A PHD finger protein involved in both the vernalization and photoperiod pathways in Arabidopsis

Abstract

The proper timing of flowering is critical for successful reproduction. The perception of the seasonal cues of day-length changes and exposure to cold influences flowering time in many plant species through the photoperiod and vernalization pathways, respectively. Here we show that a plant homeodomain (PHD) finger-containing protein, VIN3-LIKE 1 (VIL1), participates in both the photoperiod and vernalization pathways in Arabidopsis thaliana by regulating expression of the related floral repressors FLOWERING LOCUS C (FLC) and FLOWERING LOCUS M (FLM). In the vernalization pathway, VIL1, along with VERNALIZATION INSENSITIVE 3 (VIN3), is necessary for the modifications to FLC and FLM chromatin that are associated with an epigenetically silenced state and with acquisition of competence to flower. In addition, VIL1 regulates FLM independently of VIN3 in a photoperiod-dependent manner.

Figure 1.

(A) Domain structures of VIN3-like proteins. (Amino acid alignments are provided in Supplementary Fig. 1.). VIN3 is At5g57380; VIL1 is At3g24440; VIL2 is At4g30200; VIL3 is At2g18880; and VIL4 is At2g18870. (B) The VID region is required for the interaction between VIN3 and VIL1. The constructs either have the VID region alone as described in Supplementary Figure S1 or the remainder of the gene without the VID region. Yeast two-hybrid assays: (1) full-length VIN3 as bait and full-length VIL1 as a prey. (2) VIN3 without VID as bait and full-length VIL1 as a prey. (3) VID of VIN3 as bait and full-length VIL1 as prey. (4) Full-length VIL1 as bait and full-length VIN3 as prey. (5) Full-length VIL1 as bait and VIN3 without VID as prey. (6) Full-length VIL1 as bait and VID of VIN3 as prey. Yeast two-hybrid assay conditions are described in Materials and Methods (Supplemental Material). (C) Yeast two-hybrid assays among the VID regions. (D) Vernalization response in vil1 and vin3. (Filled bars) Leaf number at flowering of nonvernalized plants; (open bars) values for plants cold-treated for 40 d as described in Materials and Methods (Supplemental Material). (E) FLC mRNA levels in vin3 and vil1. (NV) Samples from nonvernalized plants grown at 22°C; (40V) samples prepared directly from plants grown for 40 d at 4°C; (40VT14) plants grown for 40 d at 4°C followed by growth at 22°C for 14 d.

Figure 2.

Chromatin immunoprecipitation (ChIP) analyses of vernalization-mediated histone modifications in wild type and vil1. Samples are from plants grown as follows: nonvernalized (NV); vernalized for 40 d (40V); vernalized for 40 d and subsequently grown for 14 d at 22°C (40VT14). Relative fold changes are indicated at the bottom of each ChIP assay. All assays were repeated with at least two independent chromatin preparations and three independent immunoprecipitations as described in Materials and Methods. Representative ChIP results are shown using antibodies recognizing acetylated Histone H3 (A), dimethylated Histone H3 Lys 9 (B), dimethylated Histone H3 Lys 27 (C), trimethylated Histone H3 Lys 9 (D), and trimethylated Histone H3 Lys 27 (E). The regions of FLC and FLM chromatin examined are described in Supplementary Figure S2.

Figure 3.

(A) Effect of cold exposure on mRNA levels of VIN3 family members. (B) SD flowering behavior. (Filled bars) Leaf number at flowering of nonvernalized plants; (open bars) leaf number of plants vernalized for 40 d as described in Materials and Methods (Supplemental Material). (C, left) mRNA levels of VIN3 and VIL1 in different photoperiod conditions. (CL) Continuous light; (LD) long days, 16 h light/8 h dark; (SD) short days, 8 h light/16 h dark. (Right) mRNA levels of VIN3 and VIL1 during vernalization under LD. Samples were nonvernalized (NV), vernalized for 40 d (40V), and vernalized for 40 d and subsequently grown for 14 d at 22°C (40VT14). (D) mRNA levels of FLC and FLM in wild-type Columbia (Col) and vil1 mutants grown in LD and SD. (The stronger FLC signal in LD is a result of intentional dilution of the cDNA templates from SD samples relative to LD samples to facilitate evaluation of the increase in FLM in SD-grown vil1 mutants.) (E) ChIP assay using antibodies recognizing hyperacetylated Histone H4. Chromatin was prepared from wild-type (Col) and vil1 mutants grown in SD without vernalization treatment. Three independent immunoprecipitation were performed, and a representative result is shown. Relative fold changes are indicated at the bottom of each ChIP assay.

Figure 4.

(A) Effect of vernalization on mRNA levels of FLC and FLM in wild-type (Col) and vil1 mutants grown in SD. Samples in all panels unless noted are nonvernalized (−) or cold-treated for 40 d and subsequently grown for 10 d at 22°C under SD (+). (Note: In Scortecci et al. [2001], we reported that FLM mRNA levels were not reduced by vernalization in a FRI-containing line. However, we now reproducibly detect a vernalization-mediated decrease of FLM with or without FRI.) (B) ChIP assays using antibodies recognizing trimethylated Histone H3 Lys 9 at FLM chromatin in wild type (Col), vin3, and vil1. (C) ChIP assay using antibodies recognizing trimethylated Histone H3 Lys 27 at FLM chromatin in wild type (Col), vin3, and vil1. (D) ChIP assay using antibodies recognizing hyperacetylated Histone H4 at FLM chromatin in vil1 mutants. (E) mRNA expression patterns of floral integrator genes in wild-type (Col), flm, and flc mutants grown in SD. cDNA templates prepared from mutant seedlings were intentionally diluted relative to wild-type samples to emphasize the increased mRNA levels in mutants. (F) Relationship of VIN3 family genes to the regulatory network controlling flowering time in response to environmental cues.