values of the 16 OAs ranged from 0.46 to 1.23 in the post-milling group (Table 1). In particular, five OAs (succinic acid, malic acid, cis-aconitic acid, citric acid, and isocitric acid) were increased by 11–23%, whereas four OAs (3-hydroxypropionic acid, malonic acid, oxaloacetic acid, and α-ketoglutaric acid) were decreased by 11–54% in the post-milling group (Figure 2a).

In Hinchalssal, the normalized values of the 16 OAs ranged from 0.20 to 1.79 in the post-milling group (Table 2). In particular, four OAs (succinic acid, malic acid, citric acid, and isocitric acid) increased by 16–79%, whereas eight OAs (lactic acid, glycolic acid, 2-hydroxybutyric acid, 3-hydroxypropionic acid, malonic acid, fumaric acid, oxaloacetic acid, and 4-hydroxyphenylacetic acid) decreased by 10–80% in the post-milling group (Figure 2b).

In the Yeongbaekchal, normalized values of 20 AAs ranged from 0.25 to 1.89 in the post-milling group (Table 3). Especially, four AAs (pyroglutamic acid, glutamic acid, as-paragine, and glutamine) were increased by 24–89%, whereas 13 AAs (alanine, glycine, valine, leucine, proline, pipecolic acid, serine, threonine, γ-aminobutyric acid, phenylalanine, cysteine, lysine, and tyrosine) was decreased by 15–75% in the post-milling group (Figure 2c).

3.4. Multivariate Analysis of Three Hordeum vulgare Cultivars (Saechalssal, Hinchalssal, and Yeongbaekchal)

3.4.1. Monitored Metabolites Altered by Milling

In Saechalssal, the PCA score plot explained 59.7% of the total variance in PC1 and PC2 (Figure 3a). To discriminate between the pre- and post-milling groups, PCA loading scores were evaluated using loadings 1 and 2. The variable that had the greatest influence on loading 1, the main component of the PCA, was proline, and that on loading 2 was lysine (Table 1). Although the pre- and post-milling groups were not completely separated in the PCA score plots (Figure 3a), they were separated in the PLS-DA (Figure 3b). Nevertheless, in the PLS-DA score plot, the pre- and post-milling groups were completely separated with R², accuracy, and Q² of 0.94, 0.7, and 0.486, respectively. According to PLS-DA, 11 metabolites (glycine, pipecolic acid, pyroglutamic acid, phenylalanine, aspartic acid, glutamine, malonic acid, succinic acid, oxaloacetic acid, α-ketoglutaric acid, and cis-aconitic acid) with a VIP score > 1.0 were evaluated as major contributing metabolites for discrimination of the pre- and post- milling groups (Table 1). However, because component 1 explains the variance lower than component 2, the variations among the 5 samples may be more subtle than the differences in milling for the Saechalssal.

In Hinchalssal, the PCA score plot explained 64.5% of the total variance in PC1 and PC2 (Figure 4a). To discriminate between the pre- and post-milling groups, PCA loading scores were evaluated using loadings 1 and 2. The variable that had the greatest influence on loading 1 was serine, and that on loading 2 was malic acid (Table 2). The pre- and post-milling groups are completely separated in the PCA and PLS-DA score plots (Figure 4a,b). In the PLS-DA score plot, the pre- and post-milling groups were completely separated with R², accuracy, and Q² values of 0.92, 1.0, and 0.706, respectively. According to the PLS-DA, 16 metabolites (pyroglutamic acid, phenylalanine, cysteine, aspartic acid, asparagine, glutamine, lysine, lactic acid, 3-hydroxypropionic acid, malonic acid, fumaric acid, oxaloacetic acid, 4-hydroxyphenylacetic acid, malic acid, citric acid, and isocitric acid) with a VIP score > 1.0 were evaluated as the major metabolites contributing to the discrimination between the pre- and post-milling groups (Table 2).

In Yeongbaekchal, the PCA score plot explained 66.9% of the total variance in PC1 and PC2 (Figure 5a). To discriminate between the pre- and post-milling groups, PCA loading scores were evaluated using loadings 1 and 2. Pipecolic acid had the greatest influence on loading 1, and pyroglutamic acid had the greatest influence on loading 2 (Table 3). The pre- and post-milling groups were completely separated in the PCA and PLS-DA score plots (Figure 5a,b). In the PLS-DA score plot, the pre- and post-milling groups were completely separated with R², accuracy, and Q² values of 0.91, 1.0, and 0.793, respectively. According to PLS-DA, 17 metabolites (glycine, pipecolic acid, γ-aminobutyric acid, glutamic acid, glutamine, lysine, tyrosine, lactic acid, glycolic acid, 2-hydroxybutyric acid, succinic acid, α-ketoglutaric acid, 4-hydroxyphenylacetic acid, malic acid, 2-hydroxyglutaric acid, cis-aconitic acid, and citric acid) with a VIP score > 1.0 were evaluated as major contributing metabolites for discrimination of the pre- and post-milling groups (Table 3).

3.4.2. Tracking Metabolites That Can Distinguish Hordeum vulgare Cultivars

In the pre-milling groups, the PCA score plot explained 61.3% of the total variance in PC1 and PC2 (Figure 6a). To discriminate among Saechalssal, Hinchalssal, and Yeongbaekchal, PCA loading scores were evaluated using Loading 1 and Loading 2. The variable that had the greatest influence on loading 1, the main component of the PCA, was asparagine, and that on loading 2 was pyroglutamic acid (Table 4). Although Saechalssal and Yeongbaekchal were not completely separated in the PCA and PLS-DA score plots, Hinchalssal was separated (Figure 6a,b). In the PLS-DA score plot, Saechalssal, Hinchalssal, and Yeongbaekchal were completely separated, with R², accuracy, and Q² of 0.88, 0.93, and 0.753, respectively. According to PLS-DA, 13 metabolites (glycine, leucine, serine, cysteine, aspartic acid, glutamic acid, asparagine, lysine, α-ketoglutaric acid, 4-hydroxyphenylacetic acid, cis-aconitic acid, citric acid and isocitric acid) with a VIP score > 1.0 were evaluated as major contributing metabolites for discrimination among the Saechalssal, Hinchalssal, and Yeongbaekchal (Table 4).

In the post-milling group, the PCA score plot explained 62.2% of the total variance in PC1 and PC2 (Figure 7a). To discriminate among Saechalssal, Hinchalssal, and Yeongbaekchal, PCA loading scores were evaluated using Loading 1 and Loading 2. The variable that had the greatest influence on loading 1, the main component of the PCA, was citric acid, and that on loading 2 was valine (Table 4). Although the three Hordeum vulgare cultivars were not completely separated in the PCA and PLS-DA score plots (Figure 7a,b), Hinchalssal was separated in the PLS-DA score plot (Figure 7b). In the PLS-DA score plot, Saechalssal, Hinchalssal, and Yeongbaekchal were completely separated with R², accuracy, and Q² of 0.92, 0.87, and 0.854, respectively. According to the PLS-DA, 12 metabolites (glycine, pipecolic acid, cysteine, aspartic acid, glutamic acid, asparagine, glutamine, lactic acid, 2-hydroxybutyric acid, 4-hydroxyphenylacetic acid, citric acid, and isocitric acid) with a VIP score > 1.0 were evaluated as the major metabolites contributing to the discrimination among Saechalssal, Hinchalssal, and Yeongbaekchal (Table 4).

4. Discussion

In this study, the developed AA and OA profiling methods were suitable for quantification of these metabolites in the three Hordeum vulgare cultivars.

First, AA and OA profiling methods were applied to compare barley pre- and post- milling, and the results showed that in the three types of barley, the majority of metabolites decreased after milling. Overall, owing to the reduction in concentration caused by milling and the concentration differences between the varieties, comparisons were made based on composition.

The metabolite composition of Saechalssal was not significantly altered by milling. Regarding Hinchalssal, the composition of two AAs (cysteine and lysine) and two OAs (malic acid and citric acid) increased, whereas asparagine and three OAs (lactic acid, 3-hydroxypropionic acid, and malonic acid) decreased. Increased lysine through milling is an essential amino acid that plays an important role in maintaining muscle mass and bone health [22]. Cysteine is a non-essential amino acid that is used nutritionally for health improvement [23]. Malic acid naturally exists in body cells and has the ability to stimulate metabolism and increase energy production. Citric acid is associated with the flavor of food [24]. Therefore, it is attributed to the increase of amino acids and malic acid as nutrients, along with changes in taste and flavor, depending on the milling process.

In previous studies, asparagine has a slightly bitter taste [25], lactic acid affects the taste and preservation of various foods [26], and 3-hydroxypropionic acid has been reported to be a widely used chemical ingredient in agriculture, food and ingredients [27]. Thus, the reduction of metabolites (asparagine, lactic acid, 3-hydroxypropionic acid) due to milling may be associated with changes in the taste and quality of Hinchalssal. In addition, malonic acid, which is known to increase cholesterol levels during ingestion [28], was found to be reduced after crushing in Hinchalssal. Therefore, these results suggest that the changes in metabolites by milling are related to taste and nutrition.

In Yeongbaekchal, the composition of two AAs (glutamic acid and glutamine) and three OAs (succinic acid, α-ketoglutaric acid, and malic acid) increased (p-value < 0.05, Table 3), whereas four AAs (glycine, pipecolic acid, γ-aminobutyric acid, and lysine) and five OAs (lactic acid, glycolic acid, 2-hydroxybutyric acid, 4-hydroxyphenylacetic acid, and 2-hydroxyglutaric acid) decreased (p-value < 0.05, Table 2). Especially, two AAs (glutamic acid and glutamine) and three OAs (succinic acid, α-ketoglutaric acid, and malic acid) significantly increased in the post-milling groups (Q-value < 0.05, VIP score > 1.0), whereas three AAs (glycine, pipecolic acid, and lysine) and two OAs (lactic acid and 4-hydroxyphenylacetic acid) significantly decreased in the post-milling groups (Q-value < 0.05, VIP score > 1.0).

For decreasing milling, these results show that the compositions of the amino acids lysine and glycine, which promote browning [29,30], were found to decrease with milling which is considered to be the result of reduced browning. 4-hydroxyphenylacetic acid derivatives have antioxidant activity [31]. This suggests that the decrease in the levels of browning-promoting amino acids in Youngbaekchal due to the milling process suggests that milling can reduce the browning. Additionally, pipecolic acid has an effect of suppressing food intake [32], and a reduction in pipecolic acid may decrease its appetite-suppressing effect.

For increasing milling, glutamic acid has various flavors and also serves as a food additive. [33]. Organic acid including succinic acid, α-ketoglutaric acid, and malic acid play an important role in maintaining the nutritional value and organoleptic quality of foods and are widely used as preservatives, acidity regulators, and antioxidants [12]. Glutamine is important in various key metabolic processes of immune cells and intestinal cells [34], so the increased glutamic acid after milling may be nutritionally beneficial.

These results indicate that milling can alter the composition of metabolites in barley, which may affect both its nutritional value and flavor.

Second, we applied AA and OA profiling methods to identify metabolites that could distinguish the three Hordeum vulgare. As a result, the six metabolites (glycine, cysteine, glutamic acid, asparagine, 4-hydroxyphenylacetic acid, and citric acid) were selected as Q-value < 0.05, VIP score > 1.0. Among the six metabolites, four (glycine, glutamic acid, asparagine, and citric acid) are associated with food flavor [24,25,33] and two (glycine and cysteine) are associated with browning [29,35]. In addition, 4-hydroxyphenylacetic acid derivatives exhibit antioxidant activities [31].

Thus, these results showed that AA and OA profiling methods helped track metabolites related to taste and browning and distinguish among Hordeum vulgare.

5. Conclusions

In this study, AA and OA profiling methods were developed and validated. Under optimal conditions, these showed good linearity (r ≥ 0.995) with LOD of 0.1 and 21.2 ng and LOQ of 0.3 and 63.6 ng, respectively. Repeatability varied from 0.1 to 12.4 (% RSD) and accuracy varied from –12.3 to 14.8 (% RE), respectively. A metabolomic study was performed by OA and AA profiling analyses using GC-MS/MS in three Hordeum vulgare of pre- and post-milling to monitor changes in nutritional components according to variety and milling. Two OAs and four AAs showed significant differences in the three Hordeum vulgare cultivars pre- and post-milling, which was useful for monitoring altered levels of metabolites following milling and discriminating between varieties. These results will be useful for monitoring metabolite changes during milling and detecting metabolite differences among Hordeum vulgare.