The dominance model for heterosis explains culm length genetics in a hybrid sorghum variety

Abstract

Heterosis helps increase the biomass of many crops; however, while models for its mechanisms have been proposed, it is not yet fully understood. Here, we use a QTL analysis of the progeny of a high-biomass sorghum F₁ hybrid to examine heterosis. Five QTLs were identified for culm length and were explained using the dominance model. Five resultant homozygous dominant alleles were used to develop pyramided lines, which produced biomasses like the original F₁ line. Cloning of one of the uncharacterised genes (Dw7a) revealed that it encoded a MYB transcription factor, that was not yet proactively used in modern breeding, suggesting that combining classic dw1or dw3, and new (dw7a) genes is an important breeding strategy. In conclusion, heterosis is explained in this situation by the dominance model and a combination of genes that balance the shortness and early flowering of the parents, to produce F₁ seed yields.

Figure 1

Tentaka is an F₁ hybrid variety that shows typical heterosis, (a–c) Plant stature of the F₁ and its parental lines. (a) MS79 (seed parent). (b) Tentaka (F₁); and (c) 74LH3213 (pollen parent). Scale bars mean 1 m. (d) Flowering date (FD) and (e) Culm length (CL) after flowering were investigated in the plants from the parental lines and the F₁ generation (n ≥ 6, means ± SD). (f) An F₂ population from a cross between MS79 and 74LH3213 was analysed for FD and CL. QTLs for FD and CL with P values < 0.05 are presented on the graphical map of the chromosomes, above which the chromosome numbers were mentioned. The positions and names of the DNA markers used for the analysis are indicated on the left and right sides, respectively. The QTL name is indicated in red.

Figure 2

Allelic effects for FD and CL in the F₂ population, and the candidate corresponding genes for the QTLs. (a) Allelic effects of FD and CL in the F₂ population, as determined from the nearest markers of the identified QTLs. Each plotted point indicates an individual plant in the F₂ population. A, H, and B on the x-axis indicate the homozygous MS79 allele, the heterozygous MS79/74LH3213 allele, and the homozygous 74LH3213 allele, respectively. Only the data from 2015 are shown as a representative sample. (b–e) Exons and introns are indicated with boxes and black lines, respectively. The dark blue boxes indicate protein coding sequences; the grey boxes indicate untranslated regions (UTR). The alleles in SbPhyB (b), SbGhd7, (c), Dw3 (d), and Dw1 (e) are shown.

Figure 3

Culm stature of the pyramided lines with five dominant alleles (i5). (a) The five dominant alleles (qFD-1/qCL-1, qFD-6/qCL-6, qCL-7a/Dw7a qCL-7b/Dw3, and qCL-9/Dw1) were pyramided by backcrossing and self-pollination, resulting in the development of the i5 lines (BC₁F₃ or BC₁F₄) (for details, see materials and methods). MAS indicates DNA marker-assisted selection for the pyramiding of the five homozygous dominant alleles. (b) Elongation patterns for the internode of an F₁ hybrid variety ‘Tentaka’ (left) and the pyramided line with five dominant alleles (i5; right). Scale bar, 1 m. (c–f) Flowering date (c,d) and culm length (e,f) of the MS79 (seed parent), 74LH3213 (pollen parent), Tentaka, and i5. Evaluations of the BC₁F₃ in 2018 (c,e) and BC₁F₄ in 2019 (d,f) are shown. The red numbers in panels (e) and (f) indicate percentages of the i5 when each score of Tentaka is 100.

Figure 4

Plant stature and the elongation patterns of the BIL-Dw7a and BIL-dw7a internodes. (a) Plant stature of BIL-Dw7a and BIL-dw7a. Scale bars = 1 m. (b,c) Patterns of internode elongation. (b) Photograph of the internodes of BIL-Dw7a (left) and BIL-dw7a (right). The white arrowhead indicates the position of the node and the Roman numerals indicate the internode number. Scale bar = 30 cm. (c) The length of each internode of BIL-Dw7a and BIL-dw7a. Four representative plants are shown from each line.

Figure 5

Positional cloning of Dw7a and phenotypes of the CRISPR/Cas9 mediated knock-out lines of OsDw7a in rice. (a) Physical position of Dw7a (corresponding gene of qCL-7a). The uppermost grey bar indicates chromosome 7. On the upper horizontal line, the vertical lines indicate the physical positions of the DNA markers (Mb), and the number of recombinants is shown in parentheses between the markers. The bottom horizontal line represents the candidate region of 21 kb. Exons are represented as black boxes. (b) Gene structure of Dw7a (Sobic.007G137101). Exons are represented as dark blue boxes and the 3ʹ or 5ʹ-untranslated regions (UTR) are indicated with grey boxes. The bold lines indicate the R2R3 MYB binding domain. (c) Four SNPs (indicated by nucleotides), three In/Dels (black/white triangles, respectively), genomic fragment substitution (260 bp to 217 bp; striped bar) in the promoter region (dotted and striped bares), and one SNP and one deletion in the 5ʹ-UTR region (grey bar) are shown. (d) The expression levels of the Dw7a in the BIL-Dw7a and BIL-dw7a (two-tailed Student’s t-test, *P < 0.05, n = 3). (e) The expression levels of Dw7a (in cultivar SIL-05) in different organs. Scale bar in the photograph = 2 cm. The positions of the two internodes ‘before elongation’ and ‘elongating’ are shown in the photograph. (f) Phenotype at the heading stage of the WT and two knockout lines. Scale bar = 20 cm. (g) Length of each internode of the WT and knockout lines (two-tailed Student’s t-test, **P < 0.01, n = 6).

Figure 6

Haplotype analyses of Dw7a and Dw1. (a,b) Regional distribution of each haplotype of the Dw7a (a) and Dw1 (b). (c) Haplotype network of Dw7a and Dw1 using 187 accessions. Hatch marks between each haplotype indicate the mutational step. In Dw7a, Hap1 to Hap11 were used for analysis (147 accessions). (d,e) Extended haplotype homozygosity (EHH) decay of Dw7a (d) and Dw1 (e). Hap1 (blue), Hap2 (orange), and Hap3 (green) are shown. Dotted lines indicate the positions of Dw7a (d) and Dw1 (e).