The soybean genome and GFF3 annotation files were obtained from the Soybase website. Seven ATCAM and fifty Arabidopsis ATCML proteins from the TAIR database were searched against protein sequences from the soybean genome database using BLASTP (E value < 1E-6). Additionally, hidden Markov model (HMM) profiles of the EF-hand domain (PF13499, PF13405, PF13202, PF13833) were downloaded from the PFAM database and utilized for HMM analysis with the TBtools tool. The results of BLASTP and HMM were compared and manually analyzed. The CAM/CML proteins were further evaluated using SMART (http://smart.embl-heidelberg.de/) and Interpro (http://www.ebi.ac.uk/interpro/), confirming the presence and integrity of the EF-hand domain without any other domains. The molecular weight (Mw) and isoelectric point (pI) of the GmCAM/CML proteins were predicted using the Protein Parameter Calc tool in TBtools.

The chromosomal positions of CAM/CML genes were established by downloading the GFF file for the soybean genome and the CAM/CML family members from the Phytozome website using TBtools (Chen et al., 2023). Collinearity analysis of the CAM/CML genes was carried out with the one-step MCScanX plug-in of TBtools, and the results were visualized using the Advanced Circos plug-in (Chen et al., 2023). The Ka/Ks calculator in TBtools was utilized to compute nonsynonymous and synonymous substitution rates (Chen et al., 2023). The conserved GmCAM/CML protein motifs were predicted using the MEME program with default parameters, with a maximum of 5 motifs.

The Arabidopsis CAM/CML protein sequences (7 ATCAMS, 50 ATCMLS) were downloaded from the TAIR website. Multiple sequence alignments were conducted using the ClustalW tool. A rooted neighbor-joining phylogenetic (NJ) tree was constructed with Mega 11 software based on the full-length CAM/CML protein sequences to study the evolutionary relationships among CAM/CML proteins. The collinearity of CAM/CML genes was analyzed using the multiple scanning tool package MCSCANX, and the collinearity between homologous proteins in the Arabidopsis and rice genomes was examined. The results were visualized with TB-Tools. The KA/KS ratios among GmCAM/CML members were calculated using Ka/Ks_calculator2.0 software.

We downloaded the promoter sequences (approximately 2 kb upstream of the transcription start sites) of soybean GmCAM/CML family genes from Phytozome, then used the PlantCARE online website (http://bioinformatics.psbugent.be/webtools/plantcare/html/) to predict and analyze promoter elements in these genes, remove the core elements (TATA-box and CAAT-box) of the promoter, and visualize them using TBtools (Chen et al., 2023).

Five resistant soybean varieties (24JD21, 24JD210, 24JD829, 24JD890, 24JD697) and five susceptible (24JD892, 24JD878, 24JD876, 24JD882, 24JD867) were selected for biotic stress treatments based on their response to C. sojina infection. The C. sojina isolate (strain CS-2018) was obtained from the National Soybean Improvement Center in Nanjing, China, and cultured on potato dextrose agar (PDA) medium at 25°C for 14 days. Spores were harvested by washing the plates with sterile water containing 0.05% Tween-20, and the concentration was adjusted to 5 × 10⁵ spores/ml. Soybean plants were grown in a greenhouse under controlled conditions (25°C day/22°C night, 16-hour photoperiod, 70% relative humidity). Plants at the V3 stage (with the third trifoliate leaf fully developed) were used for inoculation. For C. sojina inoculation, the spore suspension was sprayed onto the plants until the runoff. The inoculated plants were kept in a humidity chamber for 48 hours and then returned to normal greenhouse conditions. Disease severity was assessed 14 days after inoculation using a 0–9 scale, where 0 = no lesions and 9 = severe infection with coalescing lesions covering more than 50% of the leaf area. Varieties with average scores of 0–3 were classified as resistant, 4–6 as moderately susceptible, and 7–9 as highly susceptible ( Supplementary Table S1 ). For SMV response data, we utilized published transcriptome data from the comparative transcriptome analysis of early resistance of Glycine max to soybean mosaic virus (Li et al., 2023).

Total RNA was extracted using the RNA Extraction Kit from CWBIO (Jiangsu, China). This total RNA was reverse-transcribed to generate first-strand cDNA with the HiScript III RT SuperMix for qPCR (+gDNA wiper) provided by Vazyme (Beijing, China). Quantitative real-time PCR was conducted using a Roche instrument with ChamQ SYBR qPCR Master Mix from Vazyme (Beijing, China). The internal reference gene utilized was soybean β-Tubulin, and the relative expression levels of the target gene were calculated with the 2^-ΔΔCT method. In the expression profile analysis, significance was evaluated via the Student t-test, with levels of significance indicated by asterisks (*P < 0.05, **P < 0.01, and *** P < 0.001). The sequences of primers used in this study are listed in Supplementary Table S2 . For the C. sojina infection time course, leaf samples were gathered 14 days after inoculation. Each variety had three biological replicates, with each replicate comprising the third trifoliate leaves from three plants. The collected leaf tissues were promptly frozen in liquid nitrogen and stored at -80°C until RNA extraction.

To identify CAM/CML genes in soybean, we employed a dual screening approach. First, we conducted BLASTP searches against the soybean genome using protein sequences of seven ATCAMs and fifty ATCMLs from Arabidopsis as queries, with an e-value threshold of 1 × 10^-6. Concurrently, we performed hidden Markov model (HMM) analyses by aligning AMP domain files (PF13499, PF13405, PF13202, PF13833) with all soybean amino acid sequences. Candidate genes identified through these methods were further verified using multiple domain databases including PFAM, SMART, and NCBI’s Conserved Domain Database (CDD). This comprehensive screening approach resulted in the identification of 11 GmCAM and 102 GmCML genes. These genes were systematically named according to their chromosomal positions (GmCAM1-GmCAM11 and GmCML1-GmCML102). Subsequent characterization of the physical and chemical properties of these CAM/CML family members revealed considerable diversity. The encoded proteins ranged in length from 74 to 371 amino acids, with theoretical molecular weights between 8622.56 and 43574.75 kDa. The isoelectric points (pI) of these proteins exhibited a broad distribution from 3.94 to 9.99, indicating diverse charge properties within this gene family. Detailed information for individual family members is provided in Supplementary Table S3 .

To investigate the evolutionary relationships among CAM/CML proteins, we constructed a neighbor-joining phylogenetic tree using maximum likelihood methods with 1,000 bootstrap replicates. The analysis included 57 Arabidopsis proteins (7 AtCAMs and 50 AtCMLs) and 113 soybean proteins (11 GmCAMs and 102 GmCMLs) ( Figure 1 ). The phylogenetic analysis resolved the CAM/CML superfamily into 14 distinct monophyletic groups with strong bootstrap support (>70%). Group I emerged as the largest clade, containing 36 members (8 AtCMLs and 28 GmCMLs), suggesting an ancient expansion of CML proteins predating the Arabidopsis-soybean divergence approximately 54 million years ago. Groups II and III collectively harbored all CAM proteins, with Group II containing 15 members (3 AtCMLs, 9 GmCMLs, and 3 GmCAMs) and Group III containing 21 members (4 AtCMLs, 7 AtCAMs, 8 GmCAMs, and 2 GmCMLs). The exclusive clustering of CAM proteins within these two groups indicates their evolutionary origin from a common ancestral lineage and subsequent conservation of core calcium-sensing functions. The remaining groups (IV-XIV) comprised exclusively CML proteins, with variable sizes ranging from 2 to 22 members. Notably, Groups IV and VII represented the smallest clades, each containing only 2 members. Group IV exclusively comprises GmCML proteins (GmCML_65 and GmCML_69), while Group VII contained only AtCML proteins (AtCML_23 and AtCML_24). This species-specific clustering pattern suggests recent gene duplication events following the Arabidopsis-soybean divergence, potentially indicating lineage-specific functional adaptations. The phylogenetic distribution reveals several key evolutionary patterns. First, the extensive diversification of CML proteins (152 total) compared to CAM proteins (18 total) demonstrates significant sub functionalization following ancient duplication events. Second, the mixed species composition in most groups indicates that major gene family expansions occurred before monocot-dicot divergence. Third, the isolated positioning of the GmCML pair in Group IV, separated by long branch lengths from other clades, suggests potential neofunctionalization and acquisition of specialized calcium-sensing roles unique to legumes. Bootstrap values exceeding 85% for major nodes confirm the robustness of these phylogenetic relationships. The clear separation between CAM and CML lineages, combined with species-specific clustering patterns in smaller groups, provides a framework for understanding the evolutionary forces shaping calcium signaling diversity in flowering plants.

A comprehensive motif analysis of soybean CAM/CML genes was performed using the MEME suite to elucidate conserved structural and functional domains within this important calcium-binding protein family. Position weight matrix analysis revealed distinct amino acid conservation patterns within each motif ( Figure 2a ). The analysis identified five distinct conserved motifs across 15 CAM/CML genes from the Williams 82 reference genome (Wm82.gnm4.ann1), with statistical significance values ranging from 1.93 × 10^-40 to 1.27 × 10^-94 ( Figure 2c ). Motif distribution analysis revealed differential patterns of conservation across the gene family ( Figure 2b ). Motifs 1 and 2 demonstrated near-universal presence, occurring in 113 and 112 genes respectively, indicating their fundamental importance in CAM/CML protein structure and function. Motif 4 was present in 73 genes, while Motif 3 occurred in 60 genes. Notably, Motif 5 showed the most restricted distribution, appearing in only 24 genes, suggesting a specialized functional role within a subset of the CAM/CML family. Analysis of motif combinations revealed a modular architecture pattern across the CAM/CML gene family ( Figure 2c ). Individual proteins contained between one and five motifs in various combinations, reflecting the evolutionary plasticity of this gene family. The most statistically significant proteins, including GmCML 8, GmCML10 (both p = 1.27 × 10^-94), and GmCML9 (p = 1.23 × 10^-71), displayed complex motif arrangements with multiple domain repeats. Several distinct architectural patterns emerged: (1) proteins containing primarily Motifs 1 and 2, representing the core CAM/CML structure; (2) proteins with additional Motifs 3 and/or 4, suggesting expanded functional capacity; and (3) a specialized subset containing Motif 5, potentially representing functionally divergent family members. This modular organization indicates that CAM/CML proteins have evolved through domain duplication and recombination events, allowing for functional diversification while maintaining essential calcium-binding capabilities.

The exceptionally high statistical significance of motif conservation (all p-values < 10^-30) demonstrates strong evolutionary pressure to maintain these structural elements, consistent with their critical role in calcium-mediated cellular processes. The universal presence of Motifs 1 and 2 across nearly all analyzed genes indicates these represent the essential functional core of CAM/CML proteins, likely corresponding to canonical EF-hand calcium-binding domains. The restricted distribution of Motif 5 to a subset of genes suggests functional sub familiarization within the CAM/CML family. Proteins containing this motif may possess specialized roles in specific calcium signaling pathways or exhibit distinct regulatory mechanisms. The intermediate distribution patterns of Motifs 3 and 4 indicate additional layers of functional complexity, potentially conferring differential calcium sensitivity, protein-protein interaction capabilities, or subcellular localization signals. The identification of 15 CAM/CML genes with conserved motifs across multiple chromosomes indicates significant gene family expansion in soybean. The modular nature of motif combinations suggests that this expansion occurred through various mechanisms including tandem duplication, segmental duplication, and subsequent domain shuffling events.

Gene structure analysis of the soybean CAM/CML family revealed distinct organizational patterns across 11 GmCAM and 102 GmCML genes ( Figure 3 ). CAM genes exhibited relatively uniform structures with compact organization, while CML genes showed remarkable diversity, ranging from simple architectures to complex, multi-domain organizations extending over 8,000 base pairs. The five conserved motifs identified through MEME analysis were mapped to their genomic locations within gene structures. Motifs 1 and 2 were consistently located within coding sequences across all genes, confirming their essential functional roles. Motif 5 was predominantly found in CAM subfamily members, supporting functional specialization between CAM/CML subfamilies. Domain annotation revealed multiple EF-hand domain subtypes across the family, including EFh_PEF_Group_I, EFh_SPARC_EC superfamily, and EFh_CREC superfamily domains ( Figure 3 ). PTZ00184 and PTZ00183 superfamily domains were identified in multiple members, suggesting additional functional capabilities beyond calcium binding. The FRQ1 domain was detected in a subset of genes, potentially conferring specialized regulatory functions. Most genes contained multiple EF-hand domains arranged in tandem, consistent with the typical architecture of calcium-binding proteins, which require multiple coordination sites. The modular arrangement supports evolution through domain duplication and recombination mechanisms. CAM genes demonstrate uniform structures with consistent domain arrangements, reflecting their conserved role as primary calcium sensors. The specific presence of Motif 5 in CAM genes provides a molecular basis for their unique regulatory properties. CML genes exhibited greater structural diversity with varying EF-hand domain combinations and accessory motifs, correlating with their functional specialization for distinct cellular processes. The structural diversity observed across the CAM/CML family demonstrates the evolutionary mechanisms that have shaped functional diversification, with CAM genes under stronger purifying selection while CML genes have undergone extensive structural innovation to generate the current functional versatility of plant calcium signaling networks.

A genome-wide analysis revealed the presence of 11 CAM and 102 CML genes distributed across all 20 chromosomes of the soybean genome. The distribution pattern exhibited significant variation, with chromosomes Gm11, Gm17, and Gm04 containing the highest number of CML genes, while the CAM genes were confined to only 7 chromosomes (Gm02, Gm03, Gm05, Gm13, Gm14, Gm19, and Gm20). Notably, we identified multiple clustering patterns of CML genes in specific chromosomal regions, suggesting potential tandem duplication events during evolution. The substantial expansion of the CML gene family (102 members) compared to the more conserved CAM family (11 members) suggests functional diversification following the whole-genome duplication events known to have occurred in soybean. This comprehensive chromosomal mapping provides important insights into the evolutionary history and potential functional specialization of calcium-signaling genes in Glycine max ( Figure 4 ).

Comprehensive analysis of the 2 kb upstream sequences of GmCaM/CML genes revealed a striking enrichment of hormone-responsive cis-acting elements ( Figure 5a ). Abscisic acid (ABA) responsive elements constituted the largest proportion (254 elements, 31.2%), followed closely by methyl jasmonate (MeJA) responsive elements (225 elements, 27.7%). Gibberellin-responsive elements (93 elements, 11.4%) and auxin-responsive elements (64 elements, 7.9%) were also well-represented. Elements associated with biotic and abiotic stress responses comprised a notable fraction, with defense/stress-responsive elements and salicylic acid-responsive elements accounting for 6.5% and 6.3%, respectively. MYBHv1 binding sites (5.3%) and meristem expression elements (3.4%) were less abundant but present across multiple genes. Hierarchical clustering analysis revealed distinct cis-regulatory patterns among GmCaM/CML family members ( Figure 5b ), suggesting potential functional specialization. Several genes exhibited strong enrichment for specific hormone-responsive elements, while others displayed more diverse regulatory signatures. This regulatory diversity likely underpins the multifaceted roles of calmodulin and calmodulin-like proteins in calcium-mediated signaling networks, particularly in hormone signaling pathways and stress responses. The prevalence of hormone-responsive elements highlights the pivotal role of GmCaM/CML proteins as integration nodes between calcium signaling and hormone-regulated processes in soybean.

Collinearity analysis revealed substantial duplication events contributing to the expansion of the CaM/CML gene family in soybean ( Figure 6a ). We identified 117 pairs of homologous GmCaM/CML genes distributed across the soybean genome, visualized as connecting lines in the circular diagram. This extensive duplication pattern likely resulted from the two whole-genome duplication events in soybean’s evolutionary history. Comparative genomic analysis between soybean and model plant species uncovered significant evolutionary relationships in the CaM/CML gene family ( Figure 6b ). We identified 67 pairs of homologous CaM/CML genes between soybean and Arabidopsis, and 32 pairs between soybean and rice. This differential homology pattern reflects the closer evolutionary relationship between soybean and Arabidopsis (both dicots) compared to rice (monocot). To assess evolutionary selection pressure, we calculated Ka/Ks ratios for all homologous gene pairs. Notably, all gene pairs exhibited Ka/Ks ratios below 1.0, indicating strong purifying selection ( Supplementary Table S4 ). This evolutionary constraint suggests functional conservation of CaM/CML proteins across species, highlighting their fundamental importance in calcium-mediated signaling processes throughout plant evolution. The greater number of homologous relationships between soybean and Arabidopsis compared to rice aligns with established phylogenetic relationships between these species and suggests lineage-specific expansion and conservation patterns in the plant CaM/CML gene family.

Analysis of RNA-Seq data across soybean tissues revealed distinct spatial expression patterns among GmCaM/CML family members ( Supplementary Figure S1 ). Hierarchical clustering identified two major expression clusters: one predominantly showing low expression across most tissues (upper cluster including GmCML15 through GmCAM9), and another exhibiting moderate to high expression in various tissue combinations (lower cluster including GmCML19 through GmCML81) ( Figure 7 ). Several genes displayed striking tissue-specific expression profiles. GmCML81 demonstrated robust expression across vegetative tissues but minimal activity in seeds. GmCML83 showed high expression across multiple tissues, particularly in roots. Conversely, GmCML22 exhibited preferential expression in flowers and roots, while GmCML94 showed enhanced expression in leaves and flowers, suggesting specialized functions in these tissues. Notably, while phylogenetically related genes often shared similar expression patterns, we observed several instances of expression divergence among duplicated genes. For example, GmCML83 maintained high expression across all tissues examined, whereas its homolog GmCML94 was expressed at substantially lower levels. This expression divergence following gene duplication suggests neo- or sub-functionalization events contributing to the functional diversification of the GmCaM/CML gene family in soybean.

Transcriptome analysis of 19 CML genes in Glycine max revealed distinct expression patterns correlated with resistance to C. sojina infection ( Figure 8 ). Expression profiles clustered into three categories across the five resistant and five susceptible varieties: high-magnitude (GmCML32, GmCML65, GmCML77), moderate (GmCML7, GmCML15, GmCML66, GmCML83, GmCML94), and low-variable (GmCML19, GmCML22, GmCML23, GmCML41, GmCML58, GmCML69). GmCML77 exhibited the highest expression amplitude (10 relative units) in resistant varieties, particularly in samples 24JD210, 24JD697, and 24JD890. Interestingly, GmCML32 and GmCML65 exhibited substantial dynamic ranges (10 and 7.5 units, respectively), with significant upregulation in resistant varieties following infection. Conversely, GmCML23 maintained a relatively stable expression (± 2 units) regardless of the resistance phenotype, suggesting a constitutive rather than an inducible role. Temporal expression patterns following C. sojina challenge fell into three categories: progressive increase (GmCML15, GmCML32), bell-shaped (GmCAM9, GmCML27, GmCML30), and oscillatory (GmCML22, GmCML69, GmCML77). The resistant varieties showed distinctive early upregulation of GmCML94 and GmCML99, while late-stage infection was characterized by elevated expression of GmCML7, GmCML32, and GmCML65 in these same varieties ( Figure 8 ). Five genes (GmCML19, GmCML22, GmCML41, GmCML58, GmCML69) demonstrated negative expression values in susceptible varieties relative to resistant ones, indicating potential suppression during pathogen infection. This was most pronounced in GmCML22, which exhibited the widest expression range (-5 to +5 units) between resistant and susceptible varieties. Notably, these repression patterns were primarily associated with early infection stages in susceptible varieties, suggesting pathogen-induced transcriptional interference. Sample-specific expression analysis revealed differential responses among resistant varieties (24JD210, 24JD697, and 24JD890) compared to susceptible varieties (24JD829, 24JD882, and 24JD892) ( Figure 8 ). This divergence suggests functional specialization among CML family members may contribute to C. sojina resistance. Particularly, the coordinated expression of GmCML77, GmCML81, and GmCML83 in resistant varieties points to a potential calcium-signaling cascade essential for mounting effective defense responses against this economically significant pathogen.

Our comprehensive analysis identified 113 CAM/CML genes (11 GmCAMs and 102 GmCMLs) in the soybean genome, representing one of the most prominent CAM/CML families reported in plants. This extensive gene family size compared to Arabidopsis (7 AtCAMs and 50 AtCMLs) (Mccormack et al., 2005), and rice (5 OsCAMs and 32 OsCMLs) (Boonburapong and Buaboocha, 2007), likely reflects the paleopolyploid nature of the soybean genome, which has undergone two whole-genome duplication events approximately 59 and 13 million years ago (Schmutz et al., 2010). The collinearity analysis revealed that most GmCAM/CML genes (117 pairs) resulted from these duplication events, significantly contributing to family expansion. The phylogenetic analysis revealed 14 distinct groups of varying sizes and compositions, indicating different gene retention rates and loss following duplication events. The presence of mixed species composition in most groups suggests that significant diversification of the CAM/CML family occurred before the divergence of monocots and dicots, estimated at 140–150 million years ago (Chaw et al., 2004; Liu et al., 2024). However, the species-specific clustering observed in some groups (e.g., Group IV, which contains only GmCMLs) suggests more recent lineage-specific duplication and potential neofunctionalization. The extremely low Ka/Ks ratios (<1.0) for all homologous gene pairs indicate strong purifying selection pressure, suggesting functional conservation despite extensive duplication. This is consistent with the essential role of calcium signaling in fundamental cellular processes and stress responses (Ranty et al., 2006). Interestingly, the Ka/Ks ratios were generally lower for GmCAM genes compared to GmCML genes, reflecting the higher evolutionary conservation of canonical calmodulins, which are likely to maintain core calcium-sensing functions.

The structural analysis revealed significant diversity in gene architecture and protein features across the GmCAM/CML family. While most GmCAMs (10/11) contain introns, a majority of GmCMLs (71/102, 69.6%) are intron-less. This pattern is consistent with observations in Arabidopsis and rice (Mccormack et al., 2005; Boonburapong and Buaboocha, 2007), suggesting that the intron-less structure of most CMLs may represent an ancestral state or result from retro position events during evolution. The intron-less nature could potentially allow for more rapid expression in response to stress stimuli, as intron splicing can be rate-limiting during transcription (Jeffares et al., 2008). At the protein level, all identified GmCAMs contain four canonical EF-hand domains, consistent with their role as primary calcium sensors. In contrast, GmCMLs exhibit greater diversity in the number of EF-hand domains (1-6) and sequence, potentially reflecting functional diversification. The conservation of specific motifs (particularly Motif 1 present in all members and Motif 5 exclusive to GmCAMs) suggests distinct functional modules within these proteins. The structural divergence between CAMs and CMLs likely underlies their differential calcium-binding properties and target protein interactions, contributing to the specificity and versatility of calcium signaling responses (Gifford et al., 2007).

The striking structural divergence between GmCAMs and GmCMLs provides compelling evidence for neofunctionalization following gene duplication events in the soybean CAM/CML gene family (Yang et al., 2024). The contrasting intron-exon organization patterns, with 90.9% of GmCAMs retaining introns versus 69.6% of GmCMLs being intronless, suggest asymmetric evolutionary trajectories in which duplicated genes acquired distinct functional properties (Glick et al., 2024). The predominant intron-less structure of GmCMLs likely represents a key adaptive innovation that enabled specialization for rapid pathogen responses by eliminating time-consuming splicing processes during transcription (Johnson, 2019). This structural streamlining, coupled with diversification in EF-hand domain numbers (1–6 domains in CMLs vs. canonical 4 in CAMs) and sequence divergence in calcium-binding regions, indicates coordinated neofunctionalization that enabled CML proteins to acquire specialized pathogen-sensing capabilities (Zhao et al., 2013). The evolutionary success of this neofunctionalization is demonstrated by the massive expansion of the CML subfamily (102 genes) compared to the conserved CAM subfamily (11 genes), suggesting that the acquisition of pathogen defense functions provided significant adaptive advantages. The identification of 15 GmCML genes responding to both viral and fungal pathogens further support the hypothesis that neofunctionalization enabled the evolution of broad-spectrum pathogen recognition capabilities, representing a classic example of how gene duplication followed by functional divergence can generate evolutionary novelty in plant defense systems.

The analysis of cis-acting elements in promoter regions revealed extensive enrichment of hormone-responsive elements, particularly those associated with abscisic acid (31.2%) and methyl jasmonate (27.7%) responses. This suggests that GmCAM/CML genes are integrated into hormone signaling networks mediating stress responses, consistent with previous reports in Arabidopsis and rice (Poovaiah et al., 2013). The significant presence of defense and stress-responsive elements (6.5%) and salicylic acid-responsive elements (6.3%) further supports their involvement in biotic stress responses. Interestingly, the dual pathogen-responsive genes identified in our study showed significant enrichment of specific cis-elements compared to the overall family, particularly W-box elements (WRKY binding sites) and TCA elements (salicylic acid-responsive). This enrichment suggests that these genes may be directly regulated by defense-related transcription factors such as WRKYs, which are known to play central roles in plant immunity (Pandey and Somssich, 2009; Wang et al., 2024). The co-occurrence of multiple hormone-responsive elements in these promoters indicates complex hormonal regulation, potentially involving crosstalk between salicylic acid, jasmonic acid, and abscisic acid pathways during pathogen defense.

A major finding of our study is the identification of 19 GmCAM/CML genes responsive to both SMV and C. sojina infection, suggesting their involvement in a common calcium-dependent defense mechanism. This is particularly noteworthy given the distinct nature of these pathogens—SMV is an RNA virus that propagates through the phloem, while C. sojina is a necrotrophic fungal pathogen that directly penetrates and kills host cells. The identification of calcium signaling components responsive to both pathogens suggest a role in broad-spectrum disease resistance. Among these dual-responsive genes, GmCAM4, GmCML23, and GmCML47 showed the strongest correlation with resistance phenotypes across multiple varieties. GmCAM4 has been previously implicated in soybean defense responses (Park et al., 2004), but our study is the first to demonstrate its role in resistance to both viral and fungal pathogens. The early and transient expression pattern of GmCAM4 suggests its involvement in initial signal perception and transmission, consistent with the role of canonical calmodulins in rapid calcium signal decoding (Ranty et al., 2016). The consistent upregulation of GmCML23 and GmCML47 in resistant varieties in response to both pathogens, coupled with their strong correlation with resistance phenotypes, indicates their potential roles as key regulators in soybean immune responses. The sustained expression of GmCML23 throughout the infection time course suggests its involvement in longer-term defense responses, possibly including transcriptional reprogramming and metabolic adjustments. In contrast, the intermediate expression pattern of GmCML47 may reflect its role in the transition from early to late defense responses. The differential expression patterns of these genes between resistant and susceptible varieties could be attributed to several factors. First, resistant varieties may possess specific allelic variants of these genes that confer enhanced expression in response to pathogen infection. Second, regulatory networks controlling the expression of these genes may differ between resistant and susceptible varieties, possibly due to variations in transcription factor activity or chromatin accessibility. Third, the timing and magnitude of calcium signatures induced by pathogen perception may vary between resistant and susceptible varieties, affecting the downstream activation of CAM/CML genes. The potential regulatory relationship among GmCAM4, GmCML23, and GmCML47, as suggested by their sequential activation patterns, warrants further investigation. GmCAM4, being the earliest responder, might directly or indirectly regulate the expression of GmCML47 and GmCML23. Alternatively, these genes might respond to different temporal aspects of calcium signatures induced during pathogen infection.

While we observed that several CAM/CML genes exhibited transcriptional responses to both pathogens, these findings should be interpreted with caution. The apparent similarities or differences in gene expression patterns between the two pathogen treatments cannot be quantitatively assessed due to the lack of standardized experimental controls and uniform methodological approaches. Therefore, any comparative conclusions regarding differential CAM/CML gene regulation in response to SMV versus C. sojina infection remain speculative and require validation through future studies employing consistent experimental designs and standardized conditions for both pathogen treatments.