Studies on DNA-binding selectivity of WRKY transcription factors lend structural clues into WRKY-domain function

Abstract

WRKY transcription factors have been shown to play a major role in regulating, both positively and negatively, the plant defense transcriptome. Nearly all studied WRKY factors appear to have a stereotypic binding preference to one DNA element termed the W-box. How specificity for certain promoters is accomplished therefore remains completely unknown. In this study, we tested five distinct Arabidopsis WRKY transcription factor subfamily members for their DNA binding selectivity towards variants of the W-box embedded in neighboring DNA sequences. These studies revealed for the first time differences in their binding site preferences, which are partly dependent on additional adjacent DNA sequences outside of the TTGACY-core motif. A consensus WRKY binding site derived from these studies was used for in silico analysis to identify potential target genes within the Arabidopsis genome. Furthermore, we show that even subtle amino acid substitutions within the DNA binding region of AtWRKY11 strongly impinge on its binding activity. Additionally, all five factors were found localized exclusively to the plant cell nucleus and to be capable of trans-activating expression of a reporter gene construct in vivo.

Fig. 1

Binding site preferences of different Arabidopsis WRKY subgroup representatives to W-box variants. (a) Sequence of the parsley PR1-1 promoter region designated 1×W2 (Rushton et al. 1996). The W-box element is in bold letters. Box marks the region of the oligonucleotide within which base substitutions were generated to produce the variants listed below. The W2 box element is highlighted in bold whereas the base alterations are plain letters and underlined. (b) EMSA of the radioactively labeled designated W2 box variants incubated with crude bacterial extracts containing the indicated recombinant WRKY protein. Lanes marked by a slash contain only the labeled free DNA probes with no protein added. Specific retarded DNA-protein complexes are marked by open asterisks in the middle of the composite, whereas dots designate positions of the free probes

Fig. 2

Binding of WRKY factors to W-box containing promoter sequences. (a) Upper strand sequences of the oligonucleotide probes synthesized for gel shift assays. The sequences represent different W-box containing regions of four selected Arabidopsis gene promoters as described in the text. W-box motifs are highlighted in bold. (b) EMSA of the indicated radioactively labeled probes incubated with crude bacterial extracts containing WRKY11 or WRKY26 recombinant protein. Lane numbering is identical to the probe numbers in a. The specific retarded DNA-protein complexes are marked by open asterisks at the right-hand side of each gel whereas dots designate positions of the free running probes

Fig. 3

Spacing between two W-boxes does not significantly influence WRKY binding. (a) Upper strand sequences of the oligonucleotide probes synthesized for gel shift assays. (b) EMSA of the indicated radioactively labeled probes incubated with crude bacterial extracts containing the respective WRKY recombinant proteins. (c) EMSA with WRKY11 CTD protein and radioactively labeled DNA probes as indicated, in the presence or absence of non-labeled competitor DNA. A 500-fold excess of the respective competitor (marked in the lanes) was included in the binding reaction. Specific retarded DNA-protein complexes are marked by open asterisks at the right-hand side of each gel, whereas black bars designate the positions of the free running probes

Fig. 4

Defining critical amino acid residues of AtWRKY11 for DNA binding. (a) Peptide sequence of the C-terminal WRKY domain of WRKY11 (AtWRKY11-CTD). Highlighted in bold letters are those amino acid residues within the peptide for which substitution variants (m1 to m18) were generated. The exchange amino acids within the individual variants are given below. The open stars above the peptide sequence mark the positions of the invariant WRKY stretch and the cysteines and histidines in the zinc finger motif that are conserved in WRKY transcription factors. Black arrows above the sequence indicate regions of β-strands as recently determined for the AtWRKY4-CTD (Yamasaki et al. 2005). (b) EMSA of crude bacterial extracts harboring the indicated WRKY11-CTD proteins having amino acid exchanges at conserved positions with different radioactively labeled W-box containing probes. (c) EMSA of crude bacterial extracts harboring the various WRKY11-CTD protein variants given in a. with radioactively labeled AtSIRKpW11/12 probe. Specific retarded DNA-protein complexes are marked by open asterisks at the left-hand side of each gel, whereas black bars designate the positions of the free running probes

Fig. 5

Nuclear localization and in vivo transactivation functions of different WRKY subfamily representatives. (a) Transient expression of WRKY-GFP fusion proteins in onion epidermal cells. Two individual cells transfected with the indicated expression constructs and showing GFP fluorescence are shown. WRKY6-GFP was used as a positive control for nuclear localization of the fusion protein (Robatzek and Somssich 2002), whereas a 35S CaMV-intron-GFP reporter construct (-GFP) served as a control for non-targeted localization. (b) Transient co-transfections of 4-week-old Arabidopsis leaves using the −928 AtSIRK promoter-GUS reporter construct (AtSIRKp∷GUS) together with the various effectors indicated below each cut-out. All effector constructs were under the control of the constitutive 35S CaMV promoter. For each construct low magnification leaf sections and higher magnifications of individual bombarded areas is shown. Enlarged cut-out at the bottom right-hand side originates from similar experiments using a 4×W2 synthetic promoter-GUS reporter (Rushton et al. 2002) and the WRKY11-GFP as effector. Bar scale shown in high magnification picture of HA-GFP corresponds to 100 μm and the same for all pictures of the series

Fig. 6

Motif analysis of WRKY binding sites. (a) Sequence logos (Crooks et al. 2004) depicting nucleotide distribution for the WRKY site as derived from our study. The consensus (A) and (B) given below represent a more stringent and a more relaxed version, respectively. (b) Fold enrichment of the two consensus sequences within; (a) the dataset 600 bp upstream region (relative to the ATG) of all annotated genes versus entire genome sequence, and (B) the dataset 1500 bp upstream region (relative to the ATG) of all annotated genes versus entire genome sequence using the program Motif Mapper. (c) Pie chart display of the GO functional classification of annotated genes showing an enrichment of WRKY binding sites in their 600 bp upstream sequences identified in b. The percentage for each category is shown. (d) Fold enrichment of closely spaced WRKY binding sites, ranging from 0 to 30 bp apart, within our 600 bp promoter dataset identified using the Motif Mapper program. The dashed line on the x-axis represents the cut-off value for the statistical background distribution. A 1.4-fold enrichment over the genomic background frequency corresponds to a significant over-representation of the motifs (P < 0.001). W = A,T; Y = C,T; D = A,G,T; H = A,C,T; N = A,C,T,G