Abstract

   Cereal crops are important staple foods, and their defense metabolites
   hold significant research importance. In this study, we employed
   LC-MS-based untargeted and widely-targeted metabolomics to profile the
   leaf metabolome of nine cereal species, including rice, wheat, maize,
   barley, sorghum, common oat, foxtail millet, broomcorn millet, and
   adlay. A total of 9869 features were detected, among them, 1131 were
   annotated, encompassing 18 classes such as flavonoids, lipids, and
   alkaloids. Results revealed that 531 metabolites were detected in all
   species, while each cereal crop possessed 4 to 12 unique metabolites.
   Focusing on defense metabolites, we identified eight benzoxazinoids
   uniquely present in maize, wheat, and adlay. Hierarchical clustering
   based on metabolite abundance divided all metabolites into nine
   clusters, and subsequent pathway enrichment analysis revealed that the
   stress-related flavonoid biosynthesis pathway was enriched in multiple
   clusters. Further analysis showed that four downstream compounds of
   HBOA (2-hydroxy-1,4-benzoxazin-3-one) in the benzoxazinoid biosynthesis
   pathway were enriched in maize. Wheat uniquely accumulated the
   4′-methylated product of tricin, trimethoxytricetin, whereas adlay
   accumulated the tricin precursor tricetin in the flavonoid biosynthesis
   pathway. In summary, this study elucidates the metabolic diversity in
   defense metabolites among various cereal crops, providing valuable
   background information for the improvement of stress resistance in
   cereal crops.

   Keywords: cereal crops, defense metabolite, metabolomic, flavonoid,
   benzoxazinoid

1. Introduction

   Cereal crops, including rice (Oryza sativa L.), wheat (Triticum
   aestivum L.), maize (Zea mays L.), barley (Hordeum vulgare L.), and
   sorghum (Sorghum bicolor L.), were domesticated as early as ten
   thousand years ago and represent a primary source of human calories
   [[44]1,[45]2]. In addition to providing essential food energy, certain
   cereals exhibit high nutritional value. For instance, adlay (Coix
   lacryma-jobi L.) contains a higher protein content compared to rice,
   wheat, and maize [[46]3], while common oats (Avena sativa L.) are among
   the richest sources of not only proteins but also fats and vitamin B1
   [[47]4]. Due to their significant importance to humans, cereal crops
   have become the focus of numerous research efforts [[48]5,[49]6]. Among
   these, research on stress tolerance in cereal crops has emerged as one
   of the key focal areas [[50]7,[51]8].

   Existing reports indicate that cereal crops exhibit diversity in terms
   of stress tolerance: for example, rice is a high-water-consuming plant
   and is highly sensitive to drought [[52]9,[53]10]. In contrast, crops
   such as foxtail millet (Setaria italica L.), broomcorn millet (Panicum
   miliaceum L.), and sorghum exhibit good drought tolerance, supporting
   their distribution in arid regions [[54]6,[55]10,[56]11]. Additionally,
   crops have varying temperature requirements. Wheat, barley, and common
   oats being cool-season crops, are susceptible to high temperatures
   [[57]4,[58]12,[59]13], while rice is more sensitive to cold stress
   compared to other cereal crops like wheat and barley [[60]7]. Some
   cereal crops that grow in harsh environments, such as Tibetan hulless
   barley (Hordeum vulgare L. var. nudum), can resist high-intensity UV-B
   radiation and low atmospheric pressure [[61]14]. Furthermore, plants
   are threatened by various pests and diseases, including the fungus
   Magnaporthe oryzae (anamorph Pyricularia oryzae) which causes rice
   blast disease [[62]15], and rust fungi of the genus Puccinia which
   threaten wheat, barley, and common oat, leading to various rust
   diseases [[63]16]. Clearly, different cereal crops possess distinct
   stress response mechanisms.

   In cereal crops, the diversity of defense metabolites may influence the
   diversity of their stress responses. Some defense metabolites that
   share core structures are widely present across various plant species,
   although their diversity largely arises from chemical modifications of
   core structures [[64]17,[65]18]. For example, metabolites with diverse
   structures within the phenylpropanoid pathway are extensively involved
   in the biotic and abiotic stress defense processes of various plant
   species [[66]19,[67]20]. For instance, sakuranetin, a flavanone, is
   induced under phosphorus deficiency in rice and contributes to
   enhancing rice resistance to Magnaporthe oryzae under
   phosphorus-deficient conditions [[68]21]. Additionally, the
   accumulation of several flavonoids can improve crop tolerance to UV-B
   radiation in rice [[69]22]. Notably, the increasing content of
   flavonoids and anthocyanins enhances drought stress tolerance in
   Tibetan hulless barley [[70]23]. Moreover, phenolamines, another type
   of phenylpropanoids, have been found in tomatoes to enhance drought
   tolerance [[71]24]. CPH, a phenolamine in tobacco, has been found to
   play an important role in tobacco’s chemical defense against
   leafhoppers [[72]25]. Beyond phenylpropanoids, other substances also
   widely participate in plant stress responses, including terpenoids
   [[73]26,[74]27,[75]28,[76]29], alkaloids [[77]30,[78]31], and
   branched-chain amino acids [[79]32,[80]33]. Similarly, several
   taxonomically restricted compound classes also have important roles in
   stress resistance. For example, glucosinolates in Brassicaceae plants
   are involved in defense against herbivores and pathogens [[81]34];
   steroidal glycoalkaloids (SGAs) in Solanaceae plants help protect
   against predators [[82]35]; and isoflavones enhance leaf resistance to
   soybean (Glycine max L.) mosaic virus [[83]20]. In some cereal crops,
   such as maize, wheat, and rye, a class of nitrogen-containing
   heterocyclic compounds called benzoxazinoids is present [[84]36].
   Benzoxazinoids act as allelochemicals, participating in early growth
   competition with other plants in maize, wheat, and rye, and they are
   actively involved in regulating plant resistance to pests and diseases
   [[85]37]. Additionally, benzoxazinoids produced by cereal crops can be
   absorbed by neighboring plants and utilized to enhance their resistance
   against pests and pathogens [[86]38]. In recent years, many defense
   metabolites in plants have been discovered, particularly within
   individual species, and the natural variation of these metabolites
   within species has led to differences in their defense capabilities
   [[87]25,[88]39]. However, analyses of the differences in defense
   metabolites among different cereal crops remain insufficient.

   In this study, we employed metabolomics by LC-MS to analyze the leaf
   metabolomes of nine cereal crops, examining the similarities and
   differences in the metabolomes of cereal crops. We focused on analyzing
   the differences in defense substances such as benzoxazinoids and
   flavonoids among various cereal crops, thereby revealing potential
   differences in stress resistance mechanisms at the metabolic level. Our
   study provides a metabolic basis for understanding the resistance
   differences mediated by defense metabolites among cereal crops, thus
   offering valuable background information for the breeding and genetic
   improvement of stress resistance in cereal crops.

2. Results

2.1. Metabolomics Analysis of Nine Cereal Crops

   To comprehensively evaluate the metabolic differences among cereal
   crops, we conducted an untargeted metabolomic analysis using LC–TOF–MS
   on the leaves of nine cereal crops. To avoid bias toward specific
   metabolites introduced by the mass spectrometry scanning mode, we
   performed full scans of the pooled samples from three biological
   replicates for each species in both positive and negative modes
   separately. By comparing the total ion chromatogram (TIC), we observed
   significant differences among the various cereal crops, particularly in
   the signal-dense region between 2 and 4 min ([89]Figure 1A).

Figure 1.

   [90]Figure 1
   [91]Open in a new tab

   Overview of metabolic variation in nine cereal crops. (A) Total ion
   chromatography of metabolites in nine cereal crops based on untargeted
   metabolomics. (B) Principal component analysis (PCA) of all features
   detected in nine cereal crops. F, foxtail millet; Bm, broomcorn millet;
   S, sorghum; A, adlay; M, maize; W, wheat; B, barley; C, common oat; R,
   rice.

   Next, we performed peak detection and alignment on the untargeted data
   using MS-DIAL [[92]40], resulting in a total of 9869 signals. Based on
   these signals, we conducted an unsupervised principal component
   analysis (PCA) on the nine cereal crops ([93]Figure 1B). The results
   indicated that PC1 and PC2 explained 51.3% and 8.8% of the total
   variance, respectively. The PCA score plot revealed that adlay and
   barley were distinctly separated from the other crops along PC1, and
   adlay was also distinctly separated from all crops except barley along
   PC2, indicating significant metabolic differences between these two
   species and the others. In contrast, seven crops—rice, wheat, maize,
   sorghum, common oats, foxtail millet, and broomcorn millet—clustered
   closely together in the PCA score plot, particularly foxtail millet and
   broomcorn millet.

   To address the limitations in sensitivity and quantitative accuracy
   inherent to untargeted metabolomics [[94]41] and to obtain more precise
   quantitative results, we constructed an MS² spectral tag (MS2T) library
   based on the untargeted metabolomics data, and developed a widely
   targeted metabolomics method [[95]42] to perform the quantitative
   analysis.

2.2. Metabolite Identification of Nine Cereal Crops

   We conducted a qualitative analysis based on the fragmentation patterns
   of the metabolites: benzoxazinoids are a class of nitrogen-containing
   heterocyclic compounds that typically exist in plants as glycosides
   [[96]37]. For common benzoxazinoid aglycones, fragmentation of the
   deprotonated DIMBOA (2,4-dihydroxy-7-methoxy1,4-benzoxazin-3-one)
   typically yields fragments at m/z 166 and m/z 164, corresponding to
   neutral losses of CO[2] (44 Da) and CH[2]O[2] (46 Da). DIBOA
   (2,4-dihydroxy-1,4-benzoxazin-3-one) exhibits losses of CH[2]O[2] (46
   Da) to form a fragment at m/z 134. Typical fragments described for the
   [M − OH]^− ion of HBOA (2-hydroxy-1,4-benzoxazin-3-one) and HMBOA
   (2-hydroxy-7-methoxy-1,4-benzoxazin-3-one) are fragments corresponding
   to the loss of CO (28 Da), CH[2]O[2] (46 Da), and 2CO (56 Da) [[97]43].
   For example, metabolite CL1148 (RT 3.39 min) is present in maize,
   wheat, and adlay ([98]Figure 2A). CL1148 yielded a precursor ion [M −
   H]^− at m/z 398.1102. The tandem mass spectrum of CL1148 showed a
   characteristic fragment Y[0]^− ion at m/z 194.046, along with fragments
   of Z[0]^− ion at m/z 166.0504 and Z[1]^− ion at m/z 138.0557, which
   indicated the presence of an HMBOA skeleton ([99]Figure 2B). The main
   neutral loss observed was 204 Da at m/z 194.0463, and we hypothesize
   that this loss corresponds to an acetyl-hexose (42 + 162 Da).
   Consequently, we identified CL1148 as HMBOA-acetyl-hexose ([100]Figure
   2C). Through fragmentation analysis of benzoxazinoids, we identified a
   total of eight benzoxazinoids.

Figure 2.

   [101]Figure 2
   [102]Open in a new tab

   Annotation of CL1148 and CL0463 as HMBOA-acetyl-glucose and C-glucoside
   luteolin O-pentose, respectively. (A) Extracted ion chromatogram (EIC)
   of CL1148, a specific metabolite in maize, wheat, and adlay. F, foxtail
   millet; Bm, broomcorn millet; S, sorghum; A, adlay; M, maize; W, wheat;
   B, barley; C, common oat; R, rice; Sb, soybean; (B) MS/MS spectra of
   CL1148, which was identified as HMBOA-acetyl-glucose. (C) Putative
   fragmentation rules of HMBOA-acetyl-glucose. (D) EIC of CL0463, a
   specific metabolite in foxtail millet. F, foxtail millet; Bm, broomcorn
   millet; S, sorghum; A, adlay; M, maize; W, wheat; B, barley; C, common
   oat; R, rice; Sb, soybean; (E) MS/MS spectra of CL0463, which was
   identified as C-glucoside luteolin O-pentose. (F) Putative
   fragmentation rules of C-glucoside luteolin O-pentose.

   Flavonoids are synthesized through the condensation of hydroxycinnamic
   acid (carbon atoms 2, 3, and 4 of the B and C rings) with a propylenol
   residue (A ring) and are primarily classified into classes such as
   flavones, flavonols, flavanones, flavanols, and anthocyanins based on
   variations in the C ring [[103]44]. In mass spectrometry, the 1 and 3
   positions of the flavonoid C ring are the primary cleavage sites.
   Additionally, hydroxyl substitutions frequently occur at the 5 and 7
   positions of the A ring, resulting in [M + H]^+ ions that typically
   produce a characteristic fragment at m/z 153. Certain flavonoids, due
   to the direct C-C bond linkage between the sugar moiety and the
   flavonoid aglycone at the A ring, undergo cleavage at the sugar
   substituent, generating characteristic ions such as [M + H − 150]^+, [M
   + H − 120]^+, and [M + H − 90]^+. For example, metabolite CL0463 is a
   compound specifically present in foxtail millet ([104]Figure 2D).
   CL0463 yielded a precursor ion [M + H]^+ at m/z 581.1522. The Y[0]^+
   ion at m/z 449.1108 [M + H − 132]^+ corresponds to the loss of a
   pentose. Additionally, the Z[0]^+ ion at m/z 329.0677 and the Z[2]^+
   ion at m/z 299.0554 correspond to the neutral losses of 120 Da and 150
   Da, respectively. Subsequently, a series of fragments of m/z 431.0979
   and m/z 413.0871 corresponded to the loss of H[2]O. Therefore, we
   propose that this compound is a C-glucoside luteolin O-pentose
   ([105]Figure 2F). As indicated by the untargeted data, a substantial
   number of flavonoids are present in cereal crops, and based on existing
   standards, we annotated a total of 311 flavonoids.

   Using a similar approach and comparing to the standard, we annotated a
   total of 1131 metabolites by LC–MS/MS analysis, encompassing a wide
   range of metabolic classes in plants. In addition to the flavonoids and
   benzoxazinoids mentioned above, we also annotated 172 lipids, 101 amino
   acids, 64 alkaloids, 59 hydroxycinnamoyl derivatives, 57 organic acids
   and derivatives, 48 nucleotides and derivates, 48 alcohols and polyols,
   46 phenolamides, 39 coumarins and lignans, 33 phenolic acids, 32
   vitamins, 27 phytohormones, 26 terpenoids, 16 quinate and derivatives,
   14 benzoic acids and derivatives, among 30 other compounds that did not
   fit into these 17 main classes ([106]Figure 3B, [107]Table S1).

Figure 3.

   [108]Figure 3
   [109]Open in a new tab

   Classification and species distribution of 1131 annotated metabolites.
   (A) An upset plot of the number of annotated metabolites in nine cereal
   crops. (B) Classification of annotated metabolites. AA, amino acid; BA,
   Benzoic acid; HCA, hydroxycinnamic acid. (C) Venn diagram analysis of
   annotated metabolites. F, foxtail millet; Bm, broomcorn millet; S,
   sorghum; A, adlay; M, maize; W, wheat; B, barley; C, common oat; R,
   rice.

2.3. Differentially Analysis of Metabolome Data

2.3.1. Metabolite Composition Differences Among Nine Cereal Crops

   To better understand the metabolic diversity of the nine cereal crops,
   we analyzed their metabolite compositions using LC-QTRAP-MS-based
   widely-targeted metabolomics method. Among the 1131 targeted
   metabolites, wheat had the highest number of detected metabolites
   (890), followed by rice and maize (880). The numbers of detected
   metabolites in the other cereal crops were as follows: 860 in broomcorn
   millet, 852 in common oat, 844 in sorghum, 837 in foxtail millet, 831
   in barley, and 789 in alday ([110]Figure 3A). A total of 530
   metabolites were present in all nine species (shared metabolites),
   accounting for 46.5% of the total metabolites ([111]Figure 3C). Among
   these, there were 315 primary metabolites (including lipids, amino
   acids, nucleotides, organic acids, and alcohols), representing 59.3% of
   the shared metabolites. And there were 216 shared secondary
   metabolites, representing 40.7%. We analyzed the proportion of each
   metabolite class relative to the total detected metabolites within that
   class among the shared metabolites to preliminarily assess the
   diversity of these metabolite classes across different species. Within
   the secondary metabolites, flavonoids accounted for the lowest
   proportion of the detected flavonoids at 12.5% (39 out of 311),
   followed by phenolamines at 15.2%. Other secondary metabolites ranged
   from 30.8% to 64.1%, while shared primary metabolites ranged from 54.2%
   to 84.3% of the detected metabolites ([112]Table S2).

   Additionally, we further analyzed the 600 metabolites that were missing
   in at least one cereal crop (non-shared metabolites). We first
   identified metabolites that appeared exclusively in a single cereal
   crop (species-specific metabolites). There were 12 metabolites detected
   exclusively in sorghum, and the fewest species-specific metabolites
   were found in maize and common oats, with 4 each. The number of
   species-specific metabolites detected in other species was as follows:
   9 in wheat, 6 in barley, 5 in foxtail millet, 9 in broomcorn millet,
   and 9 in adlay ([113]Figure 3C). Among the species-specific
   metabolites, flavonoids were the most abundant, with each species
   having between 1 and 7 specific flavonoids ([114]Table S3).
   Additionally, we analyzed metabolites that were not detected in
   individual species (species-missing metabolites) to identify potential
   pathway deficiencies specific to those species. There were 43
   species-missing metabolites in adlay, including 21 flavonoids and 1 to
   4 metabolites from other classes. Subsequently, 16 species-missing
   metabolites were found in barley, 12 in foxtail millet, 8 each in
   maize, wheat, and rice, 7 in common oat, 6 in sorghum, and 3 in
   broomcorn millet ([115]Figure 3A, [116]Table S4).

   To further analyze the variation of major defense metabolites among the
   nine cereal crops, we integrated the detection proportions of major
   secondary metabolite classes for each crop (the proportion of detected
   metabolites within each class relative to the total metabolites in that
   class) and mapped them onto the phylogenetic tree ([117]Figure 4,
   [118]Table S5). We calculated the coefficient of variation (CV) for the
   detection proportions of each metabolite class across the species to
   quantify the differences in the number of these metabolites among
   cereal crops. The analysis revealed that benzoxazinoids had the highest
   CV at 151% among the nine cereal crops, followed by phenolamines at
   30%, terpenoids at 17%, and flavonoids at 11%. Further analysis showed
   that eight benzoxazinoids were uniquely present in maize, adlay, and
   wheat, with each species detecting at least six of these compounds and
   five being present in all three species ([119]Figure 4). Additionally,
   phenolamines had lower detection proportions in sorghum, common oat,
   and maize (35%, 39%, and 41%, respectively), while they were detected
   at higher proportions in rice and adlay (85% and 72%, respectively).
   Terpenoids had lower detection proportions in adlay and maize (38% and
   58%, respectively).

Figure 4.

   [120]Figure 4
   [121]Open in a new tab

   Treemap visualization of major secondary metabolite classes
   distribution. The lineage divergence time is adopted from the TimeTree
   database ([122]http://www.timetree.org, accessed on 20 December 2024).
   F, foxtail millet; Bm, broomcorn millet; S, sorghum; A, adlay; M,
   maize; W, wheat; B, barley; C, common oat; R, rice; Sb, soybean; MYA,
   millions of years; N.D., non-detected.

2.3.2. Metabolite Abundance Differences Among Nine Cereal Crops

   To identify the patterns of metabolite abundance differences among the
   species, we performed hierarchical clustering analysis on the abundance
   of the 1131 metabolites across the nine crops, and a total of the
   metabolites were divided into nine clusters ([123]Figure 5A). Analysis
   revealed that the metabolites enriched in these nine clusters were
   predominantly those with higher abundances in their respective crops
   compared to other crops: including 115 metabolites in maize, 128
   metabolites in adlay, 153 metabolites in rice, 131 metabolites in
   foxtail millet, 154 metabolites in wheat, 105 metabolites in common
   oat, 124 metabolites in barley, 124 metabolites in broomcorn millet,
   and 101 metabolites in sorghum, respectively ([124]Figure 5B–J).
   However, some metabolites, although enriched in specific clusters
   (corresponding to certain crops), exhibited higher abundances in other
   crops, such as 4 benzoxazinoids in cluster II (corresponding to adlay)
   showing higher levels in maize and to a lesser extent in adlay.

Figure 5.

   [125]Figure 5
   [126]Open in a new tab

   Hierarchical cluster analysis of 1131 annotated metabolites. (A)
   Hierarchically clustered heatmap of the 1131 annotated metabolites from
   nine cereal crops. Metabolic data were Z-score standardized (color
   scale from blue to red). AA, amino acid; BA, Benzoic acid; HCA,
   hydroxycinnamic acid. (B–J) Accumulation patterns of clusters I to IX.
   Clusters I to IX were specifically accumulated in maize (115
   metabolites), adlay (127 metabolites), rice (152 metabolites), foxtail
   millet (130 metabolites), wheat (154 metabolites), common oat (105
   metabolites), barley (124 metabolites), broomcorn millet (124
   metabolites), and sorghum (100 metabolites). The x-axis depicts nine
   cereal crops, and the y-axis depicts the Z score standardized per
   metabolite.

   To identify the metabolic pathways enriched with the metabolites in
   each cluster, we performed enrichment analysis using the Kyoto
   Encyclopedia of Genes and Genomes (KEGG) for the metabolites in the
   nine clusters ([127]Figure S1). Here, we focused on the enrichment of
   defense-related metabolic pathways in each cluster. Among the nine
   clusters, six were enriched in phenylpropanoid biosynthesis pathways,
   of which three were statistically significant (p < 0.05), and five were
   enriched in flavonoid biosynthesis pathways, of which three were
   statistically significant (p < 0.05). Additionally, benzoxazinoid
   biosynthesis pathway was enriched in cluster II (p = 0.098).

   To better investigate metabolite abundance patterns in nine cereal
   crops, we compiled the classification of metabolites in each cluster
   ([128]Table 1). Among the crops corresponding to the nine clusters,
   adlay (cluster II) was enriched in benzoxazinoids, phenolic acids,
   phytohormones, coumarins, and lignans. In rice (cluster III),
   phenolamines were found to be enriched. Foxtail Millet (cluster IV) was
   enriched in hydroxycinnamoyl derivatives, vitamins, phenolic acids,
   benzoic acids, and derivatives. Wheat (cluster V) was enriched in
   quinate and derivatives, flavonoids, and benzoxazinoids. The common oat
   (cluster VI) was enriched in lipids. Barley (cluster V) was enriched in
   benzoic acids and derivatives, coumarins and lignans, organic acids and
   derivatives, and alcohols. Broomcorn millet (cluster VII) was enriched
   in phenolamides, benzoic acids, and derivatives. Maize accumulated
   metabolites across multiple classes without a specific enrichment, and
   sorghum exhibited similar results.

Table 1.

   Distribution of the metabolites among nine clusters.
   Cluster I Cluster II Cluster III Cluster IV Cluster V Cluster VI
   Cluster VII Cluster VIII Cluster IX In Total
   Maize Adlay Rice Foxtail Millet Wheat Common Oat Barley Broocorn Millet
   Sorghum
   AA and derivatives 7 7 36 8 10 4 14 7 8 101
   Alcohols and polyols 3 9 2 6 6 5 2 10 * 5 48
   Alkaloids 9 4 15 6 6 4 10 4 6 64
   BA and derivatives 0 2 0 3 * 2 2 3 * 1 1 14
   Benzoxazinoids 0 6 ** 0 0 2 * 0 0 0 0 8
   Coumarins and lignans 3 8 * 1 5 4 0 2 9 * 7 39
   Flavonoids 35 36 32 34 69 * 24 22 23 36 311
   HCA derivatives 3 8 7 15 4 2 8 9 3 59
   Lipids 31 3 9 14 16 43 * 30 20 6 172
   Nucleotides and derivates 8 4 9 3 6 3 5 3 7 48
   Organic acids 2 10 6 9 3 6 3 12 * 6 57
   Others 4 5 4 2 6 * 1 4 3 1 30
   Phenolamides 0 7 15 * 1 2 1 9 11 * 0 46
   Phenolic acids 3 7 * 1 7 * 5 1 1 4 4 33
   Phytohormones 4 6 * 2 4 3 2 1 2 3 27
   Quinate and derivatives 2 1 2 1 4 * 2 0 2 2 16
   Terpenoids 0 2 5 5 2 3 6 * 2 1 26
   Vitamins 1 2 6 7 * 4 2 4 2 4 32
   In total 115 127 152 130 154 105 124 124 100 1131
   [129]Open in a new tab

   AA, amino acid; BA, Benzoic acid; HCA, hydroxycinnamic acid. *
   Accounted for more than 20%. ** Accounted for more than 40%.

2.4. Defense Metabolites Analysis

   To better understand the reasons behind the differences in defense
   metabolites among the nine cereal crops, we integrated established
   pathways and proposed pathways to reconstruct detailed pathway maps of
   defense metabolites. These maps, combined with our metabolomic results,
   expanded our understanding of the current metabolic pathways in these
   crops.

   We detected eight benzoxazinoids in maize, wheat, and adlay, among
   which the synthesis pathways for four compounds—HMBOA-acetyl-glucoside,
   HMBOA—glucoside-rhamnose, DIBOA-glucoside-hexose, and
   DIMBOA-glucoside-hexose—had not been identified. Therefore, we
   reconstructed a pathway (including established and proposed pathways)
   that includes the eight benzoxazinoids we detected ([130]Figure 6).
   Considering the difference of one acetyl group, we hypothesize that
   HMBOA-acetyl-glucoside may be synthesized from HMBOA-glucose through
   the catalysis of an acetyltransferase. Similarly,
   HMBOA-glucoside-rhamnose, DIBOA-glucoside-hexose, and
   DIMBOA-glucoside-hexose are likely formed from their corresponding
   mono-glycosylated substrates through a single glycosylation reaction.
   By analyzing the concentrations of metabolites within these pathways,
   we found that HBOA-glucoside, DIBOA-glucoside, DIMBOA-glucoside-hexose,
   and DIMBOA-glucoside have relatively higher abundances in maize
   compared to adlay and wheat. These four metabolites can be synthesized
   from HBOA through 1–5 reaction steps ([131]Figure 6). Wheat exhibited
   higher levels of DIBOA-glucoside-hexose and HMBOA-glucoside-rhamnose,
   while adlay showed higher levels of HBOA and HMBOA-acetyl-glucoside. In
   conclusion, based on our metabolomic results, we proposed previously
   unreported steps in the benzoxazinoid biosynthetic pathway and revealed
   both the similarities and differences of this pathway among maize,
   adlay, and wheat.

Figure 6.

   [132]Figure 6
   [133]Open in a new tab

   Metabolic pathway analysis for benzoxazinoid biosynthesis. BX1-14,
   benzoxazineless1 to benzoxazineless14, reported benzoxazinoid synthesis
   enzymes; Hex, hexose; Glc, glucoside; Rha, rhamnose.

   Additionally, we mapped the abundance of all the flavonoids we detected
   to the flavonoid biosynthesis pathway ([134]Figure 7). Analysis
   revealed that adlay and maize contain more anthocyanins. Similarly,
   flavonol compounds with hydroxylation at the 3-position were enriched
   in adlay and maize, although this enrichment is not very apparent.
   Furthermore, flavonoid C-glycosides accumulated more in wheat and
   maize, whereas adlay showed the least accumulation. Analysis of all
   flavonoid O-glycosides indicated that adlay had the least accumulation,
   followed by maize, while no significant enrichment was observed in
   other cereal crops. Notably, unlike most flavonoids, tricetin—a highly
   hydroxylated flavonoid—was significantly enriched in adlay. Further
   analysis of tricetin-related compounds revealed that tricin (tricetin
   3′,5′-dimethyl ether) and its derivatives were present at lower levels
   in adlay but accumulated more in wheat. Additionally, two
   trimethoxytricetin (TMT, 4′-methylated tricin) glycosides also
   accumulated heavily in wheat. In summary, we highlighted the
   differences in metabolites involved in each step of the flavonoid
   biosynthesis pathway among the nine cereal crops.

Figure 7.

   [135]Figure 7
   [136]Open in a new tab

   Metabolic pathway analysis for flavonoid biosynthesis. PAL,
   phenylalanine ammonialyase; 4CL, 4-coumarate-CoA ligase; C4H, cinnamic
   acid 4-hydroxylase; CHS, chalcone synthase; CHI, chalcone isomerase;
   F3H, flavonoid 3-hydroxylase; FH, flavonoid hydroxylase; F2H, flavonoid
   2-hydroxylase; DH, dehydratase; DFR, dihydroflavonol 4-reductase; ANS,
   anthocyanin synthase; FLS, flavonol synthase; FNS, flavone synthase;
   OMT, O-methyltransferase; 3′5′OMT, 3′5′O-methyltransferase; 4′OMT,
   4′O-methyltransferase; OGT, O-glycosyltransferase; CGT,
   C-glycosyltransferase. Parentheses indicate that only a part of the
   products requires this enzyme.

3. Discussion

   Defense metabolites play crucial roles in protecting cereal crops from
   both biotic and abiotic stresses, thereby maintaining normal growth and
   yield. Throughout evolution, plants have developed diverse metabolic
   pathways, with specific metabolites playing significant roles in stress
   response [[137]34,[138]35,[139]36]. In this study, comparative
   metabolomic analysis of nine cereal crops has revealed differences in
   the composition and abundance of defense-related metabolites, providing
   valuable insights into the variations in metabolism-mediated defense
   responses among different cereals.

   As important defense metabolites, flavonoids are universally present in
   all plants [[140]45]. With advancements in metabolome profiling
   methods, an increasing number of flavonoid derivatives have been
   discovered, and the functions of flavonoid compounds in plant defense
   are receiving heightened attention [[141]18]. Previous studies have
   suggested that different subclasses of flavonoids play distinct roles
   in plants. In this study, we identified 311 flavonoids, and subsequent
   analyses revealed that flavonoids exhibit considerable diversity in
   both composition and abundance among cereal crops. The distinct
   metabolic profiles of different plant species may influence their
   adaptability to environmental stresses [[142]46]. Thus, we hypothesize
   that flavonoids may be one of the primary contributors to the
   differences in environmental adaptability among cereal crops.
   Furthermore, flavone C-glycosides, a class of important phytoalexins,
   have been identified in various cereal crops, including rice, maize,
   wheat, sorghum, and barley [[143]14,[144]47,[145]48]. Our study further
   explores the distribution of flavone C-glycosides in minor cereal
   crops, revealing that adlay millet has lower flavone C-glycoside
   content compared to other cereal crops. The distribution of tricetin
   and its derivative tricin in adlay, as well as the substantial
   accumulation of trimethoxytricetin in wheat, serve as exemplary cases.
   Notably, Zhu et al. [[146]49] discovered a specific flavonoid
   4′O-methyltransferase in wheat, enabling the synthesis of
   trimethoxytricetin, which is consistent with our findings. Similarly,
   we speculate that adlay accumulates tricetin but only a small amount of
   tricin, or even undetectable levels of TMT, likely due to the
   inhibition of enzymes responsible for converting tricetin to tricin. A
   recent study suggested that the presence of fiber in adlay seeds limits
   the bioavailability of its active ingredients, whereas microbial
   fermentation can enhance the content of various components in adlay
   seeds, including flavonoids [[147]50]. The lower accumulation of
   flavonoid derivatives in adlay may be attributed to similar reasons.
   Considering that only a few flavonoids are species-specific and these
   flavonoids do not belong to specific subsets, we hypothesize that the
   majority of flavonoid modification pathways are present across all nine
   cereal crops. Whether these pathways function consistently across
   different species requires additional evidence.

   In contrast to flavonoids, some metabolites are widely recognized as
   species-specific, such as glucosinolates, steroidal glycoalkaloids
   (SGAs), and isoflavones [[148]20,[149]34,[150]35]. In this study, we
   also identified benzoxazinoids that are specifically present in maize,
   wheat, and adlay. The presence of these compounds in maize, wheat, and
   adlay is supported by extensive research [[151]36,[152]51]. Previous
   studies have primarily focused on the presence or absence of
   benzoxazinoids in various plants, their biosynthesis, and their
   bioactivities [[153]37,[154]52,[155]53]. Our study addresses the
   abundance differences of benzoxazinoids in cereal crops, thereby
   complementing previous knowledge gaps. Specifically, four
   benzoxazinoids were found at high concentrations in maize, while wheat
   and adlay each had two benzoxazinoids with the highest concentrations.
   We found that DIMBOA-glucoside accumulated to higher levels in maize
   and adlay millet, a compound that has been reported to enhance the
   resistance of wheat and maize to aphids [[156]54]. The benzoxazinoid
   biosynthesis pathway has been elucidated in maize, whereas only partial
   enzymes within this pathway have been reported in sorghum and foxtail
   millet. In contrast, genomic studies in rice and barley suggest the
   absence of genes required for benzoxazinoid biosynthesis [[157]55].
   There are no available reports for cultivated common oats and broomcorn
   millet. In the current understanding of benzoxazinoid pathways, the
   synthesis enzymes for the four benzoxazinoids we detected have not been
   reported, and our speculations on unknown pathways and abundance
   information provide references for the future discovery of these