Natural variation in ARF18 gene simultaneously affects seed weight and silique length in polyploid rapeseed

Abstract

Seed weight (SW), which is one of the three major factors influencing grain yield, has been widely accepted as a complex trait that is controlled by polygenes, particularly in polyploid crops. Brassica napus L., which is the second leading crop source for vegetable oil around the world, is a tetraploid (4×) species. In the present study, we identified a major quantitative trait locus (QTL) on chromosome A9 of rapeseed in which the genes for SW and silique length (SL) were colocated. By fine mapping and association analysis, we uncovered a 165-bp deletion in the auxin-response factor 18 (ARF18) gene associated with increased SW and SL. ARF18 encodes an auxin-response factor and shows inhibitory activity on downstream auxin genes. This 55-aa deletion prevents ARF18 from forming homodimers, in turn resulting in the loss of binding activity. Furthermore, reciprocal crossing has shown that this QTL affects SW by maternal effects. Transcription analysis has shown that ARF18 regulates cell growth in the silique wall by acting via an auxin-response pathway. Together, our results suggest that ARF18 regulates silique wall development and determines SW via maternal regulation. In addition, our study reveals the first (to our knowledge) QTL in rapeseed and may provide insights into gene cloning involving polyploid crops.

Keywords: ARF18; cell growth; maternal effect; seed weight; silique length.

Fig. 1.

Fine mapping of the SW QTL in rapeseed using NILs and association populations. (A) SW in parents of zy72360 and R1. (Scale bar, 0.5 cm.) **P = 0.01. (B) SL in parents of zy72360 and R1. (Scale bar, 0.5 cm.) **P = 0.01. (C) The SW locus was detected on chromosome A09 in the F2 population. Positional cloning narrowed the SW locus to a 1.1-Mb region between P534 and BrBAC248. (D) Testing of six recombinant plants (BC3F4) narrowed the SW locus to the region between markers TSNP13 and TSNP16 (147 kb). The seeds from the main inflorescence were used to determine the SW for each plant. Open bars indicate the S1 homozygous regions, gray bars indicate the S2 homozygous regions, and black bars indicate the heterozygous regions. The SWs of the progenies were significantly different. (E) PCR identification narrowed the SW locus to the region between SNP3 and SNP5 (120 kb). (F) Scanning of the association of SW with 11 marker loci on the A09 linkage group in a rapeseed association population. Eleven maker loci are arranged on the horizontal axis according to their physical positions on chromosome A9.

Fig. 2.

Structural characterization and functional identification of ARF18. (A) ARF18 structure and mutation sites, including nucleotide substitutions and deletions, in zy72360 and R1. (B and C) Comparison of SW (B) and the SL (C) in Arabidopsis (vector control) and ARF18 transgenic lines in Arabidopsis. (Scale bars, 0.5 mm in B and 1 mm in C.) All data are expressed as mean ± SD. (D and E), Comparison of SW (D) and SL (E) in the zy72360 line with a promoter-driven expression of R1-ARF18. (Scale bars, 0.5 cm.) All data are expressed as mean ± SD. **P = 0.01.

Fig. S1.

SW analysis in transgenic Arabidopsis lines overexpressing BnaA09g55580D (A), BnaA09g55570D (B), BnaA09g55560D (C), and BnaA09g55520D (D). **P < 0.01.

Fig. S2.

Comparison of the ARF18 CDS sequences from zy72360 and R1.

Fig. S3.

(A) Real-time PCR analysis of ARF18 expression levels in the siliques of transgenic Arabidopsis lines and WT Arabidopsis control. (B) Real-time PCR analysis of ARF18 expression levels in the siliques of transgenic rapeseed lines overexpressing R1-ARF18 gene and zy72360 control. Data presented are mean values of three biological replicates; error bars indicate SDs. (C and D) Comparison of SW (C) and SL (D) in the Arabidopsis control and the arf18 mutant (CS719283). (Scale bars, 0.5 mm in C; 1 mm in D.) All data are shown as mean ± SD. **P < 0.01.

Fig. S4.

(A) Analysis of SW and SL in rapeseed lines overexpressing R1-ARF18 showing obvious phenotypes. *P < 0.05; **P < 0.01. (B) Analysis of SW and SL in transgenic Arabidopsis overexpressing R1-ARF18 (without exon 7). Values are shown as means ± SD. (C) Prediction of ARF18 protein structure using SWISS-MODEL. ARF18 from zy72360 was predicted to be a monomer, and ARF18 from R1 was predicted to be a homodimer. (D) SW of NIL (SW) × NIL (SW) F1, NIL (SW) × R1 F1, R1 × R1 F1, and R1 × NIL (SW) F1. Values are shown as means ± SD.

Fig. 3.

Comparative analysis of ARF18 proteins from zy72360 and R1. (A) Amino acid alignments of ARF18 proteins from other species with rapeseed ARF18. The black lines indicate the three conserved domains. (B) Nuclear localization of the ARF18-GFP fusion protein in tobacco epidermal cells. Cells transformed with the plasmid were viewed under a fluorescent filter to show GFP (Left), under a bright field for cell morphology (Center), and in combination (Right). (Scale bar, 50 μm.) (C) Putative conserved domains detected in ARF18 proteins from the alleles of zy72360 and R1.

Fig. 4.

Identification of incorrect splicing in ARF18 from zy72360 and comparison of inhibitory activity in ARF18 alleles. (A) Comparison of the ARF18 partial CDS and genomic sequences originating from zy72360 and R1. (B) The sixth intron from R1 and the corresponding sequence from the zy72360 gene were inserted into the GFP gene, which was driven by the CaMV 35S promoter and transformed into tobacco leaves. A is the intron from zy72360, and B is the intron from R1. (Scale bar, 100 μm.) (C) Effector genes that consisted of the CaMV 35S promoter encoding full-length ARF18 proteins were cotransfected into Arabidopsis suspension cell protoplasts with a DR5(7×)-GUS reporter gene. GUS activities were measured from protoplasts treated with 10 μM NAA. The data are representative of three independent experiments. All data are expressed as mean ± SD. **P = 0.01.

Fig. 5.

Comparison of inhibitory and binding activity of two ARF18 proteins. (A) Detection of ARF18 transcriptional activity in a yeast one-hybrid system. ARF18 genes from zy72360 and R1 were constructed as bait vector. +, positive control; −, empty control vector. (B) Detection of interaction in ARF18 proteins by yeast two-hybridization. (C) EMSA indicating that ARF18 from R1 binds to the biotin-labeled DR5 probe. The free probe is visible at the bottom of the gel, and complexes with the ARF18 protein are shifted upward.

Fig. 6.

ARF18 expression patterns by RT-qPCR and GUS assays. (A) Analysis of ARF18 expression in selected tissues in R1 using RT-qPCR. ENTH was used as the reference. Data are expressed as the means of three biological replicates; error bars indicate SDs. (B) GUS expression patterns(blue staining) at various stages in the ARF18 promoter-GUS transgenic line. GUS was strongly expressed in the cotyledon (cn), seedling (sl), older leaves (lf), and silique wall (sw). Moderate expression levels were detected in the bud (bd) and embryo (sd). No expression was observed in the seed coat (sd). (Scale bars, 5 mm for cotyledon, seedling, and older leaves; 0.8 mm for silique wall and bud; 0.1 mm for embryo.)

Fig. 7.

Analysis of silique phenotype and DEGs in the silique wall and seed of the NIL (SW) and R1 lines. (A) Comparisons of seed number per silique and silique length. (Scale bar, 0.5 cm.) **P = 0.01. (B) Comparison of cell length. (Scale bar, 50 μm.) **P = 0.01. (C) Differentially expressed genes analyses in the silique wall and seed between the NIL (SW) and R1 lines. (D) KEGG enrichment analysis of the regulated genes in the silique wall and seed.

Fig. S5.

Comparison of leaf size in seedlings (A), plant height (B), silique number (C), and effective branch number in the main inflorescence (D) in NIL (SW) and R1 plants. Data are shown as mean ± SD. **P < 0.01.

Fig. S6.

Comparative analysis of BnaA.ARF18.a and BnaC.ARF18.a. (A) Expression levels of BnaA.ARF18.a and BnaC.ARF18.a in silique wall and seed in NIL (SW) and R1 lines. (B) Comparison of the promoter sequences of BnaA.ARF18.a and BnaC.ARF18.a from the R1 line.