Identification and Genetic Dissection of Resistance to Red Crown Rot Disease in a Diverse Soybean Germplasm Population

Red crown rot (RCR) disease caused by Calonectria ilicicola negatively impacts soybean yield and quality. Unfortunately, the knowledge of the genetic architecture of RCR resistance in soybeans is limited. In this study, 299 diverse soybean accessions were used to explore their genetic diversity and resistance to RCR, and to mine for candidate genes via emergence rate (ER), survival rate (SR), and disease severity (DS) by a multi-locus random-SNP-effect mixed linear model of GWAS. All accessions had brown necrotic lesions on the primary root, with five genotypes identified as resistant. Nine single-nucleotide polymorphism (SNP) markers were detected to underlie RCR response (ER, SR, and DS). Two SNPs colocalized with at least two traits to form a haplotype block which possessed nine genes. Based on their annotation and the qRT-PCR, three genes, namely Glyma.08G074600, Glyma.08G074700, and Glyma.12G043600, are suggested to modulate soybean resistance to RCR. The findings from this study could serve as the foundation for breeding RCR-tolerant soybean varieties, and the candidate genes could be validated to deepen our understanding of soybean response to RCR.


Introduction
Soybeans [Glycine max (L.) Merrill] are grown worldwide mainly for their high values in oil and protein, yielding over 360 million metric tons in 2020 [1].The need for soybean production continues increasing as it is essential in feeding humans, animals, and the industries for producing biofuel, ethanol, among others [2][3][4].Unfortunately, several factors arising from biotic and abiotic stress are implicated in limiting the potential of soybeans to attain high yield and quality.Pathogens are well noted for causing significant economic losses in soybeans.For instance, in soybean production globally, bacteria, nematodes, viruses, and fungi are noted to cause 3, 5, 9, and 26 diseases, respectively [5].
The main causative agent of soybean red crown rot (RCR), Calonectria ilicicola Boedijn and Reitsma, is a soil-borne pathogen [6], becoming one of the most prevalent soybean diseases.It is alternatively called Calonectria theae Loos var.crotalariae Loos; Calonectria crotalariae (Loos) The results from ANOVA showed that the 299 accessions varied significantly (p < 0.05) in terms of ER, SR, and DS (Table S1).Except for DS, the ER and SR followed a continuous distribution (Figure 1A-C).This implies a broad range of diversity of resistance to RCR in the accessions used.Among the three parameters for soybean RCR, there is a significant (p < 0.05) correlation (Figure 1D).For instance, ER and SR positively correlated with a correlation coefficient (r) of 0.85, while DS negatively correlated with either ER (r = −0.83)or SR (r = −0.89).Our data indicated that DS directly affects the soybean's emergence, survival rate, and, consequently, the crop yield.

Identification of Resistance to Soybean Red Crown Rot Strain
We observed high variability among the soybean panel for C. ilicicola resistance.The genotypes were screened using 0-5 scales for DS with six varied degrees of resistance in this work.None of the 299-soybean panel was identified as immune or highly resistant (DS = 0); however, five genotypes, namely PI 602496, PI 567731, PI 587880A, PI 424412, and PI 407196, were identified as resistant (DS = 1) with their greater ER and SR within 93 to 100% (Table S2).Nine of the materials (PI 547885, PI 567104B, PI 598124, PI 614833, PI 468967, PI 628889, PI 590931, PI 445681, and PI 567312) were identified as highly susceptible by the DS (DS = 5), and their ER and SR were less than 50% and 45%, respectively (Table S2).However, genotypes with DS of ≤2.5, ≤3.5, and >3.5 were classified as resistant, moderately resistant, and susceptible to C. ilicicola, respectively.In summary, out of the 299-soybean panel, 69, 131, and 99 genotypes were identified as resistant, moderately resistant, and susceptible, respectively (Table S2).

Identification of Resistance to Soybean Red Crown Rot Strain
We observed high variability among the soybean panel for C. ilicicola resistance.The genotypes were screened using 0-5 scales for DS with six varied degrees of resistance in this work.None of the 299-soybean panel was identified as immune or highly resistant (DS = 0); however, five genotypes, namely PI 602496, PI 567731, PI 587880A, PI 424412, and PI 407196, were identified as resistant (DS = 1) with their greater ER and SR within 93 to 100% (Table S2).Nine of the materials (PI 547885, PI 567104B, PI 598124, PI 614833, PI 468967, PI 628889, PI 590931, PI 445681, and PI 567312) were identified as highly susceptible by the DS (DS = 5), and their ER and SR were less than 50% and 45%, respectively (Table S2).However, genotypes with DS of ≤2.5, ≤3.5, and >3.5 were classified as resistant, moderately resistant, and susceptible to C. ilicicola, respectively.In summary, out of the 299-soybean panel, 69, 131, and 99 genotypes were identified as resistant, moderately resistant, and susceptible, respectively (Table S2).

SNP Density and Distribution among the 20 Chromosomes of Soybean
SNP density for mapping is documented to affect the power of detection of quantitative trait loci /nucleotides (QTL/Ns) [28].The 299 accessions were genotyped with Illumina Infinium SoySNP50K BeadChip by [29].We conducted quality control checks; thus, SNPs with a minor allele frequency (MAF) >0.05 and a missing rate of 5% were excluded for downstream analysis, leading to a total of 37,876 SNPs across the 20 chromosomes of soybean (Figure 2A).The maximum number of 2899 SNPs was located on chromosome 18 (Chr18), with the lowest (1470) on Chr11 (Figure 2A).In addition, the longest and shortest lengths were located on Chr18 and Chr11, respectively, with varied SNP densities (Figure 2B; Table S3).

SNP Density and Distribution among the 20 Chromosomes of Soybean
SNP density for mapping is documented to affect the power of detection of qu tive trait loci /nucleotides (QTL/Ns) [28].The 299 accessions were genotyped with I Infinium SoySNP50K BeadChip by [29].We conducted quality control checks; thus with a minor allele frequency (MAF) >0.05 and a missing rate of 5% were exclud downstream analysis, leading to a total of 37,876 SNPs across the 20 chromosomes bean (Figure 2A).The maximum number of 2899 SNPs was located on chromoso (Chr18), with the lowest (1470) on Chr11 (Figure 2A).In addition, the longest and s lengths were located on Chr18 and Chr11, respectively, with varied SNP densities ( 2B; Table S3).

Worldwide Soybean Germplasm, Its Population Stratification, Genetic Diversity, and lation Structures Based on Their Origin
Population relatedness has been demonstrated to cause false positives in m trait association (MTA) mapping [30,31].Consequently, it is required to further ass extent of relatedness among the 299-soybean accession.The mapping population w timally grouped into three subpopulations (i.e., I, II, and III) based on model-base

Worldwide Soybean Germplasm, Its Population Stratification, Genetic Diversity, and Population Structures Based on Their Origin
Population relatedness has been demonstrated to cause false positives in marker-trait association (MTA) mapping [30,31].Consequently, it is required to further assess the extent of relatedness among the 299-soybean accession.The mapping population was optimally grouped into three subpopulations (i.e., I, II, and III) based on model-based analysis in ADMIXTURE software version 1.3.0(Figure 3A).The groupings from ADMIX-TURE software were largely similar to those obtained from principal component analysis (PCA) and neighbor-joining tree (Figure 3B,C).Also, the 299 accessions were grouped into three according to how the pairwise kinship coefficients are distributed.The subpopulation I/group I comprised G. max accessions from Asian countries.Meanwhile, the accessions in subpopulation II/group 2 consisted largely of G. soja.In contrast, the subpopulation III/group III largely included G. max accessions from the United States.The first two PCA axes accounted for 17.92% of variability among the 299 accessions in this study.The PCA pinpointed a high genetic diversity among the 299-mapping population.
three according to how the pairwise kinship coefficients are distributed.The subp tion I/group I comprised G. max accessions from Asian countries.Meanwhile, the sions in subpopulation II/group 2 consisted largely of G. soja.In contrast, the subp tion III/group III largely included G. max accessions from the United States.The fi PCA axes accounted for 17.92% of variability among the 299 accessions in this stud PCA pinpointed a high genetic diversity among the 299-mapping population.The minor allele frequency (MAF), expected heterozygosity (He), and polym information content (PIC) on genetic diversity of the PIs based on their origin are in Table 1.Table 1 shows that MAF ranged from 0.14 for the Thia cultivar to 0.26 Russian germplasm.A similar pattern was observed for He, varying from 0.19 to 0 with PIC ranging from 0.16 to 0.27, thus showing a pattern of variation between that is comparable to MAF.In summary, a wide genetic diversity was observed fro tivars from China and Russia, whereas cultivars from Thailand and Japan exhibit rower diversity.The minor allele frequency (MAF), expected heterozygosity (H e ), and polymorphic information content (PIC) on genetic diversity of the PIs based on their origin are shown in Table 1.Table 1 shows that MAF ranged from 0.14 for the Thia cultivar to 0.26 for the Russian germplasm.A similar pattern was observed for H e , varying from 0.19 to 0.35 and with PIC ranging from 0.16 to 0.27, thus showing a pattern of variation between nations that is comparable to MAF.In summary, a wide genetic diversity was observed from cultivars from China and Russia, whereas cultivars from Thailand and Japan exhibited narrower diversity.
Using SNP marker data, principal component analysis (PCA) was applied to evaluate the soybean accessions' population composition based on their country of origin.The first two main components explained 17.92% of the overall genetic variability (Figure 4).The major population structure was revealed by the PCA based on the cultivar's origin.

Marker-Trait Associations (MTAs)
To detect SNPs with both major and minor effects for RCR tolerance/susceptibility, we employed the 3VmrMLM model from the 3VmrMLM package [32] and detected nine SNPs across eight chromosomes (i.e., Chr05, Chr06, Chr07, Chr08, Chr12, Chr14, Chr15, and Chr17) (Table 2).Three SNPs (ss715597632, ss715602602, and ss715625925) were associated with ER on Chr07, Chr08, and Chr17 with a phenotypic variance explained (PVE) of 5.44-7.64%(Table 2; Figure 5A).Of these SNPs, ss715597632 and ss715602602 contributed positively to ER, while ss715625925 reduced ER (Table 3).Using SNP marker data, principal component analysis (PCA) was applied to eval the soybean accessions' population composition based on their country of origin.The two main components explained 17.92% of the overall genetic variability (Figure 4).major population structure was revealed by the PCA based on the cultivar's origin.

Haplotype Analysis for the Identification of Superior Haplotypes and Candidate Genes Mining
To comprehend the phenotypic variances more fully among the 299 soybean accessions carrying a specific haplotype, we conducted haplo-phenotype analysis of the two stable SNPs (ss715612097 and ss715627013 on Chr08 and Chr12, respectively) (Table 4; Figure 6A-C).The haplotype around SNPs ss715612097 and ss715627013 spanned 24 and 18 kb, respectively (Figure 6A,E).Among the two stable SNPs, there were three to four different haplotype alleles underlying each block.For example, the SNP ss715612097 possessed four different alleles (GTCT, GCTG, GCCG, and GTCG), whereas SNP ss715627013 possessed three alleles (CATTA, CGCTG, and AGTG).The effects of haplotype alleles were tested on RCR resistance traits.On SNP ss715612097 (Chr08), the alleles GTCT, GCTG, and GCCG showed significant (p < 0.05) variation in ER, SR, and DS (Figure 6A-D).On the other hand, the haplotype block of ss715627013 (Chr12) divided the 299 soybean accessions into three groups (CATTA, CGCTG, and AGTG) (Figure 6E).Only CATTA and CGCTG haplotype groups showed significant variation (p < 0.05) in terms of SR and DS (Figure 6G,H).
In addition to identifying putative candidate genes around the two stable SNPs, we applied a haplotype block size up-and downstream of the SNPs to mine for putative candidate genes for soybean RCR.From this strategy, three probable genes (Glyma.08G074500,Glyma.08G074600, and Glyma.08G074700) were found in the haplotype block of ss715602602 (Chr08).Glyma.08G074600 is annotated to be involved in plant disease responses in signaling mechanisms involved in the management of fungi (Table 3).Also, Glyma.08G074700 is related to carbohydrate metabolic and xylan catabolic processes (Table 3).Moreover, six putative genes were detected around the SNP ss715612097 (Chr12) (Table 3).Out of these, Glyma.12G043600located 10.9 kb downstream encodes for protein tyrosine kinase which is involved in protein phosphorylation and could be a candidate for regulating the 299 accessions to RCR in soybean (Table 3).Therefore, around the two stable SNPs (ss715602602 (Chr08) and ss715627013 (Chr12), we suggest that Glyma.08G074600,Glyma.08G074700, and Glyma.12G043600 may be involved in modulating soybean RCR response.The determination of the resistance level among the germplasm was based on DS with supporting parameters of SR and ER.SR > 0.90 and ER > 0.85 were treated as a standard for identifying resistance germplasm.The SR assisted in determining the seedling's mortality rate incorporated into the disease resistance rating scale.The DS uses a scale from 1 to 5 (Table 4).These were used to classify genotypes into different reactions based on RCR infection.The determination of the resistance level among the germplasm was based on DS with supporting parameters of SR and ER.SR > 0.90 and ER > 0.85 were treated as a standard for identifying resistance germplasm.The SR assisted in determining the seedling's mortality rate incorporated into the disease resistance rating scale.The DS uses a scale from 1 to 5 (Table 4).These were used to classify genotypes into different reactions based on RCR infection.In addition to identifying putative candidate genes around the two stable SNPs, we applied a haplotype block size up-and downstream of the SNPs to mine for putative candidate genes for soybean RCR.From this strategy, three probable genes (Glyma.08G074500,Glyma.08G074600, and Glyma.08G074700) were found in the haplotype block of ss715602602 (Chr08).Glyma.08G074600 is annotated to be involved in plant disease responses in signaling mechanisms involved in the management of fungi (Table 3).

Analysis of Expression of Genes Associated with RCR Resistance
We examined the pattern in transcript abundance of five selected genes (Glyma.08G074600,Glyma.08G074700,Glyma.12G043200,Glyma.12G043400, and Glyma.12G043600) in four soybean genotypes exhibiting a contrasting response under C. ilicicola inoculation.A similar gene expression was observed among the resistant genotypes (PI 587880A and PI 567731) and the susceptible genotypes (PI 547885 and PI 567104B) for genes Glyma.08G074600,Glyma.08G074700, and Glyma.12G043600.The gene expression level in resistant genotypes was up on the 7th day and decreased on the 13th day, and vice versa for the susceptible genotypes for Glyma.08G074600 and Glyma.08G074700(Figure 7A,B).However, for Glyma.12G043600, the expression level was up on the 7th and 13th day in the resistant genotypes compared to the susceptible genotypes (Figure 7E).

Analysis of Expression of Genes Associated with RCR Resistance
We examined the pattern in transcript abundance of five selected genes (Glyma.08G074600,Glyma.08G074700,Glyma.12G043200,Glyma.12G043400, and Glyma.12G043600) in four soybean genotypes exhibiting a contrasting response under C. ilicicola inoculation.A similar gene expression was observed among the resistant genotypes (PI 587880A and PI 567731) and the susceptible genotypes (PI 547885 and PI 567104B) for genes Glyma.08G074600,Glyma.08G074700, and Glyma.12G043600.The gene expression level in resistant genotypes was up on the 7th day and decreased on the 13th day, and vice versa for the susceptible genotypes for Glyma.08G074600 and Glyma.08G074700(Figure 7A,B).However, for Glyma.12G043600, the expression level was up on the 7th and 13th day in the resistant genotypes compared to the susceptible genotypes (Figure 7E).

Discussion
The effects of diseases on crops cannot be ignored in an attempt to feed the world's expanding population while combating climate change.Every year, about 40% of yield losses are attributed to pathogens and pests [33], of which fungal diseases are known to cause 10-23% of losses [34], requiring much effort in controlling fungal diseases.Soybean crop is a key source of food for humans, making it valuable [35], and it needs to be protected from fungal diseases.Hence, host resistance in controlling soybean RCR is the most economical and sustainable approach.Soybean RCR incidence and severity are common in major soybean production areas globally [10,[12][13][14].

Genetic Diversity among Soybean Germplasm
Understanding the global soybean collection's genetic diversity is essential for regional breeding initiatives seeking to discover resistance genes.We found that the highest genetic diversity value of 0.31 was from the Chinese gene pool, but a decreased value was shown for Americans (0.28), which confirms the results of earlier research [36], with 0.41 recorded from sub-Saharan Africa [37].Similarly, our results on the PIC value are comparable to findings on previously published values [36,37].The comparatively high diversity of genes and PIC values recorded by others could be attributed to genetic materials included from countries such as Africa and also to the usage of simple sequence repeat markers, as the latter is reported for high diversity [38,39].The reduced genetic diversity seen in countries like Thailand is probably due to robust selection criteria and smaller sample sizes than in other nations.A large soybean accession with different origins could offer an avenue for enhancing soybean breeding programs.

Evaluation and Identification of Resistance to Soybean Red Crown Rot Strain
Detecting novel sources of resistance in the soybean gene pool to key biotic stress, such as diseases, lays the foundation for improving productivity [5].Several works on soybean improvement via enhancing resistance to diseases have been carried out globally [5,40,41], but limited efforts are geared towards soybean RCR disease.Evaluation of the soybean gene pool for disease resistance traits usually requires much labor and is, most of the time, cost-ineffective.The best alternative is genomic selection, leveraging marker data [42,43].Accordingly, in-depth knowledge of disease resistance genetics is crucial to soybean yield development.In this circumstance, soybean genome sequences and the availability of numerous soybean SNP platforms offer crucial roles in supporting the development of the cultivar's resistance to RCR.
Previous works on soybean's reaction to RCR have detected a range of vulnerabilities.For instance, in 18 soybean cultivars evaluated in the field, none were observed with complete resistance [44].Likewise, none of the 157 soybean genotypes consisting of cultivated and wild accessions were recorded as having complete resistance [45].Also, Jiang and others [21] screened 213 soybean accessions and found none with high resistance.However, they found that wild soybean (G.soja) accessions exhibit high resistance compared to cultivated soybeans.The earlier outcomes are consistent with our finding; none of the 299-soybean panel (288 cultivated soybean or the 11 wild accessions) were recorded as highly resistant.Nevertheless, nine soybean accessions showed high resistance to C. ilicicola by DS and as resistant according to the ER and SR rating scale, and others were highly susceptible (Table S2).These variations could be utilized to develop recombinant inbred lines for QTL studies.These accessions lay the foundation for developing soybean cultivars resistant to C. ilicicola.
The identification of resistant material is the central phase towards managing RCR.Yet, understanding the complexity of inheritance governing resistance is crucial for successful breeding programs [46].Evaluation of the panel of lines can be linked with diseases affecting soybeans, such as sudden death syndrome, bacterial leaf pustule, rust, and red leaf blotch disease [26,[47][48][49].However, there is no similar study conducted on red crown disease to date.

Marker-Trait Associations (MTAs), Haplotype Analysis, and Candidate Gene Mining
A GWAS was conducted using 299 PIs in a controlled environment of RCR disease and identified nine significant SNPs linked with either ER, SR, or DS.We used mrMLM since single-locus mapping models may not be able to detect all the QTNs [50].Numerous studies confirm that resistance to root rot diseases in soybean is quantitative [51,52].Several minor genes control quantitative resistance, contributing to partial resistance and reducing disease progression and its effects on plants [53].Thus, there is a need for researchers to identify more QTL/QTNs linked with RCR disease resistance.
Additionally, the haplo-phenotype analysis revealed two stable significant SNPs for RCR disease traits.Specifically, the two SNPs (ss715602602 and ss715612097 on Chr08 and Chr12, respectively) were associated with at least two indices used to assess soybean reaction towards RCR.We identified four and three haplotype alleles on SNP ss715602602 and SNP ss715612097, respectively.The results revealed that the haplotype alleles GTCT, GCTG, and GCCG possessed by SNP ss715612097 showed significant variation in ER, SR, and DS, whereas CATTA and CGCTG possessed by SNP ss715627013 showed significant variation in SR and DS.The haplotype alleles that control the various genotypes' RCR-resistance attributes allow breeders to alter soybean characteristics to suit their needs.Three putative genes were found in the haplotype block of ss715602602 (Chr08), and six putative genes were detected around the SNP ss715612097, making up nine (9) genes underlying RCR resistance.Glyma.12G043600belonged to the protein kinase family with a leucine-rich repeat (LRR) domain; Glyma.08G074600 was an arginine/serine-rich protein that is engaged in signaling mechanisms involved in the management of fungi [54][55][56].The Glyma.08G074700 encoding the glycosyl hydrolase family, as well as protein tyrosine, is reported to promote resistance to fungus-causing leaf spot in tomatoes [57] and rice blast [58,59].Thus, Glyma.12G043600,Glyma.08G074600, and Glyma.08G074700 may be involved in modulating soybean RCR response based on the annotation in plant disease responses.The expression level of the predicted genes Glyma.08G074600,Glyma.08G074700, and Glyma.12G043600 were appreciably upregulated in the resistant accession compared to the susceptible accession on the 7th day, which implies a possibility of its involvement in soybean's reaction to C. ilicicola resistance.For instance, the Glyma.08G074700homolog in Arabidopsis, AT5G64570, encodes a secreted beta-d-xylosidase that enhances resistance to Botrytis cinerea [60] as well as boosts signaling related to systemic immunity in Arabidopsis thaliana [61].
The predicted candidate genes should be further validated to confirm their key roles in regulating C. ilicicola resistance.Also, there is limited understanding about the mechanisms underlying resistance to RCR among soybean genotypes.Previous works have been geared efforts towards developing efficient screening approaches called fresh-weightbased methods [62] and inoculum-soil mixtures [21], and towards identifying resistance sources [21,44,45].Others have investigated the responses of tissue-specific expression to the C. ilicicola infection and the genes involved [63].It has also been revealed that silicon enhances soybean's resistance to RCR [64].

Seed Source, Planting Preparation, and Growth Conditions
The association mapping panel used consisted of 299 plant introductions (PIs), of which 288 are cultivated soybeans (G.max (L.) Merr.) and 11 are wild soybeans (G.soja sieb.& Zucc) from diverse countries across the globe (Table S2).The seeds were obtained from the National Center for Soybean Improvement, Nanjing Agricultural University, Nanjing, China.

Pathogen Culture, Inoculation, Planting, and Growth Conditions
The C. ilicicola strain Y62 was provided by the College of Plant Protection, Nanjing Agricultural University, Nanjing, China.The fungi were maintained on vegetable juice (V8) media plates (90 mm) at 26 • C for short-term use by subculturing and stored on a V8 slant at 5 • C for a more extended period (1 year) [65].The mycelia of C. ilicicola Y62 strain was cultured on V8 media on Petri plates (9 cm) at 25 • C for 6 days.Pathogen inoculation was carried out following the protocol by [21].Briefly, six mycelium plugs (~5 mm cubes) of V8 with actively growing C. ilicicola mycelia were placed in a 500 mL flask containing 200 g of wheat bran-vermiculite medium (wheat bran/vermiculite/water 1:1:3, w/w/v).It was then incubated at 26 • C for 14-21 days when the fungus had colonized the flask entirely.This formed the inoculum and was used to prepare inoculum-soil mixtures by mixing it with vermiculite soil to obtain a strength of 2% (w/v) and filling it into the plastic pot.
Seed coats of wild soybeans were scraped on its distal end towards the hilum to support water permeability.The media were composed of vermiculite and nutritive soil at 1:1 (v/v) and were autoclaved and filled in a plastic pot with drainage holes.The planting media were allowed to cool for two days at room temperature after which they were mixed with inoculum-soil mixtures to obtain a strength of 2% (w/v) and were filled into the plastic pot.Ten seeds were sown per pot, and the top was covered with a two-millimeter layer of the media with three biological replicates per line.Pots were placed in a container, and water was supplied to the pots via their drainage holes to ensure they were thoroughly wet in a greenhouse at 26 • C and 50% relative humidity.A supply of water to the container was performed when needed to maintain the soil wetness until the end of the assay.The pots were rotated every two days within the greenhouse to reduce any effects arising from the location of the plastic pots.The experiment was laid out as a completely randomized design with three replications.The means and the standard deviation were calculated using Microsoft Excel 2019.The soybean genotypes were scored for emergence rate (ER) on the 5th day after planting (DAP) and are expressed as the total number of seeds that emerged out of the total number of seeds planted expressed in percentage.The survival rate (SR) was taken on the 12th DAP and was calculated as the total number of plants alive out of the total number of plants that emerged expressed in percentage.

Evaluation of Soybean for Resistance to Calonectria ilicicola and Statistical Analysis
Genotypes were scored for disease severity (DS) using the 0-5 scale on the 14th DAP (Table 4).Data collected on the DS, ER, and SR were subjected to analysis of variance (ANOVA) using GenStat software, version 12 (VSN International Ltd., UK).Pearson correlation analysis was performed among ER, SR and DS, and visualized in R with Corrplot package p < 0.05 [66].

Genotyping, Quality Control, and Population Structure Analysis
The SNP data were genotyped with Illumina Infinium SoySNP50K BeadChip [29].SNP data were downloaded from the Soybase database (https://soybase.org/dlpages/#snp50k,accessed on 14 June 2023).A total of 42,506 SNPs were filtered using PLINK V1.9 [67], excluding SNPs with missing values exceeding 20% and a minor allele frequency (MAF) of less than 5% for quality control.This resulted in retaining 37,876 high-quality SNPs for subsequent analysis and investigation.Due to differences in the number of accessions used compared to the SoySNP50k dataset, we re-evaluated the population structure of the 299 soybean accessions using Admixture 1.3.0(http://dalexander.github.io/admixture/download.html,accessed on 16 June 2023).

Genetic Diversity among the Soybean Accession Based on Their Origin and Statistical Accessed Analysis
The genetic materials were classified according to their origin (Table S2).Analysis of the principal component was executed in R package "popgen" [68] to generate data on genetic structure, variation, and diversity.Only countries with at least ten cultivars were considered in the computation of genetic diversity metrics by country of origin.Applying Nei's genetic distances to serve as a basis, Ward's minimum variance approach and the R package "stats" were used for grouping all the cultivars.

Multi-Locus Genome-Wide Association Analysis
Principal component analysis (PCA) and the kinship matrix were computed internally within R package GAPIT version 3. A threshold of −log 10 (p) ≥ 3 [69] was used to select 37,876 SNPs markers from 299 PIs significantly associated with the study traits (ER, SR, and DS).Using phenotypes of the three study traits, GWAS was implemented using the mrMLM package [70].We conducted GWAS for all the study traits using the 3VmrMLM model from the 3VmrMLM package [32].The default threshold LOD value of three was used for all the study traits in detecting significant QTNs.By aligning each significant SNP's reference sequence to a soybean reference genome Wm82.a2.v1 from the SoyBase (http://www.soybase.org,accessed on 18 August 2023), the physical map placements of each SNP were found.

Haplotype Analysis and Candidate Gene Analysis
Haplotype analysis was carried out using Haploview software 4.2 [71].The stable SNP marker blocks identified were used as the reference markers.To detect the possible candidate genes around the stable SNPs significantly associated with study trait, the haplotype block size was applied up-and downstream of the SNPs to mine for putative candidate genes for RCR resistance.Candidate genes were retrieved from the reference annotation of the soybean reference genome Wm82.a2.v1 from the SoyBase (http://www.soybase.org,accessed on 30 June 2023).

RNA Extraction and qRT-PCR
Five potential candidate genes around peak SNPs were selected for qRT-PCR to assess their transcript abundance under RCR conditions.We used two resistant (PI 587880A and PI 567731) and two susceptible (PI 547885 and PI 567104B) lines from the phenotypic screening.The planting preparation, growth conditions, pathogen culture, and inoculation are elaborated above.Root samples under RCR infection and control treatment were taken on the 7th and 13th DAPs with three biological and technical replicates.Total RNA was extracted from the roots and further synthesized into cDNA using Ultrapure RNA Kit (CWBIO, Taizhou, China) and HiScript II QRT SuperMix for qPCR (+gDNA wiper) (Vazyme, Nanjing, China), respectively.Primer 5 software was used to design the qPCR primer (Table S4) and the soybean actin (Glyma.18G290800)gene was used as the internal reference for standardization [72].The ChamQSYBR qPCR master Mix Kit (Vazyme, Nanjing, China) was used for the qRT-PCR assay using the Light Cycler 480 system (Roche, Roche Diagnostic, Basel, Switzerland).The 2 −∆∆Ct method was used to calculate expressions [73].Data were analyzed using the R package through the least significant difference (LSD) test at p < 0.05 and graphs were made using GraphPad Prism software 9.5.0.

Conclusions
We found five distinct soybean accessions (PI 602496, PI 567731, PI 587880A, PI 424412, and PI 407196) with high levels of partial resistance to C. ilicicola.Also, the current study presents the first report on marker-trait associations (MTAs) and stable SNPs for soybean RCR disease coupled with its genetic diversity based on cultivar origin.We employed the GWAS, haplotype analysis, and candidate gene mining to unravel the genetic architecture for soybean RCR resistance.We used the mrMLM model to detect nine significant SNPs and two stable SNPs (ss715612097 and ss715627013 on Chr08 and Chr12, respectively).Additionally, nine (9) genes underlying these two SNPs were identified, of which we speculate three of them to be prioritized as potential candidate genes.This study provides insights into the genomic regions of RCR traits.The MTAs identified could facilitate the breeding of new soybean varieties with resistance to RCR disease through the application of MAS after validation and testing in soybean germplasm.The candidate genes identified should be validated and employed for developing RCR-resistant soybeans.The studies could contribute to finding novel ways to develop soybeans against red crown disease.Our study critically analyzed soybean accessions and detected novel SNPs for soybean disease improvement programs.

Figure 1 .
Figure 1.Phenotypic diversity of the 299 soybean accessions to RCR. (A) Emergence rate (ER, %), (B) survival rate (SR, %), (C) disease severity (DS), and (D) heatmap of Pearson correlation coefficients (r) among the ER, SR, and DS were significant at p < 0.05.The red curve on the frequency plots (A−C) represents a normal distribution line.* Significant at p < 0.05

Figure 1 .
Figure 1.Phenotypic diversity of the 299 soybean accessions to RCR. (A) Emergence rate (ER, %), (B) survival rate (SR, %), (C) disease severity (DS), and (D) heatmap of Pearson correlation coefficients (r) among the ER, SR, and DS were significant at p < 0.05.The red curve on the frequency plots (A-C) represents a normal distribution line.* Significant at p < 0.05.

Figure 2 .
Figure 2. Distribution of high-quality SNPs across the chromosomes of soybean.(A) Num SNPs per chromosome.(B) Number of SNPs within 1 Mb window size of each chromosome

Figure 2 .
Figure 2. Distribution of high-quality SNPs across the chromosomes of soybean.(A) Number of SNPs per chromosome.(B) Number of SNPs within 1 Mb window size of each chromosome.

Figure 3 .
Figure 3. Population stratification of the 299 worldwide soybean collections used in this stud on the 37,876 SNPs across the 20 chromosomes.(A) Population structure obtained from A TURE software 1.3.0 with 2−10 runs.The I, II, and III represent subpopulations I, II, and III, tively.Three colors (blue, green, and red) stand for three subpopulations.Each color represe inferred ancestral population.A single individual is represented by each vertical column, percentage of each column's colored segments reflects the individual's presumed ancestral tion among the 299 accessions.(B) Principal component analysis plot.The x-and y-axis re PC1 and PC2, respectively, with their contribution to the total variability.(C) Neighbor-join obtained from TASSEL software version 5.2 grouped the soybean collections into three cluste tified as I, II and III.

Figure 3 .
Figure 3. Population stratification of the 299 worldwide soybean collections used in this study based on the 37,876 SNPs across the 20 chromosomes.(A) Population structure obtained from ADMIXTURE software 1.3.0 with 2−10 runs.The I, II, and III represent subpopulations I, II, and III, respectively.Three colors (blue, green, and red) stand for three subpopulations.Each color represents one inferred ancestral population.A single individual is represented by each vertical column, and the percentage of each column's colored segments reflects the individual's presumed ancestral population among the 299 accessions.(B) Principal component analysis plot.The x-and y-axis represent PC1 and PC2, respectively, with their contribution to the total variability.(C) Neighbor-joining tree obtained from TASSEL software version 5.2 grouped the soybean collections into three clusters identified as I, II and III.

Figure 4 .
Figure 4. Principal component analysis of 299-soybean accession based on their country of orig

Figure 4 .
Figure 4. Principal component analysis of 299-soybean accession based on their country of origin.

Figure 5 .
Figure 5. Manhattan plots for soybean red crown rot response via by emergence rate (A), survival rate (B), and disease severity (C) under a controlled environment in this study.The black dotted horizontal lines represent the threshold for significance at the logarithm of odd (LOD) = 3 and its corresponding −log (p-value).The blue and green dots fall below the threshold.The pink dots represent significantly linked SNPs to each of the studied indices.The vertical line in the (A-C) shows traits linked to a particular chromosome.

Figure 5 .
Figure 5. Manhattan plots for soybean red crown rot response via by emergence rate (A), survival rate (B), and disease severity (C) under a controlled environment in this study.The black dotted horizontal lines represent the threshold for significance at the logarithm of odd (LOD) = 3 and its corresponding −log (p-value).The blue and green dots fall below the threshold.The pink dots represent significantly linked SNPs to each of the studied indices.The vertical line in the (A-C) shows traits linked to a particular chromosome.

Figure 6 .
Figure 6.Haplo-phenotype analysis of stable single-nucleotide polymorphisms (SNPs) linked to at least two evaluated indices of soybean RCR response (shaded SNP).(A) Haplotype block around SNP ss715602602 on chromosome 8. (B) Phenotype (emergence rate, ER) grouping based on the four haplotype groups.(C) Phenotype (survival rate, SR) grouping based on the four haplotype groups.(D) Phenotype (disease severity, DS) grouping based on the four haplotype groups.(E) Haplotype block around SNP ss715612097 on chromosome 12.(F) Phenotype (emergence rate, ER) grouping based on the three haplotype groups.(G) Phenotype (survival rate, SR) grouping based on the three haplotype groups.(H) Phenotype (disease severity, DS) grouping based on the three haplotype groups.Means among haplotype groups were compared by a one-tailed t-test at 95% confidence level.* Significant at p < 0.05

Figure 6 .
Figure 6.Haplo-phenotype analysis of stable single-nucleotide polymorphisms (SNPs) linked to at least two evaluated indices of soybean RCR response (shaded SNP).(A) Haplotype block around SNP ss715602602 on chromosome 8. (B) Phenotype (emergence rate, ER) grouping based on the four haplotype groups.(C) Phenotype (survival rate, SR) grouping based on the four haplotype groups.(D) Phenotype (disease severity, DS) grouping based on the four haplotype groups.(E) Haplotype block around SNP ss715612097 on chromosome 12.(F) Phenotype (emergence rate, ER) grouping based on the three haplotype groups.(G) Phenotype (survival rate, SR) grouping based on the three haplotype groups.(H) Phenotype (disease severity, DS) grouping based on the three haplotype groups.Means among haplotype groups were compared by a one-tailed t-test at 95% confidence level.* Significant at p < 0.05

4. 3 .
Data Collection and Analysis 4.3.1.Determination of Emergence Rate, Survival Rate, and Classification for Resistance to RCR

Table 1 .
Genetic diversity of accessions based on the origin.
a minor allele frequency.bexpected heterozygosity, c polymorphic information content.Plants 2024, 13, x FOR PEER REVIEW 6

Table 1 .
Genetic diversity of accessions based on the origin.
a minor allele frequency.b expected heterozygosity, c polymorphic information content.

Table 2 .
Nine SNPs detected to associate with parameters used to measure red crown rot am 299-soybean panel by mrMLM.

Table 2 .
Nine SNPs detected to associate with parameters used to measure red crown rot among 299-soybean panel by mrMLM.
a Red crown rot emergency rate (ER), severity rate (SR), and disease severity (DS).bSingle-nucleotide polymorphism (SNP); those with single and double underlines represent SNPs detected for two (SR and DS) and three (ER, SR, and DS) traits, respectively.cChromosome (Chr).dSNP positions in base pair (Wm82 genome, version 2).e Logarithm of odd (LOD) with a threshold of 3. f Phenotypic variance explained (PVE).gQuantitative trait nucleotides' effect; those with positive and negative effects represent increase and decrease in the traits, respectively.hMinor allele frequency (MAF) of the significant SNPs.iSignificantly associated genotypes' alleles.aRed crown rot emergency rate (ER), severity rate (SR), and disease severity (DS).b Single-nucleotide polymorphism (SNP); those with single and double underlines represent SNPs detected for two (SR and DS) and three (ER, SR, and DS) traits, respectively.c Chromosome (Chr).d SNP positions in base pair (Wm82 genome, version 2).e Logarithm of odd (LOD) with a threshold of 3. f Phenotypic variance explained (PVE).g Quantitative trait nucleotides' effect; those with positive and negative effects represent increase and decrease in the traits, respectively.h Minor allele frequency (MAF) of the significant SNPs.i Significantly associated genotypes' alleles.

Table 3 .
Putative genes within the two stable single-nucleotide polymorphism markers linked to at least two parameters used to access the 299 accessions to soybean red crown rot.Chr_SNP_Position Gene ID/Name a Position (bp) Annotation Descriptions bChr08_ ss715602602_Glyma.08G0745005684546-5692295 BRI1-associated receptor kinase, protein phosphorylation, leucine-rich repeat

Table 3 .
Putative genes within the two stable single-nucleotide polymorphism markers linked to at least two parameters used to access the 299 accessions to soybean red crown rot.

Table 4 .
Disease resistance rating scale for DS.