Natural temperature fluctuations promote COOLAIR regulation of FLC

Abstract

Plants monitor many aspects of their fluctuating environments to help align their development with seasons. Molecular understanding of how noisy temperature cues are registered has emerged from dissection of vernalization in Arabidopsis, which involves a multiphase cold-dependent silencing of the floral repressor locus FLOWERING LOCUS C (FLC). Cold-induced transcriptional silencing precedes a low probability PRC2 epigenetic switching mechanism. The epigenetic switch requires the absence of warm temperatures as well as long-term cold exposure. However, the natural temperature inputs into the earlier transcriptional silencing phase are less well understood.Here, through investigation of Arabidopsis accessions in natural and climatically distinct field sites, we show that the first seasonal frost strongly induces expression of COOLAIR, the antisense transcripts at FLC. Chamber experiments delivering a constant mean temperature with different fluctuations showed the freezing induction of COOLAIR correlates with stronger repression of FLC mRNA. Identification of a mutant that ectopically activates COOLAIR revealed how COOLAIR up-regulation can directly reduce FLC expression. Consistent with this, transgenes designed to knockout COOLAIR perturbed the early phase of FLC silencing. However, all transgenes designed to remove COOLAIR resulted in increased production of novel convergent FLC antisense transcripts. Our study reveals how natural temperature fluctuations promote COOLAIR regulation of FLC, with the first autumn frost acting as a key indicator of autumn/winter arrival.

Vernalization-triggered expression of the antisense transcriptCOOLAIRis mediated byCBFgenes

“…These results strongly indicated that CBF3 can activate COOLAIR transcription through DRE1 and 2 located in the promoter, but not via the first exon To further determine whether the 3′-end region of FLC is required for the vernalization response of COOLAIR, we measured COOLAIR expression in the wild-type and FLC∆COOLAIR mutant, which lacks a 324 bp portion of the COOLAIR promoter region (Luo et al, 2019), before and after vernalization. Although the common COOLAIR transcriptional start site was eliminated in FLC∆COOLAIR, the proximal COOLAIR isoform (type I) was still detected, probably due to alternative transcription start sites (Figure 5B; Zhao et al, 2021). As expected, the expression of the proximal COOLAIR variant in the FLC∆COOLAIR mutant was not induced at 20V, while the wild-type showed increased type I COOLAIR level at 20V (Figure 5B).…”

“…Accordingly, vernalizationtriggered acceleration of flowering was comparable in the wild-type and cbfs (Figure 7B-C), indicating that cbfs show a normal vernalization response. It has recently been proposed that the strong induction of COOLAIR by the first seasonal frost is correlated with stronger FLC repression (Zhao et al, 2021). However, hyper-repression of FLC upon exposure to freezing temperature for 8 h after 20V (20V+8hF) was not observed under our growth conditions, in both the wild-type and cbfs (Figure 7D), although both genotypes showed stronger COOLAIR induction upon 8hF treatment (Figure 4C).…”

“…Although a recent report has shown that an NAC domain-containing protein, NTM1-LIKE 8 (NTL8), can bind to the COOLAIR promoter and activate its expression, it has not been thoroughly addressed whether NTL8 is necessary for vernalization-induced COOLAIR activation (Zhao et al, 2021). In the present study, we identified that CRT/DREs at the 3′-end of FLC are required for the long-term cold response of COOLAIR.…”

“…This process is linked to H3K4 demethylation at the FLC locus through the interaction of FCA and FLOWERING LOCUS D (FLD), an AP protein that acts as an H3K4 demethylase, thus suppressing FLC (Liu et al, 2010). In contrast with such a role of COOLAIR for FLC repression at warm temperatures, there is some controversy regarding the function of COOLAIR in vernalization-induced epigenetic silencing of FLC (Swiezewski et al, 2009;Helliwell et al, 2011;Li et al, 2018;Luo et al, 2019;Luo and He, 2020;Zhao et al, 2021).…”

“…The difference between the two experimental schemes is fluctuating versus constant cold conditions. The other reason may be due to the small level of CAS (convergent antisense transcript of COOLAIR) produced at the 5′-end of FLC in transgenic lines defective of COOLAIR promoter or even in the FLC∆COOLAIR mutant (Zhao et al, 2021), which may somehow affect the epigenetic silencing of FLC. Otherwise, the effect of COOLAIR on FLC silencing may be obscured due to the functional redundancy between COOLAIR and other lncRNA, such as COLDAIR in Arabidopsis, as suggested previously (Castaings et al, 2014).…”