The role of recently derived FT paralogs in sunflower domestication

Abstract

Gene duplication provides an important source of genetic raw material for phenotypic diversification, but few studies have detailed the mechanisms through which duplications produce evolutionary novelty within species. Here, we investigate how a set of recently duplicated homologs of the floral inducer FLOWERING LOCUS T (FT) has contributed to sunflower domestication. We find that changes in expression of these duplicates are associated with differences in flowering behavior between wild and domesticated sunflower. In addition, we present genetic and functional evidence demonstrating that a frameshift mutation in one paralog, Helianthus annuus FT 1 (HaFT1), underlies a major QTL for flowering time and experienced a selective sweep during early domestication. Notably, this dominant-negative allele delays flowering through interference with action of another paralog, HaFT4. Together, these data reveal that changes affecting the expression, sequence, and gene interactions of HaFT paralogs have played key roles during sunflower domestication. Our findings also illustrate the important role that evolving interactions between new gene family members may play in fostering phenotypic change.

Figure 1. Recent Duplications of Sunflower FT-like Genes

Maximum likelihood phylogeny based on amino acid sequences of plant FT proteins. Bootstrap percentages > 50% shown above branches. Species abbreviations: Antirrhinum majus, Am; Arabidopsis thaliana, At; Citrus unshiu, Ci; Cucurbita maxima, Cm; Helianthus annuus, Ha; Ipomea nil, In; Lactuca sativa, Ls; Malus x domestica, Md; Oryza sativa, Os; Populus nigra, Pn; Solanum lycopersicum, Sl; and Vitis vinifera, Vv. See also Table S1 and Table S6.

Figure 2. Gene Structure, Map Position, and Spatial Expression of HaFT paralogs

(A) Exon-intron structure of HaFT coding sequences shown to scale. The open rectangle in HaFT1 denotes an alternatively spliced exon. Numbered sites mark the locations of the frameshift in HaFT1 (1) and putative loss-of-function mutations in HaFT3 segregating in natural populations (2–5). HaFT3 mutations include a 17.25 kb insert in the third intron (2), a 7 bp deletion (3), a 1 bp deletion (4), and two cosegregating premature stop mutations (5). (B) Genetic map of LG6 indicating the map positions of HaFT1, HaFT2, and HaFT3 relative to previously mapped QTL region (black) [35], and relative to the wild introgression into a domesticated background in NILs (blue, introgression end points occur within light blue regions). (C) Spatial expression of HaFT1, HaFT2, and HaFT4 in wild sunflower assayed by RT-PCR. See also Figure S1 and Table S3.

Figure 3. Frameshift-Carrying Domesticated HaFT1 Allele Has Dominant-Negative Effect

(A) Overexpression of in-frame wild allele of HaFT1 complements the A. thaliana ft mutant whereas overexpression of the frameshift-carrying domesticated allele does not. Plants photographed 36 days after germination. (B and C) Days to budding (B) and rosette leaf number (C) of plants overexpressing 0, 1, or 2 HaFT paralogs in an ft background. 35S::HaFT1-W plants overexpressed the wild, in-frame HaFT1 allele; 35S::HaFT1-D plants overexpressed the domesticated, frameshift-carrying HaFT1 allele. Means for each genotype were compared with a general linear model, and pairwise comparisons were performed with Tukey's multiple comparison test. Different letters above the 95% confidence intervals denote significantly different phenotype distributions. See also Figure S2 and Table S4.

Figure 4. Flowering Time, HaFT1 Expression, and HaFT2 Expression in Parental and NIL Plants

(A) Days to budding of wild (black) and domesticated (orange) parents in three photoperiods. Mean ± SE shown. (B) Days to budding of domesticated (CMS, orange) and NIL plants (blue) homozygous for the domesticated LG6 QTL region (DD), heterozygous (DW), and homozygous wild (WW) in short and long days. 95% confidence intervals shown. Differences among genotypes tested by general linear model corrected with Tukey’s multiple comparison test, **, p < 0.02; *, p = 0.058. (C) HaFT1 shoot apex expression in developing wild and domesticated parents in long days. Age measured as days after sowing. T8 plants experienced 20 long days followed by 10 short days. (D) HaFT1 shoot apex expression in CMS, DD, DW, and WW plants 15 and 25 days after sowing in short days. T16 plants experienced 15 short days followed by 10 long days. (E) HaFT2 leaf expression in wild (black) and domesticated (orange) parents every 4 hr on the 30th day after sowing. Mean ± SE for three biological replicates shown. (F) HaFT2 leaf expression in long-day grown DD, DW, and WW plants 4 hr after dawn, 30 days after sowing. Mean of three technical replicates per biological replicate (open circles) and mean ± SE for three biological replicates (filled circles) shown. Relative expression expressed as delta-delta-Ct normalized to 60S rRNA and scaled to the highest individual measurement. See also Figure S3 and Table S5.

Figure 5

HaFT1 Frameshift Distinguishes Wild From Domesticated Accessions and Experienced a Selective Sweep During Domestication. (A) Median joining haplotype network constructed from 711bp region of HaFT1 sequenced from elite-bred (red), landrace (yellow), and wild (green) H. annuus and wild H. argophyllus (black). TG→C frameshift mutation (*) and a noncoding SNP define a branch separating nearly all domesticated lines from nearly all wild accessions. The number of hatchmarks on a branch indicates the number of substitutions. (B) Average pairwise nucleotide diversity (π) for HaFT1 and five putative neutral loci in samples of elite-bred, Native American landrace, and wild H. annuus. See also Table S2.