The transcriptional regulator VAL1 promotes Arabidopsis flowering by repressing the organ boundary genes BOP1 and BOP2

Abstract

The transition to reproductive development is a critical step in the plant lifecycle and relies on the integration of intrinsic and environmental signals. Several different pathways controlling flowering time function downstream of the perception of environmental cues such as day length (photoperiodic pathway) and seasonal temperature (vernalization and ambient temperature pathways). In addition, the phytohormone gibberellin (GA) induces the floral transition under noninductive photoperiod. In the model plant Arabidopsis (Arabidopsis thaliana), the transcriptional repressor VIVIPAROUS1/ABSCISIC ACID INSENTIVE3 (ABI3)-LIKE1 (VAL1) triggers the stable repression of the floral repressor FLOWERING LOCUS C (FLC) during vernalization. However, the involvement of VAL1 in other flowering pathways remains unclear. In this work, we combined genetic and transcriptomic approaches to investigate the requirement of VAL1 for flowering activation under different day lengths. We found that VAL1, but not its sister protein VAL2, is required to induce the floral transition both under long and short days. The delayed flowering time of val1 mutant plants was fully bypassed by exogenous GA application. We demonstrated that VAL1-mediated induction of flowering occurs partially via the direct epigenetic repression of the organ boundary genes BLADE-ON-PETIOLE1 (BOP1) and BOP2. Our work thus expands the repertoire of VAL target genes and further demonstrates the pleiotropic role of VAL factors in regulating Arabidopsis development.

Figure 1.

VAL1 function is required to induce flowering under inductive and noninductive photoperiod. A) and D) Flowering phenotype of 30-d-old Col-0 (WT), val1 and val2 plants grown under long-day photoperiod (A; LD, 16 h light/8 h dark) and of 73-d-old plants grown under short-day photoperiod (D; SD, 8 h light/16 h dark). Scale bar: 2 cm. B) and C) Rosette leaf number at bolting of WT, val1 and val2 plants grown under either LD (B) or SD (C). B) and C), n ≥ 10. E) Rosette leaf number of WT, val1, val1-5, val2, val2-3 and val1-5 val2-3 double mutants grown under SD (n = 10). F) Rosette leaf number of WT, val1, val2, and VAL1 complementation lines (pVAL1:VAL1-3xHA-62) grown under SD (n = 10). Values in B) to C) and E) to F) are means ± SEM. Different letters indicate significant differences as determined using Ordinary one-way ANOVA Multiple comparisons (P  <  0.05).

Figure 2.

VAL1 accelerates flowering in SD partially through the transcriptional repression of FLC. A) Relative (Rel.) FLC transcript abundance tested by RT-qPCR in Col-0 (WT), val1 and val2 mutant plants grown under SD. Seeds were stratified during 72 h in the dark at 4 °C, and then transferred to the light for germination at 22 °C in SD. Plants were incubated during 2, 3, and 7 wk before sample collection. Leaf material was collected for total RNA isolation. Values are means ± SEM (n ≥ 3). ACTIN2 was used as normalization control. Different stars indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (*P  <  0.05, **P  <  0.01, ****P  <  0.0001; ns, not significant). B) Flowering time of WT, val1, val2 and flc-3 single mutants, and flc val1 and flc val2 double mutant plants grown under SD, measured as rosette leaf number. Values are means ± SEM (n ≥ 8). Different letters indicate significant differences as determined using Ordinary one-way ANOVA Multiple comparisons (P  <  0.05).

Figure 3.

Exogenous supplementation of gibberellin (GA) overcomes the late-flowering phenotype of val1 under SDs. A) Representative images of 55-d-old Col-0 (WT), val1 and val2 plants grown under SDs with (+GA₃, bottom panel) or without (Mock, top panel) the supplementation of GA₃. +GA₃ plants were treated with 100 μM GA₃ plus 0.02% Silwet-77 twice per week, whereas Mock plants were treated with 1% ethanol plus 0.02% Silwet-77. Scale bar: 2 cm. B) Rosette leaf number at flowering measured under SDs. Values are means ± SEM (n ≥ 9). Different letters indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (P  <  0.05). C) Endogenous GA₄ content was measured in the SAM of 6-wk-old plants under SDs. Values are means ± SEM of three biological replicates. y axis represents hormone (H) content per gram of fresh weight (FW; ng H/g FW). Different letters indicate significant differences as determined using Ordinary one-way ANOVA Multiple comparisons (P  <  0.05).

Figure 4.

VAL1 deficiency causes misregulation of flowering genes in the Arabidopsis shoot apex. A) Expression of pVAL1:VAL1-GUS in 6-wk-old seedlings grown under SD. The inset shows magnification of the shoot area. Scale bar: 1 cm. B) and C) Volcano map of number of DEGs is upregulated in val1  (B) and val2  (C) compared to Col-0 (WT). Dashed lines in B) and C) show the FC = ±2 threshold. FC, log2 fold change; FDR, false discovery rate. D) GO analysis of upregulated (left, n = 938) and downregulated (right, n = 1674) genes in val1 mutant. Top 10 representative terms are listed and ranked by P-value. E) Heat-map of upregulated and downregulated genes expression in WT, val1 and val2 mutant by RNA-Seq. Z-scores of the Transcripts Per Million (TPM) values for all samples are represented. F) Transcription level of BOP1 and BOP2 in WT, val1 and val2 by RT-qPCR in 6-wk-old shoot apical meristems (SAM) under SDs. Values are means ± SEM (n = 3). Different stars indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (*P  <  0.05, ****P  <  0.0001; ns, not significant). G)  FD mRNA expression in WT, val1-5 and val1 at 15-d-old seedlings grown under LDs by in situ hybridization. Stars indicate the SAM. Scale bar: 50 µm.

Figure 5.

VAL1 triggers epigenetic silencing of the organ boundary genes BOP1 and BOP2. A) Relative (Rel.) transcript levels of BOP1, BOP2 and FD in 2-wk-old Col-0 (WT), val1 and val2 plants grown under LD. Transcript abundance was measured by RT-qPCR. Values are means ± SEM of three biological replicates (n = 9). ACTIN2 was used as normalization control. Different stars indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (****P  <  0.0001; ns, not significant). B) and C) Schematic depicting BOP1  (B) and BOP2  (C) genomic loci. Exonic regions and VAL1-specific DNA motifs (RY motifs, TGCATG) are shown in the schematic. Line bars below each gene structure depict the position of the primer pairs (P1-6) used in ChIP experiments. D) and E) ChIP-qPCR analysis using anti-HA antibody reveals the binding of VAL1 to the BOP1  (D) and BOP2  (E) gene body in 2-wk-old seedlings of WT and VAL1-HA grown under LDs. ACTIN2 and STM were included as background controls. Values are means ± SEM of 3 biological replicates (n = 9). Different stars indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (*P < 0.05, ****P < 0.0001; ns, not significant). F) EMSA testing GST-VAL1B3 binding to the genomic regions of BOP1 and BOP2. Two fluorescently (TYE665) labeled dsDNA probes were designed: BOP1-TYE665-dsDNA and BOP2-TYE665-dsDNA. The approximate genomic locations of BOP1-TYE665-dsDNA and BOP2-TYE665-dsDNA probe sequences are depicted in B) and C), respectively (P3, in red). Competition experiments were performed using decreasing concentration of unlabeled probes (50× and 25× molar excess). G) and H) ChIP-qPCR analysis of H3K27me3 enrichment at BOP1  (G) and BOP2  (H) loci in 2-wk-old WT, val1 and val2 plants grown under LD. STM and ACTIN2 were used as positive and negative controls of H3K27me3 enrichment, respectively. Values are means ± SEM of the three technical replicates in one qPCR experiment. An independent biological replicate of this H3K27me3 ChIP experiment is presented in Supplementary Fig. S6, E and F. Different stars indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (*P < 0.05, **P < 0.01, ****P < 0.0001; ns, not significant).

Figure 6.

VAL1 regulates flowering time by repressing BOP genes. A) and B) Rosette leaf number at bolting measured for Col-0 (WT), val1, bop1-5, bop2-2, bop1 bop2, val1 bop1, val1 bop2, and val1 bop1 bop2 plants grown in either LD (A, n ≥ 9) or SD (B, n ≥ 8). Values are means ± SEM. Different letters indicate significant differences as determined using Ordinary one-way ANOVA Multiple comparisons (P  <  0.05). C) to G) Relative (Rel.) transcript abundance of flowering genes FT (C), FD  (D), SOC1  (E), FUL  (F), and LFY  (G) assayed by RT-qPCR. WT, val1, bop1 bop2 and val1 bop1 bop2 seeds were stratified during 72 h in the dark at 4 °C, and then transferred to the light for germination at 22 °C in LD. Plants were incubated during 2 and 3 wk before sample collection. Values are means ± SEM (n = 6). ACTIN2 was used as normalization control. Different stars indicate significant differences as determined using Ordinary two-way ANOVA Multiple comparisons (*P < 0.05, **P < 0.01, ****P < 0.0001; ns, not significant). H) Schematic model depicting VAL1 function in the regulation of Arabidopsis flowering time. In leaves, VAL1 promotes the floral transition by promoting FLC downregulation in response to winter cold and both in LD and SD under standard ambient temperature conditions (22 °C), thereby releasing FT from FLC block. In the shoot apex, VAL1 directly represses FLC gene, which results in induction of FD mRNA. Accumulation of FD protein in the SAM forms a complex with the FT protein translocated from leaves to induce expression of IM identity genes, thus initiating reproductive development. In parallel, VAL1 also induces flowering by directly repressing BOP1/2 in a PRC2-dependent manner, which results in the induction of FUL mRNA. The mechanism whereby BOP genes control FUL remains unknown (dotted line). Likewise, the role of GA in bypassing the delay flowering phenotype of val1 mutant needs further exploration. Solid black lines indicate mechanisms either previously known or demonstrated in this work, while dotted black lines depict unknown mechanisms. Orange arrow (dotted line) depicts protein relocalization from leaf to SAM.