QTL underlying resistance to two HG types of Heterodera glycines found in soybean cultivar 'L-10'

Soybean (Glycine max (L.) Merr.) is the most important economic and nutritional legume crop for both oil and protein products worldwide [1]. However, a variety of abiotic stresses and biotic stress are threatening the world's soybean production every year. After scarcity of water, seed yield losses in soybean are mainly due to pest and pathogen infections [2]. One of the most destructive pests of soybean is the soybean cyst nematode (SCN; Heterodera glycines Ichinohe). It can cause various symptoms after infection such as chlorotic patches within leaflets, root necrosis and suppression of root growth. Yield losses in the world due to SCN infection have approached $2 billion a year [3]. Once established in a field, the infestations have been difficult to eradicate. Eight races of SCN are found (1, 2, 3, 5, 6, 7, 9 and 14) [4] in China. Race 3 is predominant in the Northeastern Provinces of China (including Heilongjiang, Jilin, Liaoning and Inner Mongolia).

Although biological controls and transgenic technology have some promise, breeding the pest-resistant materials was still the only effective control method by 2010 [5]. There were more than 100 soybean accessions from Asia that contained SCN resistance genes, but only a few of them had been utilized for developing US cultivars and nearly all the resistance genes in developed varieties were from two sources, 'Peking' and 'PI88788' [6].

Inheritance of resistance to SCN was first reported in Peking when three resistance genes (rhg1,rhg2 and rhg3) were identified [7]. The rhg1 locus has been shown to have the greatest impact on the development of SCN from all HG Types in several resistance sources including Peking, PI437654 and PI88788. Although originally reported as a recessive locus, the rhg1 locus has more recently been characterized as incomplete dominant or dominant depending on the HG Type 0 used [8]. The fourth resistance gene Rhg4 had been confirmed as being necessary for full resistance to some races of SCN in Peking but not PI88788. It behaved as a dominant gene in Peking-derived sources. Both Rhg1 and Rhg4 were necessary for the full resistance to SCN race 3 and 14. An additional dominant gene was identified in PI88788 and was designated Rhg5 that also had been characterized as a dominant gene [9].

HG Types and plant resistance were both shown to be temperature dependent. As the traditional breeding process is time consuming and labor intensive, new cultivar breeding usually could not adapt with the speed of drift among pathogen populations driven by mutations within massive field populations. Therefore, new methodologies for SCN resistance breeding were required. Genetic marker technology has facilitated the identification, localization and characterization of QTL associated with important agronomic traits including SCN resistance [7]. DNA-based techniques such as micro satellites (SSR) have been used extensively for soybean gene mapping, because they are abundant, uniformly distributed, highly polymorphic, codominant, rapidly produced by PCR, relatively simple to interpret, and easily accessed by other laboratories via published primer sequences. Further, they have been used to anchor the soybean genome sequence [10]. Consequently, numerous publications have appeared on the identification and localization of QTL underlying resistance to SCN [7]. The marker assisted selection provided the potential to develop SCN resistance in soybean cultivars tested in single environments, often in greenhouse assays. For instance, Soybean JTN-5503, JTN-5303 (high yield, with resistance to multiple nematode populations), were the first soybean lines developed using MAS for nematode resistance [11, 12].

The SCN resistance gene Rhg4 was inferred to be linked with pBLT24, pBLT56 and I gene for seed coat color on linkage group A2 (chromosome 8) [13]. QTL identified and localized for SCN resistance based on both genotypic and phenotypic data in a segregating population was first reported by Concibido et al. (1994)[14]. During the following decade, a total of 60 SCN resistant QTL were reported on the following linkage groups: A1, A2, B1, B2, C1, C2, D1a, D2, E, F, G, H, I, J, L, M, and N [7]. QTL on the LG G near rhg1 was detected in most studies and was considered to be one of the major genes underlying resistance to SCN. Nearly 80% of these studies used the SCN HG Type 0 (race 3) resistance because it was the main cause of yield reduction among soybeans grown in many areas, including Northeast of China. A minor QTL S16-5 on LG K provided 5% additive effects for resistance to SCN race 1 had been reported in 2009, this was the first report of a QTL on linkage group K associated with SCN resistance [15]. Another QTL was detected on LG D2 (Satt574; P = 0.001, R2 = 11%) and associated with the resistance to HG Types 1.3- and 1.2.5- [6].

Candidate genes underlying QTL were identified and one, a lecine-rich repeat (LRR) trans-membrane (TM) receptor like kinase (RLK) was tested in transgenic plants [8]. Resistance was provided to SCN and a fungal pathogen, Fusarium virguliforme. In contrast, another RLK in the Rhg4 interval was shown not to underlie the resistance locus [16]. Candidate genes underlying other loci were not reported to date.

Cultivars with broad-spectrum resistance to SCN were available in China and may be more useful than those that had specific resistance. 'Heinong 37' was shown to be genetically distinct from SCN resistant cultivars with high yielding under continuous culture [17]. 'L-10' was resistant to different races of SCN (by our unpublished data). The objectives of the present study were to confirm the previously reported QTL, to identify new QTL associated with broad-spectrum resistance to multiple SCN races and to examine candidate genes underlying selected QTL using the cross of 'Heinong 37' and 'L-10'.

The SCN resistance line 'L-10' had a broad-spectrum resistant to SCN. There was great interest in transferring the SCN resistance of this cultivar into other cultivars. Traditional identification processes were time consuming and labor intensive. Increasing the selection intensity by marker-assisted selection of genotypes and subsequently more limited phenotypic selection will lead to improved the selection gains.

Segregation distortion had been reported repeatedly in soybean inbred line populations segregating for the soybean cyst nematode resistance locus rhg1, but the frequency of the SCN resistance allele was lower than expected [19]. In this study, QTL Qscn3-2 that was within the same region as the rhg1 locus was only detected in about one fourth of the population based on the flanking markers Satt275 and Satt309. The flanking markers Sat_350 and Satt244 for Qscn3-3 were only detected in nearly one third of the population. It was reported by Kopisch-Obuch et al. (2006) that the association between the resistance allele and reduced field emergence contributed to the segregation distortion at the SCN resistance loci [19]. In contrast, Webb et al. (1995) reported that either gametic or zygotic selection driven by a second locus on LG M that had to be inherited in phase were responsible [20]. In the present study, the RIL population performed a Skewness distribution due to non-typical quantitative trait of cyst nematode resistance. However, using the SAS normal logarithm conversion of the data resulted in normal distributions for these female index data, which could be used for QTL mapping.

Two major QTL (Qscn3-2 on LG G and Qscn3-3 on LG J) were identified to be significantly associated with the resistance to SCN race 3. The two QTL jointly explained about 40% of the total phenotypic variation. The QTL Qscn3-2 fell within a similar marker interval (Satt275-Satt309) previously reported associated with resistance to SCN race 3. This region was considered to be the rhg1 locus. The rhg1 locus had been shown to have an impact on SCN development in several resistance lines and provided resistance to many common SCN populations such as race 3 and 14. Many research groups have mapped the rhg1 locus on chromosome 18, to a location approximately 0.4 cM from Satt309 [21, 22].

The interval on LG J contained both Qscn3-3 and Qscn14-4 (including rhg5) might underlie multi-races resistance to SCN in soybean [14]. In this study, no QTL was identified on the LG A2. The Rhg4 locus was considered to be another main QTL for SCN resistance, tightly linked to I locus [13]. Four alleles were known at I, the dominant allele inhibited seed coat pigmentation, leading to possession of a yellow seed coat, whereas the recessive i allele conferred a pigmented seed coat. Two other alleles, iⁱ and i^k, gave rise to restricted pigmentation of the hilum and the saddle-shaped region, respectively [23]. In our work, there was no significant correlation between the color of seed coat (or hilum) and SCN resistance. This result indicated that the line 'L-10' had another locus for resistance to SCN race 3 independent of the Rhg4 locus. This locus may be similar to the LG J locus found in PI 88788 [7].

The SCN resistance alleles were previously associated with maturity, plant height, and lodging scores [19]. In contrast there were some rhg1 SCN resistance alleles from G. soja either had no effect or significantly (P < 0.05) enhanced yield compared with the susceptible alleles [24]. Here the correlation analysis suggested that some morphological traits such as plant height, color of pod coat, 100-seed weight and maturity were significantly associated with SCN resistance. Soybeans with tall stems, large seed size, late maturity and low protein content were correlated with SCN resistance for both race 3 and 14 (P < 0.05). The use of the 'L-10' SCN resistance loci would provide breeders with an alternative source of SCN resistance that was not associated with reduced yield.

Two SSR markers, Satt275 and Satt309 on MLG G, were tightly linked to the major QTL Qscn3-2. Sat_305 and Satt244 on MLG J were tightly linked to the QTL Qscn3-3 and Qscn14-4. The total phenotypic variation explained by Qscn3-2 and Qscn3-3 together was 39.0% and 41.3% for the year 2007 and 2008, respectively, and these four markers had potential use for marker-assisted selection of lines resistant to the SCN race 3. It was showed that nearly 80% of lines with the alleles of these four markers that were contributed by 'L-10' provided high resistance or completely immunity (FI = 0). As the rhg1 locus has been characterized as incompletely dominant, soybean lines heterozygous at this locus often allowed cyst formation at a rate intermediate between the genotype of homozygous resistant and homozygous susceptible [14, 25]. Here some lines were still heterozygous at the rhg1 locus after MAS, suggesting a more accurate gene based selection was needed for the greater accuracy [8, 26].

Qscn14-4 might underlie multi-races resistance in soybean. Kanazin et al. (1996) identified a gene cluster that contained eight RGAs on LG J in the similar region [27]. Further, six members of the RLK family were involved in one cluster on Chromosome J. Among these the RLK gene Glyma16g30760.1 was found to accumulate transcripts in the SCN resistance cultivar 'L-10' but not in 'Heinong 37' and the transcript abundance was higher in root than in leaf for 'L-10', suggesting that the gene might play a role in the SCN resistance to race 3 and race 14. Melito et al. (2010) suggested that the RLK at rhg1 was involved in root development but not nematode development as shown by others [25]. Unfortunately, the study of Melito et al. (2010) was flawed by the lack of an analysis of the target RLK protein abundance despite the available antibodies. Further studies of the RLKs at rhg1 and rhg5 will be needed for the clarification of the resistance mechanism and involvement of the candidate genes.

Glyma16g30760.1 was tightly linked to the QTL Qscn3-3 and Qscn14-4, indicating that this gene might be associated with both SCN race 3 and race 14 resistance. Therefore, a real-time PCR combined with molecular marker-based selection could be used for the future breeding of SCN resistant lines.

The four main SSR markers, Satt275, Satt309, Sat_305 and Satt244, which were tightly linked to the major QTL for resistance to SCN on LG G and J, would be candidates for marker-assisted selection of lines resistant to the SCN race 3. Among the six RLK genes, Glyma16g30760.1 was found to accumulate transcripts in the SCN resistance cultivar 'L-10' but not in 'Heinong 37'. The transcript abundance was higher in root than in leaf for 'L-10', suggesting that the gene might play a role in the SCN resistance to race 3 and race 14.