A Vernalization Response in a Winter Safflower (Carthamus tinctorius) Involves the Upregulation of Homologs of FT, FUL, and MAF

Abstract

Safflower (Carthamus tinctorius) is a member of the Asteraceae family that is grown in temperate climates as an oil seed crop. Most commercially grown safflower varieties can be sown in late winter or early spring and flower rapidly in the absence of overwintering. There are winter-hardy safflower accessions that can be sown in autumn and survive over-wintering. Here, we show that a winter-hardy safflower possesses a vernalization response, whereby flowering is accelerated by exposing germinating seeds to prolonged cold. The impact of vernalization was quantitative, such that increasing the duration of cold treatment accelerated flowering to a greater extent, until the response was saturated after 2 weeks exposure to low-temperatures. To investigate the molecular-basis of the vernalization-response in safflower, transcriptome activity was compared and contrasted between vernalized versus non-vernalized plants, in both 'winter hardy' and 'spring' cultivars. These genome-wide expression analyses identified a small set of transcripts that are both differentially expressed following vernalization and that also have different expression levels in the spring versus winter safflowers. Four of these transcripts were quantitatively induced by vernalization in a winter hardy safflower but show high basal levels in spring safflower. Phylogenetic analyses confidently assigned that the nucleotide sequences of the four differentially expressed transcripts are related to FLOWERING LOCUS T (FT), FRUITFUL (FUL), and two genes within the MADS-like clade genes. Gene models were built for each of these sequences by assembling an improved safflower reference genome using PacBio-based long-read sequencing, covering 85% of the genome, with N50 at 594,000 bp in 3000 contigs. Possible evolutionary relationships between the vernalization response of safflower and those of other plants are discussed.

Keywords: PacBio; flowering time; genome; safflower; vernalization.

FIGURE 1

Schematic overview of the subfamilies of the Asteraceae family based on Mandel et al. (2019). Central Asteraceae crop species are listed in their respective subfamilies. Gray stars indicate species that have a documented vernalization response in flowering. Subfamily assignments are based on Fu et al. (2016) and Mandel et al. (2019). Note that two of the subfamilies (Carduoideae and Cichorioideae) are paraphyletic and are thus marked with boxes.

FIGURE 2

The effect of vernalization on flowering time for S317 and C311. Germinated seeds were exposed to increasing longer durations of a vernalization (4°C) treatment. In C311, there is a steep decrease in the time to flowering as the length of vernalization increases and the relationship between duration of vernalization and days to heading was significant. This slope levels out at approximately 17 days and does not decrease the time to flowering after further exposure to vernalization conditions.

FIGURE 3

Differential expression of key safflower transcripts with vernalization treatment and genotype. Plotted on the y-axes are the average normalized transcript read counts, from three biological replicates for key transcripts from transcriptomes of non-vernalized plants (V0) or plants that had been vernalized for 5, 10, 15, or 20 days. Data are presented for the winter safflower accession (C311, blue line), showing the impact of increasing durations or vernalization pre-treatment (top row). Then, gene expression in the winter type is contrasted with the spring cultivar (S317, gray line). Error bars show standard error. Statistical tests include Student’s t-test comparison to the non-vernalized control for time course analysis of expression in the winter safflower (when plotted alone) or comparison between the spring versus winter safflower when genotypes are compared (NS, non-significant, *P < 0.05, **P < 0.01, ***P < 0.001). Expression not detected in some samples (ND), resulting in an absolute or presence/absence contrast (+).

FIGURE 4

Maximum likelihood phylogeny of the MIKC-MADS gene family in Arabidopsis thaliana (in green) and Helianthus annum (in red). Transcripts of Carthamus tinctorius (in blue) found to be differentially expressed in response to vernalization were included. Bootstrap support values are shown for nodes with higher than 50% support. CtMAF1 sequences from spring safflower (S317_tr33367.4 and S317_tr33367.5) appear to be isoforms of the same transcript, with variations found in the 3′ end of S317_tr33367.4. Similarly, the CtMAF1 sequences from the winter safflower transcriptome (C311_tr20021.6 and C311_tr23886.4) also appear to be isoforms of the same transcript, with a 5′ truncation and single codon gap in C311_tr23886.4. Information about the genes can be found in Supplementary Table 2.

FIGURE 5

Maximum likelihood phylogeny of the FT gene family in Arabidopsis thaliana (in green) and Helianthus annum (in red). Transcripts of Carthamus tinctorius (in green) found to be differentially expressed in response to vernalization were included. Bootstrap support values are shown for nodes with higher than 50% support. Regarding the two transcripts from the winter safflower transcriptomic assembly, namely C311_tr57705.6 and C311_tr93957.5. The former is a transcript containing a truncation at the 3′ end that closely resembles the CtFT1 from the spring safflower transcriptome (S317_tr32761.4). The latter closely resembles C311_tr57705.6 but also contains a 5′ truncation. Information about the genes can be found in Supplementary Table 2.

FIGURE 6

Gene models for (A) CtFT1, (B) CtFUL1, (C) CtMAF1, and (D) CtMAF2 based on alignments of transcript with the genome assembly of S317.

FIGURE 7

Relationships between phylogeny and seasonal gene expression signatures. Inferred seasonal gene-expression states of flowering regulators in plants related to phylogenetic relationships. Boxes filled green indicates whether a gene is actively expressed in either autumn or spring. Gray filled boxes indicate a gene is repressed in autumn or spring. White box denotes that the gene has not been found in a plant lineage or that the expression state is unknown. CtMAF refers to both CtMAF1 and CtMAF2, which show similar seasonal gene expression patterns.