Nuclear events in ethylene signaling: a transcriptional cascade mediated by ETHYLENE-INSENSITIVE3 and ETHYLENE-RESPONSE-FACTOR1

Abstract

Response to the gaseous plant hormone ethylene in Arabidopsis requires the EIN3/EIL family of nuclear proteins. The biochemical function(s) of EIN3/EIL proteins, however, has remained unknown. In this study, we show that EIN3 and EILs comprise a family of novel sequence-specific DNA-binding proteins that regulate gene expression by binding directly to a primary ethylene response element (PERE) related to the tomato E4-element. Moreover, we identified an immediate target of EIN3, ETHYLENE-RESPONSE-FACTOR1 (ERF1), which contains this element in its promoter. EIN3 is necessary and sufficient for ERF1 expression, and, like EIN3-overexpression in transgenic plants, constitutive expression of ERF1 results in the activation of a variety of ethylene response genes and phenotypes. Evidence is also provided that ERF1 acts downstream of EIN3 and all other components of the ethylene signaling pathway. The results demonstrate that the nuclear proteins EIN3 and ERF1 act sequentially in a cascade of transcriptional regulation initiated by ethylene gas.

Figure 1

Cloning and ethylene inducibility of ERF1. (A) Sequence alignment of ERF1 and EREBP proteins from tobacco (EREBP1, EREBP2, EREBP3, and EREPB4; Ohme-Takagi and Shinshi 1995) and Arabidopsis [AtEBP (Buttner and Singh 1997) and AtTINY (Wilson et al. 1996)]. (B) RNA blot analysis of the induction of ERF1 mRNA expression by ethylene gas and comparison with the expression of PDF1.2. Total RNA was isolated from 4-week-old wild-type Col-0 (W) or ein3-1 (M) plants grown in air and exposed to ethylene gas for different times (0–48 hr). (C) RNA blot analysis of the expression of ERF1 in EIN3-overexpressing plants. Thirty micrograms of total RNA were loaded per lane in B and 60 μg in C.

Figure 2

EIN3 is a sequence-specific DNA-binding protein. (A) EMSA of in vitro-translated EIN3 protein binding to the −1238 to −950 fragment of the ERF1 promoter. A control protein (control) or mock-translated reticulocyte lysates (RL) were used in the indicated lanes. (B) EMSA of EIN3-3 mutant protein, EIN3, and several carboxy-terminal deletion derivatives bound to fragments −1238 to −1204 (1) and −1213 to −1178 (2) of the ERF1 promoter. (C) Binding of EIL proteins to the EIN3 target site in the ERF1 promoter. EMSA was performed with in vitro-translated EIN3, EIL1, EIL2, and EIL3 proteins and the wild-type EIN3-binding site (W) or a mutant version (M) corresponding to the mutant G17 to C in Fig. 2D. (D; top) Scan mutagenesis of the EBS. Wild-type EBS is shown with the palindromic repeats indicated by arrows. Base changes in the mutants tested are indicated on the lines below. Dashed lines indicate mismatches. Dots indicate similar bases as in the wild-type EBS. (D; bottom) Sequence alignment of the EBS and a fragment of the promoters of the E4 and GST1 genes (including the ERE). (E) Competition of the EIN3–EBS complex formation by addition of an excess of unlabeled EBS or two mutant versions, EBSm1 and EBSm2 (see Materials and Methods). No competitor was added in the lanes labeled as 0. Black wedges represent increasing amounts of competitor (20, 60, and 200 ng). One nanogram of labeled EBS was used per lane. (F) Summary of EIN3 structural features and mutants used in EMSA experiments (adapted from Chao et al. 1997).

Figure 3

EIN3 homodimerization. (A) EMSA of full-size EIN3 and deletion derivatives binding to the EBS. Proteins were translated in vitro alone or in pair-wise combinations. (See Fig. 2F for description of EIN3 deletion derivatives.) (B) EIN3–EIN3 interactions assayed by the yeast two-hybrid system. Yeast cells transformed with the indicated constructs in the top left plate, were grown on synthetic complete (SC) medium (+HIS) or in SC medium without histidine (−HIS) and with 50 mm 3-aminotriazole (3-AT, Sigma) to repress basal activity of the his3 reporter gene. β-Galactosidase activity of the colonies grown in −HIS medium (LacZ) was determined by the filter-lift assay. SNF4/SNF1 were used as a positive control. Colonies from two independent transformation experiments are shown (BD-EIN3a and BD-EIN3b).

Figure 4

ERF1 is a GCC box DNA-binding protein. EMSAs were performed using in vitro-translated ERF1 protein and promoter fragments from the Arabidopsis basic-chitinase (b-CHI) and bean chitinase5B (CH5B) genes. DNA fragments containing the GCC box or mutated versions (b-CHIm and CH5Bm) of these same elements were incubated with mock translated rabbit reticulocyte lysates (control) or those containing ERF1 protein.

Figure 5

Constitutive activation of ethylene response phenotypes in 35S::ERF1-expressing seedlings. Three-day-old etiolated seedlings overexpressing ERF1 in wild-type (Col-0) and ein3 mutant backgrounds grown in agar plates with or without 10 μm acc (1-aminocyclopropane-d-carboxylic acid). Untransformed wild-type and ein3 mutant plants are also shown for comparison.

Figure 6

ERF1 acts downstream of EIN2 and EIN3 in the ethylene signaling pathway. Transgenic plants overexpressing ERF1 in wild-type (Col-0), ein2, ein3-1, and ein3-3 mutant backgrounds grown in continuous flowthrough cambers with hydrocarbon-free air for 5 weeks. Untransformed wild-type, ctr1-1, and EIN3-overexpressing plants are shown for comparison.

Figure 7

Transcriptional activation of ethylene-responsive genes by ERF1 (A) RNA blot analysis of the expression of ethylene-inducible genes in transgenic lines overexpressing ERF1 in wild-type (Col-0) and ethylene-insensitive mutant backgrounds. Five independent transgenic lines in Col-0 and two independent lines in each of the mutants are shown. Five micrograms of total RNA from 5-week-old plants grown in air were loaded per lane in the middle and right panels, and 50 μg in the left panel. The same blot was probed with ERF1, a loading control probe (rDNA), and ethylene-inducible genes PDF1.2 and basic-chitinase. (B) Constitutive activation of the ethylene-inducible CH5B–GUS reporter gene in ERF1-overexpressing transgenic plants. Three-week-old plants grown on agar plates were assayed for GUS activity.

Figure 8

Nuclear events in the ethylene gas signaling pathway. Model depicts the transcriptional regulatory cascade that mediates ethylene responses. Binding of ethylene (C₂H₄) to membrane receptors activates EIN3, and most likely EIL1 and EIL2, through a signaling cascade described elsewhere (Chao et al. 1997). EIN3 directs the expression of ERF1 and other primary target genes by binding directly, as a dimer to the E4-like PERE present in their promoters. ERF1 and probably other EREBPs bind to the GCC box (SERE) and activate the expression of secondary ethylene response genes such as basic-chitinase and defensin (PDF1.2). Although we favor this simple model, the results presented in this work do not exclude other more complicated models that may involve the existence of an intermediate between EIN3 and ERF1 or the existence of EIN3-interacting proteins that modulate EIN3 activity.