Volume 9, Issue 9, Pages 1272-1285 (September 2016) ABF2, ABF3, and ABF4 Promote ABA-Mediated Chlorophyll Degradation and Leaf Senescence by Transcriptional Activation of Chlorophyll Catabolic Genes and Senescence-Associated Genes in Arabidopsis  Shan Gao, Jiong Gao, Xiaoyu Zhu, Yi Song, Zhongpeng Li, Guodong Ren, Xin Zhou, Benke Kuai  Molecular Plant  Volume 9, Issue 9, Pages 1272-1285 (September 2016) DOI: 10.1016/j.molp.2016.06.006 Copyright © 2016 The Author Terms and Conditions

Figure 1 ABF2, ABF3, ABF4 Bind to NYE1 Promoter In Vitro and In Vivo. (A) Physical interactions of the ABFs with NYE1 promoter in Y1H assays. Three yeast expression plasmids pGADT7-ABF2/ABF3/ABF4 were reintroduced into yeast strain Y1HGold carrying the reporter gene AbAr under the control of the 1254-bp NYE1 promoter. The transformants were screened for their growth on the yeast synthetic-defined media (SD/-Leu) in the presence of 100 ng ml−1 Aureobasidin A (AbA) for stringent selection. The empty vector pGADT7 was included as a negative control. All suspensions of yeast cells in this assay were adjusted to OD600 = 0.1. (B) Identification of functional G-box/ABRE motifs in the NYE1 promoter. Dual-luciferase assays were carried out to test the ABA responsiveness of a series of 5′ truncated NYE1 promoter fragments in Arabidopsis protoplasts. The length (in base pairs) of each truncated fragment is indicated on the left side of each figure. Triangles indicate the relative positions of the putative ABRE motifs in the promoter fragments. Empty triangles represent wild-type putative ABRE motifs, and solid triangles represent mutated ABRE motifs. The relative luciferase activity of each fragment without ABA treatment was arbitrarily set to 1. Data are means ± SE of three biological repeats. Asterisks indicate significant differences from mock treatment as determined by Student's t-test, *P < 0.05, **P < 0.005. (C) EMSA of in vitro binding of ABF2, ABF3, and ABF4 to the ABRE motif of the NYE1 promoter. His-tagged ABF proteins were purified using Ni-agarose beads and incubated with the biotin-labeled wild-type probe NYE1-P1. Competition experiments were performed by adding excessive amounts (400×) of unlabeled NYE1-P1. A probe (mNYE1-P1) with the G-box/ABRE motif mutated was used to test binding specificity. Shifted bands, suggesting the formation of DNA–protein complexes, are indicated by arrows. “−” represents absence, “+” represents presence. Sequences of both the wild-type and mutated probes are shown at the bottom of the images, with the G-box/ABRE motif boxed. (D) ChIP assay of in vivo binding of ABF4-GFP to the NYE1 promoter. Rosette leaves from 4-week-old 35S:ABF4-GFP and 35S:GFP plants were sprayed with 100 μM ABA and harvested 24 h later. Fold enrichment is calculated as the ratio of 35S:ABF4-GFP to 35S:GFP signal. P1 to P3 indicate defined regions examined in the ChIP assay. Triangles represent two functional ABRE motifs in the promoter of NYE1. Primers used for qPCR are listed in Supplemental Table 3. A coding region was examined as a negative control. Data are means ± SE of three repeats. Asterisks indicate statistically significant differences from controls as determined by Student's t-test, *P < 0.05, **P < 0.005. Molecular Plant 2016 9, 1272-1285DOI: (10.1016/j.molp.2016.06.006) Copyright © 2016 The Author Terms and Conditions

Figure 2 Phenotypic Characterization of nye1, nye1nye2, and abf2abf3abf4 Leaves after ABA Treatment. (A) Stay-green phenotypes of nye1, nye1nye2, and abf2abf3abf4 leaves after ABA treatment. Detached leaves from 4-week-old plants were treated with water (mock) or 100 μM ABA for 2 days under dim light. Bar represents 5 mm. (B) Measurements of Chl content in the leaves shown in (A). Data are means ± SE of three biological replicates. Asterisks indicate statistically significant differences compared with Col-0 determined by Student's t-test, *P < 0.05, **P < 0.005. (C) qRT–PCR quantification of NYE1 and NYE2 transcript levels, with ACTIN2 as a reference. RNAs were isolated from the leaves shown in (A). Transcript levels of NYE1 and NYE2 in untreated Col-0 plants were arbitrarily set to 1. Data are means ± SE of three biological replicates. Asterisks indicate statistically significant differences compared with Col-0 determined by Student's t-test, **P < 0.005. (D) Transactivation of NYE1 promoter by ABF2, ABF3, and ABF4 proteins in protoplasts. The top panel shows the schematic structures of the reporter and effector constructs used in the transient expression assays. The empty effector construct 35S:GFP was used as a control. Relative luciferase activity was determined 24 h after transfection with or without ABA treatment. Data are means ± SE of three biological repeats. Asterisks indicate significant differences from control as determined by Student's t-test, **P < 0.005. Molecular Plant 2016 9, 1272-1285DOI: (10.1016/j.molp.2016.06.006) Copyright © 2016 The Author Terms and Conditions

Figure 3 Phenotypic Characterization of abi1-1, snrk2.2/2.3/2.6, and ABF4OE Leaves after ABA Treatment. (A) Stay-green phenotypes of the detached leaves of abi1-1 and snrk2.2/2.3/2.6 mutants after ABA treatment. Detached leaves from 4-week-old plants were treated with water (mock) or 100 μM ABA for 2 days under dim light. Ler and Col-0 plants were included as corresponding controls of abi1-1 and snrk2.2/2.3/2.6, respectively. Bar represents 5 mm. (B) Chl contents in the leaves shown in (A). Data are means ± SE of three biological replicates. Asterisks indicate significant differences from Ler or Col-0 as determined by Student's t-test, **P < 0.005. (C) qRT–PCR quantification of NYE1 and NYE2 transcript levels in Col-0 and snrk2.2/2.3/2.6, with ACTIN2 as a reference. RNAs were isolated from the leaves shown in (A). Transcript levels of NYE1 and NYE2 in untreated Col-0 were arbitrarily set to 1. Data are means ± SE of three biological repeats. Asterisks indicate significant differences from Col-0 as determined by Student's t-test, **P < 0.005. (D) Senescence phenotypes of 35S:ABF4 overexpression in Col-0 (ABF4OECol-0). Detached leaves from 4-week-old plants were treated with water (mock) or 50 μM ABA, as indicated. Homozygous transgenic plants were obtained from the T3 generation. In total, 23 ABF4OECol-0 transgenic lines were generated. qRT–PCR was adopted to quantify the transcript level of the transgene in independent transgenic lines; the three transgenic lines with the highest levels of ABF4 transcript were selected for phenotypic analysis. A representative line is shown. Bar represents 5 mm. (E) Measurements of Chl content in the leaves shown in (D). (F) Senescence phenotypes of 35S::ABF4 overexpression lines in nye1 (ABF4OEnye1). Detached leaves from 4-week-old plants were treated with water (mock) or 100 μM ABA, as indicated. Homozygous transgenic plants were obtained from the T3 generation. In total, 26 ABF4OEnye1 transgenic lines were generated. qRT–PCR was adopted to quantify the transcript level of the transgene in independent transgenic lines; the three transgenic lines with the highest levels of ABF4 transcript were selected for phenotypic analysis. One representative line is shown. Bar represents 5 mm. (G) Chl contents in the leaves shown in (F). Data are means ± SE of three biological repeats. Asterisks indicate significant differences from Col-0 as determined by Student's t-test (E), **P < 0.005. Molecular Plant 2016 9, 1272-1285DOI: (10.1016/j.molp.2016.06.006) Copyright © 2016 The Author Terms and Conditions

Figure 4 ABF Proteins Positively Regulate the Expression of PAO and NYC1 by Binding to their Promoters. (A) Stay-green phenotypes of acd1-20 and nyc1-1 leaves after ABA treatment. Detached leaves from 4-week-old plants were treated with water (mock) or 100 μM ABA for 2 days under dim light. Bar represents 5 mm. (B) Transcript levels of PAO and NYC1 before or after ABA treatment in Col-0 and abf2abf3abf4 leaves. The transcript level of PAO or NYC1 in untreated Col-0 samples was arbitrarily set to 1. Data are means ± SE of three biological repeats. Asterisks indicate significant differences from Col-0 as determined by Student's t-test, *P < 0.05, **P < 0.005. (C) Transactivation of PAO or NYC1 promoter by ABF proteins in protoplasts in the transient dual-luciferase assay. The empty vector 35S:GFP was used as a control. Data are means ± SE of three repeats. Asterisks indicate statistically significant differences from controls as determined by Student's t-test, *P < 0.05, **P < 0.005. (D) EMSA of the binding of ABF4 to PAO and NYC1 promoters. His-tagged ABF4 protein (ABF4-His) was incubated with biotin-labeled probes PAO-P1 or NYC1-P1. Excess amount (400×) of unlabeled PAO-P1 or NYC1-P1 probes was used for testing binding specificity. Arrows indicate the position of shifted bands. “−” represents absence, “+” represents presence. Sequences of PAO-P1 and NYC1-P1 probes are shown at the bottom, with the G-box/ABRE motif boxed. (E) ChIP assay of in vivo binding of ABF4-GFP to PAO and NYC1 promoters. The same materials as described in Figure 1D were used. Fold enrichment was calculated as the ratio of 35S:ABF4-GFP to 35S:GFP signals. The numbers at the bottom of each figure indicate the positions (in base pairs) relative to the start codons of PAO and NYC1, respectively. “−” represents upstream of ATG. A coding region was examined as a negative control. Data are mean ± SE of three replicates. Asterisks indicate statistically significant differences from controls as determined by Student's t-test, *P < 0.05, **P < 0.005. Molecular Plant 2016 9, 1272-1285DOI: (10.1016/j.molp.2016.06.006) Copyright © 2016 The Author Terms and Conditions

Figure 5 abf2abf3abf4 Is a Functional Stay-Green Mutant and ABF Proteins Positively Regulate the Expression of SAG29. (A) Demonstration of abf2abf3abf4 as a functional stay-green mutant by a delayed decrease in the Fv/Fm ratio. Detached leaves from Col-0 and abf2abf3abf4 were treated in darkness for the number of days indicated before measurement. (B) EMSA of in vitro binding of ABF4 to the promoters of SAG29 and SAUR36. Purified ABF4 protein (ABF4-His) was incubated with biotin-labeled probes SAG29-P1, SAUR36-P1, and MKK9-P1. Competition experiments were performed by adding excess amounts (400×) of unlabeled probe fragments. Shifted bands, suggesting the formation of DNA–protein complexes, are indicated by arrows. “−” represents absence, “+” represents presence. Probe sequences are shown at the bottom of the EMSA images, with putative G/ABRE motifs boxed. (C) Transactivation of SAG29 promoter by ABF2, ABF3, and ABF4 in protoplasts. Relative luciferase activities were determined 24 h after transfection with or without ABA treatment. The empty vector 35S:GFP was used as a control. (D) Transcript levels of SAG29 in abf2abf3abf4 and Col-0 leaves before or after ABA treatment. qRT–PCR was used to analyze the level of SAG29 transcripts, with ACTIN2 level as a reference. The transcript level of SAG29 in the untreated Col was arbitrarily set to 1. Means ± SE are shown for three biological replicates. In (A) and (D), asterisks indicate statistically significant differences in comparison with Col-0 as determined by Student's t-test, **P < 0.005. (E) ChIP assay of in vivo binding of ABF4-GFP to SAG29 promoter. The same batch of leaf tissues used in Figure 1D was used for the assay. Fold enrichment was calculated as the ratio of 35S:ABF4-GFP to 35S:GFP signal. The numbers at the bottom of each figure indicate the positions (in base pairs) relative to the start codon of SAG29. “−” represents upstream of ATG. A coding region was examined as a negative control. Data are means ± SE of three repeats. In (C) and (E), asterisks indicate statistically significant differences from controls as determined by Student's t-test, *P < 0.05, **P < 0.005. Molecular Plant 2016 9, 1272-1285DOI: (10.1016/j.molp.2016.06.006) Copyright © 2016 The Author Terms and Conditions

Figure 6 A Regulatory Network of ABA-Triggered Chl Degradation and Leaf Senescence in Arabidopsis. ABF2/3/4 directly activate the expression of NYE1/2 as well as two major CCGs, NYC1 and PAO; and they also mediate ABA-triggered leaf senescence by regulating the expression of SAG29 and possibly other SAGs, such as SAG12 (Zhao et al., 2016). ABI5 family proteins, ABI5 and EEL, somehow coordinately upregulate NYE1 and NYC1 (Sakuraba et al., 2014). The coordinated mediation depends on the common core-sensing system (PYLs-PP2C-SnRK2). Gray color font indicates the expected yet unidentified target genes. Molecular Plant 2016 9, 1272-1285DOI: (10.1016/j.molp.2016.06.006) Copyright © 2016 The Author Terms and Conditions