The female parent GH06 was derived from the selfing generations of yellow-seeded B. napus, with 73.34 μmol g^-1 of glucosinolate content. The male parent, Pl74, was derived from selfing generation of Youyan 2, with 18.75 μmol g^-1 of glucosinolate content. The recombinant inbred lines (RILs) were the ninth generation by successively selfing from the GH06 × P174 cross. In this study, 172 RILs and their parents were planted in a randomized complete block design with three replicates in experimental field of Southwest University in 2011 and 2012. Each plot contained 20 plants in two rows, with 0.5 m between rows and 0.2 m between plants.

Selfing seeds in each line from three plants were mixed as a sample. The total seed glucosinolate content and its components were tested using near infrared spectroscopy (NIR) by means of WinISI II software of an NIR System 6500. Each assay had three technical replicates, and the mean value was considered as a measurement.

The Brassica 60 K SNP BeadChip Array was used to genotype 172 lines and parental plants. This array, which successfully assays 52 157 Infinium Type II SNP loci in B. napus, was developed by an international consortium using preferentially single-locus SNPs contributed from genomic and transcriptomic sequencing in genetically diverse Brassica germplasm (Isobel Parkin, Agriculture and AgriFood Canada, unpublished data).

Total genomic DNA was extracted from well- expanded leaves of RIL and the two parental lines, following the methodology of DP321-03 DNA extraction kits (Tiangen, Beijing, China). DNA sample preparation, hybridisation to the BeadChip, washing, primer extension and staining were performed according to the work flow described in the Infinium HD Assay Ultra manual. Imaging of the arrays was performed using an Illumina HiSCAN scanner after BeadChip washing and coating. Allele calling for each locus was performed using the GenomeStudio genotyping software V.2011 (Illumina, Inc.). SNP markers were named using SNP plus index numbers assigned by GenomeStudio, followed by the chromosome number.

A high-density genetic linkage map for the complex allotetraploid species B. napus (oilseed rape) was constructed using the above-referenced RIL population. This map was composed of genome-wide single nucleotide polymorphism (SNP) markers assayed by the Bra s sica 60 K Infinium BeadChip Array. Composite interval mapping (CIM) was carried out using the software package Windows QTL Cartographer 2.5^{[ 14]}. The LOD thresholds for QTL significance were determined by a permutation test (1000 replications) with a genome-wide significance level P=0.01 to judge whether there exist QTLs. Running result of software can show additive effects of QTLs and phenotypic contribution rate. QTL nomenclature followed the protocol of McCouch et al.^{[ 15]}. Take “ qGSL- A01-11” as an example, “q” refers to QTL; “GSL” refers to glucosinolate content, “A01” indicates the linkage group, and the number “11” indicates the year of field experiment (the first “1”, short form of 2011) and the series number of the QTL (the second “1”). This number is in one-digit for some QTLs, which means the QTL was identified in both years.

The total glucosinolate content of the parents (GH06 and P174) were 73.34 and 18.75 μmol g^-1, respectively, with significant difference in B. napus that provided ideal material for QTLs analysis. The content of the RIL population was distributed from 12.00 to 117.75 μmol g^-1 with an average of 63.9 μmol g^-1 in 2011, and from 13.50 to 128.80 μmol g^-1 with an average content was 70.80 μmol g^-1 in 2012 (Table 1). As expected, the two years total glucosinolate contents were significantly correlated ( P< 0.01, R²= 0.997). The seed glucosinolate content from 50 to 100 μmol g^-1 accounted for 80.6% of the RIL population (Fig. 1). The RIL population showed a continuous and bell-shaped distribution for glucosinolate content in two years, suggesting the multi-genic control of glucosinolate in B. napus. However, transgressive segregations in the RIL population were observed, indicating that glucosinolate synthesis in seeds was controlled by some genes in the two parents.

Table 1

Table 1

Table 1 Variation of glucosinolate content in the RIL population of B. napus Year Max (μmol g-1) Min (μmol g-1) Average (μmol g-1) Standard deviation Coefficient of variation (%) 2011 117.5 12.0 63.9 21.4 33.5 2012 128.8 13.5 70.8 24.9 35.1

Table 1 Variation of glucosinolate content in the RIL population of B. napus

Fig. 1

	Figure Option ViewDownloadNew Window
	Fig. 1 Frequency distribution of seed glucosinolate content in the RIL populations over two years

The SNP linkage map contains 9164 SNP markers covering 1832.9 cM, with 1232 bins accounting for 7648 of the markers. A subset of 2795 SNP markers, with an average distance of 0.66 cM between adjacent markers, was applied for QTL mapping of seed glucosinolate content. Five significant QTLs for the total glucosinolate content in B. napus seed were detected on chromosome A03, C02, and A09 by analysis of the RIL population harvested in 2011 and 2012 (Fig. 2). LOD threshold values for significant QTLs in 2011 and in 2012 were determined to be 3.01-10.43 by the permutation test (Table 2). These QTLs explained 56.9% and 55.1% of the total phenotypic variance in 2011 and 2012, respectively. Locus qGSL-A09-2 detected in 2011 shared the highest contribution value (26.2%), while the QTL qGSL-C02-1 detected in 2011 explained the lowest phenotypic variance (4.9%). Four of these QTLs had the positive additive effect, indicating the female parent GH06 contributed to a strong increase in seed glucosinolate content except for the qGSL-C02-1. The qGSL-A09-2 and qGSL-A09-3 detected in 2011 explained the total phenotypic variances of 26.2% and 11.1%, respectively, and in 2012 were 19.0% and 16.5%, respectively.

Besides, five QTLs for total glucosinolate content were identified in either 2011 or 2012, which were mapped on chromosomes A01, A07, A09, C02, and C06. A single QTL explained 4.1%-7.9% of the phenotypic variance and the LOD threshold value varied from 2.18 to 3.93 (Table 2). Three QTLs ( qGSL-A01- 11, qGSL-A07-12, and qGSL-C02-12) had negative effect, showing the additive alleles from parent P174.

Fig. 2

	Figure Option ViewDownloadNew Window
	Fig. 2 QTLs associated with total seed glucosinolate content detected in the RIL from a cross of GH06×P174The bar length on the right represents one LOD supporting interval for each QTL and only QTL region markers are displayed on the linkage map.

Table 2

Table 2

Table 2 Quantitative traits loci involved in the seed gluconsinolate content of RIL populations QTL Chr. Marker interval 2011 RILs 2012 RILs LOD Additive value R2(%) LOD Additive value R2(%) qGSL-A03-1 A03 SNP5306A03-SNP5100A03 3.01 9.03 8.80 4.92 10.91 6.90 qGSL-A09-2 A09 SNP19550A09-SNP20943A09 10.43 18.39 26.20 7.01 18.12 19.00 qGSL-A09-3 A09 SNP20854A09-SNP20745A09 3.94 11.65 11.10 5.35 15.44 16.50 qGSL-C02-1 C02 SNP30935-SNP38136C02 3.23 -9.01 4.90 3.44 -6.24 5.90 qGSL-A01-11 A01 SNP2662A01-SNP2481A01 3.30 -5.81 5.20 — — — qGSL-A07-12 A07 SNP16102A07-SNP16204A07 — — — 3.53 -5.07 4.10 qGSL-A09-11 A09 SNP21773A09-SNP20222A09 3.85 11.98 7.90 — — — qGSL-C02-12 C02 SNP46003C02-SNP34478 — — — 3.93 -9.02 5.70 qGSL-C06-11 C06 SNP24582C06-SNP28486C06 3.18 5.06 5.40 — — — A: additive effect; R2: phenotypic variation explained.

Table 2 Quantitative traits loci involved in the seed gluconsinolate content of RIL populations