Phenomics, RNA sequencing and weighted gene co-expression network analysis reveals key regulatory networks and genes involved in the determination of seed hardness in Vicia sativa

Phenomics, RNA sequencing and weighted gene co-expression network analysis reveals key regulatory networks and genes involved in the determination of seed hardness in Vicia sativa | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Phenomics, RNA sequencing and weighted gene co-expression network analysis reveals key regulatory networks and genes involved in the determination of seed hardness in Vicia sativa Honglin Wang, Zizhou Wu, Yanchun Zuo, Xu Yan, Bangxing Zou, Yu Chen, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6974115/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 23 Oct, 2025 Read the published version in BMC Genomics → Version 1 posted 12 You are reading this latest preprint version Abstract Background Wild-type Vicia sativa L. demonstrates superior agronomic traits, including high yield, elevated crude protein content, enhanced reproductive efficiency, and tolerance to nutrient-poor soils, rendering it a valuable genetic resource for crop improvement. However, its utilization in germplasm enhancement is severely constrained by seed hardness, whose underlying physiological and molecular regulatory mechanisms remain poorly characterized. Results Phenomics studies reveal that wild-type seeds exhibited significantly higher hardness indices and lignin content compared to cultivated varieties, whereas cultivated seeds contain significantly greater levels of soluble sugars, soluble proteins, and starch. During the same developmental period, cultivated seeds demonstrate significantly larger diameters than wild-type seeds. However, wild-type seeds show significantly thicker palisade layers and higher palisade-layer-to-seed-coat thickness ratios. Notably, the cuticle thickness of wild-type seeds is significantly greater. Scanning electron microscopy further indicated that cultivated varieties have significantly larger hilum width and length, whereas wild-type seeds display significantly deeper micropyle depth. Through WGCNA analysis, three key candidate genes ( TRINITY_DN3402_c0_g1 , TRINITY_DN13607_c0_g1 , and TRINITY_DN6606_c0_g1 ) involved in regulating seed hardness were identified. The expression pattern analysis results indicated that, the expression levels of TRINITY_DN3402_c0_g1 , TRINITY_DN13607_c0_g1 , and TRINITY_DN6606_c0_g1 in the wild hard-seeded type vicia sativa (W12) were significantly higher than those in the cultivated type vicia sativa (W30). Conclusions ​Our study elucidates key physical determinants and molecular mechanisms underlying seed hardness in Vicia sativa L., providing critical insights for crop improvement. Vicia sativa L Hard-seededness Cuticula RNA-seq Ethylene response Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction The Vicia sativa L., an annual leguminous crop, is a high-quality dual-purpose species with integrated functions of forage production, green manure, and ecological restoration. Renowned for its broad ecological adaptability (cold-tolerant, drought-resistant, and adaptable to poor soils), efficient biological nitrogen fixation capacity, and outstanding nutritional properties, it has become a critical grass-arable crop rotation species in temperate and alpine regions worldwide [ 1 – 3 ]. In China's agro-pastoral ecotone, the mixed cropping of Vicia sativa with Avena sativa for sun-cured hay production leverages the legume-grass synergy to significantly enhance forage quality and optimize resource utilization efficiency. This system serves as a pivotal technological framework to address seasonal shortages of spring forage and ensure balanced nutritional supply for livestock, thereby supporting sustainable animal husbandry [ 4 , 5 ]. Currently, Vicia sativa faces constraints germplasm innovation due to outdated traditional breeding techniques, a narrow genetic base, and varietal degeneration arising from prolonged cultivation. Addressing these challenges hinges on the evaluation and innovation of germplasm resources. Wild Vicia sativa L., with its unique biological traits and environmental adaptability, has been identified as a high-quality gene pool for improving cultivated varieties, offering critical genetic diversity to enhance stress resistance, yield stability, and agronomic performance [ 6 , 7 ]. Breeding practices have revealed that many desirable genes in wild Vicia sativa cannot be effectively utilized, primarily due to the varying degrees of hard-seededness (dormancy) in its seeds. While this trait enhances population persistence and stress resistance in natural ecosystems, it introduces significant genetic constraints in agricultural production, manifesting as uneven seedling emergence, delayed growth cycles, and reduced yields [ 8 ]. Long-standing research on seed hard-seededness, both domestically and internationally, has predominantly focused on elucidating its underlying mechanisms and developing dormancy-breaking methods, with a heavy emphasis on staple crops such as soybeans. However, limited research has been conducted on the hard-seededness trait in Vicia sativa L., particularly from physiological and molecular perspectives, leaving critical gaps in understanding its regulatory networks and evolutionary adaptations [ 9 ]. Seed coat hardness, a key domestication related trait in legumes, is primarily governed by the synergistic interaction between the palisade layer and hilum. The palisade layer-composed of lignin- and suberin-enriched macrosclereids-forms the major barrier to water penetration in wild soybean seeds, while the hilum serves as the dominant water entry pathway. Crucially, the closure of the hilum fissure under low humidity effectively blocks external moisture infiltration, thereby enhancing seed hardness [ 10 – 12 ]. This mechanism is genetically regulated by GmHs1-1-mediated calcium signaling and PG031-directed pectin degradation [ 13 , 14 ]. The finding that "the removal of soybean seed coat results in equivalent water absorption rates among embryos confirms seed coat permeability as the primary cause of seed hardness" is supported by research on seed coat structural characteristics [ 8 ]. Another critical factor contributing to seed hardness is the presence of an impermeable cuticle on the seed coat. Significant differences exist in cuticle structure between hard seeds and imbibing seeds: the cuticle of imbibing seeds exhibits microcracks primarily distributed on the dorsal side, whereas hard seeds maintain an intact cuticular structure without such fissures [ 15 ]. The calcium ion content, pectin levels, xylan composition, and cuticular wax layer in the seed coat of hard seeds are key factors influencing seed coat permeability and contributing to physical dormancy [ 16 , 17 ]. Environmental variations significantly influence the formation of soybean seed hard-seededness. Key contributors include changes in natural ecological climate during the late growth stages, storage and transportation conditions, as well as environmental circumstances from post-harvest to long-term preservation periods [ 18 ]. Notably, subtle variations in water management directly influence soybean seed hardening: under constant irrigation frequency, reduced water application volume leads to a corresponding increase in hard seed proportion [ 19 ]. Furthermore, the intensity and duration of sunlight exposure must not be overlooked. Shorter radiation periods and lower light intensity significantly inhibit the formation of hard soybean seeds [ 20 ]. During storage and transportation, low environmental relative humidity accelerates seed hardening. For instance, phaseolus vulgaris, declining seed moisture content during storage leads to significant increases in hard seed rates [ 21 ]. Seed dormancy is considered one of the most enigmatic phenomena in seed biology. Although extensive research on seed dormancy has been conducted to date with fruitful outcomes, numerous questions and uncertainties remain unresolved. Seed hardness, a relatively stable genetic trait, is inherited by progeny plants from their parental lines. Studies reveal that the genetic basis of seed hardness appears to involve variations in a limited number of genes, with only a few critical genes driving selective differentiation between hard and non-hard seeds. Most evidence supports seed hardness as a dominant trait [ 22 , 23 ]. In Arabidopsis thaliana, the cloned DOG1 gene represents the first identified quantitative trait locus (QTL) associated with regulating seed hardness. It was initially proposed that DOG1 influences seed hardness by participating in the ABA (abscisic acid)-mediated sugar signaling pathway [ 24 ]. Subsequently, researchers identified two novel seed dormancy regulators, DESPIERTO and ATHB20 . Both regulators modulate seed dormancy through adjusting ABA sensitivity, with mutations in the DESPIERTO gene resulting in a complete loss of seed dormancy [ 25 ]. The seed hardness trait in soybeans was governed by the GmHS1-1 gene , a dominant monogenic locus. Mutations in this gene reduce calcium content in the seed coat and thin the palisade layer, leading to seed hardness breakdown [ 26 ]. The Glyma02g43680 gene, identified as a key regulator, exhibits amino acid substitutions in the substrate-binding region of 1,4-β-glucan endo-hydrolase in non-hardy soybean varieties. These mutations diminish 1,4-β-glucan synthesis in the seed coat, suggesting its critical role in seed coat impermeability [ 27 ]. Phenotypic analysis of non-hardy mutants in alfalfa ( Medicago sativa ) revealed structural alterations in seed cuticles, characterized by modified lipid monomer composition. These changes induced micro-fissures in the cuticular layer, significantly enhancing water permeability. Genetic investigations identified that mutations in the KNOX4 gene-encoding a KNOTTED -like homeobox protein—disrupt its regulatory control over CYP86A , a cytochrome P450 gene responsible for fatty acid hydroxylation. This epistatic interaction critically impairs cuticular component biosynthesis, establishing KNOX4 mutation as the primary determinant of rapid water uptake in mutant seeds [ 28 ]. In vegetable soybean ( Glycine max ), comparative genomic studies by pinpointing a pectin acetylesterase-coding gene under positive selection during domestication [ 29 ]. Hard-seeded cultivars exhibit functional alleles, while water-imbibing varieties carry frameshift mutations that abolish enzymatic activity. Despite extensive studies on plant dormancy, the molecular mechanisms governing how dormancy factors regulate this trait remain poorly understood. Weighted Gene Co-expression Network Analysis (WGCNA), a systems biology approach first established by Zhang et al [ 30 ], has emerged as a pivotal tool for integrating multi-omics data (transcriptomics, metabolomics, proteomics) to uncover gene functions and network-trait relationships. Based on 508 rice affymetrix microarray datasets, a co-expression network was constructed. Using Random Matrix Theory (RMT) and Weighted Gene Co-expression Network Analysis (WGCNA) methods, 45 functional modules were automatically identified, among which 6 modules (e.g., dark green and dark grey modules) were significantly associated with seed dormancy [ 31 ].​Wang et al. [ 32 ] analyzed 216 soybean landraces from 26 Chinese provinces, selecting contrasting genotypes for seed hardness, and transcriptome sequencing across developmental stages combined with WGCNA pinpointed GmSWEET2 as a central regulator. Functional validation confirmed that overexpression of GmSWEET2 enhances seed hardness by modulating sugar transport dynamics, providing a genetic basis for improving vegetable soybean texture. Despite extensive studies on seed hardness in legumes, no prior research has employed transcriptome sequencing coupled with WGCNA to identify hardness-associated genes in Vicia sativa . This study pioneers the integration of comparative physiology and multi-omics approaches to dissect the wild-domesticated divergence in seed coat impermeability. Results The agronomic trait analysis and ploidy identification of two vicia sativa. L In this study, the analysis of agronomic traits revealed that the cultivated type vicia sativa (W30) exhibited significantly large leaf length-width ratios, petal sizes, and seed dimensions compared to the wild hard-seeded type vicia sativa (W12) (Fig. 1 -B to E). Additionally, W30 demonstrated robust stem growth and a more spreading plant architecture (Fig. 1 -A). Cytological identification further indicated that both the wild hard-seeded type (W12) and cultivated type (W30) shared identical diploid chromosome numbers (2n = 2x = 12), as evidenced by karyotype analysis (Fig. 1 -F, G). Seed hardness and chemical composition contents At the seed full maturity stage, the seed nutritional components and hardness index were determined for the cultivated type of material (W30) and the wild hard-seeded type material (W12). The results showed no significant difference in crude protein content between the cultivated type of material (W30) and the wild hard-seeded type material (W12) (Fig. 2 -A). However, the soluble sugar, soluble protein, and starch contents of the cultivated type of material (W30) were significantly higher than those of the wild hard-seeded type material (W12) (Fig. 2 -B-D). Notably, the seed hardness index and lignin content of the wild hard-seeded type material (W12) were significantly higher than those of the cultivated type of material (W30) (Fig. 2 -E-F). Additionally, the determination of 20 mineral elements in the two materials revealed that the seed coats of the wild hard-seeded type material (W12) exhibited significantly higher levels of potassium (K), aluminum (Al), and manganese (Mn) compared to the cultivated type material (W30). However, the cultivated type material (W30) showed significantly higher levels of calcium (Ca), iron (Fe), sodium (Na), nickel (Ni), copper (Cu), strontium (Sr), cadmium (Cd), antimony (Sb), barium (Ba), and thallium (Tl) than the wild hard-seeded type material (W12) (Table S1 ). Dynamic monitoring of seed coat and seed development across various developmental stages of two vicia sativa In this study, the seed coat structures of the wild hard-seeded type vicia sativa (W12) and cultivated type vicia sativa (W30) were identical, consisting sequentially from the outer to inner layers as follows: the cuticle layer, palisade layer, osteosclereid layer, sclerenchyma layer, and parenchyma cells (Fig. 3 -A-B). The seed diameter of both types exhibited an initial increase followed by a decrease during development. However, throughout the seed development process, the seed diameter of the cultivated type vicia sativa (W30) was significantly larger than that of the wild hard-seeded type vicia sativa (W12) (Fig. 3 -C-I). Additionally, the study revealed that the seed coat thickness of the wild hard-seeded type vicia sativa (W12) was significantly greater than that of the cultivated type vicia sativa (W30) (Fig. 3 -C-II). As seed development progressed, the palisade tissue thickness of the wild hard-seeded type vicia sativa (W12) and cultivated type vicia sativa (W30) gradually decreased. However, at all developmental stages, the palisade tissue thickness of the wild hard-seeded type vicia sativa (W12) remained significantly greater than that of the cultivated type vicia sativa (W30) (Fig. 3 -C-III). Notably, the ratio of palisade to seed coat of the wild hard-seeded type vicia sativa (W12) exhibited an initial increase followed by a decline, whereas that of the cultivated type vicia sativa (W30) showed a continuous upward trend (Fig. 3 -C-IV). Notably, at the wax ripening stage (WRS), the cuticle thickness of the wild hard-seeded type vicia sativa (W12) was significantly greater than that of the cultivated type vicia sativa (W30) (Fig. 3 -D). Scanning electron microscopy (SEM) analysis of seed coat Scanning electron microscopy (SEM) results revealed that the seed structures of the cultivated type vicia sativa (W30) and the wild hard-seeded type vicia sativa (W12) were similar, both possessing a cuticle layer with comparable morphology, and identical longitudinal section structures of the seed coats (Fig. 4 -A). Further observations of the hilar traits of the two materials revealed significant differences between the cultivated type vicia sativa (W30) and the wild hard-seeded type vicia sativa (W12). The wild hard-seeded type vicia sativa (W12) exhibited a funnel-shaped hilum, positioned farther from the tracheary elements and occupying a smaller proportion of the seed diameter (Fig. 4 -A-I-III). In contrast, the cultivated type vicia sativa (W30) displayed a significantly wider hilum that was oriented perpendicular to the tracheary elements and tightly connected to them (Fig. 4 -A-IV-VI). Subsequent measurements of cuticle thickness demonstrated that the wild hard-seeded type vicia sativa (W12) exhibited a significantly thicker cuticle layer compared to the cultivated type vicia sativa (W30). As the cuticle serves as a critical barrier to water penetration into the seed coat, this structural divergence is hypothesized to correlate with seed hardness (Fig. 3 -D). Furthermore, the research also indicates that the wild hard-seeded type vicia sativa (W12) has a narrow and short hilum, with a small hilum width. The hilar fissure, hilum length, seed width, seed diameter, hilum length ratio, and hilar fissure ratio were also smaller than that of the cultivated type vicia sativa (W30) (Fig. 4 -B-I-VII). However, the hilar fissure depth in the wild hard-seeded type (W12) was significantly greater than the cultivated type vicia sativa (W30) (Fig. 4 -B-VIII). RNA-Seq analyses developing seeds of two vicia sativa. L with contrasting seed hardness In this study, 18 samples generated approximately 6.09 Gb of raw sequencing data (Raw reads). After filtering, approximately 6.00 Gb of high-quality data (Clean reads) were obtained, with each sample containing over 3.6 Gb of Clean bases, indicating a high sequencing depth (Table 1 ). Additionally, the base error rate of the sequenced reads across the 18 samples was below 0.01%, while Q20, Q30, and GC content values were above 97%, 93%, and 35%, respectively (Table S2 ). These results demonstrate that the sequences obtained from the transcriptome were accurate and reliable, making them suitable for subsequent analyses. Subsequently, the clean reads were assembled into transcripts using trinity software for further analysis. The results showed that the total length of transcripts was 261,468,601 bp, with a total of 171,109 sequences. The maximum transcript length was 16,812 bp, and the average length was 1,528.08 bp, with N50 and N90 values of 2,213 bp and 718 bp, respectively. For Unigenes, the total sequence length was 75,122,054 bp, comprising 62,729 sequences. The maximum Unigene length was 16,812 bp, with an average length of 1,197.56 bp, and N50 and N90 values of 2,032 bp and 468 bp, respectively (Table S3). Additionally, the mapping rates for all 18 samples exceeded 90%, confirming their suitability for subsequent analyses (Table S4). Functional annotation of unigenes revealed that 37,247 unigenes were successfully annotated in the NR database, 21,931 in the GO database, 13,669 in the KEGG database, 19,523 in the Pfam database, 34,086 in the eggNOG database, and 27,435 in the SwissProt database. Among these, the NR database annotated the highest number of Unigenes, accounting for 59.38% of the total (Table S5; Figure S1 -A). By aligning with the NR database, functional information of the species' genes and their similarity to homologous genes in closely related species were obtained. The NR annotation results indicated that the highest sequence similarity was to Medicago sativa (29.78%), followed by Trifolium pratense (16.39%) and Trifolium subterraneum (12.46%). Additionally, 24.14% of the annotations corresponded to other species (Figure S1 -B). The GO functional annotation results revealed that, in the biological process category, unigenes were predominantly annotated to pathways such as cellular process and metabolic process. Under cellular component, unigenes were significantly annotated to the cellular anatomical entity pathway. For molecular function, unigenes were primarily associated with binding and catalytic activity pathways (Figure S1 -C). The KEGG functional annotation results demonstrated that unigenes were prominently annotated to pathways including metabolism, genetic information processing, environmental information processing, cellular processes, and organismal systems (Figure S1 -D; Table S6). Table 1 Data filtering statistics Sample Clean Reads No. Clean Data (bp) Clean Reads % Clean Data % W12_1_1 44040178 6641055704 98.91 98.78 W12_1_2 40317258 6080154234 99.10 98.97 W12_1_3 44641220 6731921523 98.85 98.72 W30_2_1 42574096 6419387846 99.24 99.09 W30_2_2 38678104 5832232560 99.02 98.88 W30_2_3 38355094 5783190296 99.18 99.04 W12_2_1 36808734 5549679363 99.11 98.96 W12_2_2 40296592 6074116464 99.02 98.85 W12_2_3 40956608 6174278615 99.09 98.92 W30_3_1 39222454 5913565498 99.07 98.92 W30_3_2 37984666 5727832498 99.13 98.99 W30_3_3 36480664 5500403766 99.10 98.95 W12_3_1 40247532 6065676435 99.26 99.07 W12_3_2 40829154 6152134690 99.03 98.82 W12_3_3 41610942 6268440345 99.14 98.91 W30_4_1 37764994 5689771390 99.00 98.78 W30_4_2 37012072 5576123662 99.04 98.82 W30_4_3 37929466 5716276542 99.23 99.04 Identification of DEGs during seed development Pearson correlation coefficient analysis was performed on the gene expression values (FPKM) of the 18 samples. The results indicated that correlation coefficients closer to 1 reflect higher similarity in expression patterns between samples (Figure S2 -A). Principal Component Analysis (PCA) was conducted using the DESeq package in R based on expression levels, revealing high similarity in expression patterns within groups and significant differences between groups. Samples from different treatment groups were dispersed, while samples within the same group clustered together (Figure S2 -B). At the filling stage, there were 9,936 differentially expressed genes (DEGs) between the wild hard-seeded material W12 and the cultivated type material W30, with 4,056 genes up-regulated and 5,880 genes down-regulated (Figure S3-A). At the milk-ripe stage, 8,609 DEGs were identified between W12 and W30, including 3,467 up-regulated genes and 5,142 down-regulated genes (Figure S3-B). During the wax ripening stage, 9,459 DEGs were observed, with 4,200 genes up-regulated and 5,259 genes down-regulated (Table S7; Figure S3-C). GO and KEGG enrichment analysis of DEGs At the filling stage, GO enrichment analysis of the 9,936 DEGs revealed significant enrichment in 33 Biological Processes (BP), 9 Cellular Components (CC), and 17 Molecular Functions (MF) (Table S8; Figure S4-A). Among these, the top three enriched BPs were cell wall organization or biogenesis (245 DEGs), secondary metabolite biosynthetic process (105 DEGs), and external encapsulating structure organization (203 DEGs). The top three enriched CCs were intrinsic component of membrane (1,135 DEGs), integral component of membrane (1,090 DEGs), and extracellular region (385 DEGs). Additionally, KEGG enrichment analysis of the 9,936 DEGs revealed significant enrichment in 11 biological pathways. The top three enriched pathways were Biosynthesis of other secondary metabolites (81 DEGs), Lipid metabolism (14 DEGs), and Biosynthesis of other secondary metabolites (23 DEGs) (Table S9; Figure S4-D). At the milk-ripe stage, GO enrichment analysis of the 8,609 DEGs revealed significant enrichment in 5 Biological Processes (BP), 2 Cellular Components (CC), and 10 Molecular Functions (MF) (Table S10; Figure S4-B). The top three enriched BPs were RNA modification (192 DEGs), nucleic acid phosphodiester bond hydrolysis (275 DEGs), and meiotic chromosome segregation (32 DEGs). The top two enriched CCs were intrinsic component of membrane (1,014 DEGs) and integral component of membrane (969 DEGs). Furthermore, KEGG enrichment analysis of the 8,609 DEGs demonstrated significant enrichment in 12 biological pathways. The top three enriched pathways were Brassinosteroid biosynthesis (8 DEGs), DNA replication (30 DEGs), and Mismatch repair (25 DEGs) (Table S11; Figure S4-E). At the wax ripening stage, GO enrichment analysis of the 9,459 DEGs revealed significant enrichment in 8 Biological Processes (BP), 1 Cellular Component (CC), and 5 Molecular Functions (MF) (Table S12; Figure S4-C). The top three enriched BPs were RNA modification (234 DEGs), ncRNA processing (183 DEGs), and nucleic acid phosphodiester bond hydrolysis (303 DEGs). Additionally, KEGG enrichment analysis of the 9,459 DEGs demonstrated significant enrichment in 7 biological pathways. The top three enriched pathways were Ribosome biogenesis in eukaryotes (42 DEGs), DNA replication (34 DEGs), and Flavonoid biosynthesis (23 DEGs) (Table S13; Figure S4-F). In addition, we conducted a comprehensive analysis of DEGs across the three comparison groups. The results revealed 2,554 DEGs commonly enriched in all three groups (Figure S5-A). Subsequent GO enrichment analysis of these 2,554 DEGs showed that the top three enriched Biological Processes (BP) were meiotic nuclear division (19 DEGs), RNA phosphodiester bond hydrolysis, endonucleolytic (25 DEGs), and meiotic chromosome segregation (12 DEGs) (Figure S5-B). KEGG enrichment analysis indicated that the 2,554 DEGs were associated with 101 biological pathways, with only the RNA polymerase pathway (11 DEGs) showing significant enrichment (Figure S5-C-D; Table S14). Further analysis of the expression patterns of these 11 DEGs revealed that only TRINITY_DN15782_c0_g2 , TRINITY_DN3450_c1_g1 , TRINITY_DN21678_c0_g1 , TRINITY_DN200_c0_g2 , and TRINITY_DN24823_c0_g1 exhibited significantly higher expression levels in the wild hard-seeded material W12 compared to the cultivated type material W30 (Figure S5-E). Weight gene co-expression network analysis (WGCNA), identification of the key candidate gene and establishing a regulatory network of key candidate genes The WGCNA was performed using the FPKM values of 4,999 DEGs and four phenotypic traits (seed diameter, seed coat thickness, palisade tissue thickness, and ratio of palisade to seed coat). First, hierarchical clustering of all samples was conducted based on gene expression levels, with each sample treated as a cluster, and distances between clusters were calculated (Figure S6-A-B). Next, a gene clustering tree was constructed based on pairwise gene expression correlations, and modules were defined by dynamically cutting branches of the tree, grouping genes with similar expression patterns into the same module (Figure S6-C-D). The results showed that the 4,999 DEGs were partitioned into 22 modules (Fig. 5 -A). Notably, the grey60 module exhibited the highest correlation with seed diameter ( P < 0.001 ), the purple module with seed coat thickness ( P < 0.001 ), the salmon module with palisade tissue thickness ( P < 0.001 ), and the cyan module with ratio of palisade to seed coat ( P 0.70 and P-value < 0.01, combined with FPKM expression values, 7 candidate genes were identified: TRINITY_DN3202_c0_g3 , TRINITY_DN3663_c0_g2 , TRINITY_DN1963_c0_g1 , TRINITY_DN17484_c0_g1 , TRINITY_DN12479_c0_g1 , TRINITY_DN818_c0_g1 , and TRINITY_DN16154_c0_g1 (Fig. 5 -C; Table S15). Based on functional annotation, TRINITY_DN3663_c0_g2 (annotated as short-chain alcohol dehydrogenase A) was preliminarily identified as a key candidate gene (Table 2 ). Further analysis of the purple module revealed 183 DEGs within this module. Using thresholds of Pearson correlation coefficient > 0.70 and P-value < 0.01, combined with FPKM expression values, 7 key candidate genes were identified: TRINITY_DN4437_c0_g1 , TRINITY_DN477_c0_g1 , TRINITY_DN5404_c0_g2 , TRINITY_DN26784_c0_g1 , TRINITY_DN4058_c0_g1 , TRINITY_DN2362_c0_g1 , and TRINITY_DN3402_c0_g1 (Fig. 5 -D; Table S15). Based on functional annotation, TRINITY_DN3402_c0_g1 (annotated as ethylene response 2) was preliminarily identified as a key candidate gene (Table 2 ). Further analysis of the salmon module revealed 120 DEGs within this module. Using thresholds of Pearson correlation coefficient > 0.70 and P-value < 0.01, combined with FPKM expression values, 6 candidate genes were identified: TRINITY_DN1542_c0_g2 , TRINITY_DN108_c0_g1 , TRINITY_DN7853_c0_g1 , TRINITY_DN13607_c0_g1 , TRINITY_DN4244_c0_g1 , and TRINITY_DN1127_c0_g1 (Fig. 5 -E; Table S15). Based on functional annotation, TRINITY_DN13607_c0_g1 (annotated as ethylene response factor 8) was preliminarily identified as a key candidate gene (Table 2 ). Further analysis of the cyan module revealed 115 DEGs within this module. Using thresholds of Pearson correlation coefficient > 0.70 and P-value < 0.01, combined with FPKM expression values, 8 candidate genes were identified: TRINITY_DN1476_c1_g1 , TRINITY_DN12417_c1_g1 , TRINITY_DN3075_c0_g1 , TRINITY_DN3979_c0_g1 , TRINITY_DN6069_c0_g1 , TRINITY_DN6606_c0_g1 , TRINITY_DN1196_c0_g3 , and TRINITY_DN4975_c0_g2 (Fig. 5 -F; Table S15). Based on functional annotation, TRINITY_DN6606_c0_g1 (annotated as glycine-rich cell wall structural protein 1.0-like) was preliminarily identified as a key candidate gene (Table 2 ). Table 2 Key candidate genes identified in this study Gene Description TRINITY_DN3663_c0_g2 BAC81652.1 short-chain alcohol dehydrogenase A TRINITY_DN3402_c0_g1 AOD74920.1 ethylene response 2 TRINITY_DN13607_c0_g1 AEQ64868.1 ethylene response factor 8 TRINITY_DN6606_c0_g1 XP_039023342.1 glycine-rich cell wall structural protein 1.0-like Analysis of expression patterns of key candidate genes The candidate gene expression pattern analysis results indicated that during three stages of seed development, the expression levels of TRINITY_DN6606_c0_g1 and TRINITY_DN3402_c0_g1 in the wild hard-seeded type vicia sativa (W12) were significantly higher than those in the cultivated type vicia sativa (W30) (Fig. 5 -G-I-II). Notably, at both the FS and WRS, the expression level of TRINITY_DN13607_c0_g1 in the wild hard-seeded type vicia sativa (W12) was significantly higher than that in the cultivated type vicia sativa (W30), while no significant difference was observed at the MRS (Fig. 5 -G-III). Furthermore, during three stages of seed development, the expression levels of TRINITY_DN3663_c0_g2 in the wild hard-seeded type vicia sativa (W12) was significantly lower than these in the cultivated type vicia sativa (W30) (Fig. 5 -G-IV). This study ultimately identified TRINITY_DN6606_c0_g1 , TRINITY_DN3402_c0_g1 , and TRINITY_DN13607_c0_g1 as key candidate genes regulating seed hardness, warranting further investigation. Effects of exogenous ethylene on seed germination In previous studies, we hypothesized that ethylene response factors ( TRINITY_DN3402_c0_g1 and TRINITY_DN13607_c0_g1 ) regulated seed hardness. To validate this hypothesis, we investigated the effects of soaking seeds in varying ethylene concentrations on germination. The results demonstrated that as ethylene concentration increased, the cultivated type vicia sativa (W30) seeds exhibited a decreasing trend in germination rate, with germination rates under 100 mg/L and 150 mg/L treatments being significantly lower than those of the control group (Fig. 6 -A; B-I). Additionally, under the 150 mg/L ethylene treatment, the wild hard-seeded type vicia sativa (W12) exhibited a significantly lower seed germination rate compared to the control group (Fig. 6 -A; B-IV). Notably, increasing ethylene concentrations significantly reduced both root length and shoot length in the cultivated type vicia sativa (W30), demonstrating a clear inhibitory effect (Fig. 6 -B-II-III). However, at the 50 mg/L treatment, the wild hard-seeded type vicia sativa (W12) showed significantly greater root and shoot lengths than the control group and the other two treatment groups (Fig. 6 -B-V-VI). These results indicate that the 50 mg/L ethylene treatment significantly promotes root and shoot growth in the wild hard-seeded type vicia sativa (W12). Discussion Physical factors affecting seed hardness in Vicia sativa The results of this study indicate that the hardness index of wild Vicia sativa seeds is significantly higher than that of cultivated varieties. This phenomenon may be closely associated with their unique physical structural features and cell wall composition. First, the thicker cuticle layer in wild seeds likely enhances seed coat mechanical resistance, directly contributing to increased hardness. As the outermost protective barrier of seeds, the thickened cuticle effectively resists external mechanical stress and water penetration. This mechanism aligns with the functional model of the cuticle proposed by wherein increased cuticle thickness enhances physical resistance by improving the density of the cutin-wax complex [ 33 ]. Wherein increased cuticle thickness enhances physical resistance by improving the density of the cutin-wax complex. Second, the higher lignin content in wild seeds may serve as a critical biochemical basis for increased hardness. Lignin, a secondary metabolite in the cell wall, forms a rigid network structure by cross-linking cellulose and hemicellulose [ 34 ]. Its accumulation significantly enhances the mechanical strength of the seed coat and cotyledon cells. Additionally, the structural features of the hilum in wild seeds, such as a narrow and short hilum, smaller hilar fissure width, and deeper hilar fissure, may indirectly influence hardness by restricting rapid water absorption. The constricted hilum likely slows the imbibition rate, thereby reducing the risk of microcracks in internal tissues caused by rapid expansion [ 35 ]. This mechanism helps maintain seed integrity and prolongs seed hardness. Notably, although cultivated varieties exhibit higher levels of metabolites such as soluble sugars and starch, their lower hardness suggests that seed hardness is primarily governed by physical structures rather than storage compounds. This finding contrasts with studies on common bean ( Phaseolus vulgaris ) by [ 32 ], but aligns with conclusions on hardseededness in pea ( Pisum sativum ) [ 36 ], which emphasize that seed coat structural traits, such as lignification degree and cuticle thickness are the core determinants of hardness, while storage components predominantly influence post-germination metabolic activity. Additionally, calcium (Ca) is a key element influencing seed hardness, as it participates in pectin cross-linking and stabilizes cell wall structure. Ca²⁺ form Ca²⁺-pectin cross-links with pectic acids in the cell wall, enhancing the mechanical strength of the seed coat and reducing water permeability [ 37 ]. However, in this study, hard-seeded genotypes exhibited significantly lower calcium (Ca) content compared to non-hard-seeded genotypes, while showing higher levels of potassium (K), manganese (Mn), and aluminum (Al), alongside significantly higher hardness indices. This apparent contradiction to the traditional view of calcium promoting seed coat hardening suggests that the regulation of seed hardness in Vicia sativa may involve elemental interactions, structural compositional differences in the seed coat, or functional specificity of calcium. Manganese (Mn) is a cofactor for key enzymes in lignin biosynthesis. Elevated Mn levels may enhance the activity of lignin-synthesizing enzymes, such as manganese peroxidase, promoting the oxidative polymerization of lignin monomers [ 38 ]. Aluminum (Al³⁺) can induce peroxidase activity to accelerate lignin deposition [ 39 ] and may directly bind to carboxyl groups of pectin in the cell wall, strengthening polysaccharide network cross-linking [ 40 ]. The observed low calcium content may trigger compensatory mechanisms for lignin synthesis, while high Mn and Al levels directly promote lignin accumulation. Ultimately, this leads to lignin-dominated cell wall reinforcement, achieving high seed hardness. These findings challenge the traditional “calcium-centric” model and reveal the complexity of a multi-element-lignin network, offering new directions for improving seed permeability or storage tolerance. Transcriptome analysis revealed the key pathway involved in the regulation of vicia sativa seed hardness This study revealed the dynamic molecular network underlying seed hardness formation in common vetch through transcriptome sequencing, with the core mechanisms involving cell wall reinforcement, secondary metabolite accumulation, and coordinated hormonal signaling. During the filling stage, DEGs were significantly enriched in biological pathways such as cell wall organization or biogenesis (245 DEGs) and secondary metabolite biosynthesis (105 DEGs) (Table S8; Table S9). The high expression of cell wall-related regulatory genes may be a key driver for the deposition of cellulose, hemicellulose, and pectin in the seed coat [ 41 ]. Concurrently, the activation of secondary metabolic pathways (e.g., lignin precursor synthesis) suggests early regulation of phenylalanine ammonia-lyase (PAL) and peroxidase [ 42 ]. This stage likely establishes the initial mechanical strength of the seed coat by coordinating the supply of cell wall polysaccharides and secondary metabolites. During the milk-ripe stage, DEGs were significantly enriched in biological pathways such as brassinosteroid biosynthesis (8 DEGs) and DNA replication (30 DEGs) (Table S10; Table S11). Brassinosteroids (BRs) have been shown to promote the expression of cell wall-loosening enzymes (e.g., expansins) by activating the BZR1 transcription factor; however, the enrichment of BR pathways at this stage may indirectly maintain cell wall stability by antagonizing ethylene signaling [ 43 ]. Additionally, the high expression of genes associated with RNA modification (192 DEGs) and hydrolysis of nucleic acid phosphodiester bonds (275 DEGs) may fine-tune the spatiotemporal expression of cell wall synthesis enzymes by regulating mRNA stability or translation efficiency [ 44 ] (Table S10; Table S11). During the wax-ripening stage, DEGs were significantly enriched in biological pathways such as flavonoid biosynthesis (23 DEGs) and eukaryotic ribosome biogenesis (42 DEGs) (Table S12; Table S13). Flavonoids may enhance seed coat impermeability through covalent binding with cell wall polysaccharides [ 45 ], while the upregulation of ribosome biogenesis-related genes likely supports the synthesis of abundant cell wall structural proteins (e.g., glycine-rich proteins, GRPs) [ 46 ]. Concurrently, the sustained activity of nucleic acid phosphodiester bond hydrolases may optimize metabolic resource allocation toward cell wall reinforcement pathways by degrading redundant RNA molecules. Studies have also shown that seed hardness formation exhibits stage-specific characteristics: the filling stage initiates cell wall skeleton construction, the milk-ripe stage maintains structural plasticity through hormonal balance, and the wax-ripening stage achieves final hardening via secondary metabolite deposition (e.g., flavonoids, lignin) and cross-linking of structural proteins. This process shares high similarity with seed coat development mechanisms reported in pomegranate and rice [ 47 , 48 ]. However, the unique flavonoid biosynthesis pathway in common vetch suggests species-specific stress resistance strategies in its seed coat, reflecting both conserved regulatory networks and species-specific adaptations. These findings provide critical molecular targets for breeding vicia sativa with improved seed hardness traits. Functional analysis of key candidate genes Studies have shown that short-chain alcohol dehydrogenases (SDRs) are a class of oxidoreductases widely present in living organisms, catalyzing dehydrogenation or reduction reactions of substrates such as alcohols, steroids, and lipids. Members of this family typically rely on NAD(P)+/NAD(P)H as cofactors and are involved in the synthesis of secondary metabolites and detoxification processes [ 49 , 50 ]. In plants, SDRs may participate in the biosynthesis of defensive compounds such as flavonoids and terpenoids, or respond to oxidative stress (e.g., by reducing toxic aldehydes). For instance, the SDR gene At5g16970 in Arabidopsis has been demonstrated to regulate the metabolism of abscisic acid (ABA) [ 51 ]. In this study, we identified three key candidate genes regulating seed hardness (Table 2 ). Among them, ethylene response 2 ( TRINITY_DN3402_c0_g1 ) belongs to the ethylene receptor family and serves as the initial component of the ethylene signal transduction pathway. In Arabidopsis , ETR1 and ETR2 regulated downstream signaling cascades by binding ethylene molecules, influencing seed germination, fruit ripening, and stress responses [ 52 ]. ETR2 likely transmits signals via its histidine kinase activity, exerting negative regulation on ethylene responses (e.g., suppressing the activity of EIN3/EIL1 transcription factors). Studies have shown that ETR2 mutants exhibit increased ethylene sensitivity, highlighting its role in suppressing ethylene signaling [ 53 ]. The function of this gene may relate to seed hardness, as ethylene is involved in regulating the expression of cell wall modification genes. Additionally, ethylene response factor 8 ( TRINITY_DN13607_c0_g1 ) belongs to the AP2/ERF transcription factor family and directly binds to the ethylene-responsive element (GCC-box) to activate or repress downstream target gene expression. ERF8 plays a critical role in plant stress resistance (e.g., drought, salinity) and pathogen defense [ 54 ]. For example, SlERF.B3 in tomato ( Solanum lycopersicum ) regulated the expression of cell wall-modifying enzymes, thereby influencing fruit softening [ 55 ]. In vicia sativa , we hypothesize that TRINITY_DN13607_c0_g1 may enhance seed coat mechanical strength and thus improve seed hardness by regulating the expression of lignin biosynthesis genes (e.g., PAL, 4CL) [ 56 ]. Studies have also shown that cell wall structural proteins (GRPs), characterized by glycine-rich repeat motifs, enhance cell wall mechanical stability by interacting with cell wall polysaccharides (e.g., cellulose, hemicellulose) via hydrogen bonds. In Arabidopsis , AtGRP3 has been demonstrated to participate in secondary cell wall thickening, and its expression is induced under stress conditions (e.g., low temperature, pathogen infection) [ 57 ]. Similarly, OsGRP1 in rice ( Oryza sativa ) improves stem lodging resistance by regulating cell wall cross-linking These proteins may further reinforce seed coat structure and influence seed hardness by binding to lignin-polysaccharide complexes [ 58 , 59 ]. Based on these findings, this study hypothesizes that the key candidate genes TRINITY_DN3402_c0_g1 , TRINITY_DN13607_c0_g1 , and TRINITY_DN6606_c0_g1 may form a synergistic regulatory network in seed hardness control. Specifically, ETR2 ( TRINITY_DN3402_c0_g1 ) likely suppresses ethylene signaling, indirectly reducing the expression of cell wall-loosening enzyme genes (e.g., polygalacturonases), thereby maintaining cell wall integrity [ 60 , 61 ]. ERF8 ( TRINITY_DN13607_c0_g1 ), as an AP2/ERF transcription factor, directly activates the high expression of lignin biosynthesis genes (e.g., PAL, peroxidases), promoting lignin deposition [ 55 , 56 ]. GRP ( TRINITY_DN6606_c0_g1 ) enhances cell wall rigidity through physical cross-linking, synergizing with lignin to construct a mechanical stress-resistant barrier [ 57 , 58 ]. These results provide potential molecular targets for deciphering the mechanisms underlying seed hardness in vicia sativa , highlighting the interplay between ethylene signaling, lignin biosynthesis, and structural protein-mediated reinforcement in shaping seed coat mechanical properties. Materials and methods Plant materials and sampling The wild material W12, characterized by complete seed hardness, was provided by the Germplasm Bank of wild species in Southwest China (Accession Number: 868710143316; Chongqing, China) and plant materials were identified by germplasm bank administrator shaofa Qin. The cultivated variety W30 (internally assigned identifier) is a local landrace collected from Pingwu County, Mianyang City, Sichuan Province (Collection ID: 2019511248) and plant materials were identified by Dr. yongqun Zhu (Institute of agricultural resources and environment, sichuan academy of agricultural sciences). The experimental materials can be obtained in accordance with relevant agreements. From 2019 to 2022, both accessions were cultivated for three consecutive years under natural conditions in Yingxi Town, Nanchong City, China (106°12′N, 31°12′E). The planting layout included 3-meter-wide rows with plant and row spacing 50 cm, arranged in a completely randomized block design with three replicates. At the bud formation stage, individual plants were labeled. Seeds were collected at three developmental stages: filling stage (W12-1; W30-2), milk-ripe stage (W12-2; W30-3), and wax ripening stage (W12-3; W30-4). Pods were manually opened to extract seeds, and seed coats were immediately separated. The samples were flash-frozen in liquid nitrogen and stored at -80°C for subsequent analysis. Chromosome number was determined through cytological analysis Seeds were germinated in Petri dishes lined with double-layered filter paper and placed in a constant-temperature incubator at 25°C. For hard-seeded seeds, the seed coat was scarified to facilitate water uptake. When root tips reached approximately 1–2 cm in length, they were excised and fixed in α-bromonaphthalene for 3.5 hours. Following fixation, root tips were rinsed in distilled water for 30 minutes, then transferred to a 3:1 (v/v) ethanol-acetic acid solution for overnight fixation and stored at 4°C until further processing. For enzymatic maceration, root tips were washed in distilled water for 10 minutes, and 2–3 mm of the white meristematic tissue was excised from the tip. The excised tissue was incubated in a mixed enzyme solution containing cellulase and pectinase at 37°C for 1.5 h (20 µL per root tip). Microscope slides were prepared according to the method described by Kato et al. [ 62 ], involving squashing and staining for chromosomal observation. Seed hardness and chemical composition contents The seeds to be tested were dried in an oven at 60 ℃ then were treated separately for one second by a small grinder (ZT-150, Yongkang Zhanfan Industry and Trade Co., Ltd.​) to get the seed coat debris. 1 g seed coat from each of 10 random individuals for each seed coat type were harvested and then mixed. The seed coat debriswas collected and fully crushed into fine powder by grinder, passed through an 40-mesh sieve Save as backup [ 63 ]. Used for elemental analysis by Inductively Coupled Plasma Mass Spectrometer (ICP-MS) (Thermo Fisher Scientific, America). Similarly, The seeds to be tested were dried in an oven at 60 ℃ then were treated separately for one second by a small grinder (ZT-150, Yongkang Zhanfan Industry and Trade Co., Ltd.) to get the seed fragments for future use. The contents of crude protein (CP), soluble protein (SP), soluble sugar (SS), starch (St), and lignin (Lg) in mature seeds were determined using kits from Quanzhou Ruixin Biotechnology Co., Ltd. In this study, we used the national standard method GB/T21304-2007 (Determination of wheat hardness-Hardness index method) to measure seed hardness indicators. The seed hardness index was determined in accordance with GBT 21304. The analysis of structural changes in seed coat cells during various developmental stages and the process of seed development To study the dynamic changes of seed coat and seeds during seed development, wild hardness material W12 and non-hardness material W30 with high hard seed rate were selected as the research objects. 18 samples were collected at various stages of seed development (filling stage, milk-ripe stage, and wax ripening stage). The seeds were separately fixed with FAA fixative and made into paraffin sections. After rinsing the fixative solution, the solvent was dehydrated using a gradient of ethanol concentrations and subsequently stained with Safranin O. The material achieved transparency by employing a mixture of absolute ethanol and xylene in a 1:1 volume fraction ratio, followed by immersion in a xylene solution. Upon achieving transparency, the material was then immersed in wax and embedded. Then, the embedded material was sectioned into 8 µm thick slices using a Leica RM2245 semi-automatic rotary microtome. Following drying, the sections were deparaffinized and made transparent with a xylene solution and a mixture of absolute ethanol and xylene. Subsequently, they were rehydrated using an ethanol solution with varying concentration gradients against the original gradient, before being restained with toluidine blue. Finally, after being transparent again, it was sealed with a sealing tablet. Observations were made using a light microscope OLYMPUS BX53, photographed, and recorded. The seeds were randomly selected from each batch, with 18 samples in total. Furthermore, these samples were photographed and analyzed using Digimizer version 5.4.4, a software specifically designed for image analysis. Scanning electron microscopy (SEM) analysis of seed coat At seeds full-ripe stage, a total of 10 seeds with hard material (W12) and 10 seeds with non-hard material (W30) were randomly selected for the purpose of observing the surface structure (such as cracks and waxy appearance) as well as the longitudinal structural morphology of their seed coat. The initial step involved the utilization of an IB-5 ion plating device for the application of a 2.3 nm Au film coating [ 64 ]. Subsequently, the prepared samples were captured using an S-570 scanning electron microscope (Hitachi LTD, Japan). Treating seeds with exogenous ethephon Randomly selected wild-type seeds in sufficient quantity were scarified by soaking in a 98% concentrated sulfuric acid solution for 30 minutes. After treatment, the seeds were rinsed with purified water for 3 minutes, and the surface moisture was blotted dry using absorbent paper. The seeds were then air-dried for subsequent use. After sterilization, wild hard-seeded and cultivated seeds were separately soaked in exogenous ETH solutions (purity ≥ 85%) at concentrations of 0 (CK), 100, 150, and 200 mg/L for 24 h at room temperature. Following treatment, seeds were retrieved, rinsed thoroughly with distilled water, and air-dried for subsequent use. For germination assays, 100 seeds per treatment group were randomly selected and placed in sterile Petri dishes lined with two layers of moist filter paper. The dishes were incubated in an artificial climate chamber (AS-R600L2N, Xunneng Instruments Beijing Co., Ltd.) under controlled conditions: temperature: 25 ± 1°C, light cycle: 16-h light/8-h dark photoperiod, replicates: three independent replicates per treatment. Germination was monitored daily for 7 d. The number of germinated seeds was recorded each day, and germination potential and germination rate were calculated on the 7 d. After the germination period, shoot height and root length of all seedlings were measured using a vernier caliper. Transcriptome sequencing and identification of differentially expressed genes (DEGs) The wild hardness material W12 and non-hardness seed W30 were chosen as the subjects of this study, and 18 samples were collected at various stages of seed development (filling stage, milk-ripe stage, and wax ripening stage). Extracted total RNA from each sample, tested the purity, concentration, and integrity, and send 0.5-2 µg RNA for library construction and RNA-seq (Shanghai Personal Biotechnology Co., Ltd., China). After library construction, PCR amplification was employed for library fragment enrichment. Subsequently, library selection was conducted based on the desired fragment size of 450 bp. The library's quality was subsequently assessed using the Agilent 2100 Bioanalyzer, which allowed for the determination of both the total concentration and effective concentration of the library. Then, the libraries containing different Index sequences were proportionally mixed based on the effective concentration and data requirement. The pooled libraries were uniformly diluted to a concentration of 2 nM, and single-stranded libraries were formed through alkali denaturation. Following RNA extraction, purification, and library construction, Next-Generation Sequencing (NGS) technology was utilized for paired-end (PE) sequencing of these libraries using the Illumina sequencing platform [ 65 ]. Raw data underwent filtration, removing reads with connectors, length less than 50bp, and average sequence quality below Q20. Subsequently, the resulting high-quality sequences were de novo spliced to obtain transcript sequences. The splicing transcript was sequenced based on its length from long to short, and the length of transcript was added to the length of splicing transcript, so that it was not < 50%/90% of the total length, namely N50/N90, to measure the continuity of de novo assembly. Subsequently, the transcripts were clustered, and the longest transcript was selected as the Unigene, and unigene was utilized for subsequent annotation processes including GO, KEGG, eggNOG, SwissProt, Pfam annotation, ORF prediction, SSR prediction [ 66 ]. The filtered sequences were simultaneously aligned to Unigene to obtain the Reads Count of each unigene, and the samples underwent further analysis for differential expression, enrichment, cSNP and InDel [ 67 ]. Subsequently, the read count was converted to an FPKM, and the level of gene expression level was evaluated using the FPKM value [ 68 , 69 ]. ‘DEseq2’ R package was used for the identification of DEGs which were filtered with|log 2 (fold change) | ≥ 1 and false discovery rate (FDR) < 0.05. GO and KEGG enrichment analysis of DEGs The heatmaps of gene expression, gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) pathway analyses were performed using the OmicShare tools, a free online platform for data analysis ( http://www.omicshare.com/tools ), and padj < 0.01 was used as the threshold of significant enrichment in both analyses described above. The correlation and principal Component analysis (PCA) were performed with R package gmodels ( http://www.r-project.org/ ) in this study. Weighted gene co-expression network analysis (WGCNA) and identification of the key candidate genes After screening out undetectable or low expression genes, genes with FPKM of more than 0.5 were utilized for WGCNA using the ‘WGCNA’ R package [ 70 ]. The power value was set to 11 to make the gene co-expression network conform to a scale-free network. The ‘merge Cut Height’ was set to 0.25, and the ‘min Module Size’ was set to 25 to further classify and merge the gene modules [ 32 ]. The module eigengene E, was calculated to identify the modules related to seed hardness. In each module, the for which with Person Correlation Value > 0.70 and P _value<0.01 were regarded as hub genes. The hub genes heat map was performed using the OmicShare bioinformatics learning platform ( www.omicshare.com/tools )[ 71 ]. Subsequently, the gene co-expression network map was constructed by cytoscape software [ 72 ]. Analysis of expression patterns of key candidate genes To further identify key candidate genes, qRT-PCR was used to determine the relative expression levels of four candidate genes in each variety and stage. Primer Premier 5 software was used to design the gene-specific primers, which were listed in Table S16. The Vicia sativa gene VSACT11 (GenBank: GU946218) is the internal control gene and fold change was calculated using the 2 −ΔΔCT method [ 73 ]. Three independent biological replicates were performed on each sample to ensure statistical reliability. Statistical analysis Microsoft Excel (Microsoft Corp., Redmond, WA, USA) and SPSS (IBM, Inc., Armonk, NY, USA) were used to calculate descriptive statistics, and to test statistical significance, respectively. The bar chart was drawn by graphpad prism 9.5.1, and the Box plot was drawn by the OmicShare tools, an online platform for data analysis ( https://www.omicshare.com/tools ). Conclusions The results of this study indicate that a thicker cuticle and narrower hilum fissure width constitute the first barrier affecting rapid water absorption. Excessive lignin content leads to hardened seed coats, which further impacts seed water permeability and gas permeability, thereby inducing seed hardness (hard-coated seeds). Furthermore, through WGCNA analysis, three key candidate genes involved in regulating seed hardness were identified ( TRINITY_DN3402_c0_g1 , TRINITY_DN13607_c0_g1 , and TRINITY_DN6606_c0_g1 ). The candidate gene expression pattern analysis results indicated that, the expression levels of TRINITY_DN3402_c0_g1 , TRINITY_DN13607_c0_g1 , and TRINITY_DN6606_c0_g1 in the wild hard-seeded type vicia sativa (W12) were significantly higher than those in the cultivated type vicia sativa (W30). Declarations Acknowledgements We gratefully acknowledge Dr. Ruyu He for his expert guidance in the transcriptome data analysis, which was critical to this study. Thanks for germplasms provided by Germplasm Bank of wild species in Southwest. Author contributions HW performed the experiments, analyzed the data and prepared the figures and tables. ZW approved the final draft. YZ, XY and BZ processed the data. HW and ZD assisted in completing part of the experiment. YC and ZY conceived and designed the experiments, authored or reviewed drafts of the article, and approved the final draft. ZD revised the manuscript and provided financial support for this research. All authors have read and agreed to the published version of the manuscript. Funding This work was supported by the Open Competition Mechanism to Select the Best Candidates from the Sichuan Academy of Agricultural Sciences (no. 1+9KJGG004), and the Sichuan Province Research Grant (nos. 2022ZZCX086, 2022ZZCX084). Data availability Data is provided within the manuscript or supplementary information files. The raw sequence data of RNA-seq is available in the National Genomics Data Center (NGDC) with the accession number CRA044150. Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Author details 1 Institute of Special Economic Animals and Plants, Sichuan Academy of Agricultural Sciences, Nanchong, 637000, China. 2 Forage Crops Germplasm Innovation and Production Management Key Laboratory of Nanchong City, Sericulture Research Institute, Sichuan Academy of Agricultural Sciences, Nanchong, 637000, China. 3 Maize Research Institute, Sichuan Agricultural University, Chengdu 611130, China. References Zhou Q,Cui Y,Dong S,Luo D,Fang L, Shi Z, Liu W,Wang Z, Nan Z. Integrative physiological, transcriptome, and metabolome analyses reveal the associated genes and metabolites involved in cold stress response in common vetch (Vicia sativa L.). Food and Energy Security.2023;12(4):22.https://doi.org/10.1002/fes3.484. Sun Y,Zhao N,Sun H,Xu S,Lu Y,Xi H,Guo Z,Shi H.Transcriptome profiling reveals molecular responses to salt stress in Common Vetch ( Vicia sativa L.). Plants (2223-7747).2024;13(5). https://doi.org/10.3390/p222. https://doi.org/10.1002/fes Grela E,Samolińska W,Rybiński W,Kiczorowska B,Kowalczuk-Vasilev E,Matras J, Wesoowska S.Nutritional and anti-nutritional factors in vicia sativa l.seeds and the variability of phenotypic and morphological characteristics of some vetch accessions cultivated in european countries. Animals. 2021;11(1). https://doi.org/10.3390/ani11010044. Li R,Zhang Z,Tang W,Huang Y,Nan Z.Common vetch cultivars improve yield of oat row intercropping on the Qinghai-Tibetan plateau by optimizing photosynthetic performance. European Journal of Agronomy.2020;117:126088.https://doi.org/10.1016/j.eja.2020.126088. Lai X,Yang X,Wang Z,Shen Y,Ma L.Productivity and water use in forage-winter wheat cropping systems across variable precipitation gradients on the Loess Plateau of China. Agricultural Water Management.2022;259.https://doi.org/10.1016/j.agwat.2021.107250. Xu C,Kang N,Xu H-L,Yu D-H,Bi Y-X.Research progress of wild soybean germplasmsand utilization.Agricultural Science & Technology.2017;18:812-817. https://doi.org/10.16175/j.cnki.1009-4229.2017.05.016. Avramidou E,Sarri E,Papadopoulou E,Petsoulas C,Tigka E,Tourvas N,et al.Phenotypic, genetic, and metabolite variability among genotypes of Vicia sativa L.Applied Sciences-basel.2024;14(20):9272.https://doi.org/10.3390/app14209272. Meyer CJ,Steudle E,Peterson CA.Patterns and kinetics of water uptake by soybean seeds. Journal of Experimental Botany.2007;58(3):717-732.https://doi.org/10.1093/jxb/erl244. Shu K,Liu X-D,Xie Q,He Z-H.Two faces of one seed: Hormonal regulation of dormancy and germination.Molecular Plant.2016;9(1):34-45.https://doi.org/10.1016/j.molp.2015.08.010. Rancisco H, Dübbern DS,Júlio M.The seed coat as a modulator of seed-environment relationships in Fabaceae.Brazilian Journal of Botany.2001;24(4):365-375. https://doi.org/10.1590/S0100-84042001000400002. Wen Z-Z,Lu XO-R,Wen J-Q,Wang Z-Y,Chai M-F.Physical seed dormancy in legumes: molecular advances and perspectives. Plants. 2024;13(11):1473. https://doi.org/10.3390/plants13111473. Taylor GB.Hardseededness in Mediterranean annual pasture legumes in Australia:a review. Australian Journal of Agricultural Research.2005;56(7):645-661. https://doi.org/10.1071/AR04284. Wang F-F,Sun X,Liu B-H,Kong F-J,Pan X-W, Zhang H-Y. A polygalacturonase gene PG031 regulates seed coat permeability with a pleiotropic effect on seed weight in soybean. Theoretical and Applied Genetics.2022;135(5):1603-1618.https://doi.org/10.1007/s00122-022-04057-4. Sun L-J,Miao Z-Y,Cai C-M,Zhang D-J,Zhao M-X,Wu Y-Y,Zhang X-L,Stephen AS,Zhou L-W, et al. GmHs1-1, encoding a calcineurin-like protein,controls hard-seededness in soybean. Nature Genetics.2015;47(8):939. https://doi.org/10.1038/ng.3339. Ma F, Cholewa E, Mohamed T, Peterson CA, Gijzen M.Cracks in the palisade cuticle of soybean seed coats correlate with their permeability to water.Annals of Botany.2004;(2):213-228. https://doi.org/10.1093/aob/mch133. Shao S-Q,Chris J.Meyer,Ma F-S,Carol A,Peterson,Mark AB et al.The outermost cuticle of soybean seeds: chemical composition and function during imbibition.Journal of Experimental Botany.2007;58(5):1071-1082.https://doi.org/10.1093/jxb/erl268. Wang F,Sun X,Liu B,Kong F,Pan X,Zhang H. A polygalacturonase gene PG031 regulates seed coat permeability with a pleiotropic effect on seed weight in soybean. Theoretical and Applied Genetics.2022;135(5):1603-1618. https://doi.org/10.1007/s00122-022-04057-4. Kebede H,Smith JR,Ray JD.Identification of a single gene for seed coat impermeability in soybean PI 594619.Theoretical and Applied Genetics.2014;127(9):1991. https://doi.org/10.1007/s00122-014-2355-2. Ma F-S,Cholewa E,Mohamed T,Peterson CA,Gijzen M.Cracks in the palisade cuticle of soybean seed coats correlate with their permeability to water.Annals of Botany.2004(2):213-228. https://doi.org/10.1093/aob/mch133. Yu B,He X-M,Tang Y,Chen Z-H,Zhou L-M,Li X-M,Zhang C-Y,Huang X,Yang Y-H,Zhang W-B.Photoperiod controls plant seed size in a CONSTANS-dependent manner.Nature Plants.2023;9:343-354.https://doi.org/10.1038/s41477-023-01350-y. Ellis R H, Hong TD, Roberts EH.Moisture content and the longevity of seeds of phaseolus vulgaris.Annals of Botany.1990;(3):341-348.https://doi.org/10.2307/2444634. Gladstones J,Atkins C,Hamblin J.Lupins as crop plants: biology, production and utilization. Field Crop Abstracts.1998;123-148. https://doi.org/10.1079/9780851992242.0000 Niu J, Cao D, Li H-X, Xue H, Chen L-N, Liu B-B, Cao S-Y.Quantitative proteomics of pomegranate varieties with contrasting seed hardness during seed development stages.Tree Genetics & Genomes.2018;14: 14.https://doi.org/10.1007/s11295-018-1229-1. Bentsink L,Jowett J,Hanhart CJ,Koornneef M. Cloning of DOG1 , a quantitative trait locus controlling seed dormancy in Arabidopsis . Proceedings of the National Academy of Sciences of the United States of America.2006;103(45). https://doi.org/www.pnas.org/cgi/doi/10.1073/pnas.0607877103. 2006;103:17042-17047. Barrero JM,Millar AA,Griffiths J,Czechowski T,Scheible WR.Gene expression profiling identifies two regulatory genes controlling dormancy and ABA sensitivity in Arabidopsis seeds. Plant Journal.2010;61(4):611-622. https://doi.org/10.1111/j.1365-313X.2009.04088.x.. Xia F,Chen L,Sun Y,Mao P.Relationships between ultrastructure of embryo cells and biochemical variations during ageing of oat ( Avena sativa L.) seeds with different moisture content.Acta Physiologiae Plantarum.2015;37(4):1-11.https://doi.org/10.1007/s11738-015-1825-8. Jang S-J,Sato M,Sato K,Jitsuyama Y,Fujino K,Mori H, et al. A Single-Nucleotide Polymorphism in an Endo-1,4-β-Glucanase Gene Controls Seed Coat Permeability in Soybean.Plos One.2015;10(6):1-19.https://doi.org/10.1371/journal.pone.0128527. Chai M, Zhou C, Molina I, Fu C, Nakashima J, Li G, et al. A class II KNOX gene, KNOX4 , controls seed physical dormancy. Proc Natl Acad Sci U S A.2016;113(25):6997-7002 http://dx.doi.org/10.1073/pnas.1601256113. Soltani A,Walter KA,Wiersma AT,Santiago JP,Lowry DB. The genetics and physiology of seed dormancy, a crucial trait in common bean domestication. BMC Plant Biology.2021;21(1):58. https://doi.org/10.1186/s12870-021-02837-6. Zhang B, Horvath S. A general framework for weighted gene co-expression network analysis. .Statistical Applications in Genetics & Molecular Biology.2005;4(1):1-10. https://doi.org/10.2202/1544-6115.1128. Ficklin S, Luo F, Feltus F. The association of multiple interacting genes with specific phenotypes in rice using gene coexpression networks.Plant Physiology.2010;154(1):13-24.https://doi.org/10.1104/pp.110.159459. Wang C-C, Lin J-Y, Bu Y-P, Sun R-D, Lu Y, Gai J-Y, Xing H, Guo Na, Zhao J-M. Genome-wide transcriptome analysis reveals key regulatory networks and genes involved in the determination of seed hardness in vegetable soybean. Horticulture Research.2024;11(5):1-10.https://doi.org/10.1093/hr/uhae084. Gerhard K. Cuticular water permeability and its physiological significance.Journal of Experimental Botany.1996;12:1813-1832.https://doi.org/10.1093/jxb/47.12.1813. Baucher M, Monties B, Montagu M, Boerjan W. Biosynthesis and Genetic Engineering of Lignin. Critical Reviews in Plant Sciences.1998;17(2):125-197.https://doi.org/10.1080/07352689891304203. Rolston MP.Water impermeable seed dormancy.Botanical Review.1978;44(3):365-396. https://doi.org/10.1007/bf02957854. Smýkal P, Vernoud V, Blair MW, Aleš S,Thompson RD. The role of the testa during development and in establishment of dormancy of the legume seed.Frontiers in Plant Science.2014;5(351):1-44.https://doi.org/10.3389/fpls.2014.00351. Haughn G, Chaudhury A. Genetic analysis of seed coat development in Arabidopsis. Trends in Plant Science.2005;10(10):472-477. https://doi.org/10.1016/j.tplants.2005.08.005. Sasaki S, Nishida T, Tsutsumi Y, Kondo R. Lignin dehydrogenative polymerization mechanism: a poplar cell wall peroxidase directly oxidizes polymer lignin and produces in vitro dehydrogenative polymer rich in β-O-4 linkage.FEBS Lett.2004;562(1-3). https://doi.org/10.1016/S0014-5793(04)00224-8. Tahara K, Norisada M, Hogetsu T, Kojima K. Aluminum tolerance and aluminum-induced deposition of callose and lignin in the root tips of Melaleuca and Eucalyptus species. Journal of Forest Research.2005;10(4):325-333.https://doi.org/10.1007/s10310-005-0153-z. Horst W, Wang Y, Eticha D. The role of the root apoplast in aluminium-induced inhibition of root elongation and in aluminium resistance of plants:a review.Annals of Botany.2010; 106(1):185. https://doi.org/10.1093/aob/mcq053. Buckeridge SM.Seed cell wall storage polysaccharides:models to understand cell wall biosynthesis and degradation.Plant Physiology.2010;154(3):1017-1023. https://doi.org/10.1104/pp.110.158642. Vanholme R, Demedts B, Morreel K, Ralph J, Boerjan W. Lignin biosynthesis and structure. Plant Physiology.2010;153(3): 895-905. https://doi.org/10.1104/pp.110.155119. Ye Q, Zhu W, Li L, Zhang S,Yin Y,Ma H,Chory WJ. Brassinosteroids control male fertility by regulating the expression of key genes involved inArabidopsisanther and pollen development. Proceedings of the National Academy of Sciences of the United States of America.2010;107(13):6100-6105. https://doi.org/10.2307/25665121. Chantarachot T, Bailey SJ. Polysomes, stress granules, and processing bodies: a dynamic triumvirate controlling cytoplasmic mRNA fate and function. Plant Physiology.2018;176(1): 254-269. https://doi.org/10.1104/pp.17.01468. Pourcel L, Routaboul JM, Cheynier V, Lepiniec L,Debeaujon I. Flavonoid oxidation in plants: from biochemical properties to physiological functions. Trends in Plant Science.2007; 12(1):29-36. https://doi.org/10.1016/j.tplants.2006.11.006. Mangeon A, Junqueira R, Sachetto MG. Functional diversity of the plant glycine-rich proteins superfamily. Plant Signaling & Behavior.2010;5(2):99-104 https://doi.org/10.4161/psb.5.2.10336. Chen L,Wang H,Xu T-T,Liu R-T, Zhu J-L,Li H-X,Zhang HW, Tang L-Y,et al. A telomere-to-telomere gap-free assembly integrating multi-omics uncovers the genetic mechanism of fruit quality and important agronomic trait associations in pomegranate.Plant Biotechnology Journal. Advance online publication. https://doi.org/10.1111/pbi.70107. Gao X, Yin C, Liu X, Peng J, Chen D, He D, et al. A glycine-rich protein MoGrp1 functions as a novel splicing factor to regulate fungal virulence and growth in Magnaporthe oryzae. Phytopathology Research.2019;1(1):2-15.https://doi.org/10.1186/s42483-018-0007-1. Kallberg Y, Oppermann U, Jörnvall H, Persson B. Short-chain dehydrogenase/reductase (SDRs). European Journal of Biochemistry.2002;269(18):4409-4417.https://doi.org/10.1046/j.1432-1033.2002.03130.x. Priego‐Cubero S, Liu Y, Toyomasu T, Gigl M, Hasegawa Y, Nojiri H, et al. Evolution and diversification of the momilactone biosynthetic gene cluster in the genus Oryza.New Phytologist.2025;245(6):2681-2697.https://doi.org/10.1111/nph.20416. Cheng W-H. A unique short-chain dehydrogenase/reductase in arabidopsis glucose signaling and abscisic acid biosynthesis and functions.The Plant Cell.2002;14(11):2723-2743. https://doi.org/doi:10.1105/tpc.006494. Cancel DJ. Loss-of-function mutations in the ethylene receptoretr1 cause enhanced sensitivity and exaggerated response to ethylene in arabidopsis.Plant Physiology.2002;129(4):1557. https://doi.org/10.1104/pp.003780. Hua J, Meyerowitz EM.Ethylene responses are negatively regulated by a receptor gene family in arabidopsis thaliana.Cell.1998;94(2):261-271.https://doi.org/10.1016/S0092-8674(00)81425-7. Müller M, Munné-Bosch S. Ethylene response factors: a key regulatory hub in hormone and stress signaling.Plant physiology.2015;169(1):32-41.https://doi.org/10.1104/pp.15.00677. Liu M, Yao C, Chen Y, Shin J, Mila I, Audran C, Zouine M, Pirrello J, Bouzayen M. The tomato ethylene response factor Sl‐ERF.B3 integrates ethylene and auxin signaling via direct regulation of Sl‐Aux/IAA27. The New phytologist.2018;219(2):631-640. https://doi.org/10.1111/nph.15165. Licausi F, Ohme TM, Perata P. APETALA2/Ethylene responsive factor (AP2/ERF) transcription factors: mediators of stress responses and developmental programs. The New phytologist .2013; 199 (3): 639-649.https://doi.org/10.1111/nph.12291. Ganar K,Turbina P,Nandy M,Chen C,Suylen D,Beelen S,Pascoe E,Koenraadt C,Dijkgraaf I, Deshpande S.Phase separation and ageing of glycine-rich protein from tick adhesive.Nature Chemistry.2025;17(2):186-197.https://doi.org/10.1038/s41557-024-01686-8. Wang Y, Wang M-H, Yan X, Chen K-X, Tian F-H, Yang X, Cao L-Y, et al. The DEP1 mutation improves stem lodging resistance and biomass saccharification by affecting cell wall biosynthesis in rice. Rice (New York, N.Y.). 2024;17(1):1-13. https://doi.org/10.1186/s12284-024-00712-0. Sowadan O, Xu S, Li Y, Muleke EM, Sitoe HM, Dang X, et al. Genome-wide association analysis unravels new quantitative trait loci (qtls) for eight lodging resistance constituent traits in Rice (Oryza sativa L.). Genes.2024;15(1):2-18.https://doi.org/10.3390/genes15010105. Wang J, Sun Q, Ma C, Wei M, Wang C, Zhao Y, et al. MdWRKY31-MdNAC7 regulatory network: orchestrating fruit softening by modulating cell wall-modifying enzyme MdXTH2 in response to ethylene signalling. Plant Biotechnology Journal.2024;22(12):3244-3261. https://doi.org/10.1111/pbi.14445. Sakai H, Hua J, Chen Q-G, Chang C, Medrano LJ, Bleecker AB, Meyerowitz EM. ETR2 is an ETR1-like gene involved in ethylene signaling in Arabidopsis. Proceedings of the National Academy of Sciences of the United States of America.1998;95(10):5812-5817 https://doi.org/10.1073/pnas.95.10.5812. Kato T. Metaphase chromosome preparation and classification of chromosomal aberrations. Methods in molecular biology (Clifton, N.J.).2023;2519:9-15.https://doi.org/10.1007/978-1-0716-2433-3_2. Sun Q, Zhu L-W,Zhang W-J,Wang J-H. Physical and chemical difference of seed coat between hard and soft seeds of Licorice (Glycyrrhiza Uralensis Fisch). Legume Research.2018; 41:441-446.https://doi.org/10.18805/LR-376. Chen X-Y, Xia Y-Y, MaoY-F, HuangY, Zhu J,Xu J,Zhu R,Shi L,WuW-G. A programmable nanofabrication method for complex 3d meta-atom array based on focused-ion-beam stress-induced deformation effect.Micromachines. 2020;11(1):95. https://doi.org/10.3390/mi11010095. Bolger A,Marc L,Bjoern U.Trimmomatic:a flexible trimmer for Illumina sequence data. Bioinformatics.2014;30(15):2114-2120.https://doi.org/10.1093/bioinformatics/btu170. Altschul SF,Gish W,Miller W, Myers EW,Lipman DJ.Basic local alignment search tool (BLAST). Journal of molecular biology.1990;215(3):403-410.https://doi.org/10.1016/S0022-2836(05)80360-2. Grabherr M, Haas B, Yassour M, Levin J, Thompson D, Amit I, et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nature biotechnology.2011;29(7): 644-652.https://doi.org/10.1038/nbt.1883. Xu J, Yang J, Xu Z, Zhao D, Hu X. Exogenous spermine-induced expression of SlSPMS gene improves salinity–alkalinity stress tolerance by regulating the antioxidant enzyme system and ion homeostasis in tomato. Plant physiology and biochemistry.2020;157:79-92. https://doi.org/10.1016/j.plaphy.2020.09.033. Trapnell C, Williams B, Pertea G, Mortazavi A, Kwan G, Baren M, et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nature biotechnology. 2010;28(5):511-515.https://doi.org/10.1038/nbt.1621. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics.2008;9:559.https://doi.org/10.1186/1471-2105-9-559. Huang Y, Qin M, Lai J, Liang J, Luo X, Li C. Assessing obt formation and enrichment: ROS signaling is involved in the radiation hormesis induced by low- dose tritium exposure in algae.Journal of Hazardous Materials.2023;443:130159.https://doi.org/10.2139/ssrn.4179079. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T.Cytoscape:A software environment for integrated models of biomolecular interaction networks. Genome Research.2003;13(11):2498-2504.https://doi.org/10.1101/gr.1239303. Livak K, Schmittgen T. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods. 2001;25(4):402-408. https://doi.org/10.1006/meth.2001.1262. Additional Declarations No competing interests reported. Supplementary Files SupplementaryFigure.docx SupplementaryTable.docx Cite Share Download PDF Status: Published Journal Publication published 23 Oct, 2025 Read the published version in BMC Genomics → Version 1 posted Editorial decision: Revision requested 21 Jul, 2025 Reviews received at journal 18 Jul, 2025 Reviews received at journal 15 Jul, 2025 Reviews received at journal 14 Jul, 2025 Reviewers agreed at journal 14 Jul, 2025 Reviewers agreed at journal 13 Jul, 2025 Reviewers agreed at journal 13 Jul, 2025 Reviewers invited by journal 01 Jul, 2025 Editor assigned by journal 01 Jul, 2025 Editor invited by journal 30 Jun, 2025 Submission checks completed at journal 29 Jun, 2025 First submitted to journal 29 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6974115","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":479623672,"identity":"148e4f44-d2ec-4511-9f9d-4db5b72376c9","order_by":0,"name":"Honglin Wang","email":"","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Honglin","middleName":"","lastName":"Wang","suffix":""},{"id":479623673,"identity":"974e128a-4cb1-44fc-b977-aba0f7b248a3","order_by":1,"name":"Zizhou Wu","email":"","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Zizhou","middleName":"","lastName":"Wu","suffix":""},{"id":479623674,"identity":"e3af5d37-68e8-495e-8ed2-2e7574c148d7","order_by":2,"name":"Yanchun Zuo","email":"","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Yanchun","middleName":"","lastName":"Zuo","suffix":""},{"id":479623676,"identity":"0062824f-ad79-4ace-ba16-c45f59c4b8e4","order_by":3,"name":"Xu Yan","email":"","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Xu","middleName":"","lastName":"Yan","suffix":""},{"id":479623679,"identity":"03eda683-d2ea-4038-b89b-7688e3c4d0e6","order_by":4,"name":"Bangxing Zou","email":"","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Bangxing","middleName":"","lastName":"Zou","suffix":""},{"id":479623681,"identity":"c6e696f7-84fa-4cfd-b4c4-63b8fd7836c2","order_by":5,"name":"Yu Chen","email":"","orcid":"","institution":"Sichuan Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Chen","suffix":""},{"id":479623683,"identity":"a7cd5f06-f9e0-417e-9cb3-782f8bca1d43","order_by":6,"name":"Zhengcai Yuan","email":"","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Zhengcai","middleName":"","lastName":"Yuan","suffix":""},{"id":479623686,"identity":"ac13f0fb-9bdb-4308-a81b-0c709c923bd2","order_by":7,"name":"Zhouhe Du","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwUlEQVRIiWNgGAWjYDCCAwxsYJKfdC2SDSRrMThArA6+G+nPHnxsuyNnfP7w4w8/auoY+GcTsE/yRkK64cy2Z8ZmN9LMJHuOHWaQuEPAPoMbCcekedsOJ267wcPGwMMGdKFEAiEtiW0gLfWb+88wf/zzr44YLclsIC0JBgw5DEAGM2EtkmeesUnOOHfYcAbQL9KyfYd5JG4Q0MJ3PP2ZxIeyw/L8/Ycff3zzrU6OfwYBLRiAh0T1o2AUjIJRMAqwAQDJgEbfSCcfkwAAAABJRU5ErkJggg==","orcid":"","institution":"Sichuan Academy of Agricultural Sciences","correspondingAuthor":true,"prefix":"","firstName":"Zhouhe","middleName":"","lastName":"Du","suffix":""}],"badges":[],"createdAt":"2025-06-25 11:23:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6974115/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6974115/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12864-025-12138-z","type":"published","date":"2025-10-23T16:16:46+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":85921634,"identity":"c34ce933-ce9a-405b-ac2d-b82d1c848127","added_by":"auto","created_at":"2025-07-03 07:59:44","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2508501,"visible":true,"origin":"","legend":"\u003cp\u003eAgronomic trait analysis and chromosome number identification of two plant materials. \u003cstrong\u003eA\u003c/strong\u003e. Cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30). \u003cstrong\u003eB\u003c/strong\u003e. Wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12). \u003cstrong\u003eC\u003c/strong\u003e. leaf morphology. \u003cstrong\u003eD\u003c/strong\u003e: Leaf morphology. \u003cstrong\u003eE\u003c/strong\u003e. seed dimensions. \u003cstrong\u003eF\u003c/strong\u003e. Chromosome number identification of wild hard-seeded type (W12). \u003cstrong\u003eG\u003c/strong\u003e. Chromosome number identification of cultivated type (W30).\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/7a9295a4bffc27cc16220a2d.png"},{"id":85921874,"identity":"0b82117f-1f41-4067-a083-3c952249cb11","added_by":"auto","created_at":"2025-07-03 08:07:44","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":259244,"visible":true,"origin":"","legend":"\u003cp\u003eDetermination of seed nutrient content and hardness index. \u003cstrong\u003eA\u003c/strong\u003e.\u003cstrong\u003e \u003c/strong\u003eCrude protein content between the cultivated type material and the wild hard-seeded type. \u003cstrong\u003eB-D\u003c/strong\u003e. Soluble sugar, soluble protein, and starch contents of the cultivated type material and wild hard-seeded type material. \u003cstrong\u003eE\u003c/strong\u003e. The difference in seed hardness index between hard-seeded type and cultivated type. \u003cstrong\u003eF\u003c/strong\u003e. The difference in lignin between hard-seeded type and cultivated type.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: **indicated significance at 0.01 level, and ns indicated not significance.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/fa78f021376a04c8c777bc28.png"},{"id":85921639,"identity":"3a2be007-69af-43ec-a433-45d1d25b9628","added_by":"auto","created_at":"2025-07-03 07:59:44","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":7141208,"visible":true,"origin":"","legend":"\u003cp\u003eDevelopment dynamics of seeds (pods) and seed coat. \u003cstrong\u003eA\u003c/strong\u003e. Hard-seeded type (W12) and cultivated type (W30) seeds and pod morphology at different developmental stages. \u003cstrong\u003eB\u003c/strong\u003e. Hard-seeded type (W12)and cultivated type (W30) seed coat structures at different developmental stages. \u003cstrong\u003eC\u003c/strong\u003e. Differences in seed and seed coat structure between hard-seeded type (W12)and cultivated type (W30) at different development stages.\u003cstrong\u003e D\u003c/strong\u003e. Thickness of cuticle layer in ripening stage of hard-seeded type (W12) and cultivated type (W30).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNote: \u003c/strong\u003eFS (filling stage), MRS (milk-ripe stage), WRS (wax ripening stage), CL (cuticula), PL (palisade layer), SC (sclerenchyma cell), OS (osteogenic stone cells), PC (Parenchyma cell), LI (light line). *Indicated significance at 0.05 level, and **indicated significance at 0.01 level.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/1f38a547305ddc687cedf847.png"},{"id":85921878,"identity":"0df20706-ebd2-40d9-a19e-ebe526fdbfd3","added_by":"auto","created_at":"2025-07-03 08:07:44","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":8380464,"visible":true,"origin":"","legend":"\u003cp\u003eThe scanning electron microscopy (SEM) analysis of seed coat and seed morphology. \u003cstrong\u003eA\u003c/strong\u003e. Differences in hilum structure between hard-seeded type (W12) and cultivated type (W30) seeds. \u003cstrong\u003eB\u003c/strong\u003e. Quantitative indicators of hard-seeded type (W12) and cultivated type (W30) seed morphology. \u003cstrong\u003eNote\u003c/strong\u003e: Hilum (HU), hilar fissure (HF), stellate tissue (ST), tracheid (TB). *Indicated significance at 0.05 level, **indicated significance at 0.01 level, and ***indicated significance at 0.001 level.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/e247e6ed4e5db244c281b4d5.png"},{"id":85921875,"identity":"5b70e117-2aaa-44a9-8657-1a0fb14bde7d","added_by":"auto","created_at":"2025-07-03 08:07:44","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1602634,"visible":true,"origin":"","legend":"\u003cp\u003eWGCNA of expressed genes in W12 and W30 and their expression levels at different stages of seed coat development. \u003cstrong\u003eA\u003c/strong\u003e. The gene cluster dendrogram constructed by gene correlation coefficients. \u003cstrong\u003eB\u003c/strong\u003e. Relationships between modules and traits. \u003cstrong\u003eC-F\u003c/strong\u003e. Gene heatmaps within different modules. \u003cstrong\u003eG. \u003c/strong\u003eGene expression analysis of key candidate genes at different developmental stages.\u003c/p\u003e\n\u003cp\u003eNote: **indicated significance at 0.01 level, ***indicated significance at 0.001 level, ****indicated significance at 0.0001 level, and ns indicated not significance.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/c1f0c7af3b0bac375d7f2c5d.png"},{"id":85922633,"identity":"e8580c00-989d-4118-97cf-eca2e232e1de","added_by":"auto","created_at":"2025-07-03 08:15:44","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":3432477,"visible":true,"origin":"","legend":"\u003cp\u003eEffects of ethylene solutions at different concentrations on seed germination of \u003cem\u003eVicia sativa\u003c/em\u003e. \u003cstrong\u003eA\u003c/strong\u003e. Phenotypic performance differences between hard-seeded type and cultivated types under different concentrations. \u003cstrong\u003eB\u003c/strong\u003e. Effects of ethylene at varying concentrations on germination rate, root length, and shoot length in hard-seeded type and cultivated seeds.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/a28ba7a9fa5cfec7c443bf1c.png"},{"id":94490506,"identity":"9eba2844-269c-4abb-a39f-aa7ac85b8189","added_by":"auto","created_at":"2025-10-27 17:11:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":27309006,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/703dc061-b882-4842-9826-fa91c7111493.pdf"},{"id":85921654,"identity":"b85d07ff-f71b-4e1b-831a-a0b71b66de04","added_by":"auto","created_at":"2025-07-03 07:59:45","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":10372334,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure.docx","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/776eace36f63de31b1177037.docx"},{"id":85921635,"identity":"e5e89018-c743-4666-b9b0-5e0f6d1f7975","added_by":"auto","created_at":"2025-07-03 07:59:44","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":55503,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable.docx","url":"https://assets-eu.researchsquare.com/files/rs-6974115/v1/47e708b5b4ab0dcc11928a31.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Phenomics, RNA sequencing and weighted gene co-expression network analysis reveals key regulatory networks and genes involved in the determination of seed hardness in Vicia sativa","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe \u003cem\u003eVicia sativa\u003c/em\u003e L., an annual leguminous crop, is a high-quality dual-purpose species with integrated functions of forage production, green manure, and ecological restoration. Renowned for its broad ecological adaptability (cold-tolerant, drought-resistant, and adaptable to poor soils), efficient biological nitrogen fixation capacity, and outstanding nutritional properties, it has become a critical grass-arable crop rotation species in temperate and alpine regions worldwide [\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In China's agro-pastoral ecotone, the mixed cropping of \u003cem\u003eVicia sativa\u003c/em\u003e with Avena sativa for sun-cured hay production leverages the legume-grass synergy to significantly enhance forage quality and optimize resource utilization efficiency. This system serves as a pivotal technological framework to address seasonal shortages of spring forage and ensure balanced nutritional supply for livestock, thereby supporting sustainable animal husbandry [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Currently, \u003cem\u003eVicia sativa\u003c/em\u003e faces constraints germplasm innovation due to outdated traditional breeding techniques, a narrow genetic base, and varietal degeneration arising from prolonged cultivation. Addressing these challenges hinges on the evaluation and innovation of germplasm resources. Wild \u003cem\u003eVicia sativa\u003c/em\u003e L., with its unique biological traits and environmental adaptability, has been identified as a high-quality gene pool for improving cultivated varieties, offering critical genetic diversity to enhance stress resistance, yield stability, and agronomic performance [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Breeding practices have revealed that many desirable genes in wild \u003cem\u003eVicia sativa\u003c/em\u003e cannot be effectively utilized, primarily due to the varying degrees of hard-seededness (dormancy) in its seeds. While this trait enhances population persistence and stress resistance in natural ecosystems, it introduces significant genetic constraints in agricultural production, manifesting as uneven seedling emergence, delayed growth cycles, and reduced yields [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Long-standing research on seed hard-seededness, both domestically and internationally, has predominantly focused on elucidating its underlying mechanisms and developing dormancy-breaking methods, with a heavy emphasis on staple crops such as soybeans. However, limited research has been conducted on the hard-seededness trait in \u003cem\u003eVicia sativa\u003c/em\u003e L., particularly from physiological and molecular perspectives, leaving critical gaps in understanding its regulatory networks and evolutionary adaptations [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSeed coat hardness, a key domestication related trait in legumes, is primarily governed by the synergistic interaction between the palisade layer and hilum. The palisade layer-composed of lignin- and suberin-enriched macrosclereids-forms the major barrier to water penetration in wild soybean seeds, while the hilum serves as the dominant water entry pathway. Crucially, the closure of the hilum fissure under low humidity effectively blocks external moisture infiltration, thereby enhancing seed hardness [\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. This mechanism is genetically regulated by GmHs1-1-mediated calcium signaling and PG031-directed pectin degradation [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. The finding that \"the removal of soybean seed coat results in equivalent water absorption rates among embryos confirms seed coat permeability as the primary cause of seed hardness\" is supported by research on seed coat structural characteristics [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Another critical factor contributing to seed hardness is the presence of an impermeable cuticle on the seed coat. Significant differences exist in cuticle structure between hard seeds and imbibing seeds: the cuticle of imbibing seeds exhibits microcracks primarily distributed on the dorsal side, whereas hard seeds maintain an intact cuticular structure without such fissures [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The calcium ion content, pectin levels, xylan composition, and cuticular wax layer in the seed coat of hard seeds are key factors influencing seed coat permeability and contributing to physical dormancy [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eEnvironmental variations significantly influence the formation of soybean seed hard-seededness. Key contributors include changes in natural ecological climate during the late growth stages, storage and transportation conditions, as well as environmental circumstances from post-harvest to long-term preservation periods [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Notably, subtle variations in water management directly influence soybean seed hardening: under constant irrigation frequency, reduced water application volume leads to a corresponding increase in hard seed proportion [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Furthermore, the intensity and duration of sunlight exposure must not be overlooked. Shorter radiation periods and lower light intensity significantly inhibit the formation of hard soybean seeds [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. During storage and transportation, low environmental relative humidity accelerates seed hardening. For instance, phaseolus vulgaris, declining seed moisture content during storage leads to significant increases in hard seed rates [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSeed dormancy is considered one of the most enigmatic phenomena in seed biology. Although extensive research on seed dormancy has been conducted to date with fruitful outcomes, numerous questions and uncertainties remain unresolved. Seed hardness, a relatively stable genetic trait, is inherited by progeny plants from their parental lines. Studies reveal that the genetic basis of seed hardness appears to involve variations in a limited number of genes, with only a few critical genes driving selective differentiation between hard and non-hard seeds. Most evidence supports seed hardness as a dominant trait [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn Arabidopsis thaliana, the cloned \u003cem\u003eDOG1\u003c/em\u003e gene represents the first identified quantitative trait locus (QTL) associated with regulating seed hardness. It was initially proposed that \u003cem\u003eDOG1\u003c/em\u003e influences seed hardness by participating in the ABA (abscisic acid)-mediated sugar signaling pathway [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Subsequently, researchers identified two novel seed dormancy regulators, \u003cem\u003eDESPIERTO\u003c/em\u003e and \u003cem\u003eATHB20\u003c/em\u003e. Both regulators modulate seed dormancy through adjusting ABA sensitivity, with mutations in the \u003cem\u003eDESPIERTO\u003c/em\u003e gene resulting in a complete loss of seed dormancy [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. The seed hardness trait in soybeans was governed by the \u003cem\u003eGmHS1-1 gene\u003c/em\u003e, a dominant monogenic locus. Mutations in this gene reduce calcium content in the seed coat and thin the palisade layer, leading to seed hardness breakdown [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. The \u003cem\u003eGlyma02g43680\u003c/em\u003e gene, identified as a key regulator, exhibits amino acid substitutions in the substrate-binding region of 1,4-β-glucan endo-hydrolase in non-hardy soybean varieties. These mutations diminish 1,4-β-glucan synthesis in the seed coat, suggesting its critical role in seed coat impermeability [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Phenotypic analysis of non-hardy mutants in alfalfa (\u003cem\u003eMedicago sativa\u003c/em\u003e) revealed structural alterations in seed cuticles, characterized by modified lipid monomer composition. These changes induced micro-fissures in the cuticular layer, significantly enhancing water permeability. Genetic investigations identified that mutations in the \u003cem\u003eKNOX4\u003c/em\u003e gene-encoding a \u003cem\u003eKNOTTED\u003c/em\u003e-like homeobox protein\u0026mdash;disrupt its regulatory control over \u003cem\u003eCYP86A\u003c/em\u003e, a cytochrome \u003cem\u003eP450\u003c/em\u003e gene responsible for fatty acid hydroxylation. This epistatic interaction critically impairs cuticular component biosynthesis, establishing \u003cem\u003eKNOX4\u003c/em\u003e mutation as the primary determinant of rapid water uptake in mutant seeds [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. In vegetable soybean (\u003cem\u003eGlycine max\u003c/em\u003e), comparative genomic studies by pinpointing a pectin acetylesterase-coding gene under positive selection during domestication [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Hard-seeded cultivars exhibit functional alleles, while water-imbibing varieties carry frameshift mutations that abolish enzymatic activity. Despite extensive studies on plant dormancy, the molecular mechanisms governing how dormancy factors regulate this trait remain poorly understood. Weighted Gene Co-expression Network Analysis (WGCNA), a systems biology approach first established by Zhang \u003cem\u003eet al\u003c/em\u003e [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], has emerged as a pivotal tool for integrating multi-omics data (transcriptomics, metabolomics, proteomics) to uncover gene functions and network-trait relationships. Based on 508 rice affymetrix microarray datasets, a co-expression network was constructed. Using Random Matrix Theory (RMT) and Weighted Gene Co-expression Network Analysis (WGCNA) methods, 45 functional modules were automatically identified, among which 6 modules (e.g., dark green and dark grey modules) were significantly associated with seed dormancy [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].​Wang \u003cem\u003eet al.\u003c/em\u003e [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e] analyzed 216 soybean landraces from 26 Chinese provinces, selecting contrasting genotypes for seed hardness, and transcriptome sequencing across developmental stages combined with WGCNA pinpointed \u003cem\u003eGmSWEET2\u003c/em\u003e as a central regulator. Functional validation confirmed that overexpression of \u003cem\u003eGmSWEET2\u003c/em\u003e enhances seed hardness by modulating sugar transport dynamics, providing a genetic basis for improving vegetable soybean texture. Despite extensive studies on seed hardness in legumes, no prior research has employed transcriptome sequencing coupled with WGCNA to identify hardness-associated genes in \u003cem\u003eVicia sativa\u003c/em\u003e. This study pioneers the integration of comparative physiology and multi-omics approaches to dissect the wild-domesticated divergence in seed coat impermeability.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e \u003cb\u003eThe agronomic trait analysis and ploidy identification of two\u003c/b\u003e \u003cb\u003evicia sativa.\u003c/b\u003e\u003cb\u003eL\u003c/b\u003e\u003c/p\u003e \u003cp\u003eIn this study, the analysis of agronomic traits revealed that the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) exhibited significantly large leaf length-width ratios, petal sizes, and seed dimensions compared to the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-B to E). Additionally, W30 demonstrated robust stem growth and a more spreading plant architecture (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-A). Cytological identification further indicated that both the wild hard-seeded type (W12) and cultivated type (W30) shared identical diploid chromosome numbers (2n\u0026thinsp;=\u0026thinsp;2x\u0026thinsp;=\u0026thinsp;12), as evidenced by karyotype analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-F, G).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSeed hardness and chemical composition contents\u003c/h2\u003e \u003cp\u003eAt the seed full maturity stage, the seed nutritional components and hardness index were determined for the cultivated type of material (W30) and the wild hard-seeded type material (W12). The results showed no significant difference in crude protein content between the cultivated type of material (W30) and the wild hard-seeded type material (W12) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-A). However, the soluble sugar, soluble protein, and starch contents of the cultivated type of material (W30) were significantly higher than those of the wild hard-seeded type material (W12) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-B-D). Notably, the seed hardness index and lignin content of the wild hard-seeded type material (W12) were significantly higher than those of the cultivated type of material (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-E-F).\u003c/p\u003e \u003cp\u003eAdditionally, the determination of 20 mineral elements in the two materials revealed that the seed coats of the wild hard-seeded type material (W12) exhibited significantly higher levels of potassium (K), aluminum (Al), and manganese (Mn) compared to the cultivated type material (W30). However, the cultivated type material (W30) showed significantly higher levels of calcium (Ca), iron (Fe), sodium (Na), nickel (Ni), copper (Cu), strontium (Sr), cadmium (Cd), antimony (Sb), barium (Ba), and thallium (Tl) than the wild hard-seeded type material (W12) (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cb\u003eDynamic monitoring of seed coat and seed development across various developmental stages of two\u003c/b\u003e \u003cb\u003evicia sativa\u003c/b\u003e\u003c/p\u003e \u003cp\u003eIn this study, the seed coat structures of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) and cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) were identical, consisting sequentially from the outer to inner layers as follows: the cuticle layer, palisade layer, osteosclereid layer, sclerenchyma layer, and parenchyma cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-A-B). The seed diameter of both types exhibited an initial increase followed by a decrease during development. However, throughout the seed development process, the seed diameter of the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) was significantly larger than that of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C-I).\u003c/p\u003e \u003cp\u003eAdditionally, the study revealed that the seed coat thickness of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) was significantly greater than that of the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C-II). As seed development progressed, the palisade tissue thickness of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) and cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) gradually decreased. However, at all developmental stages, the palisade tissue thickness of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) remained significantly greater than that of the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C-III). Notably, the ratio of palisade to seed coat of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) exhibited an initial increase followed by a decline, whereas that of the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) showed a continuous upward trend (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C-IV). Notably, at the wax ripening stage (WRS), the cuticle thickness of the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) was significantly greater than that of the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-D).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eScanning electron microscopy (SEM) analysis of seed coat\u003c/h3\u003e\n\u003cp\u003eScanning electron microscopy (SEM) results revealed that the seed structures of the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) and the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) were similar, both possessing a cuticle layer with comparable morphology, and identical longitudinal section structures of the seed coats (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-A). Further observations of the hilar traits of the two materials revealed significant differences between the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) and the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12). The wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) exhibited a funnel-shaped hilum, positioned farther from the tracheary elements and occupying a smaller proportion of the seed diameter (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-A-I-III). In contrast, the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) displayed a significantly wider hilum that was oriented perpendicular to the tracheary elements and tightly connected to them (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-A-IV-VI).\u003c/p\u003e \u003cp\u003eSubsequent measurements of cuticle thickness demonstrated that the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) exhibited a significantly thicker cuticle layer compared to the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30). As the cuticle serves as a critical barrier to water penetration into the seed coat, this structural divergence is hypothesized to correlate with seed hardness (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-D). Furthermore, the research also indicates that the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) has a narrow and short hilum, with a small hilum width. The hilar fissure, hilum length, seed width, seed diameter, hilum length ratio, and hilar fissure ratio were also smaller than that of the cultivated type vicia sativa (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-B-I-VII). However, the hilar fissure depth in the wild hard-seeded type (W12) was significantly greater than the cultivated type \u003cem\u003evicia\u003c/em\u003e sativa (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e-B-VIII).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eRNA-Seq analyses developing seeds of two\u003c/b\u003e \u003cb\u003evicia sativa.\u003c/b\u003e \u003cb\u003eL with contrasting seed hardness\u003c/b\u003e\u003c/p\u003e \u003cp\u003eIn this study, 18 samples generated approximately 6.09 Gb of raw sequencing data (Raw reads). After filtering, approximately 6.00 Gb of high-quality data (Clean reads) were obtained, with each sample containing over 3.6 Gb of Clean bases, indicating a high sequencing depth (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Additionally, the base error rate of the sequenced reads across the 18 samples was below 0.01%, while Q20, Q30, and GC content values were above 97%, 93%, and 35%, respectively (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). These results demonstrate that the sequences obtained from the transcriptome were accurate and reliable, making them suitable for subsequent analyses. Subsequently, the clean reads were assembled into transcripts using trinity software for further analysis. The results showed that the total length of transcripts was 261,468,601 bp, with a total of 171,109 sequences. The maximum transcript length was 16,812 bp, and the average length was 1,528.08 bp, with N50 and N90 values of 2,213 bp and 718 bp, respectively. For Unigenes, the total sequence length was 75,122,054 bp, comprising 62,729 sequences. The maximum Unigene length was 16,812 bp, with an average length of 1,197.56 bp, and N50 and N90 values of 2,032 bp and 468 bp, respectively (Table S3). Additionally, the mapping rates for all 18 samples exceeded 90%, confirming their suitability for subsequent analyses (Table S4).\u003c/p\u003e \u003cp\u003eFunctional annotation of unigenes revealed that 37,247 unigenes were successfully annotated in the NR database, 21,931 in the GO database, 13,669 in the KEGG database, 19,523 in the Pfam database, 34,086 in the eggNOG database, and 27,435 in the SwissProt database. Among these, the NR database annotated the highest number of Unigenes, accounting for 59.38% of the total (Table S5; Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-A). By aligning with the NR database, functional information of the species' genes and their similarity to homologous genes in closely related species were obtained. The NR annotation results indicated that the highest sequence similarity was to \u003cem\u003eMedicago sativa\u003c/em\u003e (29.78%), followed by \u003cem\u003eTrifolium pratense\u003c/em\u003e (16.39%) and \u003cem\u003eTrifolium subterraneum\u003c/em\u003e (12.46%). Additionally, 24.14% of the annotations corresponded to other species (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-B). The GO functional annotation results revealed that, in the biological process category, unigenes were predominantly annotated to pathways such as cellular process and metabolic process. Under cellular component, unigenes were significantly annotated to the cellular anatomical entity pathway. For molecular function, unigenes were primarily associated with binding and catalytic activity pathways (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-C). The KEGG functional annotation results demonstrated that unigenes were prominently annotated to pathways including metabolism, genetic information processing, environmental information processing, cellular processes, and organismal systems (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-D; Table S6).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eData filtering statistics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSample\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClean Reads No.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eClean Data (bp)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eClean Reads %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eClean Data %\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_1_1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e44040178\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6641055704\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e98.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.78\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_1_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e40317258\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6080154234\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.97\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_1_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e44641220\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6731921523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e98.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.72\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_2_1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e42574096\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6419387846\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e99.09\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_2_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e38678104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5832232560\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.88\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_2_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e38355094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5783190296\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e99.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_2_1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e36808734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5549679363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.96\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_2_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e40296592\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6074116464\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.85\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_2_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e40956608\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6174278615\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.92\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_3_1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e39222454\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5913565498\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.92\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_3_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e37984666\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5727832498\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_3_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e36480664\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5500403766\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_3_1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e40247532\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6065676435\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e99.07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_3_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e40829154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6152134690\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.82\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW12_3_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e41610942\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6268440345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.91\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_4_1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e37764994\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5689771390\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.78\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_4_2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e37012072\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5576123662\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e98.82\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW30_4_3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e37929466\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5716276542\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e99.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e99.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch3\u003eIdentification of DEGs during seed development\u003c/h3\u003e\n\u003cp\u003ePearson correlation coefficient analysis was performed on the gene expression values (FPKM) of the 18 samples. The results indicated that correlation coefficients closer to 1 reflect higher similarity in expression patterns between samples (Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e-A). Principal Component Analysis (PCA) was conducted using the DESeq package in R based on expression levels, revealing high similarity in expression patterns within groups and significant differences between groups. Samples from different treatment groups were dispersed, while samples within the same group clustered together (Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e-B).\u003c/p\u003e \u003cp\u003eAt the filling stage, there were 9,936 differentially expressed genes (DEGs) between the wild hard-seeded material W12 and the cultivated type material W30, with 4,056 genes up-regulated and 5,880 genes down-regulated (Figure S3-A). At the milk-ripe stage, 8,609 DEGs were identified between W12 and W30, including 3,467 up-regulated genes and 5,142 down-regulated genes (Figure S3-B). During the wax ripening stage, 9,459 DEGs were observed, with 4,200 genes up-regulated and 5,259 genes down-regulated (Table S7; Figure S3-C).\u003c/p\u003e\n\u003ch3\u003eGO and KEGG enrichment analysis of DEGs\u003c/h3\u003e\n\u003cp\u003eAt the filling stage, GO enrichment analysis of the 9,936 DEGs revealed significant enrichment in 33 Biological Processes (BP), 9 Cellular Components (CC), and 17 Molecular Functions (MF) (Table S8; Figure S4-A). Among these, the top three enriched BPs were cell wall organization or biogenesis (245 DEGs), secondary metabolite biosynthetic process (105 DEGs), and external encapsulating structure organization (203 DEGs). The top three enriched CCs were intrinsic component of membrane (1,135 DEGs), integral component of membrane (1,090 DEGs), and extracellular region (385 DEGs). Additionally, KEGG enrichment analysis of the 9,936 DEGs revealed significant enrichment in 11 biological pathways. The top three enriched pathways were Biosynthesis of other secondary metabolites (81 DEGs), Lipid metabolism (14 DEGs), and Biosynthesis of other secondary metabolites (23 DEGs) (Table S9; Figure S4-D).\u003c/p\u003e \u003cp\u003eAt the milk-ripe stage, GO enrichment analysis of the 8,609 DEGs revealed significant enrichment in 5 Biological Processes (BP), 2 Cellular Components (CC), and 10 Molecular Functions (MF) (Table S10; Figure S4-B). The top three enriched BPs were RNA modification (192 DEGs), nucleic acid phosphodiester bond hydrolysis (275 DEGs), and meiotic chromosome segregation (32 DEGs). The top two enriched CCs were intrinsic component of membrane (1,014 DEGs) and integral component of membrane (969 DEGs). Furthermore, KEGG enrichment analysis of the 8,609 DEGs demonstrated significant enrichment in 12 biological pathways. The top three enriched pathways were Brassinosteroid biosynthesis (8 DEGs), DNA replication (30 DEGs), and Mismatch repair (25 DEGs) (Table S11; Figure S4-E).\u003c/p\u003e \u003cp\u003eAt the wax ripening stage, GO enrichment analysis of the 9,459 DEGs revealed significant enrichment in 8 Biological Processes (BP), 1 Cellular Component (CC), and 5 Molecular Functions (MF) (Table S12; Figure S4-C). The top three enriched BPs were RNA modification (234 DEGs), ncRNA processing (183 DEGs), and nucleic acid phosphodiester bond hydrolysis (303 DEGs). Additionally, KEGG enrichment analysis of the 9,459 DEGs demonstrated significant enrichment in 7 biological pathways. The top three enriched pathways were Ribosome biogenesis in eukaryotes (42 DEGs), DNA replication (34 DEGs), and Flavonoid biosynthesis (23 DEGs) (Table S13; Figure S4-F).\u003c/p\u003e \u003cp\u003eIn addition, we conducted a comprehensive analysis of DEGs across the three comparison groups. The results revealed 2,554 DEGs commonly enriched in all three groups (Figure S5-A). Subsequent GO enrichment analysis of these 2,554 DEGs showed that the top three enriched Biological Processes (BP) were meiotic nuclear division (19 DEGs), RNA phosphodiester bond hydrolysis, endonucleolytic (25 DEGs), and meiotic chromosome segregation (12 DEGs) (Figure S5-B). KEGG enrichment analysis indicated that the 2,554 DEGs were associated with 101 biological pathways, with only the RNA polymerase pathway (11 DEGs) showing significant enrichment (Figure S5-C-D; Table S14). Further analysis of the expression patterns of these 11 DEGs revealed that only \u003cem\u003eTRINITY_DN15782_c0_g2\u003c/em\u003e, \u003cem\u003eTRINITY_DN3450_c1_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN21678_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN200_c0_g2\u003c/em\u003e, and \u003cem\u003eTRINITY_DN24823_c0_g1\u003c/em\u003e exhibited significantly higher expression levels in the wild hard-seeded material W12 compared to the cultivated type material W30 (Figure S5-E).\u003c/p\u003e \u003cp\u003e \u003cb\u003eWeight gene co-expression network analysis (WGCNA), identification of the key candidate gene and establishing a regulatory network of key candidate genes\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe WGCNA was performed using the FPKM values of 4,999 DEGs and four phenotypic traits (seed diameter, seed coat thickness, palisade tissue thickness, and ratio of palisade to seed coat). First, hierarchical clustering of all samples was conducted based on gene expression levels, with each sample treated as a cluster, and distances between clusters were calculated (Figure S6-A-B). Next, a gene clustering tree was constructed based on pairwise gene expression correlations, and modules were defined by dynamically cutting branches of the tree, grouping genes with similar expression patterns into the same module (Figure S6-C-D). The results showed that the 4,999 DEGs were partitioned into 22 modules (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-A). Notably, the grey60 module exhibited the highest correlation with seed diameter (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/em\u003e), the purple module with seed coat thickness (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/em\u003e), the salmon module with palisade tissue thickness (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/em\u003e), and the cyan module with ratio of palisade to seed coat (\u003cem\u003eP\u0026thinsp;\u0026lt;\u0026thinsp;0.001\u003c/em\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-B). Subsequent analysis of the grey60 module revealed 45 DEGs within this module. Using thresholds of Pearson correlation coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.70 and P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.01, combined with FPKM expression values, 7 candidate genes were identified: \u003cem\u003eTRINITY_DN3202_c0_g3\u003c/em\u003e, \u003cem\u003eTRINITY_DN3663_c0_g2\u003c/em\u003e, \u003cem\u003eTRINITY_DN1963_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN17484_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN12479_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN818_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN16154_c0_g1\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-C; Table S15). Based on functional annotation, \u003cem\u003eTRINITY_DN3663_c0_g2\u003c/em\u003e (annotated as short-chain alcohol dehydrogenase A) was preliminarily identified as a key candidate gene (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFurther analysis of the purple module revealed 183 DEGs within this module. Using thresholds of Pearson correlation coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.70 and P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.01, combined with FPKM expression values, 7 key candidate genes were identified: \u003cem\u003eTRINITY_DN4437_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN477_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN5404_c0_g2\u003c/em\u003e, \u003cem\u003eTRINITY_DN26784_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN4058_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN2362_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-D; Table S15). Based on functional annotation, \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e (annotated as ethylene response 2) was preliminarily identified as a key candidate gene (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFurther analysis of the salmon module revealed 120 DEGs within this module. Using thresholds of Pearson correlation coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.70 and P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.01, combined with FPKM expression values, 6 candidate genes were identified: \u003cem\u003eTRINITY_DN1542_c0_g2\u003c/em\u003e, \u003cem\u003eTRINITY_DN108_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN7853_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN4244_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN1127_c0_g1\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-E; Table S15). Based on functional annotation, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e (annotated as ethylene response factor 8) was preliminarily identified as a key candidate gene (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFurther analysis of the cyan module revealed 115 DEGs within this module. Using thresholds of Pearson correlation coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.70 and P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.01, combined with FPKM expression values, 8 candidate genes were identified: \u003cem\u003eTRINITY_DN1476_c1_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN12417_c1_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN3075_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN3979_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN6069_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN1196_c0_g3\u003c/em\u003e, and \u003cem\u003eTRINITY_DN4975_c0_g2\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-F; Table S15). Based on functional annotation, \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e (annotated as glycine-rich cell wall structural protein 1.0-like) was preliminarily identified as a key candidate gene (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eKey candidate genes identified in this study\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGene\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eTRINITY_DN3663_c0_g2\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBAC81652.1 short-chain alcohol dehydrogenase A\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAOD74920.1 ethylene response 2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAEQ64868.1 ethylene response factor 8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eXP_039023342.1 glycine-rich cell wall structural protein 1.0-like\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e\n\u003ch3\u003eAnalysis of expression patterns of key candidate genes\u003c/h3\u003e\n\u003cp\u003eThe candidate gene expression pattern analysis results indicated that during three stages of seed development, the expression levels of \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e and \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e in the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) were significantly higher than those in the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-G-I-II). Notably, at both the FS and WRS, the expression level of \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e in the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) was significantly higher than that in the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30), while no significant difference was observed at the MRS (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-G-III). Furthermore, during three stages of seed development, the expression levels of \u003cem\u003eTRINITY_DN3663_c0_g2\u003c/em\u003e in the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) was significantly lower than these in the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e-G-IV). This study ultimately identified \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e as key candidate genes regulating seed hardness, warranting further investigation.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eEffects of exogenous ethylene on seed germination\u003c/h2\u003e \u003cp\u003eIn previous studies, we hypothesized that ethylene response factors (\u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e and \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e) regulated seed hardness. To validate this hypothesis, we investigated the effects of soaking seeds in varying ethylene concentrations on germination. The results demonstrated that as ethylene concentration increased, the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30) seeds exhibited a decreasing trend in germination rate, with germination rates under 100 mg/L and 150 mg/L treatments being significantly lower than those of the control group (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-A; B-I). Additionally, under the 150 mg/L ethylene treatment, the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) exhibited a significantly lower seed germination rate compared to the control group (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-A; B-IV). Notably, increasing ethylene concentrations significantly reduced both root length and shoot length in the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30), demonstrating a clear inhibitory effect (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-B-II-III). However, at the 50 mg/L treatment, the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) showed significantly greater root and shoot lengths than the control group and the other two treatment groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-B-V-VI). These results indicate that the 50 mg/L ethylene treatment significantly promotes root and shoot growth in the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003e \u003cb\u003ePhysical factors affecting seed hardness in\u003c/b\u003e \u003cb\u003eVicia sativa\u003c/b\u003e\u003c/p\u003e \u003cp\u003eThe results of this study indicate that the hardness index of wild \u003cem\u003eVicia sativa\u003c/em\u003e seeds is significantly higher than that of cultivated varieties. This phenomenon may be closely associated with their unique physical structural features and cell wall composition. First, the thicker cuticle layer in wild seeds likely enhances seed coat mechanical resistance, directly contributing to increased hardness. As the outermost protective barrier of seeds, the thickened cuticle effectively resists external mechanical stress and water penetration. This mechanism aligns with the functional model of the cuticle proposed by wherein increased cuticle thickness enhances physical resistance by improving the density of the cutin-wax complex [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Wherein increased cuticle thickness enhances physical resistance by improving the density of the cutin-wax complex. Second, the higher lignin content in wild seeds may serve as a critical biochemical basis for increased hardness. Lignin, a secondary metabolite in the cell wall, forms a rigid network structure by cross-linking cellulose and hemicellulose [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Its accumulation significantly enhances the mechanical strength of the seed coat and cotyledon cells. Additionally, the structural features of the hilum in wild seeds, such as a narrow and short hilum, smaller hilar fissure width, and deeper hilar fissure, may indirectly influence hardness by restricting rapid water absorption. The constricted hilum likely slows the imbibition rate, thereby reducing the risk of microcracks in internal tissues caused by rapid expansion [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. This mechanism helps maintain seed integrity and prolongs seed hardness. Notably, although cultivated varieties exhibit higher levels of metabolites such as soluble sugars and starch, their lower hardness suggests that seed hardness is primarily governed by physical structures rather than storage compounds. This finding contrasts with studies on common bean (\u003cem\u003ePhaseolus vulgaris\u003c/em\u003e) by [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e], but aligns with conclusions on hardseededness in pea (\u003cem\u003ePisum sativum\u003c/em\u003e) [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e], which emphasize that seed coat structural traits, such as lignification degree and cuticle thickness are the core determinants of hardness, while storage components predominantly influence post-germination metabolic activity. Additionally, calcium (Ca) is a key element influencing seed hardness, as it participates in pectin cross-linking and stabilizes cell wall structure. Ca\u0026sup2;⁺ form Ca\u0026sup2;⁺-pectin cross-links with pectic acids in the cell wall, enhancing the mechanical strength of the seed coat and reducing water permeability [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. However, in this study, hard-seeded genotypes exhibited significantly lower calcium (Ca) content compared to non-hard-seeded genotypes, while showing higher levels of potassium (K), manganese (Mn), and aluminum (Al), alongside significantly higher hardness indices. This apparent contradiction to the traditional view of calcium promoting seed coat hardening suggests that the regulation of seed hardness in \u003cem\u003eVicia sativa\u003c/em\u003e may involve elemental interactions, structural compositional differences in the seed coat, or functional specificity of calcium. Manganese (Mn) is a cofactor for key enzymes in lignin biosynthesis. Elevated Mn levels may enhance the activity of lignin-synthesizing enzymes, such as manganese peroxidase, promoting the oxidative polymerization of lignin monomers [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Aluminum (Al\u0026sup3;⁺) can induce peroxidase activity to accelerate lignin deposition [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e] and may directly bind to carboxyl groups of pectin in the cell wall, strengthening polysaccharide network cross-linking [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. The observed low calcium content may trigger compensatory mechanisms for lignin synthesis, while high Mn and Al levels directly promote lignin accumulation. Ultimately, this leads to lignin-dominated cell wall reinforcement, achieving high seed hardness. These findings challenge the traditional \u0026ldquo;calcium-centric\u0026rdquo; model and reveal the complexity of a multi-element-lignin network, offering new directions for improving seed permeability or storage tolerance.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTranscriptome analysis revealed the key pathway involved in the regulation of\u003c/b\u003e \u003cb\u003evicia sativa\u003c/b\u003e \u003cb\u003eseed hardness\u003c/b\u003e\u003c/p\u003e \u003cp\u003eThis study revealed the dynamic molecular network underlying seed hardness formation in common vetch through transcriptome sequencing, with the core mechanisms involving cell wall reinforcement, secondary metabolite accumulation, and coordinated hormonal signaling. During the filling stage, DEGs were significantly enriched in biological pathways such as cell wall organization or biogenesis (245 DEGs) and secondary metabolite biosynthesis (105 DEGs) (Table S8; Table S9). The high expression of cell wall-related regulatory genes may be a key driver for the deposition of cellulose, hemicellulose, and pectin in the seed coat [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Concurrently, the activation of secondary metabolic pathways (e.g., lignin precursor synthesis) suggests early regulation of phenylalanine ammonia-lyase (PAL) and peroxidase [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. This stage likely establishes the initial mechanical strength of the seed coat by coordinating the supply of cell wall polysaccharides and secondary metabolites. During the milk-ripe stage, DEGs were significantly enriched in biological pathways such as brassinosteroid biosynthesis (8 DEGs) and DNA replication (30 DEGs) (Table S10; Table S11). Brassinosteroids (BRs) have been shown to promote the expression of cell wall-loosening enzymes (e.g., expansins) by activating the \u003cem\u003eBZR1\u003c/em\u003e transcription factor; however, the enrichment of BR pathways at this stage may indirectly maintain cell wall stability by antagonizing ethylene signaling [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. Additionally, the high expression of genes associated with RNA modification (192 DEGs) and hydrolysis of nucleic acid phosphodiester bonds (275 DEGs) may fine-tune the spatiotemporal expression of cell wall synthesis enzymes by regulating mRNA stability or translation efficiency [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e] (Table S10; Table S11).\u003c/p\u003e \u003cp\u003eDuring the wax-ripening stage, DEGs were significantly enriched in biological pathways such as flavonoid biosynthesis (23 DEGs) and eukaryotic ribosome biogenesis (42 DEGs) (Table S12; Table S13). Flavonoids may enhance seed coat impermeability through covalent binding with cell wall polysaccharides [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], while the upregulation of ribosome biogenesis-related genes likely supports the synthesis of abundant cell wall structural proteins (e.g., glycine-rich proteins, GRPs) [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. Concurrently, the sustained activity of nucleic acid phosphodiester bond hydrolases may optimize metabolic resource allocation toward cell wall reinforcement pathways by degrading redundant RNA molecules. Studies have also shown that seed hardness formation exhibits stage-specific characteristics: the filling stage initiates cell wall skeleton construction, the milk-ripe stage maintains structural plasticity through hormonal balance, and the wax-ripening stage achieves final hardening via secondary metabolite deposition (e.g., flavonoids, lignin) and cross-linking of structural proteins. This process shares high similarity with seed coat development mechanisms reported in pomegranate and rice [\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. However, the unique flavonoid biosynthesis pathway in common vetch suggests species-specific stress resistance strategies in its seed coat, reflecting both conserved regulatory networks and species-specific adaptations. These findings provide critical molecular targets for breeding \u003cem\u003evicia sativa\u003c/em\u003e with improved seed hardness traits.\u003c/p\u003e\n\u003ch3\u003eFunctional analysis of key candidate genes\u003c/h3\u003e\n\u003cp\u003eStudies have shown that short-chain alcohol dehydrogenases (SDRs) are a class of oxidoreductases widely present in living organisms, catalyzing dehydrogenation or reduction reactions of substrates such as alcohols, steroids, and lipids. Members of this family typically rely on NAD(P)+/NAD(P)H as cofactors and are involved in the synthesis of secondary metabolites and detoxification processes [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. In plants, SDRs may participate in the biosynthesis of defensive compounds such as flavonoids and terpenoids, or respond to oxidative stress (e.g., by reducing toxic aldehydes). For instance, the SDR gene \u003cem\u003eAt5g16970\u003c/em\u003e in \u003cem\u003eArabidopsis\u003c/em\u003e has been demonstrated to regulate the metabolism of abscisic acid (ABA) [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. In this study, we identified three key candidate genes regulating seed hardness (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Among them, ethylene response 2 (\u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e) belongs to the ethylene receptor family and serves as the initial component of the ethylene signal transduction pathway. In \u003cem\u003eArabidopsis\u003c/em\u003e, \u003cem\u003eETR1\u003c/em\u003e and \u003cem\u003eETR2\u003c/em\u003e regulated downstream signaling cascades by binding ethylene molecules, influencing seed germination, fruit ripening, and stress responses [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. \u003cem\u003eETR2\u003c/em\u003e likely transmits signals via its histidine kinase activity, exerting negative regulation on ethylene responses (e.g., suppressing the activity of \u003cem\u003eEIN3/EIL1\u003c/em\u003e transcription factors). Studies have shown that \u003cem\u003eETR2\u003c/em\u003e mutants exhibit increased ethylene sensitivity, highlighting its role in suppressing ethylene signaling [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. The function of this gene may relate to seed hardness, as ethylene is involved in regulating the expression of cell wall modification genes. Additionally, ethylene response factor 8 (\u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e) belongs to the \u003cem\u003eAP2/ERF\u003c/em\u003e transcription factor family and directly binds to the ethylene-responsive element (GCC-box) to activate or repress downstream target gene expression. \u003cem\u003eERF8\u003c/em\u003e plays a critical role in plant stress resistance (e.g., drought, salinity) and pathogen defense [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. For example, \u003cem\u003eSlERF.B3\u003c/em\u003e in tomato (\u003cem\u003eSolanum lycopersicum\u003c/em\u003e) regulated the expression of cell wall-modifying enzymes, thereby influencing fruit softening [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. In \u003cem\u003evicia sativa\u003c/em\u003e, we hypothesize that \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e may enhance seed coat mechanical strength and thus improve seed hardness by regulating the expression of lignin biosynthesis genes (e.g., PAL, 4CL) [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. Studies have also shown that cell wall structural proteins (GRPs), characterized by glycine-rich repeat motifs, enhance cell wall mechanical stability by interacting with cell wall polysaccharides (e.g., cellulose, hemicellulose) via hydrogen bonds. In \u003cem\u003eArabidopsis\u003c/em\u003e, \u003cem\u003eAtGRP3\u003c/em\u003e has been demonstrated to participate in secondary cell wall thickening, and its expression is induced under stress conditions (e.g., low temperature, pathogen infection) [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. Similarly, \u003cem\u003eOsGRP1\u003c/em\u003e in rice (\u003cem\u003eOryza sativa\u003c/em\u003e) improves stem lodging resistance by regulating cell wall cross-linking These proteins may further reinforce seed coat structure and influence seed hardness by binding to lignin-polysaccharide complexes [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. Based on these findings, this study hypothesizes that the key candidate genes \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e may form a synergistic regulatory network in seed hardness control. Specifically, \u003cem\u003eETR2\u003c/em\u003e (\u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e) likely suppresses ethylene signaling, indirectly reducing the expression of cell wall-loosening enzyme genes (e.g., polygalacturonases), thereby maintaining cell wall integrity [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e]. \u003cem\u003eERF8\u003c/em\u003e (\u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e), as an \u003cem\u003eAP2/ERF\u003c/em\u003e transcription factor, directly activates the high expression of lignin biosynthesis genes (e.g., PAL, peroxidases), promoting lignin deposition [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. \u003cem\u003eGRP\u003c/em\u003e (\u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e) enhances cell wall rigidity through physical cross-linking, synergizing with lignin to construct a mechanical stress-resistant barrier [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. These results provide potential molecular targets for deciphering the mechanisms underlying seed hardness in \u003cem\u003evicia sativa\u003c/em\u003e, highlighting the interplay between ethylene signaling, lignin biosynthesis, and structural protein-mediated reinforcement in shaping seed coat mechanical properties.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003ePlant materials and sampling\u003c/h2\u003e \u003cp\u003eThe wild material W12, characterized by complete seed hardness, was provided by the Germplasm Bank of wild species in Southwest China (Accession Number: 868710143316; Chongqing, China) and plant materials were identified by germplasm bank administrator shaofa Qin. The cultivated variety W30 (internally assigned identifier) is a local landrace collected from Pingwu County, Mianyang City, Sichuan Province (Collection ID: 2019511248) and plant materials were identified by Dr. yongqun Zhu (Institute of agricultural resources and environment, sichuan academy of agricultural sciences). The experimental materials can be obtained in accordance with relevant agreements. From 2019 to 2022, both accessions were cultivated for three consecutive years under natural conditions in Yingxi Town, Nanchong City, China (106\u0026deg;12\u0026prime;N, 31\u0026deg;12\u0026prime;E). The planting layout included 3-meter-wide rows with plant and row spacing 50 cm, arranged in a completely randomized block design with three replicates. At the bud formation stage, individual plants were labeled. Seeds were collected at three developmental stages: filling stage (W12-1; W30-2), milk-ripe stage (W12-2; W30-3), and wax ripening stage (W12-3; W30-4). Pods were manually opened to extract seeds, and seed coats were immediately separated. The samples were flash-frozen in liquid nitrogen and stored at -80\u0026deg;C for subsequent analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eChromosome number was determined through cytological analysis\u003c/h2\u003e \u003cp\u003eSeeds were germinated in Petri dishes lined with double-layered filter paper and placed in a constant-temperature incubator at 25\u0026deg;C. For hard-seeded seeds, the seed coat was scarified to facilitate water uptake. When root tips reached approximately 1\u0026ndash;2 cm in length, they were excised and fixed in α-bromonaphthalene for 3.5 hours. Following fixation, root tips were rinsed in distilled water for 30 minutes, then transferred to a 3:1 (v/v) ethanol-acetic acid solution for overnight fixation and stored at 4\u0026deg;C until further processing. For enzymatic maceration, root tips were washed in distilled water for 10 minutes, and 2\u0026ndash;3 mm of the white meristematic tissue was excised from the tip. The excised tissue was incubated in a mixed enzyme solution containing cellulase and pectinase at 37\u0026deg;C for 1.5 h (20 \u0026micro;L per root tip). Microscope slides were prepared according to the method described by Kato \u003cem\u003eet al.\u003c/em\u003e [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e], involving squashing and staining for chromosomal observation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSeed hardness and chemical composition contents\u003c/h2\u003e \u003cp\u003eThe seeds to be tested were dried in an oven at 60 ℃ then were treated separately for one second by a small grinder (ZT-150, Yongkang Zhanfan Industry and Trade Co., Ltd.​) to get the seed coat debris. 1 g seed coat from each of 10 random individuals for each seed coat type were harvested and then mixed. The seed coat debriswas collected and fully crushed into fine powder by grinder, passed through an 40-mesh sieve Save as backup [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. Used for elemental analysis by Inductively Coupled Plasma Mass Spectrometer (ICP-MS) (Thermo Fisher Scientific, America).\u003c/p\u003e \u003cp\u003eSimilarly, The seeds to be tested were dried in an oven at 60 ℃ then were treated separately for one second by a small grinder (ZT-150, Yongkang Zhanfan Industry and Trade Co., Ltd.) to get the seed fragments for future use. The contents of crude protein (CP), soluble protein (SP), soluble sugar (SS), starch (St), and lignin (Lg) in mature seeds were determined using kits from Quanzhou Ruixin Biotechnology Co., Ltd. In this study, we used the national standard method GB/T21304-2007 (Determination of wheat hardness-Hardness index method) to measure seed hardness indicators. The seed hardness index was determined in accordance with GBT 21304.\u003c/p\u003e \u003cp\u003e \u003cb\u003eThe analysis of structural changes in seed coat cells during various developmental stages and the process of seed development\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo study the dynamic changes of seed coat and seeds during seed development, wild hardness material W12 and non-hardness material W30 with high hard seed rate were selected as the research objects. 18 samples were collected at various stages of seed development (filling stage, milk-ripe stage, and wax ripening stage). The seeds were separately fixed with FAA fixative and made into paraffin sections. After rinsing the fixative solution, the solvent was dehydrated using a gradient of ethanol concentrations and subsequently stained with Safranin O. The material achieved transparency by employing a mixture of absolute ethanol and xylene in a 1:1 volume fraction ratio, followed by immersion in a xylene solution. Upon achieving transparency, the material was then immersed in wax and embedded. Then, the embedded material was sectioned into 8 \u0026micro;m thick slices using a Leica RM2245 semi-automatic rotary microtome. Following drying, the sections were deparaffinized and made transparent with a xylene solution and a mixture of absolute ethanol and xylene. Subsequently, they were rehydrated using an ethanol solution with varying concentration gradients against the original gradient, before being restained with toluidine blue. Finally, after being transparent again, it was sealed with a sealing tablet. Observations were made using a light microscope OLYMPUS BX53, photographed, and recorded. The seeds were randomly selected from each batch, with 18 samples in total. Furthermore, these samples were photographed and analyzed using Digimizer version 5.4.4, a software specifically designed for image analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eScanning electron microscopy (SEM) analysis of seed coat\u003c/h2\u003e \u003cp\u003eAt seeds full-ripe stage, a total of 10 seeds with hard material (W12) and 10 seeds with non-hard material (W30) were randomly selected for the purpose of observing the surface structure (such as cracks and waxy appearance) as well as the longitudinal structural morphology of their seed coat. The initial step involved the utilization of an IB-5 ion plating device for the application of a 2.3 nm Au film coating [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. Subsequently, the prepared samples were captured using an S-570 scanning electron microscope (Hitachi LTD, Japan).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eTreating seeds with exogenous ethephon\u003c/h2\u003e \u003cp\u003eRandomly selected wild-type seeds in sufficient quantity were scarified by soaking in a 98% concentrated sulfuric acid solution for 30 minutes. After treatment, the seeds were rinsed with purified water for 3 minutes, and the surface moisture was blotted dry using absorbent paper. The seeds were then air-dried for subsequent use. After sterilization, wild hard-seeded and cultivated seeds were separately soaked in exogenous ETH solutions (purity\u0026thinsp;\u0026ge;\u0026thinsp;85%) at concentrations of 0 (CK), 100, 150, and 200 mg/L for 24 h at room temperature. Following treatment, seeds were retrieved, rinsed thoroughly with distilled water, and air-dried for subsequent use. For germination assays, 100 seeds per treatment group were randomly selected and placed in sterile Petri dishes lined with two layers of moist filter paper. The dishes were incubated in an artificial climate chamber (AS-R600L2N, Xunneng Instruments Beijing Co., Ltd.) under controlled conditions: temperature: 25\u0026thinsp;\u0026plusmn;\u0026thinsp;1\u0026deg;C, light cycle: 16-h light/8-h dark photoperiod, replicates: three independent replicates per treatment. Germination was monitored daily for 7 d. The number of germinated seeds was recorded each day, and germination potential and germination rate were calculated on the 7 d. After the germination period, shoot height and root length of all seedlings were measured using a vernier caliper.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eTranscriptome sequencing and identification of differentially expressed genes (DEGs)\u003c/h2\u003e \u003cp\u003eThe wild hardness material W12 and non-hardness seed W30 were chosen as the subjects of this study, and 18 samples were collected at various stages of seed development (filling stage, milk-ripe stage, and wax ripening stage). Extracted total RNA from each sample, tested the purity, concentration, and integrity, and send 0.5-2 \u0026micro;g RNA for library construction and RNA-seq (Shanghai Personal Biotechnology Co., Ltd., China). After library construction, PCR amplification was employed for library fragment enrichment. Subsequently, library selection was conducted based on the desired fragment size of 450 bp. The library's quality was subsequently assessed using the Agilent 2100 Bioanalyzer, which allowed for the determination of both the total concentration and effective concentration of the library. Then, the libraries containing different Index sequences were proportionally mixed based on the effective concentration and data requirement. The pooled libraries were uniformly diluted to a concentration of 2 nM, and single-stranded libraries were formed through alkali denaturation. Following RNA extraction, purification, and library construction, Next-Generation Sequencing (NGS) technology was utilized for paired-end (PE) sequencing of these libraries using the Illumina sequencing platform [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eRaw data underwent filtration, removing reads with connectors, length less than 50bp, and average sequence quality below Q20. Subsequently, the resulting high-quality sequences were de novo spliced to obtain transcript sequences. The splicing transcript was sequenced based on its length from long to short, and the length of transcript was added to the length of splicing transcript, so that it was not \u0026lt;\u0026thinsp;50%/90% of the total length, namely N50/N90, to measure the continuity of de novo assembly. Subsequently, the transcripts were clustered, and the longest transcript was selected as the Unigene, and unigene was utilized for subsequent annotation processes including GO, KEGG, eggNOG, SwissProt, Pfam annotation, ORF prediction, SSR prediction [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. The filtered sequences were simultaneously aligned to Unigene to obtain the Reads Count of each unigene, and the samples underwent further analysis for differential expression, enrichment, cSNP and InDel [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e]. Subsequently, the read count was converted to an FPKM, and the level of gene expression level was evaluated using the FPKM value [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e]. \u0026lsquo;DEseq2\u0026rsquo; R package was used for the identification of DEGs which were filtered with|log\u003csub\u003e2\u003c/sub\u003e (fold change) | \u0026ge; 1 and false discovery rate (FDR)\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eGO and KEGG enrichment analysis of DEGs\u003c/h2\u003e \u003cp\u003eThe heatmaps of gene expression, gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) pathway analyses were performed using the OmicShare tools, a free online platform for data analysis (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.omicshare.com/tools\u003c/span\u003e\u003cspan address=\"http://www.omicshare.com/tools\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and padj\u0026thinsp;\u0026lt;\u0026thinsp;0.01 was used as the threshold of significant enrichment in both analyses described above. The correlation and principal Component analysis (PCA) were performed with R package gmodels (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.r-project.org/\u003c/span\u003e\u003cspan address=\"http://www.r-project.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) in this study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eWeighted gene co-expression network analysis (WGCNA) and identification of the key candidate genes\u003c/h2\u003e \u003cp\u003eAfter screening out undetectable or low expression genes, genes with FPKM of more than 0.5 were utilized for WGCNA using the \u0026lsquo;WGCNA\u0026rsquo; R package [\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e]. The power value was set to 11 to make the gene co-expression network conform to a scale-free network. The \u0026lsquo;merge Cut Height\u0026rsquo; was set to 0.25, and the \u0026lsquo;min Module Size\u0026rsquo; was set to 25 to further classify and merge the gene modules [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. The module eigengene E, was calculated to identify the modules related to seed hardness. In each module, the for which with Person Correlation Value\u0026thinsp;\u0026gt;\u0026thinsp;0.70 and \u003cem\u003eP\u003c/em\u003e_value\u0026lt;0.01 were regarded as hub genes. The hub genes heat map was performed using the OmicShare bioinformatics learning platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.omicshare.com/tools\u003c/span\u003e\u003cspan address=\"http://www.omicshare.com/tools\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e)[\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e]. Subsequently, the gene co-expression network map was constructed by cytoscape software [\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eAnalysis of expression patterns of key candidate genes\u003c/h2\u003e \u003cp\u003eTo further identify key candidate genes, qRT-PCR was used to determine the relative expression levels of four candidate genes in each variety and stage. Primer Premier 5 software was used to design the gene-specific primers, which were listed in Table S16. The \u003cem\u003eVicia sativa\u003c/em\u003e gene \u003cem\u003eVSACT11\u003c/em\u003e (GenBank: GU946218) is the internal control gene and fold change was calculated using the 2\u003csup\u003e\u0026minus;ΔΔCT\u003c/sup\u003e method [\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e]. Three independent biological replicates were performed on each sample to ensure statistical reliability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eMicrosoft Excel (Microsoft Corp., Redmond, WA, USA) and SPSS (IBM, Inc., Armonk, NY, USA) were used to calculate descriptive statistics, and to test statistical significance, respectively. The bar chart was drawn by graphpad prism 9.5.1, and the Box plot was drawn by the OmicShare tools, an online platform for data analysis (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.omicshare.com/tools\u003c/span\u003e\u003cspan address=\"https://www.omicshare.com/tools\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThe results of this study indicate that a thicker cuticle and narrower hilum fissure width constitute the first barrier affecting rapid water absorption. Excessive lignin content leads to hardened seed coats, which further impacts seed water permeability and gas permeability, thereby inducing seed hardness (hard-coated seeds). Furthermore, through WGCNA analysis, three key candidate genes involved in regulating seed hardness were identified (\u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e). The candidate gene expression pattern analysis results indicated that, the expression levels of \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e in the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) were significantly higher than those in the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe gratefully acknowledge Dr. Ruyu He for his expert guidance in the transcriptome data analysis, which was critical to this study. Thanks for germplasms provided by Germplasm Bank of wild species in Southwest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHW performed the experiments, analyzed the data and prepared the figures and tables. ZW approved the final draft. YZ, XY and BZ processed the data. HW and ZD assisted in completing part of the experiment. YC and ZY conceived and designed the experiments, authored or reviewed drafts of the article, and approved the final draft. ZD revised the manuscript and provided financial support for this research. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the Open Competition Mechanism to Select the Best Candidates from the Sichuan Academy of Agricultural Sciences (no. 1+9KJGG004), and the Sichuan Province Research Grant (nos. 2022ZZCX086, 2022ZZCX084).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData is provided within the manuscript or supplementary information files. The raw sequence data of RNA-seq is available in the National Genomics Data Center (NGDC) with the accession number CRA044150.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor details\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1\u003c/sup\u003eInstitute of Special Economic Animals and Plants, Sichuan Academy of Agricultural Sciences, Nanchong, 637000, China.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e2\u003c/sup\u003eForage Crops Germplasm Innovation and Production Management Key Laboratory of Nanchong City, Sericulture Research Institute, Sichuan Academy of Agricultural Sciences, Nanchong, 637000, China.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e3\u003c/sup\u003eMaize Research Institute, Sichuan Agricultural University, Chengdu 611130, China.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eZhou Q,Cui Y,Dong S,Luo D,Fang L, Shi Z, Liu W,Wang Z, Nan Z. Integrative physiological, transcriptome, and metabolome analyses reveal the associated genes and metabolites involved in cold stress response in common vetch (Vicia sativa L.). Food and Energy Security.2023;12(4):22.https://doi.org/10.1002/fes3.484.\u003c/li\u003e\n\u003cli\u003eSun Y,Zhao N,Sun H,Xu S,Lu Y,Xi H,Guo Z,Shi H.Transcriptome profiling reveals molecular responses to salt stress in Common Vetch (\u003cem\u003eVicia sativa\u003c/em\u003e L.). Plants (2223-7747).2024;13(5). https://doi.org/10.3390/p222. https://doi.org/10.1002/fes\u003c/li\u003e\n\u003cli\u003eGrela E,Samolińska W,Rybiński W,Kiczorowska B,Kowalczuk-Vasilev E,Matras J, Wesoowska S.Nutritional and anti-nutritional factors in vicia sativa l.seeds and the variability of phenotypic and morphological characteristics of some vetch accessions cultivated in european countries. Animals. 2021;11(1). https://doi.org/10.3390/ani11010044.\u003c/li\u003e\n\u003cli\u003eLi R,Zhang Z,Tang W,Huang Y,Nan Z.Common vetch cultivars improve yield of oat row intercropping on the Qinghai-Tibetan plateau by optimizing photosynthetic performance. European Journal of Agronomy.2020;117:126088.https://doi.org/10.1016/j.eja.2020.126088.\u003c/li\u003e\n\u003cli\u003eLai X,Yang X,Wang Z,Shen Y,Ma L.Productivity and water use in forage-winter wheat cropping systems across variable precipitation gradients on the Loess Plateau of China. Agricultural Water Management.2022;259.https://doi.org/10.1016/j.agwat.2021.107250.\u003c/li\u003e\n\u003cli\u003eXu C,Kang N,Xu H-L,Yu D-H,Bi Y-X.Research progress of wild soybean germplasmsand utilization.Agricultural Science \u0026amp; Technology.2017;18:812-817. https://doi.org/10.16175/j.cnki.1009-4229.2017.05.016.\u003c/li\u003e\n\u003cli\u003eAvramidou E,Sarri E,Papadopoulou E,Petsoulas C,Tigka E,Tourvas N,et al.Phenotypic, genetic, and metabolite variability among genotypes of \u003cem\u003eVicia sativa\u003c/em\u003e L.Applied Sciences-basel.2024;14(20):9272.https://doi.org/10.3390/app14209272.\u003c/li\u003e\n\u003cli\u003eMeyer CJ,Steudle E,Peterson CA.Patterns and kinetics of water uptake by soybean seeds. Journal of Experimental Botany.2007;58(3):717-732.https://doi.org/10.1093/jxb/erl244.\u003c/li\u003e\n\u003cli\u003eShu K,Liu X-D,Xie Q,He Z-H.Two faces of one seed: Hormonal regulation of dormancy and germination.Molecular Plant.2016;9(1):34-45.https://doi.org/10.1016/j.molp.2015.08.010. \u003c/li\u003e\n\u003cli\u003eRancisco H, D\u0026uuml;bbern DS,J\u0026uacute;lio M.The seed coat as a modulator of seed-environment relationships in Fabaceae.Brazilian Journal of Botany.2001;24(4):365-375. https://doi.org/10.1590/S0100-84042001000400002.\u003c/li\u003e\n\u003cli\u003eWen Z-Z,Lu XO-R,Wen J-Q,Wang Z-Y,Chai M-F.Physical seed dormancy in legumes: molecular advances and perspectives. Plants. 2024;13(11):1473. https://doi.org/10.3390/plants13111473.\u003c/li\u003e\n\u003cli\u003eTaylor GB.Hardseededness in Mediterranean annual pasture legumes in Australia:a review. Australian Journal of Agricultural Research.2005;56(7):645-661. https://doi.org/10.1071/AR04284.\u003c/li\u003e\n\u003cli\u003eWang F-F,Sun X,Liu B-H,Kong F-J,Pan X-W, Zhang H-Y. A polygalacturonase gene PG031 regulates seed coat permeability with a pleiotropic effect on seed weight in soybean. Theoretical and Applied Genetics.2022;135(5):1603-1618.https://doi.org/10.1007/s00122-022-04057-4.\u003c/li\u003e\n\u003cli\u003eSun L-J,Miao Z-Y,Cai C-M,Zhang D-J,Zhao M-X,Wu Y-Y,Zhang X-L,Stephen AS,Zhou L-W, et al. GmHs1-1, encoding a calcineurin-like protein,controls hard-seededness in soybean. Nature Genetics.2015;47(8):939. https://doi.org/10.1038/ng.3339.\u003c/li\u003e\n\u003cli\u003eMa F, Cholewa E, Mohamed T, Peterson CA, Gijzen M.Cracks in the palisade cuticle of soybean seed coats correlate with their permeability to water.Annals of Botany.2004;(2):213-228. https://doi.org/10.1093/aob/mch133.\u003c/li\u003e\n\u003cli\u003eShao S-Q,Chris J.Meyer,Ma F-S,Carol A,Peterson,Mark AB et al.The outermost cuticle of soybean seeds: chemical composition and function during imbibition.Journal of Experimental Botany.2007;58(5):1071-1082.https://doi.org/10.1093/jxb/erl268. \u003c/li\u003e\n\u003cli\u003eWang F,Sun X,Liu B,Kong F,Pan X,Zhang H. A polygalacturonase gene PG031 regulates seed coat permeability with a pleiotropic effect on seed weight in soybean. Theoretical and Applied Genetics.2022;135(5):1603-1618. https://doi.org/10.1007/s00122-022-04057-4.\u003c/li\u003e\n\u003cli\u003eKebede H,Smith JR,Ray JD.Identification of a single gene for seed coat impermeability in soybean PI 594619.Theoretical and Applied Genetics.2014;127(9):1991. https://doi.org/10.1007/s00122-014-2355-2.\u003c/li\u003e\n\u003cli\u003eMa F-S,Cholewa E,Mohamed T,Peterson CA,Gijzen M.Cracks in the palisade cuticle of soybean seed coats correlate with their permeability to water.Annals of Botany.2004(2):213-228. https://doi.org/10.1093/aob/mch133.\u003c/li\u003e\n\u003cli\u003eYu B,He X-M,Tang Y,Chen Z-H,Zhou L-M,Li X-M,Zhang C-Y,Huang X,Yang Y-H,Zhang W-B.Photoperiod controls plant seed size in a CONSTANS-dependent manner.Nature Plants.2023;9:343-354.https://doi.org/10.1038/s41477-023-01350-y.\u003c/li\u003e\n\u003cli\u003eEllis R H, Hong TD, Roberts EH.Moisture content and the longevity of seeds of phaseolus vulgaris.Annals of Botany.1990;(3):341-348.https://doi.org/10.2307/2444634.\u003c/li\u003e\n\u003cli\u003eGladstones J,Atkins C,Hamblin J.Lupins as crop plants: biology, production and utilization. Field Crop Abstracts.1998;123-148. https://doi.org/10.1079/9780851992242.0000\u003c/li\u003e\n\u003cli\u003eNiu J, Cao D, Li H-X, Xue H, Chen L-N, Liu B-B, Cao S-Y.Quantitative proteomics of pomegranate varieties with contrasting seed hardness during seed development stages.Tree Genetics \u0026amp; Genomes.2018;14: 14.https://doi.org/10.1007/s11295-018-1229-1.\u003c/li\u003e\n\u003cli\u003eBentsink L,Jowett J,Hanhart CJ,Koornneef M. Cloning of \u003cem\u003eDOG1\u003c/em\u003e, a quantitative trait locus controlling seed dormancy in \u003cem\u003eArabidopsis\u003c/em\u003e. Proceedings of the National Academy of Sciences of the United States of America.2006;103(45). https://doi.org/www.pnas.org/cgi/doi/10.1073/pnas.0607877103. 2006;103:17042-17047.\u003c/li\u003e\n\u003cli\u003eBarrero JM,Millar AA,Griffiths J,Czechowski T,Scheible WR.Gene expression profiling identifies two regulatory genes controlling dormancy and ABA sensitivity in Arabidopsis seeds. Plant Journal.2010;61(4):611-622. https://doi.org/10.1111/j.1365-313X.2009.04088.x..\u003c/li\u003e\n\u003cli\u003eXia F,Chen L,Sun Y,Mao P.Relationships between ultrastructure of embryo cells and biochemical variations during ageing of oat (\u003cem\u003eAvena sativa\u003c/em\u003e L.) seeds with different moisture content.Acta Physiologiae Plantarum.2015;37(4):1-11.https://doi.org/10.1007/s11738-015-1825-8.\u003c/li\u003e\n\u003cli\u003eJang S-J,Sato M,Sato K,Jitsuyama Y,Fujino K,Mori H, et al. A Single-Nucleotide Polymorphism in an Endo-1,4-\u0026beta;-Glucanase Gene Controls Seed Coat Permeability in Soybean.Plos One.2015;10(6):1-19.https://doi.org/10.1371/journal.pone.0128527.\u003c/li\u003e\n\u003cli\u003eChai M, Zhou C, Molina I, Fu C, Nakashima J, Li G, et al. A class II KNOX gene, \u003cem\u003eKNOX4\u003c/em\u003e , controls seed physical dormancy. Proc Natl Acad Sci U S A.2016;113(25):6997-7002 http://dx.doi.org/10.1073/pnas.1601256113.\u003c/li\u003e\n\u003cli\u003eSoltani A,Walter KA,Wiersma AT,Santiago JP,Lowry DB. The genetics and physiology of seed dormancy, a crucial trait in common bean domestication. BMC Plant Biology.2021;21(1):58. https://doi.org/10.1186/s12870-021-02837-6.\u003c/li\u003e\n\u003cli\u003eZhang B, Horvath S. A general framework for weighted gene co-expression network analysis. .Statistical Applications in Genetics \u0026amp; Molecular Biology.2005;4(1):1-10. https://doi.org/10.2202/1544-6115.1128.\u003c/li\u003e\n\u003cli\u003eFicklin S, Luo F, Feltus F. The association of multiple interacting genes with specific phenotypes in rice using gene coexpression networks.Plant Physiology.2010;154(1):13-24.https://doi.org/10.1104/pp.110.159459.\u003c/li\u003e\n\u003cli\u003eWang C-C, Lin J-Y, Bu Y-P, Sun R-D, Lu Y, Gai J-Y, Xing H, Guo Na, Zhao J-M. Genome-wide transcriptome analysis reveals key regulatory networks and genes involved in the determination of seed hardness in vegetable soybean. Horticulture Research.2024;11(5):1-10.https://doi.org/10.1093/hr/uhae084.\u003c/li\u003e\n\u003cli\u003eGerhard K. Cuticular water permeability and its physiological significance.Journal of Experimental Botany.1996;12:1813-1832.https://doi.org/10.1093/jxb/47.12.1813.\u003c/li\u003e\n\u003cli\u003eBaucher M, Monties B, Montagu M, Boerjan W. Biosynthesis and Genetic Engineering of Lignin. Critical Reviews in Plant Sciences.1998;17(2):125-197.https://doi.org/10.1080/07352689891304203.\u003c/li\u003e\n\u003cli\u003eRolston MP.Water impermeable seed dormancy.Botanical Review.1978;44(3):365-396. https://doi.org/10.1007/bf02957854.\u003c/li\u003e\n\u003cli\u003eSm\u0026yacute;kal P, Vernoud V, Blair MW, Ale\u0026scaron; S,Thompson RD. The role of the testa during development and in establishment of dormancy of the legume seed.Frontiers in Plant Science.2014;5(351):1-44.https://doi.org/10.3389/fpls.2014.00351.\u003c/li\u003e\n\u003cli\u003eHaughn G, Chaudhury A. Genetic analysis of seed coat development in Arabidopsis. Trends in Plant Science.2005;10(10):472-477. https://doi.org/10.1016/j.tplants.2005.08.005.\u003c/li\u003e\n\u003cli\u003eSasaki S, Nishida T, Tsutsumi Y, Kondo R. Lignin dehydrogenative polymerization mechanism: a poplar cell wall peroxidase directly oxidizes polymer lignin and produces in vitro dehydrogenative polymer rich in \u0026beta;-O-4 linkage.FEBS Lett.2004;562(1-3). https://doi.org/10.1016/S0014-5793(04)00224-8.\u003c/li\u003e\n\u003cli\u003eTahara K, Norisada M, Hogetsu T, Kojima K. Aluminum tolerance and aluminum-induced deposition of callose and lignin in the root tips of Melaleuca and Eucalyptus species. Journal of Forest Research.2005;10(4):325-333.https://doi.org/10.1007/s10310-005-0153-z.\u003c/li\u003e\n\u003cli\u003eHorst W, Wang Y, Eticha D. The role of the root apoplast in aluminium-induced inhibition of root elongation and in aluminium resistance of plants:a review.Annals of Botany.2010; 106(1):185. https://doi.org/10.1093/aob/mcq053.\u003c/li\u003e\n\u003cli\u003eBuckeridge SM.Seed cell wall storage polysaccharides:models to understand cell wall biosynthesis and degradation.Plant Physiology.2010;154(3):1017-1023. https://doi.org/10.1104/pp.110.158642.\u003c/li\u003e\n\u003cli\u003eVanholme R, Demedts B, Morreel K, Ralph J, Boerjan W. Lignin biosynthesis and structure. Plant Physiology.2010;153(3): 895-905. https://doi.org/10.1104/pp.110.155119.\u003c/li\u003e\n\u003cli\u003eYe Q, Zhu W, Li L, Zhang S,Yin Y,Ma H,Chory WJ. Brassinosteroids control male fertility by regulating the expression of key genes involved inArabidopsisanther and pollen development. Proceedings of the National Academy of Sciences of the United States of America.2010;107(13):6100-6105. https://doi.org/10.2307/25665121.\u003c/li\u003e\n\u003cli\u003eChantarachot T, Bailey SJ. Polysomes, stress granules, and processing bodies: a dynamic triumvirate controlling cytoplasmic mRNA fate and function. Plant Physiology.2018;176(1): 254-269. https://doi.org/10.1104/pp.17.01468.\u003c/li\u003e\n\u003cli\u003ePourcel L, Routaboul JM, Cheynier V, Lepiniec L,Debeaujon I. Flavonoid oxidation in plants: from biochemical properties to physiological functions. Trends in Plant Science.2007; 12(1):29-36. https://doi.org/10.1016/j.tplants.2006.11.006.\u003c/li\u003e\n\u003cli\u003eMangeon A, Junqueira R, Sachetto MG. Functional diversity of the plant glycine-rich proteins superfamily. Plant Signaling \u0026amp; Behavior.2010;5(2):99-104 https://doi.org/10.4161/psb.5.2.10336.\u003c/li\u003e\n\u003cli\u003eChen L,Wang H,Xu T-T,Liu R-T, Zhu J-L,Li H-X,Zhang HW, Tang L-Y,et al. A telomere-to-telomere gap-free assembly integrating multi-omics uncovers the genetic mechanism of fruit quality and important agronomic trait associations in pomegranate.Plant Biotechnology Journal. Advance online publication. https://doi.org/10.1111/pbi.70107.\u003c/li\u003e\n\u003cli\u003eGao X, Yin C, Liu X, Peng J, Chen D, He D, et al. A glycine-rich protein MoGrp1 functions as a novel splicing factor to regulate fungal virulence and growth in Magnaporthe oryzae. Phytopathology Research.2019;1(1):2-15.https://doi.org/10.1186/s42483-018-0007-1.\u003c/li\u003e\n\u003cli\u003eKallberg Y, Oppermann U, J\u0026ouml;rnvall H, Persson B. Short-chain dehydrogenase/reductase (SDRs). European Journal of Biochemistry.2002;269(18):4409-4417.https://doi.org/10.1046/j.1432-1033.2002.03130.x.\u003c/li\u003e\n\u003cli\u003ePriego‐Cubero S, Liu Y, Toyomasu T, Gigl M, Hasegawa Y, Nojiri H, et al. Evolution and diversification of the momilactone biosynthetic gene cluster in the genus Oryza.New Phytologist.2025;245(6):2681-2697.https://doi.org/10.1111/nph.20416.\u003c/li\u003e\n\u003cli\u003eCheng W-H. A unique short-chain dehydrogenase/reductase in arabidopsis glucose signaling and abscisic acid biosynthesis and functions.The Plant Cell.2002;14(11):2723-2743. https://doi.org/doi:10.1105/tpc.006494.\u003c/li\u003e\n\u003cli\u003eCancel DJ. Loss-of-function mutations in the ethylene receptoretr1 cause enhanced sensitivity and exaggerated response to ethylene in arabidopsis.Plant Physiology.2002;129(4):1557. https://doi.org/10.1104/pp.003780.\u003c/li\u003e\n\u003cli\u003eHua J, Meyerowitz EM.Ethylene responses are negatively regulated by a receptor gene family in arabidopsis thaliana.Cell.1998;94(2):261-271.https://doi.org/10.1016/S0092-8674(00)81425-7.\u003c/li\u003e\n\u003cli\u003eM\u0026uuml;ller M, Munn\u0026eacute;-Bosch S. Ethylene response factors: a key regulatory hub in hormone and stress signaling.Plant physiology.2015;169(1):32-41.https://doi.org/10.1104/pp.15.00677.\u003c/li\u003e\n\u003cli\u003eLiu M, Yao C, Chen Y, Shin J, Mila I, Audran C, Zouine M, Pirrello J, Bouzayen M. The tomato ethylene response factor Sl‐ERF.B3 integrates ethylene and auxin signaling via direct regulation of Sl‐Aux/IAA27. The New phytologist.2018;219(2):631-640. https://doi.org/10.1111/nph.15165.\u003c/li\u003e\n\u003cli\u003eLicausi F, Ohme TM, Perata P. APETALA2/Ethylene responsive factor (AP2/ERF) transcription factors: mediators of stress responses and developmental programs. The New phytologist .2013; 199 (3): 639-649.https://doi.org/10.1111/nph.12291.\u003c/li\u003e\n\u003cli\u003eGanar K,Turbina P,Nandy M,Chen C,Suylen D,Beelen S,Pascoe E,Koenraadt C,Dijkgraaf I, Deshpande S.Phase separation and ageing of glycine-rich protein from tick adhesive.Nature Chemistry.2025;17(2):186-197.https://doi.org/10.1038/s41557-024-01686-8.\u003c/li\u003e\n\u003cli\u003eWang Y, Wang M-H, Yan X, Chen K-X, Tian F-H, Yang X, Cao L-Y, et al. The DEP1 mutation improves stem lodging resistance and biomass saccharification by affecting cell wall biosynthesis in rice. Rice (New York, N.Y.). 2024;17(1):1-13. https://doi.org/10.1186/s12284-024-00712-0.\u003c/li\u003e\n\u003cli\u003eSowadan O, Xu S, Li Y, Muleke EM, Sitoe HM, Dang X, et al. Genome-wide association analysis unravels new quantitative trait loci (qtls) for eight lodging resistance constituent traits in Rice (Oryza sativa L.). Genes.2024;15(1):2-18.https://doi.org/10.3390/genes15010105.\u003c/li\u003e\n\u003cli\u003eWang J, Sun Q, Ma C, Wei M, Wang C, Zhao Y, et al. MdWRKY31-MdNAC7 regulatory network: orchestrating fruit softening by modulating cell wall-modifying enzyme MdXTH2 in response to ethylene signalling. Plant Biotechnology Journal.2024;22(12):3244-3261. https://doi.org/10.1111/pbi.14445.\u003c/li\u003e\n\u003cli\u003eSakai H, Hua J, Chen Q-G, Chang C, Medrano LJ, Bleecker AB, Meyerowitz EM. ETR2 is an ETR1-like gene involved in ethylene signaling in Arabidopsis. Proceedings of the National Academy of Sciences of the United States of America.1998;95(10):5812-5817 https://doi.org/10.1073/pnas.95.10.5812.\u003c/li\u003e\n\u003cli\u003eKato T. Metaphase chromosome preparation and classification of chromosomal aberrations. Methods in molecular biology (Clifton, N.J.).2023;2519:9-15.https://doi.org/10.1007/978-1-0716-2433-3_2.\u003c/li\u003e\n\u003cli\u003eSun Q, Zhu L-W,Zhang W-J,Wang J-H. Physical and chemical difference of seed coat between hard and soft seeds of Licorice (Glycyrrhiza Uralensis Fisch). Legume Research.2018; 41:441-446.https://doi.org/10.18805/LR-376.\u003c/li\u003e\n\u003cli\u003eChen X-Y, Xia Y-Y, MaoY-F, HuangY, Zhu J,Xu J,Zhu R,Shi L,WuW-G. A programmable nanofabrication method for complex 3d meta-atom array based on focused-ion-beam stress-induced deformation effect.Micromachines. 2020;11(1):95. https://doi.org/10.3390/mi11010095.\u003c/li\u003e\n\u003cli\u003eBolger A,Marc L,Bjoern U.Trimmomatic:a flexible trimmer for Illumina sequence data. Bioinformatics.2014;30(15):2114-2120.https://doi.org/10.1093/bioinformatics/btu170.\u003c/li\u003e\n\u003cli\u003eAltschul SF,Gish W,Miller W, Myers EW,Lipman DJ.Basic local alignment search tool (BLAST). Journal of molecular biology.1990;215(3):403-410.https://doi.org/10.1016/S0022-2836(05)80360-2.\u003c/li\u003e\n\u003cli\u003eGrabherr M, Haas B, Yassour M, Levin J, Thompson D, Amit I, et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nature biotechnology.2011;29(7): 644-652.https://doi.org/10.1038/nbt.1883.\u003c/li\u003e\n\u003cli\u003eXu J, Yang J, Xu Z, Zhao D, Hu X. Exogenous spermine-induced expression of SlSPMS gene improves salinity\u0026ndash;alkalinity stress tolerance by regulating the antioxidant enzyme system and ion homeostasis in tomato. Plant physiology and biochemistry.2020;157:79-92. https://doi.org/10.1016/j.plaphy.2020.09.033.\u003c/li\u003e\n\u003cli\u003eTrapnell C, Williams B, Pertea G, Mortazavi A, Kwan G, Baren M, et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nature biotechnology. 2010;28(5):511-515.https://doi.org/10.1038/nbt.1621.\u003c/li\u003e\n\u003cli\u003eLangfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics.2008;9:559.https://doi.org/10.1186/1471-2105-9-559.\u003c/li\u003e\n\u003cli\u003eHuang Y, Qin M, Lai J, Liang J, Luo X, Li C. Assessing obt formation and enrichment: ROS signaling is involved in the radiation hormesis induced by low- dose tritium exposure in algae.Journal of Hazardous Materials.2023;443:130159.https://doi.org/10.2139/ssrn.4179079.\u003c/li\u003e\n\u003cli\u003eShannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T.Cytoscape:A software environment for integrated models of biomolecular interaction networks. Genome Research.2003;13(11):2498-2504.https://doi.org/10.1101/gr.1239303.\u003c/li\u003e\n\u003cli\u003eLivak K, Schmittgen T. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods. 2001;25(4):402-408. https://doi.org/10.1006/meth.2001.1262.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-genomics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"gics","sideBox":"Learn more about [BMC Genomics](http://bmcgenomics.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/gics","title":"BMC Genomics","twitterHandle":"#BMCGenomics","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Vicia sativa L, Hard-seededness, Cuticula, RNA-seq, Ethylene response","lastPublishedDoi":"10.21203/rs.3.rs-6974115/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6974115/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eWild-type \u003cem\u003eVicia sativa\u003c/em\u003e L. demonstrates superior agronomic traits, including high yield, elevated crude protein content, enhanced reproductive efficiency, and tolerance to nutrient-poor soils, rendering it a valuable genetic resource for crop improvement. However, its utilization in germplasm enhancement is severely constrained by seed hardness, whose underlying physiological and molecular regulatory mechanisms remain poorly characterized.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003ePhenomics studies reveal that wild-type seeds exhibited significantly higher hardness indices and lignin content compared to cultivated varieties, whereas cultivated seeds contain significantly greater levels of soluble sugars, soluble proteins, and starch. During the same developmental period, cultivated seeds demonstrate significantly larger diameters than wild-type seeds. However, wild-type seeds show significantly thicker palisade layers and higher palisade-layer-to-seed-coat thickness ratios. Notably, the cuticle thickness of wild-type seeds is significantly greater. Scanning electron microscopy further indicated that cultivated varieties have significantly larger hilum width and length, whereas wild-type seeds display significantly deeper micropyle depth. Through WGCNA analysis, three key candidate genes (\u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e) involved in regulating seed hardness were identified. The expression pattern analysis results indicated that, the expression levels of \u003cem\u003eTRINITY_DN3402_c0_g1\u003c/em\u003e, \u003cem\u003eTRINITY_DN13607_c0_g1\u003c/em\u003e, and \u003cem\u003eTRINITY_DN6606_c0_g1\u003c/em\u003e in the wild hard-seeded type \u003cem\u003evicia sativa\u003c/em\u003e (W12) were significantly higher than those in the cultivated type \u003cem\u003evicia sativa\u003c/em\u003e (W30).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003e​Our study elucidates key physical determinants and molecular mechanisms underlying seed hardness in \u003cem\u003eVicia sativa\u003c/em\u003e L., providing critical insights for crop improvement.\u003c/p\u003e","manuscriptTitle":"Phenomics, RNA sequencing and weighted gene co-expression network analysis reveals key regulatory networks and genes involved in the determination of seed hardness in Vicia sativa","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-03 07:59:39","doi":"10.21203/rs.3.rs-6974115/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-21T10:15:23+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-19T02:35:48+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-15T15:11:58+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-15T02:53:32+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"73386312785402326422216243772182883502","date":"2025-07-14T10:01:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"127882804863836532833332284609774303920","date":"2025-07-14T02:04:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"235483274596693647226764303714084329553","date":"2025-07-14T01:11:58+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-02T01:52:17+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-02T01:50:40+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-06-30T12:12:20+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-29T17:40:59+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Genomics","date":"2025-06-29T17:36:28+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-genomics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"gics","sideBox":"Learn more about [BMC Genomics](http://bmcgenomics.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/gics","title":"BMC Genomics","twitterHandle":"#BMCGenomics","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"70d9a9f0-ad06-4c5c-8165-02f4970724e8","owner":[],"postedDate":"July 3rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-10-27T16:26:35+00:00","versionOfRecord":{"articleIdentity":"rs-6974115","link":"https://doi.org/10.1186/s12864-025-12138-z","journal":{"identity":"bmc-genomics","isVorOnly":false,"title":"BMC Genomics"},"publishedOn":"2025-10-23 16:16:46","publishedOnDateReadable":"October 23rd, 2025"},"versionCreatedAt":"2025-07-03 07:59:39","video":"","vorDoi":"10.1186/s12864-025-12138-z","vorDoiUrl":"https://doi.org/10.1186/s12864-025-12138-z","workflowStages":[]},"version":"v1","identity":"rs-6974115","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6974115","identity":"rs-6974115","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}