Molecular characterization of Rht-1 dwarfing genes in hexaploid wheat

Abstract

The introduction of the Reduced height (Rht)-B1b and Rht-D1b semidwarfing genes led to impressive increases in wheat (Triticum aestivum) yields during the Green Revolution. The reduction in stem elongation in varieties containing these alleles is caused by a limited response to the phytohormone gibberellin (GA), resulting in improved resistance to stem lodging and yield benefits through an increase in grain number. Rht-B1 and Rht-D1 encode DELLA proteins, which act to repress GA-responsive growth, and their mutant alleles Rht-B1b and Rht-D1b are thought to confer dwarfism by producing more active forms of these growth repressors. While no semidwarfing alleles of Rht-A1 have been identified, we show that this gene is expressed at comparable levels to the other homeologs and represents a potential target for producing novel dwarfing alleles. In this study, we have characterized additional dwarfing mutations in Rht-B1 and Rht-D1. We show that the severe dwarfism conferred by Rht-B1c is caused by an intragenic insertion, which results in an in-frame 90-bp insertion in the transcript and a predicted 30-amino acid insertion within the highly conserved amino-terminal DELLA domain. In contrast, the extreme dwarfism of Rht-D1c is due to overexpression of the semidwarfing Rht-D1b allele, caused by an increase in gene copy number. We show also that the semidwarfing alleles Rht-B1d and Rht-B1e introduce premature stop codons within the amino-terminal coding region. Yeast two-hybrid assays indicate that these newly characterized mutations in Rht-B1 and Rht-D1 confer "GA-insensitive" dwarfism by producing DELLA proteins that do not bind the GA receptor GA INSENSITIVE DWARF1, potentially compromising their targeted degradation.

Figure 1.

Phenotypes of Rht-B1 and Rht-D1 dwarfing alleles in NILs. Wheat NILs (var Mercia) were grown to maturity. The photograph is from the John Innes Centre archives (produced by Tony Worland). [See online article for color version of this figure.]

Figure 2.

Predicted RHT-1 sequences. A, Amino acid sequence alignment of RHT-A1A, RHT-B1A, and RHT-D1A proteins. The amino acid sequences are predicted from nucleotide sequences amplified from var Cadenza. Gaps introduced to improve the sequence alignment are indicated by dots. Conserved N-terminal regulatory motifs (DELLA, LExLE, and TVHYNP) are indicated by thick solid lines above the sequences. The C-terminal (LHR1, VHIID, LHR2, PFYRE, and SAW) functional domains are indicated by thick dashed lines above the sequences. B, Comparison of the nucleotide sequences of the Rht-1 homeologs across the conserved N-terminal coding region. SNPs between the Rht-1 homeologs are indicated by lowercase letters and asterisks above the sequences. The Rht-B1b and Rht-D1b point mutations are shown, and their positions are indicated by arrows above the sequences. The predicted amino acid sequences of RHT-A1A, RHT-B1A, and RHT-D1A are identical in this region.

Figure 3.

Expression analysis of the three Rht-1 homeologs and two GA biosynthesis genes in the upper expanding tissues of the developing wheat stem. A to C, Relative expression levels of Rht-A1, Rht-B1, and Rht-D1 (A), TaGA20ox1 (B), and TaGA3ox2 (C) in different regions of the extending wheat stem at 7 weeks post germination. The P-1 internode was divided into three equal sections for tissue collection. The mRNA levels of these genes were determined by qRT-PCR from three biological replicates for each sample type, normalized against UBIQUITIN. Results are plotted as the ratio to the lowest detectable level ± se. D to F, Localization of GUS activity under the control of the Rht-A1 (D), TaGA20ox1 (E), and TaGA3ox2 (F) promoters in an elongating wheat stem at the same stage of development as analyzed by qRT-PCR. The positions of the peduncle node (PN) and the P-1 node (P-1N) are indicated. The positions of the upper (UI), middle (MI), and lower (LI) internodes are indicated in D. Bars = 5 mm.

Figure 4.

Rht-1 allele characterization. A, Amino acid alignment of RHT-B1A and RHT-B1C. The translated RHT-B1C protein contains a 30-amino acid insertion close to the DELLA motif and two upstream amino acid substitutions (indicated with arrows above the sequences; G15R and M25I). Conserved N-terminal regulatory (DELLA, LExLE, and TVHYNP) domains are indicated by lines above the sequences. B, Amino acid alignment of the N-terminal domains of Rht-B1 alleles. Rht-B1d carries the same point mutation as Rht-B1b, while Rht-B1e expresses a point mutation that introduces a premature stop codon three amino acids farther upstream. C to E, mRNA levels for Rht-1 in 2-week-old seedlings of wheat NILs in the Mercia variety, determined by qRT-PCR using primers designed to amplify all three homeologs. Values shown are means of three biological replicates, normalized against UBIQUITIN. Results are plotted as the ratio to the level in the Rht-1 control ± se. *** Significantly different from the Rht-1 control (P < 0.01). F to H, Gene copy number of the Rht-1 homeologues in the Rht-D1 allelic series (NILs in the Mercia variety) determined by qPCR of gDNA template. The relative gDNA copy number was normalized against UBIQUITIN. Results are plotted as the ratio to the level in the Rht-1 control. Values shown are means of three biological replicates ± se. *** Significantly different from the Rht-1 control (P < 0.01). I, Hybridization of D1-2855 to a Southern blot containing gDNA isolated from wheat lines containing different Rht-D1 alleles: Rht-D1a Mercia (Rht-1) control, lanes 1 and 7; Rht-D1b Mercia NIL, lanes 2 and 8; Rht-D1c Mercia NIL, lanes 3 and 9; Rht-D1d Mercia NIL, lanes 4 and 10; Rht-D1c Ai-bian 1, lanes 5 and 11; Rht-D1d Ai-bian 1a, lanes 6 and 12, digested with DraI (lanes 1–6) or EcoRI (lanes 7–12).

Figure 5.

Interaction between wild-type and mutant forms of RHT-B1 and TaGID1 in yeast two-hybrid assays. A, Schematic diagram showing full-length and mutant forms of RHT-B1, which were expressed as LexA activation domain (AD) fusions in yeast. Numbers indicate the amino acid positions within RHT-B1. The black box in the RHT-B1C schematic diagram indicates the position of the 30-amino acid insertion. The full-length TaGID1 was expressed as a LexA DNA-binding domain (DB) fusion protein in yeast. B, Interactions between activation domain and DNA-binding domain protein fusions were determined in cotransformed L40 yeast cells by scoring growth on His⁻ medium containing 3-aminotriazole (3-AT; 0, 2, 5, 10, 30, and 60 mm). Data presented are maximum concentrations of 3-AT at which growth was observed; dashes signify no growth on 2 mm 3-AT. Quantitative values for each interaction were determined by measuring β-Gal activity in the presence and absence of 100 μm GA₃. Values are means of three biological replicates ± se.