A study of the genetic diversity in the world soybean collection using microsatellite markers associated with fungal disease resistance

Almaty Region during 2018–2019, the distribution of the studied SSR markers in the population was not signi f icantly associated with resistance to the analyzed diseases under f ield conditions. Conclusion. SSR genotyping of the soybean collection helped to identify accessions that potentially pos-sess resistance-associated alleles of fungal disease resistance genes. The data obtained can be further used for the development of DNA documentation and the breeding the promising cultivars and lines of soybean. signi f icant between alleles to resistance other alleles, low three assessment of SSR genotyping a high level of polymorphism and suggested that a of the applied markers can be successfully used in DNA documentation of collections and breeding programs.

Background. Soybean (Glycine max (L.) Merr.) gradually becomes one of the leading legume crops in Kazakhstan. The area under soybeans in the country has been increasing annually and requires the development of adapted cultivars with a higher yield, improved quality characters, and resistance to emerging fungal diseases. The enlargement of the crop's gene pool also suggests the need to study and document local soybean accessions to meet the standards of the available world soybean collection by using reliable and informative types of DNA markers. Materials and methods. In this study, the soybean collection consisting of 288 accessions from different countries, including 36 cultivars and promising lines from Kazakhstan, was studied. The molecular genetic analysis was performed using nine polymorphic SSR (simple sequence repeats) markers, seven of which (Satt244, Satt565, Satt038, Satt309, Satt371, Satt570 and Sat_308) were associated with resistance to three main fungal diseases of soybean -frogeye leaf spot, fusarium root rot, and purple seed stain. Results. The average PIC (polymorphism information content) value of the analyzed SSR markers constituted 0.66 ± 0.07, confirming their highlevel polymorphism. The principal coordinate analysis suggested that the local accessions were genetically most close to the accessions from East Asia. As the collection showed a robust resistance to three studied fungal diseases in Almaty Region during 2018-2019, the distribution of the studied SSR markers in the population was not significantly associated with resistance to the analyzed diseases under field conditions. Conclusion. SSR genotyping of the soybean collection helped to identify accessions that potentially possess resistance-associated alleles of fungal disease resistance genes. The data obtained can be further used for the development of DNA documentation and the breeding the promising cultivars and lines of soybean.

Introduction
Soybean (Glycine max (L.) Merrill) is one of the most economically important legume crops, which provides plant protein for more than a quarter of all food and feed in the world (Garcia et al., 1997). Over the past ten years, the worldwide soybean production increased 1.2 times due to the ever-rising demand for this crop. In 2019, it was at the level of 358,650 million metric tons, and the major part of this yield belonged to Brazil, USA, and Argentina (U.S. Department of Agricul-ture…, 2020). Soybean is considered a highly profitable, export-oriented crop in Kazakhstan. Nowadays, over 90% of soybean production has been concentrated in Almaty Region (Makulbekova et al., 2017). In 2019, Kazakhstan harvested 229,000 metric tons of soybean. For further development of the soybean industry, the Government of Kazakhstan has declared a new program "Northern Soybean" for diversification of soybean to the north part of the country. Implementation of this program would make it possible over five years to enlarge sown areas to 1.5 million ha, raise productivity, and increase the production over 3 million tons (North Kazakhstan…, 2019).
The important factor that severely limits the soybean productivity worldwide is susceptibility to harmful diseases (Wrather et al., 1996). In Kazakhstan, more than ten fungal diseases of soybean have been identified (Maui et al., 2016;. Due to the expanding area under this crop, it is an obvious necessity to study the genetic potential of soybean associated with the tolerance to harmful pathogens. In general, the total yield loss due to susceptibility to fungal diseases can reach 40% (Wrather et al., 1996). Frogeye leaf spot (FLS, caused by Cercospora sojina Hara), fusarium root rot (FRR, caused by Fusarium solani (Mart.) Sacc. f. sp. glycines Roy), and purple seed stain (PSS, caused by Cercospora kikuchii (Tak. Matsumoto & Tomoy.) M.W. Gardner) are the most dangerous and widespread fungal diseases of soybean worldwide (Zatybekov et al., 2017).
In local breeding programs targeted for soybean resistance to fungal diseases, phenotypic analysis is mainly carried out on the basis of field observations; there are practically no studies at the molecular level. Therefore, along with field assessment of the various gene pools for resistance to fungal diseases, the analysis of genetic diversity is also required.
The genetic marker is a DNA sequence with a known chromosome localization which is linked with a particular gene or is a part of a gene of interest (Idrees et al., 2014). They are useful tools to differentiate individuals in a population or to classify individuals representing different varieties or cultivars within a species. To date, there are a number of various classes of DNA markers applied for different purposes in plant molecular genetics, breeding, biotechnology, etc. (Idrees et al., 2014;Singh et al., 2010). Simple sequence repeats (SSRs), or microsatellites, have become the most widely used markers for DNA fingerprinting or genotyping plant accessions. SSRs were applied in many studies due to their high informativeness, codominance, multiallelism, reproducibility in different laboratories, and transferability among related species (Cregan et al., 1999). In particular, SSRs are considered as a reliable tool to identify the genetic diversity and relationships among soybean genotypes in different populations (Wang et al., 2006). Assessment of genetic divergence and relatedness among genetic resources (as potential breeding material) has significant implications for the improvement of crop plants. Knowledge of genetic diversity could help soybean geneticists and breeders to understand the structure of germplasm, predict which combinations would produce the best offspring, and facilitate widening of the genetic base of breeding material for selection (Bisen et al., 2015). The survey of recent reports demonstrated the successful use of SSR markers for screening soybean in relation with resistance to fungal diseases (Zhong et al., 2018;Ghorbanipour et al., 2019).
Rouf Mian et al. (1999) used the NILs (near-isogenic lines) population created by backcrossing the Rcs3 gene from cv. Davis to the FLS-susceptible cultivar Wright. In the cultivar Davis (showing complete resistance to all isolates of FLS causative agent), the donor parent of the Rcs3 allele for FLS resistance, a 154 bp fragment of Satt244 was amplified (Rouf Mian et al., 1999). Iqbal with co-authors identified six QTLs responsible for SDS resistance and reported five of them to be related to the SSR markers Satt214, Satt309, Satt570 (on the linkage group G), Satt371 (LG C2), and Satt354 (LG I) (Iqbal et al., 2001). Closely located to the Rfs gene is the SSR marker Satt309, which produced a 165 bp band in cv. Forrest resistant to SDS. Jointly, these QTLs explained about 90% of the total variation in SDS disease incidence of RILs (recombinant inbred lines) from the Forrest × Essex cross, and they showed only the presence of additive genic action (Iqbal et al., 2001). Jackson et al. (2008) carried out genetic mapping of resistance to PSS in the F 2 population from crossing the susceptible cultivar Agripro 350 and resistant PI 80837. The candidate gene has been mapped between Sat_308 and Satt594 loci on molecular linkage group G. The Sat_308 primers produced a band of approximately 310 bp both in PI 80837 and the resistant bulk (Jackson et al., 2008). Hence, the involvement of polymorphic informative SSR markers into breeding is important for crop improvement programs, including targeting resistance to diseases. The purpose of this work was to study the genetic diversity in the soybean collection, consisting of 288 accessions, using microsatellite markers associated with resistance to common fungal diseases -FLS, FRR, and PSS.

Plant material and field observations
The objects of this study were 288 soybean accessions from different countries, including 36 released cultivars and prospective lines from Kazakhstan . The world collection was represented by the soybean accessions originated from 5 geographic regions, including Eastern Europe (n = 122), Western Europe (n = 23), East Asia (n = 57), North America (n = 50), and Kazakhstan (n = 36). Cv. 'Zhansaya' was used as a check cultivar Almaty Region. Field trials were conducted on the experimental plots of the Kazakh Research Institute of Agronomy and Plant Industry (KRIAPI) in 2018 and 2019 . The soybean collection was grown in three randomized replicates on one-meter plots. The resistance to FLS, FRR, and PSS was studied under natural infection. A nine-point scale was used, where point 1 denoted high resistance (no symptoms); 3resistance (5 -19% affected); 5 -partial resistance (20-49% affected); 7 -susceptibility (50-79% affected); and 9 -high susceptibility (up to 80% affected) (Hnetkovsky et al., 1996). The data were used to study the relationships between the disease resistance of the collection accessions and the following 8 major agronomic traits: plant height (PH), the height of lowest pod (HLP), number of lateral branches (NLB), number of fertile nodes (NFN), number of seeds per plant (NSP), thousand seeds weight (TSW), yield per plant (YP), and yield per plot (YpP). The field trials were conducted in 2018, and 2019; the analyzed traits were evaluated according to Korsakov et al. (1968).

DNA extraction and SSR genotyping
Total DNA was isolated from seedlings of all soybean accessions, according to Dellaporta et al. (1983). The soybean collection was genotyped using nine SSR markers, with appropriate optimization of polymerase chain reaction (PCR) conditions for specific primer pairs (Cregan et al., 1999). PCR was conducted in a 20 μl total volume containing 20 ng of genomic DNA, 1U Taq DNA polymerase, 0.2 mM of each dNTP, 10 pmol of each primer, 1.5 mM MgCl2 in the 1× Taq Buffer. The list of SSR markers used for the analysis is presented in Table 1.

Statistical analysis
Statistical analyses of field data were done using SPSS 22.0 and STATISTICA 13.5 software.
The effective number of alleles per locus was determined using the GenAlex, ver.6.5 software (Peakall et al., 2012). The values of the PIC index (polymorphism information content) suggested the effectiveness of the markers used, given that markers with a value of PIC > 0.5 were considered as highly informative; 0.5 > PIC > 0.25 as informative; and PIC ≤ 0.25 as marginally informative (Botstein et al., 1980). Genetic diversity was assessed on the basis of Nei's genetic diversity index and Shannon Information Index, using the GenAlex, ver.6.5 program (Peakall et al., 2012). The resulting similarity matrix was further analyzed using the unweighted pair-group method with arithmetic mean (UPGMA) clustering algorithm for the construction of the dendrogram. Variation among populations was studied using Principal Coordinate Analysis (PCoA) with the software GenAlex, ver.6.5 (Peakall et al., 2012).
Analysis of the population structure was carried out using STRUCTURE (v2.3.4.) software with a Bayesian Markov Chain Monte Carlo (MCMC) approach, based on the admixture and correlated frequency models (Evanno et al., 2005). The output from STRUCTURE was analyzed for the delta K value (ΔK) in STRUCTURE HARVESTER (Earl et al., 2012). The genetic map was drawn using MapChart v.2.3 software (Voorrips, 2002).

Phenotypic variation of the soybean resistance to fungal diseases
Results of the two-year field trials of 288 soybean accessions revealed variations in their resistance to FLS, FRR, and PSS caused by three fungal pathogens -Cercospora sojina, Fusarium solani, and Cercospora kikuchii respectively (Table 2).
In 2018, symptoms of infection by the FRR causative agent were not observed. More than 90% of the analyzed soybean collection showed high resistance to the studied fungal diseases ( Table 2). The Pearson correlation analysis based on the average values of 2018 and 2019 field trials helped to reveal  (3), 2020 • A. K. ZATYBEKOV • Y. T. TURUSPEKOV • B. N. DOSZHANOVA • S. I. ABUGALIEVA a highly positive relationship between morphological traits and yield components (P < 0.01) (Fig. 1). The disease resistance traits had negative correlations with agronomic traits. Most of them were observed in the relations between FLS resistance and major agronomic traits. The resistance to FRR had negative correlations with TSW and PH, while resistance to PSS was negatively correlated with YpP (Fig. 1).

Microsatellite analysis of soybean accessions
Soybean collection was studied using 9 SSR markers associated with resistances to frogeye leaf spot, fusarium root rot, and purple seed stain. Figures 2-4 present the results of electrophoresis of PCR products with primers Satt244, Satt309 and Sat_308. The Satt244 marker is known to be associated with resistance to FLS (Rouf Mian et al., 1999). The Satt244-154 allele linked to the Rcs3 gene on chromosome 4 (linkage group C1) was detected for 50 accessions (Fig. 2).
According to Ding et al. (2012), the Satt565 marker is closely linked to the gene Rcs1 controlling the resistance to FLS. In our study, the resistance-associated allele Satt565-208 was identified in 32 soybean access ions. Analysis of the studied collection using SSR markers associated with the resistance to FRR made it possible to identify 48 accessions with the Satt309-165 allele (Fig. 3)    Also, soybean collection was analyzed on Satt570 (Iqbal et al., 2001), the marker closely linked to the Rfs gene that controls resistance to FRR. Seventy accessions with the Satt570-110 allele were identified. The analysis of the studied soybean collection using Sat_308 (Fig. 4) associated with resistance to PSS revealed 28 accessions carrying the Sat_308-310 allele, previously described by Jackson et al. (2008).
The frequencies for the alleles of SSR loci known to be related to disease resistance were detected for each group of soybean accessions (Table 3).
The results based on using 9 SSR polymorphic markers associated with resistance to FLS, FRR, and PSS revealed 40 alleles, with the average number of 4.42 alleles per marker (Table 4). The number of alleles per locus was from     3 (Sat_308) to 7 (Satt244). Each of the four microsatellite loci (Satt038, Satt371, Satt565, and Satt570) was represented by 6 alleles. The average number of alleles ranged from 3.0 for Sat_308 to 5.4 for Satt371 (Table 4). The effective number of alleles varied from 1.38 to 3.72, with a mean value of 2.6. Nei's genetic diversity index averaged 0.54 ( Table 4). The average value of polymorphism information content (PIC) was 0.66, ranging from 0.32 for Sat_308 to 0.83 for Satt371. The PCoA analysis based on the variability data for nine microsatellite markers in soybean populations demonstrated that genotypes from Western Europe are genetically more distant from the other four groups of origin. The PCoA2 coordinate effectively separated groups of North America and Eastern Europe from groups of East Asia and Kazakhstan. The local soybean accessions were close to the East Asian genotypes (Fig. 5).
The analysis based on using the STRUCTURE HARVEST-ER led to the development of Delta K (ΔK), which showed that the collection of 288 soybean accessions had an optimal number of populations equal to 3 (Fig. 6).
Three generated populations were represented by genotypes from all five groups with different origins. Population 2 had a largest number of accessions, which mainly included genotypes from Eastern Europe and North America. Population 1 included the majority of accessions from East-ern Europe and East Asia. Most of the local genotypes were grouped in the Population 3 (Table 5).

Associations with alleles linked to disease resistance genes
Seven out of nine SSR loci were located close to the known resistance genes that controlled immune response to three studied fungal diseases (Fig. 7). The polymorphism observed at Satt244 and Satt565 associated with resistance to FLS showed that 78 accessions had at least one allele related to disease resistance. The analysis of polymorphism at four SSR loci ( Satt038, Satt309, Satt371, and Satt570) related to FRR resistance helped to identify 154 accessions with at least one allele. There were no accessions carrying alleles of all four SSR marker loci related to disease resistance. The screening suggested that 38 out of 154 accessions had at least two alleles, and 4 accessions had three alleles associated with disease resistance, respectively. The Sat_308-310 allele associated with the resistance to PSS was found in 28 soybean accessions.
In total, the study made it possible to identify four genotypes which had at least one allele associated with resistance to one of the three studied fungal diseases. The application of the t-test did not reveal statistically significant differences between accessions with positive and negative alleles of SSR markers associated with resistance to three diseases under field conditions.

Discussion
The results of field trials of the global soybean collection in the environments of Southeastern Kazakhstan showed high levels of resistance in a majority of the accessions to common fungal diseases in the tested environments. The twoyear study of the resistance to FLS, FRR and PSS suggested that more than 97% of the collection were resistant to these diseases. The correlation analysis revealed significant positive correlations between agronomic characters and their negative correlations with resistance to fungal diseases. The soybean collection consisting of 288 accessions from different parts of the world was studied using nine microsatellite markers related to the genes of resistance to three fungal diseases, FLS, FRR and PSS. The SSR markers used in this study were highly polymorphic and informative, with the average PIC equal to 0.66. Two markers linked to PSS, Sat_308 and Satt115, showed moderate PIC values: 0.32 and 0.34, respectively (Table 4). At the same time, PIC for the entire soybean collection of 288 accessions, using the other seven SSR markers (Satt244, Satt565, Satt529, Satt038, Satt570, Satt309 and Satt371) related to FLS and FRR resistances, was higher (0.75), confirming high informativeness of these markers. The SSR loci with more than five alleles and PIC over 0.6 would be informative for genetic structure analysis and marker-assisted selection.
The PCoA analysis using nine SSR markers showed that local genotypes were genetically closer to East Asian cultivars, compared with accessions from other regions. The result confirms the field test data, suggesting that they have higher adaptability to the environments of Southeastern Kazakhstan.
Screening of the global soybean collection with nine polymorphic SSR markers made it possible to identify genotypes carrying 7 marker alleles associated with resistance to three fungal diseases. They were Satt244-154 and Satt565-208 associated with resistance to FLS; Satt 038-182, Satt309-165, Satt371-272 and Satt570-110 associated with resistance to FRR; and Sat_308-310 associated with the response PSS.

Conclusion
Field trials of the soybean collection consisting of 288 accessions from different parts of the world suggested that over 97% of the collection was resistant to FLS, FRR and PSS fungal diseases in the studied period. Still, the correlation analysis showed that FLS and FRR had a negative effect on TSW, while FRR and PSS negatively correlated with yield. The evaluation of accessions using nine SSR markers associated with the resistance to FLS, FRR and PSS differentiated local samples from soybean accessions from other regions and indicated that they are genetically closer to cultivars from East Asia. The usage of the t-test did not show significant differences between the accessions with alleles linked to resistance and those with other alleles, which could be explained by low infection spreads of the three studied diseases in 2018 and 2019. The assessment of the SSR genotyping results revealed a high level of polymorphism and suggested that a majority of the applied markers can be successfully used in DNA documentation of soybean collections and breeding programs.

This research has been funded by Project AP05131592 supported by the Ministry of Education and Sciences of the Republic of Kazakhstan.
Работа финансировалась в рамках проекта AP05131592 при поддержке Министерства образования и на уки Республики Казахстан.