Independent FLC mutations as causes of flowering-time variation in Arabidopsis thaliana and Capsella rubella

Abstract

Capsella rubella is an inbreeding annual forb closely related to Arabidopsis thaliana, a model species widely used for studying natural variation in adaptive traits such as flowering time. Although mutations in dozens of genes can affect flowering of A. thaliana in the laboratory, only a handful of such genes vary in natural populations. Chief among these are FRIGIDA (FRI) and FLOWERING LOCUS C (FLC). Common and rare FRI mutations along with rare FLC mutations explain a large fraction of flowering-time variation in A. thaliana. Here we document flowering time under different conditions in 20 C. rubella accessions from across the species' range. Similar to A. thaliana, vernalization, long photoperiods and elevated ambient temperature generally promote flowering. In this collection of C. rubella accessions, we did not find any obvious loss-of-function FRI alleles. Using mapping-by-sequencing with two strains that have contrasting flowering behaviors, we identified a splice-site mutation in FLC as the likely cause of early flowering in accession 1408. However, other similarly early C. rubella accessions did not share this mutation. We conclude that the genetic basis of flowering-time variation in C. rubella is complex, despite this very young species having undergone an extreme genetic bottleneck when it split from C. grandiflora a few tens of thousands of years ago.

Figure 1

Distribution of flowering time among 20 C. rubella accessions in 16LD and16LDv. Note the differences in scale. Accessions, which are ordered by mean flowering time in each condition, are color coded for comparison. For each accession, the number of individuals that flowered on a given day is indicated.

Figure 2

Relationship of flowering-time variation and vernalization response. (A) Reaction norms of flowering time in 16LD without and with vernalization (−vern, +vern). (B) Correlation of mean flowering times with and without vernalization; same data as in Figure 2A.

Figure 3

Flowering time in 16LD among 305 F₂ individuals from the MTE × 1408 cross. Average and range of flowering times of grandparents are indicated. Darker shading indicates the 61 individuals used for bulked segregant mapping.

Figure 4

SHOREmap analysis of early flowering QTL. The homozygosity estimator is 0 at even allele frequencies of both parents, 1 when homozygous for late-flowering accession MTE, and −1 when homozygous for early flowering accession 1408 (Schneeberger et al. 2009). Sliding windows of 100 kb with step size 10 kb were used. The region of chromosome 6 enriched at over 600 markers indicative of 1408 is indicated in gray. The eight largest scaffolds of a preliminary C. rubella genome assembly are shown, corresponding to the majority of the eight chromosomes.

Figure 5

Relationship between FLC expression measured by qRT–PCR and flowering time of accessions. (A) Correlation between FLC expression and flowering without vernalization. The lowest value was arbitrarily set to 1. (B) Correlation between FLC expression and vernalization response. (C) Effect of vernalization (vern) on FLC expression. The difference between FLC and BETA-TUBULIN, ΔcT, was used to calculate FLC expression, assuming PCR efficiency was 100% for both genes. Two biological replicates, each with two technical replicates, were analyzed.

Figure 6

Sequence variation of FLC and FRI in C. rubella. In addition to the FLC point mutation at the end of the last intron in accession 1408, two insertions that are segregating in the population, are shown. Thin lines indicate introns; thick lines protein-coding sequences. The C → G substitution creates a new splice acceptor site. PCR amplification and sequencing of cDNA confirmed that 2 bases are inserted into the FLC mRNA in accession 1408. There was no apparent heterogeneity in the sequence, suggesting that the canonical splice form, if it exists at all, is rare. For FRI, amino acid residues that appear to be ancestral based in several Brassicaceae (Irwin et al. 2012) are shown in boldface type; numbers indicate position in the peptide sequence. Shaded numbers indicate affected accessions.

Figure 7

Functional analysis of C. rubella FLC alleles in A. thaliana. (A) Flowering time of plants grown in 23LD. The three left strains are controls. fri^Col-0 FLC^Col-0 is the Col-0 reference strain, which carries a functional FLC allele, but a naturally inactive FRI allele (Johanson et al. 2000). FRI^Sf-2 flc-3 carries an introgression of the functional FRI allele from the Sf-2 accession and has also an induced mutation at the FLC locus (Michaels and Amasino 1999). The three right strains are transgenic lines, expressing the indicated FLC allele from the constitutive CaMV 35S promoter. The effect of the C. rubella MTE allele is similar to that of the fully functional A. thaliana Col-0 allele. n ≥ 30. (B) Expression of FLC and of two downstream flowering regulators in transgenic and non-transgenic FRI^Sf-2 flc-3 plants, determined by qRT–PCR using the 2^−ΔΔcT method. Expression levels were normalized to those of wild-type Col-0 plants. Averages for four individuals for controls and eight for transgenic lines are shown. Error bars indicate standard errors of the mean. Differences between groups are significant at P < 0.05. (C) cDNA analysis of MTE and 1408 alleles expressed in transgenic A. thaliana plants. The splicing variant generated by the SNP in 1408 abolishes a BstXI restriction site in the amplified cDNA fragment.