Next Article in Journal
Liver Metabolomics and Inflammatory Profiles in Mouse Model of Fentanyl Overdose Treated with Beta-Lactams
Previous Article in Journal
A Solid-Phase Microextraction—Liquid Chromatography-Mass Spectrometry Method for Analyzing Serum Lipids in Psoriatic Disease
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Metabolites
Volume 13
Issue 8
10.3390/metabo13080964
metabolites-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Faba Bean (Vicia faba L. minor) Bitterness: An Untargeted Metabolomic Approach to Highlight the Impact of the Non-Volatile Fraction

by
Adeline Karolkowski
1,2,
Emmanuelle Meudec
3,4,
Antoine Bruguière
1,
Anne-Claire Mitaine-Offer
1,
Emilie Bouzidi
5,
Loïc Levavasseur
2,
Nicolas Sommerer
3,4,
Loïc Briand
1,* and
Christian Salles
1,*
1
Centre des Sciences du Goût et de L’Alimentation, CNRS, INRAE, Institut Agro, Université de Bourgogne, F-21000 Dijon, France
2
Groupe Soufflet-Invivo, F-10400 Nogent-sur-Seine, France
3
SPO, Université de Montpellier, INRAE, Institut Agro, F-34000 Montpellier, France
4
INRAE, PROBE Research Infrastructure, PFP Polyphenol Analysis Facility, F-34060 Montpellier, France
5
Vivien Paille (Groupe Avril), F-59300 Valenciennes, France
*
Authors to whom correspondence should be addressed.
Metabolites 2023, 13(8), 964; https://doi.org/10.3390/metabo13080964
Submission received: 2 August 2023 / Revised: 17 August 2023 / Accepted: 18 August 2023 / Published: 21 August 2023
(This article belongs to the Section Nutrition and Metabolism)
Browse Figures
Versions Notes

Abstract

:
In the context of climate change, faba beans are an interesting alternative to animal proteins but are characterised by off-notes and bitterness that decrease consumer acceptability. However, research on pulse bitterness is often limited to soybeans and peas. This study aimed to highlight potential bitter non-volatile compounds in faba beans. First, the bitterness of flours and air-classified fractions (starch and protein) of three faba bean cultivars was evaluated by a trained panel. The fractions from the high-alkaloid cultivars and the protein fractions exhibited higher bitter intensity. Second, an untargeted metabolomic approach using ultra-high-performance liquid chromatography–diode array detector–tandem–high resolution mass spectrometry (UHPLC–DAD–HRMS) was correlated with the bitter perception of the fractions. Third, 42 tentatively identified non-volatile compounds were associated with faba bean bitterness by correlated sensory and metabolomic data. These compounds mainly belonged to different chemical classes such as alkaloids, amino acids, phenolic compounds, organic acids, and terpenoids. This research provided a better understanding of the molecules responsible for bitterness in faba beans and the impact of cultivar and air-classification on the bitter content. The bitter character of these highlighted compounds needs to be confirmed by sensory and/or cellular analyses to identify removal or masking strategies.

Graphical Abstract

1. Introduction

In the context of climate change, it seems appropriate to reconcile the economic, social, and environmental impacts of human activities. The transition to a more plant-based diet, in particular in pulses, thus appears to be one of the main levers for improving human health and the sustainability of food systems [1]. For example, faba bean (Vicia faba L. minor) ingredients are increasingly used in the formulation of food products due to agronomic, environmental, nutritional, and functional interests [2,3]. Despite the great interest in pulses, consumer acceptability remains low due to off-flavours, such as off-notes, bitterness, and astringency [4]. However, faba beans are still of sensory interest. They are not astringent and have a lower intensity of off-notes than other pulses, such as peas or soybeans [5,6]. Research on the off-flavours of faba beans has focused on the involvement of volatile compounds in off-notes [7,8]. In addition, a few studies have been carried out on the molecules responsible for the bitterness of pulses, but they were mainly conducted on peas and soybeans [9].
Many compounds from plant defence metabolisms could be responsible for pulse bitterness [9]. Recent studies have highlighted the role of saponins, phenolic compounds, peptides, lipids, and lipid oxidation products in the bitterness of peas [10,11,12,13]. Concerning lupins, bitter perception should be more related to specific alkaloids and tannins [14]. Such research is similarly scarce for the bitter compounds in faba beans. In general, only the profile of non-volatile compounds has been studied in faba beans without making a link to their sensory perception. The main identified metabolites belonged to alkaloids, phenolic compounds, saponins, jasmonates, organic acids, and amino acids [15,16,17,18,19,20,21,22]. It has been suggested that the presence of free phenolic compounds, vicine and convicine was linked to bitterness in faba bean ingredients (flour, protein concentrate, and isolate). Moreover, saponins and tannins have been proposed to play a minor role in bitter perception [8]. This observation was partly confirmed by a recent study that demonstrated that the concentration of saponins in faba beans was too low to contribute to their bitterness. However, vicine (one of the major alkaloids) activated one of the 25 human bitter taste receptors (TAS2Rs), and its concentration in faba beans should partly account for bitterness [23].
The aim of this study was to determine the non-volatile compounds that could contribute to faba bean bitterness by linking chemical and sensory data. To this end, three cultivars and three fractions (flour (F), starch fraction (S), and protein fraction (P) (concentrate)) of faba beans were studied. First, the sensory bitterness of the fractions was evaluated by a trained panel. Second, the non-volatile profiles of the different samples were characterised by a metabolomic approach using ultra-high-performance liquid chromatography–diode array detector–tandem–high resolution mass spectrometry (UHPLC–DAS–HRMS). Third, the metabolomic and sensory data were correlated, and the compounds positively correlated with bitterness were tentatively identified. All the selected sensory and physico-chemical methods have already been validated and used to make a link between the chemical composition of mushrooms or peas with their sensory attributes [10,24].

2. Materials and Methods

2.1. Faba Bean Fractions

Three cultivars (C1, C2, and C3) of faba beans (Vicia faba L. minor) were studied. Mature seeds were harvested in 2020 and provided by Soufflet Agriculture (Groupe Soufflet-Invivo, Nogent-sur-Seine, France). C1 was cultivated under organic conditions. C1 and C2 were richer in alkaloids (vicine and convicine) than C3. The seeds were dehulled, ground, and air-classified by an external laboratory (SAS IMPROVE, Amiens, France). The F, P, and S fractions obtained for each cultivar were vacuum-packed in a glass container and stored at −20 °C before analysis.

2.2. Sensory Analysis

Faba bean fractions were incorporated into a gel made of xanthan gum and water. Twenty-one trained panellists (35–73 years old, 12 females and 9 males) evaluated the bitterness of the 9 gels along a linear scale (from 0 (absence) to 10 (very intense)). A solution of caffeine (0.5 g/L) was used as an external reference for bitterness. Additional details on the gel formulation and sensory profiling can be found elsewhere [23,25].

2.3. Non-Volatile Compound Extraction

Extraction of the non-volatile compounds was adapted from previous protocols [10,26]. For each fraction, 100 mg of sample was extracted twice with 5 mL of a methanol (Biosolve Chimie, Dieuze, France) and formic acid (Biosolve Chimie, Dieuze, France) mixture (99/1 v/v) under a 1 h constant agitation at room temperature. The suspensions were kept at −20 °C for 20 min and centrifuged (4500 rpm, 20 min, 4 °C). Then, the supernatant was evaporated at 35 °C using a centrifugal evaporator (EZ-2 Plus Evaporating System, Genevac, Ipswich, UK). After evaporation, the samples were solubilised in 200 µL of a methanol, Milli-Q® ultrapure water (Merck Millipore, USA), and formic acid mixture (49.5/49.5/1 v/v) before being centrifuged (15,000 rpm, 15 min, 4 °C). The supernatants were collected through a 0.22 µM PTFE (polytetrafluoroethylene) filter, placed in HPLC vials and stored at −80 °C in darkness until analysis. The samples were prepared in triplicate.

2.4. Non Volatile Compound Untargeted Analysis

The extracts were analysed by UHPLC (Vanquish, Thermo Scientific, Waltham, MA, USA) using an Acquity UPLC HSST3 C18 column (100 mm × 1 mm i.d., 1.7 µm; Waters, Saint-Quentin-en-Yvelines, France). The mobile phase consisted of (A) water–formic acid (99/1 v/v) and (B) acetonitrile–water–formic acid (79.5/19.5/1 v/v/v). The gradient program with a flow rate of 0.22 mL/min was set as follows: 1–1.5 min A/B, 98/2%; 4.5–7 min A/B, 88/12%; 12 min A/B, 76/24%; 15 min A/B, 72/28%; and 16 min A/B, 40/60%. The column and injector temperatures were maintained at 35 and 10 °C, respectively. The injection volume was 0.5 µL. The UHPLC system was coupled with a DAD (UV–visible diode array detector) covering the full range of acquisition (190–600 nm) and an HRMS (Orbitrap ExplorisTM 480, Thermo Scientific, Waltham, MA, USA) equipped with a heated electrospray ionisation probe. HRMS was operated in both negative and positive ion modes. The parameters for the ion source were as follows: ion transfer tube temperature—280 °C; voltage (+)—3500 V; voltage (−)—2500 V; sheath gas—40 a.u.; auxiliary gas—10 a.u.; sweep gas—2 a.u.; vaporiser temperature—300 °C; mass range—100–1800 Th, resolution—at m/z 240,000 and 480,000. The sample sequence was adapted from a previous method [27]. It included extracts from S, F, and P fractions, a quality control (QC, which was a mix of all samples) sample, and blank. Samples were injected in biological triplicate and in a random order. After equilibration, the sequence started with 3 injections of a blank, followed by 5 injections of QC samples. The blank and QC sample were then injected every 7 real samples. The sequence ended with 1 injection of QC sample and 1 injection of blank. XcaliburTM 4.4 (Thermo Scientific, Waltham, MA, USA) was used for instrument control, data acquisition, and data analysis.

2.5. Metabolomic Analysis

Raw data from HRMS were processed with Compound Discoverer™ (v 3.2.2.421, Thermo Scientific, Waltham, MA, USA), which facilitated peak recognition and dereplication of raw data (retention time alignment, adducts, and isotopic peak clustering). Details about the workflow are available in the Supplementary Materials. Then, these data were correlated (Pearson correlation) with the bitter intensity of the faba bean fractions to identify those that were positively correlated. The tentative identification of the positively correlated compounds was based on ultra-high-resolution mass spectrum, mass accuracy below 1 ppm, characteristic MS/MS fragmentation, UV spectrum, and retention time available in the literature. In particular, the isotope-ratio at ultra-high resolution of 15N vs. 13C and 18O vs. 2x13C and 34S was used to determine the chemical formula of the studied compounds (FreeStyle 1.5, Thermo Scientific, Waltham, MA, USA).

2.6. Statistical Analysis

Analyses were performed using XLSTAT (Addinsoft, Paris, France). For the bitter intensity, a one-way analysis of variance (ANOVA) was performed, and significant differences were evaluated by Tukey’s honest significant difference (HSD) post hoc test (p < 0.05). Linear and logarithmic Pearson correlations (α = 2.5%) were used to determine the relationships between the non-volatile compound areas, in both negative and positive modes, and the bitter intensity of the 9 fractions. Only compounds positively correlated with bitterness were then tentatively identified. In addition, a principal component analysis (PCA, Pearson correlation) was also carried out to visually explore the differences in the non-volatile compound profile of the studied samples.

3. Results and Discussion

3.1. Bitterness of the Faba Bean Fractions

The bitter taste of the nine gels was evaluated by a trained panel (Figure 1). No significant difference in the bitter intensity was observed for the three flours. However, the P fractions, except for P3, were perceived as more bitter than the F and S fractions. P1 was also perceived as more bitter than P2 and P3. These sensory data were correlated with those from the untargeted metabolomic approach to highlight potential bitter compounds in faba bean fractions.

3.2. Tentative Identification of Non-Volatile Compounds Correlated with Faba Bean Bitterness

A list of 355 and 155 potential compounds was established in negative and positive modes, respectively. In the negative mode, 14 non-volatile compounds were positively correlated with bitterness using linear and/or logarithmic correlations, whereas in the positive mode, 31 compounds were positively correlated with bitterness. Only three compounds were detected and positively correlated with bitterness in both positive and negative modes. Figure 2 shows the different non-volatile compounds resulting from the metabolomic analysis of the fractions and those that were positively correlated with bitterness (variables in blue for linear model, variables in dark green for logarithmic model and variables in pink for both models). Few adducts or fragments of these compounds were also positively correlated with bitterness. Those adducts should have been dereplicated by the workflow used but were not. An expert overlooked of the features after dereplication was carried out, and they were easily identified. As there were few mis-dereplications, the workflow was not modified. Those mis-dereplicated compounds were not further considered for this analysis; however, they were represented by black variables on the PCA.
Then, the positively correlated compounds were tentatively identified by comparing the UV spectrum and the MS/MS spectra from the literature. Moreover, the occurrence of tentatively identified compounds in faba beans and/or other pulses/plants was also specified. For all the proposed identifications, the use of high-mass accuracy (<1 ppm) and ultra-high-resolution MS spectra (R = 480 k at m/z 200) of the isotopic profile allowed unique and unambiguous molecular formulas. The correlated compounds are presented in Table 1. They are distributed into the following chemical classes: alkaloids, amino acids, phenolic compounds, organic acids, other compounds, and unidentified compounds.

3.2.1. Alkaloids

A total of six compounds belonging to the alkaloid family were tentatively identified. Vicine (1) had one absorption band at 274 nm and displayed a major molecular ion at m/z 305.1093 in the positive mode. The fragment at m/z 143.0564 in the MS2 spectra should indicate the presence of divicine ([C4H6N4O2 + H]+) and the loss of a hexoside residue (−162). Vicine and convicine are the main alkaloids widely spread in Vicia faba [16,28]. Although convicine had been identified in both positive and negative modes, it was not positively correlated with bitterness (linear and logarithmic models). The compounds (2)–(4) and (6) had one absorption band between 270 and 274 nm and exhibited two characteristic fragments of vicine in their MS2 spectra: m/z 305.1093 ([vicine + H]+) and 143.0564 m/z; they were tentatively identified as vicine derivatives. Only the derivative of vicine ((4), m/z 387.1521 (negative) and 389.1667 (positive)) has already been identified in faba beans [19,21]. It also demonstrated the loss of the C5H8O moiety in both positive and negative modes that could correspond to a valeric or isovaleric residue according to Kowalczyk et al. (2021) [21]. Moreover, a compound derived from convicine (5) was detected at 7.49 min and exhibited a mass of m/z 388.1361 and 390.1507 in negative and positive modes, respectively. It was characterised by two fragments: m/z 304.0791 (negative mode) corresponding to convicine and m/z 142.0264 (negative mode) and 144.0405 (positive mode) corresponding to isouramil [23,28]. The mass difference between the alkaloid derivative and the ion concerning the respective alkaloid was similar (−84) for the compounds (4) and (5), which also suggested the presence of a valeric or isovaleric residue for convicine derivative (5). Many derivatives of alkaloids have been identified in faba beans, even if they did not all correspond to those tentatively identified in this article [16,19,21].

3.2.2. Amino Acids

Many amino acids and their derivatives were detected in positive mode and positively correlated with faba bean bitterness. The compound (7), with a mass of m/z 175.1192 ([M+H]+), was also found under a potassium adduct m/z 213.0747 ([M+2K-H]+) and showed a neutral molecular formula of C6H14N4O2. The fragments m/z 158.0926, 130.0975, 116.0707, 112.0870, and 60.0557 were characteristic of L-arginine [29]. The L-arginine ion and the associated fragments were found in the fragmentation of the three following compounds (8, 9, and 17), suggesting that they were derived from this amino acid. Only N-formyl-L-arginine (9) showing the same molecular formula and its fragmentation has already been identified in another plant, the black cohosh [30]. At 1.00 min, the mass of m/z 198.0762 associated with the fragments m/z 181.0496, 152.0707, 139.0390, and 135.0441 was related to L-DOPA (12) [22]. An isomer of L-DOPA hexoside (13) was observed in its protonated (m/z 360.1286) and stacked (m/z 719.2506) forms. The fragmentation patterns revealed the loss of a hexoside residue (m/z 198.0762; −162) and an ammonia moiety (m/z 181.0495; −17); these observations were consistent with previous results [16,18]. Few isomers have already been identified in faba beans [16,18], but it was not possible to determine which isomer was positively correlated with bitterness. Moreover, three other L-DOPA derivatives were tentatively identified. They exhibited the same characteristic fragments related to L-DOPA, but only the fragmentation of the compounds (11) and (19) revealed the neutral loss of a hexoside compared to the compound (20). However, it was not possible to determine a molecular formula by the isotope-ratio method due to the presence of other signals at the same retention time for the compound (19). L-phenylalanine (16) and L-tryptophan (18) were also tentatively identified, and their fragmentations were in agreement with a previous study [29]. The compound (15) exhibited the same fragment m/z 146.0601 as L-tryptophan, suggesting that it was derived from this amino acid. Concerning the compounds (10) and (14), their fragments (m/z 224.0918, 178.0864, and 165.0547) were similar to those of N-acetyl-L-tyrosine [31]. Moreover, their fragmentation patterns revealed the loss of a hexoside residue (m/z 224.0918; −162); which suggested that these compounds were two isomers of N-acetyl-L-tyrosine hexoside. N-acetyl-L-tyrosine was produced in common bean plants during fungal pathogen attacks [32]. Finally, it is important to note that L-arginine, L-DOPA, L-tryptophan, L-phenylalanine, and L-tyrosine have already been identified in faba beans [16,18,33], which suggested that the tentatively identified compounds were indeed derived from these amino acids.

3.2.3. Phenolic Compounds

Three phenolic compounds, only detected in the negative mode, were positively correlated with bitterness. p-Hydroxybenzoic acid (21) was tentatively identified by comparing the accurate MS1 and MS2 spectra from the literature (obtained on a standard) [10]. The compound (22) had two absorption bands at 257 and 293 nm characteristic of phenolic acids [10,34]. The molecular ion at m/z 315.0721 and the fragment m/z 153.0194 should indicate the presence of a protocatechuic acid moiety and the loss of a hexoside residue [34]. Concerning the compound (23), the band at 282 nm and the fragments from MS2 spectra suggested the presence of methylfukiic acid [16,35]. This compound was identified by NMR in Piscidia Erythrina L., belonging to the Fabaceae family [35]. p-Hydroxybenzoic acid has already been identified in peas [10], whereas protocatechuic acid hexoside and methylfukiic acid were detected out in faba beans [16,17,19,20,21].
The compound (24) had two absorption bands at 270 and 340 nm corresponding to the phenolic core and the conjugated system of a flavonoid, respectively [36]. Moreover, its fragment ion at m/z 287.0551 indicated the presence of a kaempferol moiety. Many kaempferol derivatives have already been observed in faba beans (leaves, pods, and mature seeds) and peas (leaves and protein isolates) [10,15,19,20,21].

3.2.4. Organic Acids

A total of three organic acids were tentatively characterised. The fragmentation patterns in the positive mode of the compound (25) were consistent with those of hydroxy aspergillic acid, and this compound has already been identified in brown alga [37]. The compound (26) exhibited the fragments m/z 190.0499, 172.0394, and 144.0444, which should indicate that it was derived from kynurenic acid, which has been identified in olive fruit [38,39,40]. The MS1/MS2 spectra of the compound (27) were characteristic of pantothenic acid hexoside [41,42]. Pantothenic acid has been found in faba beans [16,18], whereas pantothenic acid hexoside has also been detected in chickpeas and tomatoes [41,42].

3.2.5. Terpenoids

A total of two terpenoids were detected. The compound (28) was tentatively identified as 8-β-D-glucopyranosyloxy-2,7-dimethyl-2,4-decadiene-1,10-dioic acid by comparing the fragmentation patterns with the faba bean literature [16]. Another terpenoid (29) was detected in both negative and positive modes and exhibited the same fragments as dihydrophaseic acid 4’-O-β-D-glucopyranoside; dihydrophaseic acid and its derivatives were detected in faba beans [16].

3.2.6. Other Non-Volatile Compounds

Two compounds were also tentatively identified, but they did not belong to a specific class. First, the compound (30) detected at 0.74 min exhibited fragments at m/z 104.1071, 60.0808, and 58.0657, indicating that it was derived from choline [43], which has already been detected in faba beans [44]. Choline was also tentatively identified in samples with the same MS2 spectrum but was not positively correlated with bitterness. Second, the fragments m/z 315.1812, 191.0564, 161.0456, 149.0456, 143.0347, and 131.0349 indicated that the compound (31) should be a geraniol pentoxide hexoside [45]. However, it was not possible to determine the corresponding isomer. Geraniol has already been detected in the volatile content of faba beans [7].

3.2.7. Unidentified Compounds

A total of 11 compounds were not identified among the 45 metabolites positively correlated with bitterness; however, a few comments could be made. The fragmentation patterns of the compound (35) revealed the loss of a hexoside moiety (m/z 321.1010; −162). It was not possible to tentatively identify the compounds (40) and (42) because their fragmentation data were not available in the literature. The compound (40) could be a diterpene glycoside called 19-hydroxycinnzeylanol 19-glucoside, which has already been identified in the leaf of a dicotyledon plant [46]. Concerning the compound (42), the MS1 spectrum was consistent with those of 4-chloro-oxoindole-acetic acid, which have been detected in faba bean seeds [21]. However, these hypotheses must be verified.
Table 1. Tentative identification of the non-volatile compounds tentatively identified and positively correlated with faba bean bitterness (linear and logarithmic models; Pearson correlation, α = 2.5%). The p-value is shown in bold when a positive correlation is observed (p-value < 0.025).
Table 1. Tentative identification of the non-volatile compounds tentatively identified and positively correlated with faba bean bitterness (linear and logarithmic models; Pearson correlation, α = 2.5%). The p-value is shown in bold when a positive correlation is observed (p-value < 0.025).
No.RT (min)Linear ModelLogarithmic
Model		UV (nm)ModeExperi-
mental m/z		Formula
(Neutral)		Expected m/zError (ppm)Main MS/MS
Fragment Ions		CompoundRIRef.
Rp-ValueRp-Value
ALKALOIDS
10.860.7370.0240.5490.126274POS305.1093C10H14N4O7305.10920.10143.0564 (100)Vicine1,2,3,4[23,28]
21.270.7960.0100.5980.089274POS391.1097C13H18N4O10391.11071.00305.1099 (1); 143.0564 (100)Vicine derivative1,2
36.610.830.0060.6280.070278POS613.1988C25H32N4O14613.19880.00305.1093 (1); 147.0442 (60); 143.0564 (100)Vicine derivative1,2
47.610.7850.0120.6090.082274NEG775.3110
(387.1512)		C15H24N4O8387.15210.90387.1512 (50); 303.0947 (5); 141.0418 (100)Vicine derivative (ester with valeric/isovaleric acid)1,2,4[19,21]
0.8710.0020.7010.035278POS389.1670389.1667−0.30305.1092 (5); 143.0564 (100)
57.680.8290.0060.6450.060274NEG777.2793
(388.1352)		C15H23N3O9388.13610.90388.1352 (100); 304.0791 (5); 142.0264 (5)Convicine derivative
(ester with valeric/isovaleric acid)		1,2,4[19,21]
0.8830.0020.6480.059274POS390.1509390.1507−0.20229.1070 (20); 144.0405 (100); 127.0390 (20); 85.0648 (30); 57.0699 (20)
611.560.7550.0190.5790.102274POS815.3052C32H46N8O17815.30540.20305.1093 (5); 143.0565 (100)Vicine derivative1,2
AMINO ACIDS
70.710.7360.0240.6950.038 POS213.0747
(175.1192)		C6H14N4O2175.12010.90175.1192 (100); 158.0925 (20); 130.0975 (10); 116.0707 (30); 112.0869 (5); 60.0557 (20)L-arginine1,4[29,33]
80.740.8850.0020.7920.011 POS292.1979C11H25N5O4292.19790.00175.1190 (70); 158.0926 (10); 118.0863 (100); 116.0705 (1); 60.0556 (1)L-arginine derivative1
90.850.8460.0040.8110.008 POS203.1138C7H15N4O3203.11501.20203.1143 (100); 186.0876 (10); 175.1191 (20); 158.0928 (10); 144.0657 (20); 130.0974 (1); 116.0707 (10); 112.0870 (5); 88.0870 (5); 70.0651 (1)N-formyl-L- arginine1,6[30]
100.930.7870.0120.7650.016 POS178.0863
(386.1447)		C17H23NO9386.14560.90224.0917 (100); 178.0863 (40); 85.0284 (15)N-acetyl-L-tyrosine hexoside (unknow isomer)1,4[16,18]
110.960.7530.0190.7060.033 POS568.1873
(730.2401)		C25H43NO21730.24111.00198.0761 (100); 181.0496 (40); 152.0707 (30)L-DOPA hexoside derivative1
121.000.8880.0010.8150.007 POS198.0762C9H11NO4198.9761−0.10181.0496 (40); 152.0707 (100); 139.0390 (40); 135.0441 (20)L-DOPA1,4[22]
131.020.7970.0100.6580.054 POS719.2506
(360.1286)		C15H21NO9360.12890.30360.1286 (5); 198.0760 (100); 181.0495 (10); 152.0704 (10); 139.0390 (20); 85.0284 (5)L-DOPA hexoside (unknow isomer)1,4[16,18]
141.110.7700.0150.7360.024 POS386.1446C17H23NO9386.14561.00224.0918 (100); 178.0864 (40); 165.0547 (1); 85.0284 (10)N-acetyl-L-tyrosine hexoside (unknow isomer)1
151.860.8440.0040.7280.026 POS323.0874C14H14N2O7323.08740.00146.0601 (100)L-tryptophan derivative1
162.190.8130.0080.7880.012 POS166.0863C9H12NO2166.08630.00149.0597 (5); 120.0808 (100); 103.0542 (5)L-phenylalanine1,4[29,33]
173.600.9000.0010.7170.030 POS259.1764C11H22N4O3259.17650.10259.1766 (100); 242.1499 (20); 200.1279 (10); 175.1191 (10); 158.0925 (20); 112.0871 (1). 116.0705 (1); 70.0651 (5)L-arginine derivative1
185.300.8380.0050.7940.011 POS188.0706
(205.0969)		C11H12N2O2205.09710.20205.0969; 146.0602 (100); 118.0652 (10)L-tryptophan1,4[29,33]
199.680.7340.0240.6890.040 POS222.0648
(443.1226)		ND--281.0705 (30); 252.0441 (20); 237.0800 (80); 198.0761 (50); 181.0497 (80); 152.0708 (100); 139.0392 (30); 135.0442 (20); 85.0650 (30)L-DOPA hexoside derivative1
2011.280.7340.0240.6520.057 POS444.1867C20H29NO10444.1864−0.30229.1071 (20); 198.0761 (100); 181.0496 (70); 152.0706 (60); 139.0390 (10); 135.0440 (10); 85.0648 (30)L-DOPA derivative1
PHENOLIC COMPOUNDS
211.570.8510.0040.7810.013 NEG299.0773C13H16O8299.0772−0.10137.0244 (100); 93.0345 (30)p-Hydroxybenzoic hexoside1,4,5[10,42,47]
222.630.8080.0080.7300.025257; 293NEG315.0721C13H16O9315.07220.10153.0194 (50); 152.0114 (100); 109.0295 (30); 108.0217 (40)Protocatechuic acid
Hexoside		1,2,4,5[16,34,42,47]
234.940.8600.0030.8420.004282NEG285.0616C12H14O8285.06160.00223.0613 (20); 209.0456 (40); 195.0663 (100); 137.0608 (30)3′-O-Methylfukiic acid
(3-O-methyl
(3′,4′-dihydroxybenzyl tartaric acid))		1,2,3,4,5[16,35]
2412.800.8240.0060.7420.022270; 340POS595.1658C27H30O15595.1657−0.10287.0551Kaempferol derivative1,2,4,5[10,15,48]
ORGANIC ACIDS
251.070.7320.0250.7350.024 POS241.1546C12H20N2O3241.15470.10241.1546 (100); 242.1585 (10); 196.0965 (1); 168.0365 (1); 128.1069 (5); 84.0444 (5)Hydroxy aspergillic acid1,6
262.870.9070.0010.8160.007 POS305.0768C14H12N2O6305.07791.10215.0814 (20); 190.0499 (50); 172.0394 (100); 144.0444 (10)Kynurenic acid derivative1[38,39]
274.350.9020.0010.8430.004 NEG380.1552C15H27O10N380.15621.00362.1441 (10); 308.1351 (20); 218.1036 (10); 146.0822(90)Pantothenic acid hexoside1,5,6[41,42]
TERPENOIDS
286.140.7780.0130.7690.016 NEG403.1602C18H28O10403.1599−0.30403.1602 (100); 223.0976 (5); 179.1077 (5); 161.0455 (10); 59.0138 (50)8-β-D-glucopyranosyloxy-2,7-dimethyl-2,4-decadiene-1,10-dioic acid1,4[16]
296.250.7800.0130.7510.020 NEG887.3914
(443.1916)		C21H32O10443.19230.70281.1314 (1); 237.1497 (5); 219.1391 (10); 161.0454 (10); 101.0244 (70)Dihydrophaseic acid 4’-O-β-D-glucopyranoside1,4[16,19,21]
0.8340.0050.7810.013 POS467.1887
(445.2068)		445.20791.10284.0916 (50); 143.0563 (100)
OTHERS
300.740.9210.0000.8770.002 POS221.1859C10H24N2O3221.18600.10104.1071 (100); 60.0808 (1); 58.0657 (1)Choline derivative1
3115.980.7650.0160.6540.056 NEG447.2228C21H36O10447.22360.80315.1812 (20); 191.0564 (10); 161.0456 (70); 149.0456 (5); 143.0347 (10); 131.0349 (5); 113.0244 (70); 101.0244 (100)Geraniol pentoside hexoside (unknow isomer)1,6[45]
UNIDENTIFIED COMPOUNDS
320.870.7880.0120.6830.042 NEG545.1620ND--201.0709 (20); 196.0614 (30); 142.0509 (100); 100.0404 (40)Unknown
331.560.8770.0020.6360.066 POS317.1092C11H16N4O7317.11031.10155.0564 (100)Unknown
342.160.9020.0010.7470.021 POS374,1446C16H23NO9374.14571.10212.0918 (90); 195.0652 (90); 153.0547 (20); 152.0707 (100); 85.0284 (40); 69.0335 (10)Unknown
356.460.7920.0110.7380.023 POS242.0803
(483.1538)		ND--363.1116 (30); 339.1116 (30); 321.1010 (100); 303.0902 (40)Unknown (+ hexoside)
366.460.8460.0040.7870.012 POS490.2646C23H39NO10490.26470.10462.0538 (20); 320.0827 (90); 311.0769 (100); 265.1437 (40); 247.1310 (20)Unknown
376.460.8540.0030.7790.013 POS942.3170ND--499.1261 (10); 378.0767 (80); 320.0706 (100)Unknown
386.460.8630.0030.7720.015 POS927.3515ND--483.1540 (100); 363.1117 (40)
397.860.7410.0220.7080.033 NEG161.0819C7H14O4161.08190.00117.0557 (50); 99.0451 (50); 71.0502 (20)Unknown
4011.820.7480.0210.6830.043 NEG561.2550C26H42O13561.25530.30519.2444 (80); 387.2013 (100); 207.1386 (40); 191.0561 (40); 161.0454 (40); 113.0244 (30); 101.0244 (50); 99.0087 (70); 89.0244 (50); 71.0138 (40)Unknown
4112.380.8240.0060.7770.014 NEG529.2652C26H42O11529.26540.20-Unknown
4214.210.8880.0010.8200.007 NEG224.0120C10H8ClNO3224.01200.00180.0222 (70)Unknown
RT: retention time; RI: reliability of the attempted identification (1: main MS/MS fragment ions, 2: UV, 3: NMR, 4: identified in faba beans; 5: identified in other pulses; 6: identified in other plants); ND: not determined.

3.3. Non-Volatile Compounds Potentially Responsible for Faba Bean Bitterness

The difference in the bitter intensities has been partly explained by the presence of alkaloids, including vicine and convicine, in faba bean fractions [23]. However, the aim of this research was to gain a better understanding of the non-volatile compounds responsible for bitterness. Thus, linear and logarithmic Pearson correlations (α = 2.5%) were used to determine the relationships between the non-volatile compound areas, in both negative and positive modes, and the bitter intensity of the nine fractions. Indeed, linear regression should be used when the compound concentration is above the threshold, whereas non-linear models are more suitable when its concentration is below the threshold [10,49].
A total of 6 alkaloids, 4 phenolic compounds, 14 amino acids, 3 organic acids, 2 terpenoids, 2 other compounds, and 11 unidentified compounds positively correlated with bitterness were tentatively identified (Table 1). Then, PCA was performed to explore differences in the non-volatile content of the fractions related to bitterness (Figure 2). In the negative mode (Figure 2A), component 1 (47.44%) was more related to the cultivars, whereas component 2 (29.22%) separated the samples among the fractions. Most of the compounds were positively correlated with bitterness according to the linear model alone (6/14) or to both models (8/14). Concerning the positive mode (Figure 2B), component 1 (51.27%) was related to the type of fractions, whereas the cultivars were much more discriminated for component 2 (29.30%). Only one compound (25) was positively correlated with bitterness according to the logarithmic model, whereas the majority was highlighted by linear (12/31) or both models (18/31). Unlike the PCA from the negative mode, there were two groups of positively correlated compounds with bitterness, each of which appeared to be dependent on the P1 (bottom right) or P2 fractions (top right). N-acetyl-L-tyrosine hexoside (10 and 14) and L-DOPA hexoside derivative (19) were more characteristic of P2, whereas P1 was more related to the presence of vicine (1) and alkaloid derivatives (2, 3, 4, 5, and 6), L-arginine (7) and a derivative (17), L-tryptophan (15), and L-DOPA (13) and its derivatives (11, 12, and 15).
Among the alkaloids, vicine (1) is known to activate TAS2R16 and should be partly responsible for faba bean bitterness [23]. This suggested that vicine derivatives (2, 3, 4, and 6) also activated this receptor, notably if their β-glucopyranoside moiety was free to bind to the active site of TAS2R16 [50]. Concerning amino acids, L-phenylalanine (16) has been shown to activate TAS2R1 and 8, whereas L-tryptophan (18) only activates TAS2R4 and 39 receptors [51,52]. These results were consistent with the literature, which confirms that this multidisciplinary approach was relevant for identifying potential bitter compounds. In addition, the compound (15) was a derivative of L-tryptophan and could therefore activate bitter taste receptors. L-tyrosine, also a hydrophobic amino acid, exhibits a bitter taste [53], which suggested that N-acetyl-tyrosine hexoside (10 and 14) was characterised by a bitter taste. L-arginine (7) did not activate TAS2Rs, but this amino acid enhanced the bitter intensity of 6-n-propylthiouracil (PROP) and caffeine [54]. This suggested that the presence of free L-arginine in faba bean fractions could increase the bitter perception of the bitter compounds. There was no information on the bitterness of p-hydroxybenzoic hexoside (21), protocatechuic acid hexose (22), kaempferol derivative (24), and kynurenic acid derivative (26), but the bitter characteristics of the molecules from which they were derived have been determined. p-Hydroxybenzoic acid and protocatechuic acid were perceived as slightly strong bitter and moderately bitter, respectively, at a concentration of 2 g/L in water [55]. TAS2R14 and 39 receptors have been shown to be sensitive to kaempferol at a concentration in the µM range [56]. The bitter detection threshold of kynurenic acid in distilled water was 78.1 ppm [57]. In addition, it would be interesting to identify the compounds that were negatively correlated with bitterness that would potentially mask bitter perception and behave as potential TAS2R blockers.
These results were consistent with our previous research focused on the role of saponins (DDMP soyasaponin and soyasaponin βb) and alkaloids (vicine and convicine) in faba bean bitterness. Indeed, alkaloids should be partly responsible for bitterness, whereas the soyasaponin concentration was too low to activate TAS2R receptors [23]. In this current study, soyasaponins were detected but were not positively correlated with bitterness. In addition, vicine was also correlated with the bitter taste, which suggested anew the role of this molecule in faba bean off-flavours. Unlike peas, saponins did not contribute to the bitterness of faba beans [10,12].
Most of the tentatively identified compounds were derived from plant defence metabolism. Alkaloids limit the mollusc repellency of lupin plants [58]. It has been shown that water stress promotes the generation of alkaloids in faba beans [59]. Moreover, vicine exhibits fungicidal and insecticidal properties [60]. Amino acids serve mainly as intermediates in the generation of metabolites, but sometimes they have direct defensive functions against herbivores. For example, L-DOPA, a derivative of L-tyrosine, is a precursor or mimic of neurotransmitters that have been associated with insecticidal properties [61]. There was no information on the defence properties of N-acetyl-L-tyrosine hexoside (10 and 14), but N-acetyl-L-tyrosine was produced in common beans during fungal pathogen attacks [32]. Phenolic compounds are known to protect plant tissues against UV irradiation and attacks from herbivores, fungi, and viruses [62]. Concerning organic acids, kynurenic acid, which comes from tryptophan metabolism, is involved in plant defence mechanisms [38,40]. In pine, the catabolism of abscisic acid was activated during infection by Fusarium circinate, leading to the production of dihydrophaseic acid [63]. There were two hypotheses to explain the bitterness of a compound. The first hypothesis supposed that bitter molecules in plants would repel aggressors (insects, herbivores) to allow their development for survival. The second hypothesis suggests that bitter compounds are toxic to animals and prevent them from ingesting harmful doses [64]. In this article, the first hypothesis could explain the bitter taste of these compounds, which ensures the plant’s healthy development and survival. However, vicine and convicine are dangerous for people suffering from favism [65], and the second hypothesis thus could be plausible for specific compounds to prevent them from being ingested. It should be noted that these molecules, which were potentially responsible for off-flavours, also ensured good agricultural yields and highlighted that cultivars must be carefully selected. Thus, it would be interesting to also identify compounds that were not positively correlated with bitterness and verify if they are also involved in plant defence mechanisms to better select cultivars.
Finally, combining sensory and untargeted metabolomic approaches allowed us to target potential bitter compounds among a high number of metabolites. However, it was possible that some of these compounds were specific to the P1 and P2 fractions without activating the bitter taste receptors. For most compounds, only the chemical class or main moiety from which they were derived was tentatively used to characterise them. It would be interesting to confirm the identification of these compounds by standards and then determine their bitter characteristics by sensory and/or cellular approaches [9]. However, this approach was probably not complete for identifying all the bitter compounds. Indeed, some metabolites could be present in all the samples but insufficient concentrations to be detected. If several molecules were insufficient to activate the same bitter taste receptor alone, it was possible that all the molecules together, in the same sample, activated a TAS2R at the origin of perceived bitterness.

4. Conclusions

In this study, a combination of sensory and untargeted metabolomic analyses was used to tentatively identify the potential non-volatile compounds responsible for bitterness in air-classified faba bean fractions. Bitterness was more related to high-alkaloid cultivars and P fractions. A total of 42 compounds were positively correlated with bitterness, including 6 alkaloids, 4 phenolic compounds, 14 amino acids, 3 organic acids, 2 terpenoids, and 2 other compounds. However, the tentative identification of 11 compounds remains to be determined. To our knowledge, this work is the first one illustrating the variety of compounds from faba beans that are strong markers of bitterness. It should be interesting to apply our experimental design and this untargeted metabolomic approach to other faba bean cultivars. These results showed that bitter compounds in faba beans were different from those in peas, whose bitterness was more related to flavonoids and saponins. Then, the majority of these compounds were defence metabolites that were generated in the plants or seeds when they were exposed to abiotic or biotic stresses. This suggested that faba beans produced bitter compounds to prevent them from being ingested by pests to ensure their survival. One strategy for reducing the bitterness of faba beans is to select cultivars with low levels of these compounds while ensuring the presence of other defence metabolites to guarantee sensory, agronomic, and economic benefits. It should also be interesting to confirm the identification of the highlighted compounds to verify their ability to activate TAS2Rs by in vitro cellular-based assays or to be sensory detected by panellists. Finally, this approach allowed us to select potential compounds responsible for faba bean bitterness, but it did not consider the interactions between the compounds in the fractions, which could modulate their bitter intensity. However, this approach was relevant for reducing the number of studied compounds and can be applied to other food matrices or to analyse negative perceptions related to volatile compounds in faba beans such as green, rancid, and metallic notes, for example.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/metabo13080964/s1. Supplementary Materials about the workflow.

Author Contributions

Conceptualization, A.K., N.S. and C.S.; methodology, A.K.; formal analysis, A.K.; investigation, A.K. and E.M.; resources, E.M., A.B., A.-C.M.-O. and E.B.; writing—original draft preparation, A.K.; writing—review and editing, A.K., E.M., N.S., L.B. and C.S.; supervision, N.S., L.B. and C.S.; project administration, L.L.; funding acquisition, L.B. and C.S. All authors have read and agreed to the published version of the manuscript.

Funding

The PhD of Adeline Karolkowski was funded by Groupe Soufflet-Invivo (France), Groupe AVRIL (France), and ANRT (National Agency for Research and Technology, France) (ANRT-CIFRE 2019/1744).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Inserm (French National Institute for Health and Medical Research) (No. 21-773, approved in February 2021).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The data are not publicly available due to confidential reasons.

Acknowledgments

The authors are indebted to Christophe Martin and David Pertuit (CSGA, INRAE, Dijon, France) for their helpful advice in this study. They are also thankful to Damien Florès (SPO/PFP, INRAE, Montpellier, France) for the UHPLC–DAD–HRMS experiments. INRAE, Regional Council of Bourgogne Franche-Comté and the European Regional Development Fund (FEDER) are thanked for the equipment funding.

Conflicts of Interest

Adeline Karolkowski and Loïc Levavasseur are employees of Groupe Soufflet-Invivo. Emilie Bouzidi is employee of Vivien Paille (Groupe Avril). The paper reflects the views of the scientists, and not the companies.

References

  1. Willett, W.; Rockström, J.; Loken, B.; Springmann, M.; Lang, T.; Vermeulen, S.; Garnett, T.; Tilman, D.; DeClerck, F.; Wood, A.; et al. Food in the Anthropocene: The EAT–Lancet Commission on Healthy Diets from Sustainable Food Systems. Lancet 2019, 393, 447–492. [Google Scholar] [CrossRef]
  2. Sharan, S.; Zanghelini, G.; Zotzel, J.; Bonerz, D.; Aschoff, J.; Saint-Eve, A.; Maillard, M.-N. Fava Bean (Vicia faba L.) for Food Applications: From Seed to Ingredient Processing and Its Effect on Functional Properties, Antinutritional Factors, Flavor, and Color. Compr. Rev. Food Sci. Food Saf. 2021, 20, 401–428. [Google Scholar] [CrossRef] [PubMed]
  3. Mayer Labba, I.-C.; Frøkiær, H.; Sandberg, A.-S. Nutritional and Antinutritional Composition of Fava Bean (Vicia faba L., Var. Minor) Cultivars. Food Res. Int. 2021, 140, 110038. [Google Scholar] [CrossRef] [PubMed]
  4. Roland, W.S.U.; Pouvreau, L.; Curran, J.; van de Velde, F.; de Kok, P.M.T. Flavor Aspects of Pulse Ingredients. Cereal Chem. 2017, 94, 58–65. [Google Scholar] [CrossRef]
  5. Amarowicz, R.; Troszyńska, A.; Baryłko-Pikielna, N.; Shahidi, F. Polyphenolics Extracts from Legume Seeds: Correlations Between Total Antioxidant Activity, Total Phenolics Content, Tannins Content and Astringency. J. Food Lipids 2004, 11, 278–286. [Google Scholar] [CrossRef]
  6. Khrisanapant, P.; Kebede, B.; Leong, S.Y.; Oey, I. A Comprehensive Characterisation of Volatile and Fatty Acid Profiles of Legume Seeds. Foods 2019, 8, 651. [Google Scholar] [CrossRef] [PubMed]
  7. Karolkowski, A.; Gourrat, K.; Bouzidi, E.; Albouy, J.-F.; Levavasseur, L.; Briand, L.; Guichard, E.; Salles, C. Origins of Volatile Compounds and Identification of Odour-Active Compounds in Air-Classified Fractions of Faba Bean (Vicia faba L. Minor). Food Res. Int. 2023, 163, 112260. [Google Scholar] [CrossRef]
  8. Tuccillo, F.; Kantanen, K.; Wang, Y.; Martin Ramos Diaz, J.; Pulkkinen, M.; Edelmann, M.; Knaapila, A.; Jouppila, K.; Piironen, V.; Lampi, A.-M.; et al. The Flavor of Faba Bean Ingredients and Extrudates: Chemical and Sensory Properties. Food Res. Int. 2022, 162, 112036. [Google Scholar] [CrossRef]
  9. Karolkowski, A.; Belloir, C.; Briand, L.; Salles, C. Non-Volatile Compounds Involved in Bitterness and Astringency of Pulses: A Review. Molecules 2023, 28, 3298. [Google Scholar] [CrossRef]
  10. Cosson, A.; Meudec, E.; Ginies, C.; Danel, A.; Lieben, P.; Descamps, N.; Cheynier, V.; Saint-Eve, A.; Souchon, I. Identification and Quantification of Key Phytochemicals in Peas—Linking Compounds with Sensory Attributes. Food Chem. 2022, 385, 132615. [Google Scholar] [CrossRef]
  11. Cosson, A.; Oliveira Correia, L.; Descamps, N.; Saint-Eve, A.; Souchon, I. Identification and Characterization of the Main Peptides in Pea Protein Isolates Using Ultra High-Performance Liquid Chromatography Coupled with Mass Spectrometry and Bioinformatics Tools. Food Chem. 2022, 367, 130747. [Google Scholar] [CrossRef] [PubMed]
  12. Heng, L.; Vincken, J.-P.; van Koningsveld, G.; Legger, A.; Gruppen, H.; van Boekel, T.; Roozen, J.; Voragen, F. Bitterness of Saponins and Their Content in Dry Peas. J. Sci. Food Agric. 2006, 86, 1225–1231. [Google Scholar] [CrossRef]
  13. Gläser, P.; Mittermeier-Kleßinger, V.K.; Spaccasassi, A.; Hofmann, T.; Dawid, C. Quantification and Bitter Taste Contribution of Lipids and Their Oxidation Products in Pea-Protein Isolates (Pisum sativum L.). J. Agric. Food Chem. 2021, 69, 8768–8776. [Google Scholar] [CrossRef] [PubMed]
  14. Dupont, M.S.; Muzquiz, M.; Estrella, I.; Fenwick, G.R.; Price, K.R. Relationship between the Sensory Properties of Lupin Seed with Alkaloid and Tannin Content. J. Sci. Food Agric. 1994, 65, 95–100. [Google Scholar] [CrossRef]
  15. Neugart, S.; Rohn, S.; Schreiner, M. Identification of Complex, Naturally Occurring Flavonoid Glycosides in Vicia faba and Pisum sativum Leaves by HPLC-DAD-ESI-MSn and the Genotypic Effect on Their Flavonoid Profile. Food Res. Int. 2015, 76, 114–121. [Google Scholar] [CrossRef]
  16. Abu-Reidah, I.M.; Contreras, M.d.M.; Arráez-Román, D.; Fernández-Gutiérrez, A.; Segura Carretero, A. UHPLC-ESI-QTOF-MS Based Metabolic Profiling of Vicia faba L. (Fabaceae) Seeds as a Key Strategy for Characterization in Foodomics. Electrophoresis 2014, 35, 1571–1581. [Google Scholar] [CrossRef]
  17. Abu-Reidah, I.M.; Arráez-Román, D.; Warad, I.; Fernández-Gutiérrez, A.; Segura-Carretero, A. UHPLC/MS2-Based Approach for the Comprehensive Metabolite Profiling of Bean (Vicia faba L.) by-Products: A Promising Source of Bioactive Constituents. Food Res. Int. 2017, 93, 87–96. [Google Scholar] [CrossRef]
  18. Mekky, R.H.; Thabet, M.M.; Rodríguez-Pérez, C.; Elnaggar, D.M.Y.; Mahrous, E.A.; Segura-Carretero, A.; Abdel-Sattar, E. Comparative Metabolite Profiling and Antioxidant Potentials of Seeds and Sprouts of Three Egyptian Cultivars of Vicia faba L. Food Res. Int. 2020, 136, 109537. [Google Scholar] [CrossRef]
  19. Valente, I.M.; Cabrita, A.R.J.; Malushi, N.; Oliveira, H.M.; Papa, L.; Rodrigues, J.A.; Fonseca, A.J.M.; Maia, M.R.G. Unravelling the Phytonutrients and Antioxidant Properties of European Vicia faba L. Seeds. Food Res. Int. 2019, 116, 888–896. [Google Scholar] [CrossRef]
  20. Valente, I.M.; Maia, M.R.G.; Malushi, N.; Oliveira, H.M.; Papa, L.; Rodrigues, J.A.; Fonseca, A.J.M.; Cabrita, A.R.J. Profiling of Phenolic Compounds and Antioxidant Properties of European Varieties and Cultivars of Vicia faba L. Pods. Phytochemistry 2018, 152, 223–229. [Google Scholar] [CrossRef]
  21. Kowalczyk, M.; Rolnik, A.; Adach, W.; Kluska, M.; Juszczak, M.; Grabarczyk, Ł.; Wozniak, K.; Olas, B.; Stochmal, A. Multifunctional Compounds in the Extract from Mature Seeds of Vicia faba Var. Minor: Phytochemical Profiling, Antioxidant Activity and Cellular Safety in Human Selected Blood Cells in in Vitro Trials. Biomed. Pharmacother. 2021, 139, 111718. [Google Scholar] [CrossRef] [PubMed]
  22. Purves, R.W.; Zhang, H.; Khazaei, H.; Vandenberg, A. Rapid Analysis of Medically Relevant Compounds in Faba Bean Seeds Using FAIMS and Mass Spectrometry. Int. J. Ion Mobil. Spectrom. 2017, 20, 125–135. [Google Scholar] [CrossRef]
  23. Karolkowski, A.; Belloir, C.; Lucchi, G.; Martin, C.; Bouzidi, E.; Levavasseur, L.; Salles, C.; Briand, L. Activation of Bitter Taste Receptors by Saponins and Alkaloids Identified in Faba Beans (Vicia faba L. Minor). Food Chem. 2023, 426, 136548. [Google Scholar] [CrossRef]
  24. Aisala, H.; Manninen, H.; Laaksonen, T.; Linderborg, K.M.; Myoda, T.; Hopia, A.; Sandell, M. Linking Volatile and Non-Volatile Compounds to Sensory Profiles and Consumer Liking of Wild Edible Nordic Mushrooms. Food Chem. 2020, 304, 125403. [Google Scholar] [CrossRef] [PubMed]
  25. Karolkowski, A.; Martin, C.; Bouzidi, E.; Albouy, J.-F.; Levavasseur, L.; Briand, L.; Salles, C. Heat Treatment, Cultivar and Formulation Modify the Sensory Properties and Consumer Acceptability of Gels Containing Faba Bean (Vicia faba L. Minor) Protein Concentrates. Foods 2022, 11, 3018. [Google Scholar] [CrossRef]
  26. Mané, C.; Souquet, J.M.; Ollé, D.; Verriés, C.; Véran, F.; Mazerolles, G.; Cheynier, V.; Fulcrand, H. Optimization of Simultaneous Flavanol, Phenolic Acid, and Anthocyanin Extraction from Grapes Using an Experimental Design: Application to the Characterization of Champagne Grape Varieties. J. Agric. Food Chem. 2007, 55, 7224–7233. [Google Scholar] [CrossRef] [PubMed]
  27. Arapitsas, P.; Mattivi, F. LC-MS Untargeted Protocol for the Analysis of Wine. In Metabolic Profiling: Methods and Protocols; Theodoridis, G.A., Gika, H.G., Wilson, I.D., Eds.; Methods in Molecular Biology; Springer: New York, NY, USA, 2018; pp. 225–235. ISBN 978-1-4939-7643-0. [Google Scholar]
  28. Pulkkinen, M.; Gautam, M.; Lampi, A.-M.; Ollilainen, V.; Stoddard, F.; Sontag-Strohm, T.; Salovaara, H.; Piironen, V. Determination of Vicine and Convicine from Faba Bean with an Optimized High-Performance Liquid Chromatographic Method. Food Res. Int. 2015, 76, 168–177. [Google Scholar] [CrossRef]
  29. Piraud, M.; Vianey-Saban, C.; Petritis, K.; Elfakir, C.; Steghens, J.-P.; Morla, A.; Bouchu, D. ESI-MS/MS Analysis of Underivatised Amino Acids: A New Tool for the Diagnosis of Inherited Disorders of Amino Acid Metabolism. Fragmentation Study of 79 Molecules of Biological Interest in Positive and Negative Ionisation Mode. Rapid Commun. Mass Spectrom. 2003, 17, 1297–1311. [Google Scholar] [CrossRef]
  30. Nikolić, D.; Gödecke, T.; Chen, S.-N.; White, J.; Lankin, D.C.; Pauli, G.F.; van Breemen, R.B. Mass Spectrometric Dereplication of Nitrogen-Containing Constituents of Black Cohosh (Cimicifuga racemosa L.). Fitoterapia 2012, 83, 441–460. [Google Scholar] [CrossRef]
  31. Li, H.; Qin, Q.; Shi, X.; He, J.; Xu, G. Modified Metabolites Mapping by Liquid Chromatography-High Resolution Mass Spectrometry Using Full Scan/All Ion Fragmentation/Neutral Loss Acquisition. J. Chromatogr. A 2019, 1583, 80–87. [Google Scholar] [CrossRef]
  32. Chen, L.; Wu, Q.; He, T.; Lan, J.; Ding, L.; Liu, T.; Wu, Q.; Pan, Y.; Chen, T. Transcriptomic and Metabolomic Changes Triggered by Fusarium solani in Common Bean (Phaseolus vulgaris L.). Genes 2020, 11, 177. [Google Scholar] [CrossRef] [PubMed]
  33. Alghamdi, S.S. Chemical Composition of Faba Bean (Vicia faba L.) Genotypes under Various Water Regimes. Pak. J. Nutr. 2009, 8, 447–482. [Google Scholar]
  34. Abu-Reidah, I.M.; Arráez-Román, D.; Segura-Carretero, A.; Fernández-Gutiérrez, A. Extensive Characterisation of Bioactive Phenolic Constituents from Globe Artichoke (Cynara scolymus L.) by HPLC–DAD-ESI-QTOF-MS. Food Chem. 2013, 141, 2269–2277. [Google Scholar] [CrossRef]
  35. Heller, W.; Tamm, C. Fukiinsäure Und 3′-O-Methyl-Fukiinsäure, Zwei Phenolische Hydroxycarbonsäuren Aus Piscidia Erythrina. Helv. Chim. Acta 1975, 58, 974–979. [Google Scholar] [CrossRef]
  36. Mabry, T.; Markham, K.R.; Thomas, M.B. The Systematic Identification of Flavonoids; Springer Science & Business Media: New York, NY, USA, 1970; ISBN 978-3-642-88458-0. [Google Scholar]
  37. Carvalho, L.G.; McDougall, G.; Allwood, W.; Sungurtas, J.; Verrall, S.; Stewart, D.; Neilson, K.; O’Byrne, M. Tissues and Industrial Co-Products Formed during Alginate Extraction from Laminaria Hyperborea Provide Different Metabolite Profiles Depending on Harvest Season. J. Appl. Phycol. 2023, 35, 849–865. [Google Scholar] [CrossRef]
  38. Wang, Y.; Sun, W.; Zheng, J.; Xu, C.; Wang, X.; Li, T.; Tang, Y.; Li, Z. Urinary Metabonomic Study of Patients with Acute Coronary Syndrome Using UPLC-QTOF/MS. J. Chromatogr. B 2018, 1100–1101, 122–130. [Google Scholar] [CrossRef]
  39. Zhang, Z.-H.; Zhao, Y.-Y.; Cheng, X.-L.; Lin, R.-C.; Dai, Z.; Zhou, C. Metabonomic Study of Biochemical Changes in the Rat Urine Induced by Pinellia ternata (Thunb.) Berit. J. Pharm. Biomed. Anal. 2013, 85, 186–193. [Google Scholar] [CrossRef]
  40. Novelli, S.; Gismondi, A.; Di Marco, G.; Canuti, L.; Nanni, V.; Canini, A. Plant Defense Factors Involved in Olea Europaea Resistance against Xylella Fastidiosa Infection. J. Plant Res. 2019, 132, 439–455. [Google Scholar] [CrossRef] [PubMed]
  41. Mintz-Oron, S.; Mandel, T.; Rogachev, I.; Feldberg, L.; Lotan, O.; Yativ, M.; Wang, Z.; Jetter, R.; Venger, I.; Adato, A.; et al. Gene Expression and Metabolism in Tomato Fruit Surface Tissues. Plant Physiol. 2008, 147, 823–851. [Google Scholar] [CrossRef]
  42. Mekky, R.H.; Contreras, M.d.M.; El-Gindi, M.R.; Abdel-Monem, A.R.; Abdel-Sattar, E.; Segura-Carretero, A. Profiling of Phenolic and Other Compounds from Egyptian Cultivars of Chickpea (Cicer arietinum L.) and Antioxidant Activity: A Comparative Study. RSC Adv. 2015, 5, 17751–17767. [Google Scholar] [CrossRef]
  43. Min, M.-H.; Park, J.-H.; Hur, J.-H.; Shin, H.-C.; Cho, Y.; Kim, D.-D. Formulation and Bioequivalence Studies of Choline Alfoscerate Tablet Comparing with Soft Gelatin Capsule in Healthy Male Volunteers. Drug Des. Dev. Ther. 2019, 13, 1049–1058. [Google Scholar] [CrossRef] [PubMed]
  44. Hefni, M.; McEntyre, C.; Lever, M.; Slow, S. A Simple HPLC Method with Fluorescence Detection for Choline Quantification in Foods. Food Anal. Methods 2015, 8, 2401–2408. [Google Scholar] [CrossRef]
  45. Dong, L.; Miettinen, K.; Goedbloed, M.; Verstappen, F.W.A.; Voster, A.; Jongsma, M.A.; Memelink, J.; van der Krol, S.; Bouwmeester, H.J. Characterization of Two Geraniol Synthases from Valeriana officinalis and Lippia dulcis: Similar Activity but Difference in Subcellular Localization. Metab. Eng. 2013, 20, 198–211. [Google Scholar] [CrossRef]
  46. Jayeoye, T.J.; Eze, F.N.; Olatunji, O.J.; Tyopine, A.A. Synthesis of Biocompatible Konjac Glucomannan Stabilized Silver Nanoparticles, with Asystasia Gangetica Phenolic Extract for Colorimetric Detection of Mercury (II) Ion. Sci. Rep. 2022, 12, 9176. [Google Scholar] [CrossRef] [PubMed]
  47. Saied, D.B.; Farag, M.A. How Does Maturity Stage Affect Seeds Metabolome via UPLC/MS Based Molecular Networking and Chemometrics and in Relation to Antioxidant Effect? A Case Study in 4 Major Cereals and Legumes. Food Chem. 2023, 426, 136491. [Google Scholar] [CrossRef]
  48. García-Cayuela, T.; Gómez-Maqueo, A.; Guajardo-Flores, D.; Welti-Chanes, J.; Cano, M.P. Characterization and Quantification of Individual Betalain and Phenolic Compounds in Mexican and Spanish Prickly Pear (Opuntia ficus-Indica L. Mill) Tissues: A Comparative Study. J. Food Compos. Anal. 2019, 76, 1–13. [Google Scholar] [CrossRef]
  49. Chambers, E.; Koppel, K. Associations of Volatile Compounds with Sensory Aroma and Flavor: The Complex Nature of Flavor. Molecules 2013, 18, 4887–4905. [Google Scholar] [CrossRef]
  50. Bufe, B.; Hofmann, T.; Krautwurst, D.; Raguse, J.-D.; Meyerhof, W. The Human TAS2R16 Receptor Mediates Bitter Taste in Response to β-Glucopyranosides. Nat. Genet. 2002, 32, 397–401. [Google Scholar] [CrossRef]
  51. Bassoli, A.; Borgonovo, G.; Caremoli, F.; Mancuso, G. The Taste of D- and L-Amino Acids: In Vitro Binding Assays with Cloned Human Bitter (TAS2Rs) and Sweet (TAS1R2/TAS1R3) Receptors. Food Chem. 2014, 150, 27–33. [Google Scholar] [CrossRef]
  52. Kohl, S.; Behrens, M.; Dunkel, A.; Hofmann, T.; Meyerhof, W. Amino Acids and Peptides Activate at Least Five Members of the Human Bitter Taste Receptor Family. J. Agric. Food Chem. 2013, 61, 53–60. [Google Scholar] [CrossRef]
  53. Ishibashi, N.; Sadamori, K.; Yamamoto, O.; Kanehisa, H.; Kouge, K.; Kikuchi, E.; Okai, H.; Fukui, S. Bitterness of Phenylalanine- and Tyrosine-Containing Peptides. Agric. Biol. Chem. 1987, 51, 3309–3313. [Google Scholar] [CrossRef]
  54. Melis, M.; Arca, M.; Aragoni, M.C.; Cabras, T.; Caltagirone, C.; Castagnola, M.; Crnjar, R.; Messana, I.; Tepper, B.J.; Barbarossa, I.T. Dose-Dependent Effects of L-Arginine on PROP Bitterness Intensity and Latency and Characteristics of the Chemical Interaction between PROP and L-Arginine. PLoS ONE 2015, 10, e0131104. [Google Scholar] [CrossRef] [PubMed]
  55. Ferrer-Gallego, R.; Hernández-Hierro, J.M.; Rivas-Gonzalo, J.C.; Escribano-Bailón, M.T. Sensory Evaluation of Bitterness and Astringency Sub-Qualities of Wine Phenolic Compounds: Synergistic Effect and Modulation by Aromas. Food Res. Int. 2014, 62, 1100–1107. [Google Scholar] [CrossRef]
  56. Roland, W.S.U.; van Buren, L.; Gruppen, H.; Driesse, M.; Gouka, R.J.; Smit, G.; Vincken, J.-P. Bitter Taste Receptor Activation by Flavonoids and Isoflavonoids: Modeled Structural Requirements for Activation of HTAS2R14 and HTAS2R39. J. Agric. Food Chem. 2013, 61, 10454–10466. [Google Scholar] [CrossRef]
  57. Takahashi, K.; Tadenuma, M.; Kitamoto, K.; Sato, S. L-Prolyl-l-Leucine Anhydride A Bitter Compound Formed in Aged Sake. Agric. Biol. Chem. 1974, 38, 927–932. [Google Scholar] [CrossRef]
  58. Wink, M. Chemical Defense of Lupins. Mollusc-Repellent Properties of Quinolizidine Alkaloids. Z. Für Naturforschung C 1984, 39, 553–558. [Google Scholar] [CrossRef]
  59. Ulyanych, O.; Poltoretskyi, S.; Liubych, V.; Yatsenko, A.; Yatsenko, V.; Lazariev, O.; Kravchenko, V. Effect of Surface Drip Irrigation and Cultivars on Physiological State and Productivity of Faba Bean Crop. J. Agric. Sci. 2021, 32, 139–149. [Google Scholar] [CrossRef]
  60. Rimando, A.M.; Duke, S.O. Natural Products for Pest Management. In Natural Products for Pest Management; ACS Symposium Series; American Chemical Society: Washington, DC, USA, 2006; Volume 927, pp. 2–21. ISBN 978-0-8412-3933-3. [Google Scholar]
  61. Huang, T.; Jander, G.; de Vos, M. Non-Protein Amino Acids in Plant Defense against Insect Herbivores: Representative Cases and Opportunities for Further Functional Analysis. Phytochemistry 2011, 72, 1531–1537. [Google Scholar] [CrossRef] [PubMed]
  62. Cheynier, V. Phenolic Compounds: From Plants to Foods. Phytochem. Rev. 2012, 11, 153–177. [Google Scholar] [CrossRef]
  63. Amaral, J.; Correia, B.; Escandón, M.; Jesus, C.; Serôdio, J.; Valledor, L.; Hancock, R.D.; Dinis, L.-T.; Gomez-Cadenas, A.; Alves, A.; et al. Temporal Physiological Response of Pine to Fusarium circinatum Infection Is Dependent on Host Susceptibility Level: The Role of ABA Catabolism. Tree Physiol. 2021, 41, 801–816. [Google Scholar] [CrossRef]
  64. Meyerhof, W.; Batram, C.; Kuhn, C.; Brockhoff, A.; Chudoba, E.; Bufe, B.; Appendino, G.; Behrens, M. The Molecular Receptive Ranges of Human TAS2R Bitter Taste Receptors. Chem. Senses 2010, 35, 157–170. [Google Scholar] [CrossRef] [PubMed]
  65. Yannai, S.; Marquardt, R.R. Induction of Favism-like Symptoms in the Rat: Effects of Vicine and Divicine in Normal and Buthionine Sulphoximine-Treated Rats. J. Sci. Food Agric. 1985, 36, 1161–1168. [Google Scholar] [CrossRef]
Metabolites 13 00964 g001
Figure 1. Bitter intensity (over 10) of the 9 fractions evaluated by a trained panel. Significant differences are indicated by different letters (Tukey’s HSD test, α = 5.0%). S: starch fraction; F: flour; P: protein fraction—the number after the fraction corresponds to the cultivar (1, 2, or 3).
Figure 1. Bitter intensity (over 10) of the 9 fractions evaluated by a trained panel. Significant differences are indicated by different letters (Tukey’s HSD test, α = 5.0%). S: starch fraction; F: flour; P: protein fraction—the number after the fraction corresponds to the cultivar (1, 2, or 3).
Metabolites 13 00964 g001
Metabolites 13 00964 g002
Figure 2. Biplot representation of the PCA (centred reduced variables, Pearson correlation, α = 5.0%) of the detected compound areas in the negative (A) or positive (B) modes and the perceived bitterness (as a supplementary variable in light green) of the 9 faba bean fractions. The compounds positively correlated with bitterness are related to variables in blue for the linear model, in dark green for the logarithmic model and in pink for both the linear and logarithmic models, whereas the black highlighted compounds correspond to the mis-dereplicated data of the positively correlated compounds. S: starch fraction; F: flour; P: protein fraction—the number after the fraction corresponds to the cultivar (1, 2, or 3).
Figure 2. Biplot representation of the PCA (centred reduced variables, Pearson correlation, α = 5.0%) of the detected compound areas in the negative (A) or positive (B) modes and the perceived bitterness (as a supplementary variable in light green) of the 9 faba bean fractions. The compounds positively correlated with bitterness are related to variables in blue for the linear model, in dark green for the logarithmic model and in pink for both the linear and logarithmic models, whereas the black highlighted compounds correspond to the mis-dereplicated data of the positively correlated compounds. S: starch fraction; F: flour; P: protein fraction—the number after the fraction corresponds to the cultivar (1, 2, or 3).
Metabolites 13 00964 g002
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Karolkowski, A.; Meudec, E.; Bruguière, A.; Mitaine-Offer, A.-C.; Bouzidi, E.; Levavasseur, L.; Sommerer, N.; Briand, L.; Salles, C. Faba Bean (Vicia faba L. minor) Bitterness: An Untargeted Metabolomic Approach to Highlight the Impact of the Non-Volatile Fraction. Metabolites 2023, 13, 964. https://doi.org/10.3390/metabo13080964

AMA Style

Karolkowski A, Meudec E, Bruguière A, Mitaine-Offer A-C, Bouzidi E, Levavasseur L, Sommerer N, Briand L, Salles C. Faba Bean (Vicia faba L. minor) Bitterness: An Untargeted Metabolomic Approach to Highlight the Impact of the Non-Volatile Fraction. Metabolites. 2023; 13(8):964. https://doi.org/10.3390/metabo13080964

Chicago/Turabian Style

Karolkowski, Adeline, Emmanuelle Meudec, Antoine Bruguière, Anne-Claire Mitaine-Offer, Emilie Bouzidi, Loïc Levavasseur, Nicolas Sommerer, Loïc Briand, and Christian Salles. 2023. "Faba Bean (Vicia faba L. minor) Bitterness: An Untargeted Metabolomic Approach to Highlight the Impact of the Non-Volatile Fraction" Metabolites 13, no. 8: 964. https://doi.org/10.3390/metabo13080964

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop