Lei Yang,Yu Tian, Yulin Liu, Johen C.Reif, Yinghui Li, Lijuan Qiu

aThe National Key Facility for Crop Gene Resources and Genetic Improvement (NFCRI)/Key Lab of Germplasm Utilization (MOA), Institute of Crop Sciences,Chinese Academy of Agricultural Sciences,Beijing 100081,China

bCollege of Forestry,Northwest A&F University,Yangling 712100,Shaanxi,China

cDepartment of Breeding Research,Leibniz Institute of Plant Genetics and Crop Plant Research(IPK), Gatersleben, Germany

Keywords:

ABSTRACT Soybean cyst nematode (SCN, Heterodera glycines Ichinohe) is one of the most economically destructive pathogens.The soybean line Zhongpin03-5373 (ZP), which combines resistance genes from several donors,is highly resistant to SCN race 3(SCN3).In our previous study,two QTL(rhg1 and GmSNAP11) were identified in a population of recombinant inbred lines derived from a cross between ZP and the susceptible parent Zhonghuang 13.The two QTL explained around one-third of the resistance,suggesting the presence of further QTL contributing to SCN resistance.In the present study,we used an improved version of the genetic map comprising the previously applied 1062 molecular markers and 47 newly developed InDel (insertion-deletion)markers.The improved map revealed a novel locus contributing to SCN3 resistance: qSCN3-1,flanked by InDel marker InDel1-7 and SNP marker Map-0047,explained 4.55%of the phenotypic variance for resistance to SCN3 and was not involved in digenic epistatic interaction with rhg1 and GmSNAP11.Haplotypes of Map-0047_CAPS (a CAPS marker developed for Map-0047) and InDel1-7 were significantly associated with SCN3 resistance in a panel of 209 resistant and susceptible accessions.Using further allele-combination analysis for three functional markers representing three cloned resistance genes(rhg1,Rhg4,and GmSNAP11)and two markers flanking qSCN3-1, we found that adding the resistance allele of qSCN3-1 greatly increased soybean resistance to SCN, even in diverse genetic backgrounds.The qSCN3-1 locus will be useful for marker-assisted polygene pyramid breeding and should be targeted for the future identification of candidate genes.

1.Introduction

Soybean (Glycine max (L.) Merr.) is one of the main sources of vegetable oil and plant protein.The soybean cyst nematode(SCN, Heterodera glycines Ichinohe) is one of its most economically destructive pathogens[1]and causes severe yield losses in soybean worldwide.Cultivation of SCN-resistant cultivars in combination with a suitable crop rotation system is the most sustainable and economical way to control the damage caused by SCN in soybean production.However, the limited diversity of resistance genes used in breeding has led to rapid breakdown of resistance.It is desirable to develop cultivars with more horizontal or broad-spectrum resistance by pyramiding not only major but minor SCN resistance genes.

Resistance to SCN is typically a complex quantitative trait regulated by several genes.Two major (rhg1 and Rhg4) and two minor (GmSNAP11 and NSFRAN07) resistance genes have been cloned [2-5].Classical genetic analyses suggest that the rhg1 locus confers the strongest resistance to SCN race 3(SCN3) [6,7].Two different alleles, rhg1-a (Peking-type) and rhg1-b (PI 88788-type), have been identified at the rhg1 locus and show differing mechanisms of resistance to SCN [8].The allele rhg1-b independently regulates SCN resistance by copy number variation (CNV) of a combination of three tandem genes (Glyma18g02580, Glyma18g02590, and Glyma18g02610)[2].As the copy number of rhg1-b increases, resistance increases and at least 5 or 6 copies of rhg1-b are required to ensure resistance to SCN [9].In contrast, the rhg1-a allele,which occurs at medium copy number (fewer than 5 or 6 copies), interacts epistatically with the major resistance gene Rhg4 to ensure resistance to SCN3.The Rhg4 gene(GmSHMT08)regulates SCN resistance via two non-synonymous SNPs,which lead to changes in enzymatic properties resulting in the death of syncytia [3,10].

GmSNAP18 regulates SCN resistance by encoding a dysfunctional α-soluble N-ethylmaleimide-sensitive factor (NSF)attachment protein (α-SNAP) variant that interrupts NSF function and vesicle trafficking, underlying rhg1 locus[10,11].Recently [4,5], two minor resistance genes GmSNAP11 and NSFRAN07have been shown to be involved in SCN resistance.The first is the closest paralog of GmSNAP18,a splicing site in GmSNAP11 that results in the production of a truncated form of α-SNAP lacking 50 C-terminal residues [4,12].The joint presence of GmSNAP11 and GmSNAP18 increases SCN resistance [12,13].The NSFRAN07protein bound to α-SNAP proteins and reduced rhg1 α-SNAP cytotoxicity [5].The NSFRAN07allele may contribute to this cooperative function, co-segregating with the resistance Rhg1 allele[5].

In addition to the cloned genes, more than 200 QTL(http://www.soybase.org/)representing mostly minor resistance genes have been described, enabling breeders to diversify SCN resistance.Zhongpin 03-5373 (ZP) is an elite line that is highly resistant to SCN3.ZP acquired resistance genes from several resistant resources,including,PI 437654 and Huipizhiheidou [14].In a previous study [13], we identified a new minor QTL, SCN3-11 and a known QTL,SCN3-18 in a recombinant inbred line (RIL) population derived from a cross between ZP and a susceptible parent ZH,using 1062 markers.However,the combination of SCN3-18 and SCN3-11 explained only 36.9% of phenotypic variance, suggesting that additional QTL contributing to the SCN resistance were present in the RIL population and remained undetected owing to inadequate marker density and coverage.

InDel markers offer the advantages of low cost, high accuracy and stability and are widely distributed in the genome.They have been widely used in the detection of QTL and genes influencing agronomic traits and improvement of molecular breeding in wheat [15], rice [16], and maize [17].The purpose of the present study was to identify and characterize new QTL for SCN resistance in the ZP/ZH RIL population by developing new InDel markers to enrich the genetic map for further QTL analysis.

2.Materials and methods

2.1.Plant materials

Two panels of soybean accessions used in the current study had been described in the previous study [13].One was a biparental population of 242 RILs derived from a cross between ZP (resistant) and Zhonghuang 13 (ZH, susceptible).Of the 242 RILs, 32 were resistant and 210 susceptible to SCN3.The second panel was a diverse set of 159 resistant (130 highly and 29 moderately resistant) and 50 susceptible (22 highly and 28 moderately susceptible) soybean cultivars obtained from China National Gene Bank (http://www.nationalgenebank.org/).The two panels were described in a previous report [13].Plant genomic DNA was extracted using a DNA extraction kit (Tiangen Biotech Co., Ltd., Beijing, China).

Phenotypic data were obtained from our previous study [13].The resistance of each accession was evaluated by the disease resistance grade based on the female index (FI).The calculation formula of the FI is as follows: (mean female number per plant of the identified accession/mean female number of the susceptible control (Lee))×100%.FI ＜ 10%, 10% ≤ FI ＜ 30%, 30% ≤ FI ＜ 60% and FI ≥ 60%represent highly resistant (HR), moderately resistant (MR), moderately susceptible (MS), and highly susceptible (HS), respectively.

2.2.Whole genome sequencing, alignment, and InDel identification

In the previous study, ZP and ZH were resequenced using the Illumina Solexa system and the paired-end 75-nucleotide reads were aligned to the Williams 82 soybean reference genome Glyma1.1 (https://phytozome.jgi.doe.gov).Owing to the limited length of reads, only small (1-5 bp) InDel loci had been identified among Williams 82, ZP, and ZH [18].In the present study, ZP and ZH were resequenced using the Illumina HiSeqX platform, and the 125,747,328 and 107,868,279 high-quality paired-end 150-nucleotide reads were mapped to the updated soybean reference genome Wm82.a2.v1 (cv.Williams 82) using Burrows-Wheeler Aligner software [19]with the command “mem -t 4 -k 32 -M”.Here, “mem” indicates the algorithm used, and the symbols “-t”, “-k”, and “-M” indicate the number of reads, the length of the INT Minimum seed, and Mark shorter split hits as secondary (for Picard compatibility), respectively.Duplicated reads were then removed using SAMtools (0.1.19) [20].Finally, InDel polymorphisms with gap from 1 to 15 bp in length were identified between ZP and ZH.

2.3.Development of InDel markers

We selected 47 InDel loci, focusing mainly on reported QTL underlying resistance to SCN and gaps existing in the previous soybean genetic linkage map for the parental ZP/ZH RIL population, and extracted the 300 bp upstream and downstream of the flanking sequences of InDels.We designed primers for specific amplification of the InDel loci using Primer3 (http://bioinfo.ut.ee/primer3-0.4.0/primer3/).The sizes of PCR products ranged from 100 to 500 bp.A volume of 20 μL PCR mixture contained 80 ng genomic DNA,1×PCR buffer, 2 mmol L?1primers, 2.5 mmol L?1dNTPs, and 1U EasyTaq polymerase (TransGen Biotech Co., Ltd., Beijing,China), Thermocycling was started with 94 °C initial denaturation for 5 min, followed by 32 cycles of 94 °C denaturation for 30 s, optimized annealing temperature(Table S1) for 30 s, 72 °C extension for 40 s, and 72 °C final extension for 7 min.Finally, the PCR products were separated on 6% polyacrylamide gel electrophoresis and visualized by silver staining.

2.4.Construction of genetic linkage map and QTL mapping

The construction of the genetic linkage map and the methods used for QTL mapping were described by Liu et al.[21].QTL IciMapping v.3.1 (http://www.isbreeding.net) was used to construct the soybean map using a LOD threshold of 3.0 to group markers, the Kosambi mapping function to calculate genetic distances between markers, and the default maximum genetic distance of 50 cM.QTL analysis for resistance to SCN3 was performed using the Inclusive Composite Interval Mapping (ICIM) program in QTL IciMapping 3.1[22].

2.5.Development of CAPS marker

A pair of CAPS primers(Map-0047-F/R)targeting synonymous SNP locus Map-0047 was developed using dCAPS Finder 2.0(http://helix.wustl.edu/dcaps/dcaps.html) and Primer3.The primer sequences were 5’-TCCAGTGGTTTTTGAGAGAT and 5’-CAAACCAGCATTATGGAAA.Twenty-μL PCR mixtures contained 80 ng genomic DNA, 1×PCR buffer, 2 mmol L?1primers, 2.5 mmol L?1dNTPs, and 1U EasyTaq polymerase(TransGen Biotech Co., Ltd., Beijing, China), Thermocycling was started with 94 °C initial denaturation for 5 min, and followed by 35 cycles of 94 °C denaturation for 30 s, 54 °C annealing for 30 s, 72 °C extension for 40 s, and 72 °C final extension for 7 min.The amplicons were digested with AvrII(New England BioLabs, Beijing, China) in 10 μL reaction containing 3 μL PCR product, 1.5 μL 10×NEBuffer, 0.5 μL restriction enzyme(AvrII)and 5 μL ddH2O,37°C incubated for 1 h.Finally, the products were separated by 1% agarose gel electrophoresis.

2.6.Statistical analysis

The accuracy of genotyping was calculated as follows:(number of accessions with consistent genotypes and phenotypes/number of accessions with accurate phenotypes)×100%.The significance analysis of allelic combinations was performed using Student’s t-tests in SAS 9.3 software (http://www.sas.com/).

3.Results

3.1.Polymorphisms of InDels between ZP and ZH

Resequencing of accessions ZP and ZH and mapping of the sequences to the soybean reference genome Wm82.a2.v1 revealed 102,491 InDel polymorphisms of 1-15 bp length.These InDels were distributed throughout the genome with a mean density of 10.7 InDels/kb and a range from 5.4 InDels/kb on chromosome 18 to 21.9 InDels/kb on chromosome 4 (Table S2).Among these InDel variants, those one base pair (1 bp) in length were the most common (56,155, 54.8%), followed by 2 bp(17,339, 16.9%), 3 bp (7552, 7.4%), and 4 bp (5374, 5.2%),and InDels longer than 4 bp in length were rare (0.6%-2.6%).As the length of the InDels increased, their number decreased sharply (Fig.1).

Most InDels (60,337, 58.9%) were located in intergenic regions.Of 42,155 InDels (41.1%) in genic regions, 18,363(43.6%) were in introns, 6173 (14.6%) in UTR (untranslated regions),15,295(36.3%)in upstream and downstream regions,and only 2324(5.5%)in exons.A total of 1165 InDels with large effects were identified.These led to frameshifts,changed the splicing and stop codons, and influenced the function of genes by elongating or truncating the peptide.

3.2.Development of 47 InDel markers

A total of 47 InDel loci were selected from the new and existing InDel database [18]for the development of InDel markers (Table S1).These InDel markers with a length of 2-14 bp insertions or deletions were distributed over all chromosomes but 4, 9, 13, and 19.Only six InDels were from intergenic regions.Of the remaining 41 InDels located in genic regions, 39 were from one candidate gene and two (InDel1-7 and GM6-2) from two candidate genes each.InDel1-7 was located in the upstream region of Glyma.01G054500 and Glyma.01G054600.InDel GM6-2 was located in the upstream region of Glyma.06G264500 and Glyma.06G264600.One InDel(GM15-6) with large effect resulted in an elongated or truncated peptide and consequently altered gene function.Finally, PAGE (polyacrylamide gel electrophoresis)-screened markers(Table S1)were developed for the 47 InDels.

3.3.QTL mapping for resistance to SCN3

A total of 1109 molecular markers, including 1062 previously reported [13]and 47 new InDel markers, were used to construct a genetic linkage map in the ZP/ZH RIL population(Fig.2).The genetic length of the map was 3264 cM, with a mean genetic distance between adjacent markers of 2.94 cM,and was longer than the previous 2911-cM map constructed using 1062 markers [13].Chromosome 2 was the longest at 193.4 cM and chromosome 10 the shortest at 130.7 cM.Thirteen gaps longer than 20 cM were distributed over 12 chromosomes (1, 2, 3, 4, 6, 9, 11, 12, 13, 14, 17, 19).Of these,only two gaps were longer than 30 cM,on chromosomes 1 and 4.

Fig.1–Number of InDel polymorphisms between ZP and ZH by size and annotation.

Fig.2– Soybean genetic linkage map with location of newly developed InDel markers.Red stars indicate the locations of 47 InDel markers developed in the present study.Number in parentheses indicate numbers of closely adjacent InDel markers.

Based on the genetic linkage map and the known phenotype of SCN3, three QTL, qSCN3-1, SCN3-11, and SCN3-18 were identified (Fig.3).Among them, the major resistance locus SCN3-18 was the same as the one described by Li et al.[13]and was located in the reported rhg1-b interval [2].The resistance locus SCN3-11 was a minor QTL that explained 6.74% of phenotypic variance.After addition of five InDel markers, the genomic region containing SCN3-11 was narrowed from 50 kb[13]to 1.83 kb,flanked by InDel markers GM11-1 and GM11-16.The original SCN3-11 region covered only one candidate gene, Glyma.11G234500 (GmSNAP11) [4].qSCN3-1 was a novel resistance QTL detected in this RIL population with LOD score of 3.2 that explained 4.55% of phenotypic variance.The additive effect of qSCN3-1 was 11.43,indicating that the resistance allele of qSCN3-1 was derived from the resistant parent ZP.Analysis of epistatic interactions with a LOD value of 4.0 as threshold showed significant epistatic interaction between SCN3-11 and SCN3-18.The proportion of phenotypic variance explained by epistatic SCN3-11 and SCN3-18 was 39%.In contrast, qSCN3-1 showed no significant epistatic interaction with SCN3-11 or SCN3-18.qSCN3-1 was flanked by Map-0047 and InDel1-7 and occupied a 791 kb genomic region from Chr01:7813803 to Chr01:7022927.

Fig.3– Quantitative trait locus(QTL)genome scans of logarithm of odds(LOD)score for resistance to SCN3(1109 molecular markers).

3.4.A CAPS marker targeting Map-0047 polymorphism

A CAPS marker(Map-0047_CAPS)targeting synonymous Map-0047 (A/T) locating in exon 7 of Glyma.01G059200 was developed.The sequence spanning Map-0047 contained an AvrII recognition site in the genomes carrying the T allele in Map-0047, but not in the genomes carrying the A allele (Fig.4A).The pair of Map-0047_CAPS primers produced a 502 bp amplicon from ZH, ZP and their descendant RILs.After AvrII digestion, the profiles of the susceptible parent ZH showed two fragments (302 and 200 bp), whereas the profile of the resistant parent ZP showed a single 502-bp fragment(Fig.4B).The allele classes of genotypes of RILs were the same for Map-0047.

3.5.Evaluation of qSCN3-1 in soybean germplasm

Fig.4–CAPS marker of Map-0047_CAPS.(A)The sequence spanning the synonymous SNP locus Map-0047(T/A)(shown in red).The AvrII digestion site is underlined.(B)Agarose gel electrophoresis results after digestion with AvrII of parents ZP,ZH and eight recombinant inbred lines(RILs).R, resistant; S,susceptible.

Map-0047_CAPS, and InDel1-7 were used to genotype a diverse panel of 209 soybean accessions,including 130 highly SCN3-resistant (HR) accessions, 29 moderately resistant (MR)accessions, 28 moderately susceptible(MS)accessions and 22 highly susceptible (HS) accessions [13].The mean female index(FI)of the 139 accessions carrying the A allele(FI=9.3%)was significantly (P= 8.3×10?7) lower than that of the 70 accessions carrying the T allele (FI = 49.8%) (Fig.5A).The diagnosis of Map-0047_CAPS was correct for 81.5% of HR types, 72.4% of MR types, 71.4% of MS types, and 81.8% of HS types.The mean FI of 151 accessions carrying the diseaseresistant genotype InDel1-7-INS9(FI=14.2%)was significantly(P= 0.00016) lower than that of 58 accessions carrying the disease-susceptible genotype InDel1-7-Ref (FI = 45.2%) (Fig.5B).The diagnosis of InDel1-7-INS9 was correct for 85.4%of HR types, 72.4% of MR types, 64.3% of MS types, and 59.1% of HS types.

Of four haplotypes observed in the diverse germplasm panel (Fig.5C), Map-0047_CAPS-A/InDel1-7-INS9, and Map-0047_CAPS-T/InDel1-7-Ref were the most common haplotypes, with frequencies of 64.1% and 25.4%, respectively,followed by Map-0047_CAPS-T/InDel1-7-INS9 (8.1%) and Map-0047_CAPS-A/InDel1-7-Ref haplotypes (2.4%).A significant (P＜0.001) correlation between haplotype and SCN3 resistance was detected by chi-square test.A total of 134 soybean accessions with the Map-0047_CAPS-A/InDel1-7-INS9 haplotype,which contains the two resistance alleles,showed on average 8.8% lower FI than the accessions carrying the remaining three haplotypes.Of the 134 accessions, 123 were resistant (HR/MR, 104/19) and 80.0% of HR phenotypes were correctly classified.The 53 soybean accessions carrying Map-0047_CAPS-T/InDel1-7-Ref haplotype, which contains two susceptible alleles, showed a significantly (P= 9.8×10?6)higher mean FI (47.5%) than the Map-0047_CAPS-A/InDel1-7-INS9 haplotype.Of the 53 accessions, 56.6% showed a susceptible phenotype (HS/MS, 18/12) and 81.8% of the HS phenotype were correctly classified.These results indicated that Map-0047_CAPS-A/InDel1-7-INS9 was a dominant resistance haplotype and Map-0047_CAPS-T/InDel1-7-Ref a major susceptibility haplotype.

3.6.Multiple marker-assisted selection for rhg1, Rhg4,GmSNAP11 and qSCN3-1

In the previous study, three markers, rhg1-2 KASP, Rhg4-389 CAPS, and GmSNAP11-2565, which represent rhg1, Rhg4, and GmSNAP11, respectively, were used to fingerprint 209 accessions of a diversity panel[12].Combining genotyping data for the functional markers of Map-0047_CAPS and InDel1-7 flanking qSCN3-1 allowed the identification of 10 major allelic combinations(AllelicC)(Fig.6).Of the 10,AllelicC1,consisting of five resistance alleles, was a major resistance haplotype showing the lowest FI (0.8%), which was significantly (P=2.6×10?5) lower than FI (8.6%) of AllelicC4 carrying the three resistance alleles Rhg4, rhg1, and GmSNAP11.AllelicC8 carrying resistance alleles only in qSCN3-1 showed a mean FI of 31.3%, which was significantly (P= 0.04) lower than that of AllelicC9 (64.5%) carrying five susceptible alleles.Indicated that adding qSCN3-1 increased soybean resistance to soybean cyst nematode, even in diverse genetic backgrounds.

3.7.Candidate genes for qSCN3-1

Fig.5–Effect of qSCN3-1 on SCN resistance in a diverse germplasm panel of 209 soybean accessions.(A)Flanking marker Map-0047_CAPS;(B) flanking marker InDel1-7;(C)haplotypes of Map-0047_CAPS and InDel1-7.

Based on the soybean reference genome Wm82.a2.v1, 48 candidate genes were identified in the genomic region containing qSCN3-1.Based on RNA-Seq data of the reference genome retrieved from SoyBase (https://www.soybase.org/),expression data for 31 candidate genes from various tissues and stages were examined(Fig.S1).Of 31 candidate genes,24(77.4%)showed extremely low expression.Of seven candidate genes with high expression, only Glyma.01G057100 and Glyma.01G058000 were highly expressed in roots.Based on Gene Ontology annotation (http://wego.genomics.org.cn/),Glyma.01G057100 is involved in cell wall pectin metabolic process and is involved in root hair elongation[23].AT3G29360 is the closest paralog of Glyma.01G057100.It encodes a UDPglucose 6-dehydrogenase family protein and its mutations cause plant development defects and affect pectin networks[24].Glyma.01G058000 is involved in protein domain-specific binding and protein phosphorylated amino acid binding.It is a member of the 14-3-3 gene family, which regulates plant growth and development and responds to environmental stress [25].AT1G26480 is the closest paralog of Glyma.01G058000.It acts as a general regulatory factor 12 and encodes a 14-3-3 protein (involved in multiple signaling pathways such as hormone signaling) [26].Comparison of the sequences of ZP and ZH revealed a nonsynonymous SNP Chr01:7509283 (T/C) which resulted in amino acid changes between ZP and ZH in Glyma.01G057100.No SNP or InDel in Glyma.01G058000 resulted in amino acid changes between ZP and ZH[27].

4.Discussion

ZP is an elite soybean line that shows high resistance to SCN owing to its diverse resistances derived from multiple donors,including Peking, PI 437654 and Huipizhiheidou [14].In a previous study[13],we constructed a genetic linkage map for the RIL population derived from the cross of ZP and a susceptible parent ZH using 1062 markers.One major QTL,SCN3-18, conferring to rhg1 and a minor QTL, SCN3-11,containing GmSNAP11 were identified.Increasing the density of markers can increase the power of QTL mapping[18,28].In the present study, we developed 47 new InDel markers by targeting 2-15 bp InDel polymorphism loci derived from the resequencing and alignment of the genome sequence of the parents of the RIL population, ZP and ZH.An expanded genetic linkage map of soybean was constructed using 1062 previously developed and 47 new InDel markers, and QTL mapping was performed based on the new map.

Besides two previously identified QTL (SCN3-11 and SCN3-18)[13],a new QTL qSCN3-1 flanked by Map-0047 and InDel1-7 was discovered by addition of the InDel markers.The QTL qSCN3-1 comprised a 791-kb genomic region,from 7,022,927 to 7,813,803 bp, which includes a previously reported QTL SCN 26-2 [29].SCN 26-2 was flanked by the markers Satt368 and Satt532, in a physical interval of 19.7 Mb spanning 6,926,577-26,610,094 bp on chromosome 1 [29].Three QTL on chromosome 1 were identified using a F2population derived from a cross of the resistant accession PI 438489B and the susceptible accession Hamilton, which confer resistance to SCN3, SCN5, and SCN14, respectively [30].These three QTL were located in the same genome region from 355,570 to 7,729,201 bp that contained qSCN3-1 localized to a 706-kb region from 7,022,927 to 7,729,201 bp in the soybean reference genome.This finding suggests that qSCN3-1 may contain a gene conferring broad resistance to SCN.

Planting resistant cultivars in combination with sustainable crop rotations is an effective means to reduce the damage of SCN and ensure sustainable soybean production.Markerassisted selection (MAS) offers an efficient approach to pyramiding resistance alleles in cultivars in order to broaden the genetic basis of resistance.Knowledge of epistatic interactions is essential for MAS, given that epistatic interactions play a significant role in the expression of complex traits such as SCN resistance.Previous studies showed that pyramiding of resistance alleles represented a powerful means of improving resistance to SCN.First, epistatic interaction was reported between resistance alleles at the rhg1 and Rhg4 loci in several resistant resources including PI 437654,PI 88788,PI 404198A,and Peking(Forrest)[31-36].Over 98%of the SCN3 resistance was achieved by stacking the Rhg1 and Rhg4 loci in Forrest [32].Further interaction analyses showed that Rhg4 was needed to achieve almost full resistance in combination with the Peking-type rhg1, while in the PI 88788-type rhg1 background, Rhg4 was not needed [3,10,32].Besides rhg1 and Rhg4, the combination of rhg1 and SCN3-11 could provide a higher level of SCN3 resistance than individual rhg1 or SCN3-11[13].

Recently, our study [12]suggested that pyramiding the resistance alleles of GmSNAP18 (syn.rhg1), GmSHMT (syn.Rhg4), and GmSNAP11 represents a powerful means of selecting for strong resistance to SCN3 across a highly heterogeneous set of genetic backgrounds.In the present study, we evaluated the effects of the allele combination by examining the resistance of the newly identified SCN3-11 allele together with the GmSNAP18, GmSHMT, and GmSNAP11 alleles in a diverse germplasm panel comprising 209 resistant and susceptible genetic resources.We found that when the three resistance alleles at GmSNAP11-2565CAPS, rhg1-2, and Rhg4-389 were pyramided, the FI reached 8.6%.Pyramiding the two resistance alleles at Map-0047 and InDel1-7 representing qSCN3-1, led to a decrease of the FI to 0.84%.This finding suggests that the combinations of qSCN3-1 with rhg1, Rhg4, and SCN3-11 have complementary additive and epistatic effects and yield the maximum resistance to SCN in soybean germplasm.

CRediT authorship contribution statement

Yinghui Li and Lijuan Qiu designed the study.Lei Yang and Yu Tian performed the experiments and data analysis.Yulin Liu developed the RIL population.Lei Yang, Jochen C.Reif,Yinghui Li, and Lijuan Qiu wrote the manuscript.All authors approved the manuscript.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

This research was financed by the National Key Research and Development Program of China (2016YFD0100201) and the Agricultural Science and Technology Innovation Program (ASTIP) of the Chinese Academy of Agricultural Sciences.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.08.003.