Structural and Functional Insights into WRKY3 and WRKY4 Transcription Factors to Unravel the WRKY-DNA (W-Box) Complex Interaction in Tomato (Solanum lycopersicum L.). A Computational Approach

Abstract

The WRKY transcription factors (TFs), play crucial role in plant defense response against various abiotic and biotic stresses. The role of WRKY3 and WRKY4 genes in plant defense response against necrotrophic pathogens is well-reported. However, their functional annotation in tomato is largely unknown. In the present work, we have characterized the structural and functional attributes of the two identified tomato WRKY transcription factors, WRKY3 (SlWRKY3), and WRKY4 (SlWRKY4) using computational approaches. Arabidopsis WRKY3 (AtWRKY3: NP_178433) and WRKY4 (AtWRKY4: NP_172849) protein sequences were retrieved from TAIR database and protein BLAST was done for finding their sequential homologs in tomato. Sequence alignment, phylogenetic classification, and motif composition analysis revealed the remarkable sequential variation between, these two WRKYs. The tomato WRKY3 and WRKY4 clusters with Solanum pennellii showing the monophyletic origin and evolution from their wild homolog. The functional domain region responsible for sequence specific DNA-binding occupied in both proteins were modeled [using AtWRKY4 (PDB ID:1WJ2) and AtWRKY1 (PDBID:2AYD) as template protein structures] through homology modeling using Discovery Studio 3.0. The generated models were further evaluated for their accuracy and reliability based on qualitative and quantitative parameters. The modeled proteins were found to satisfy all the crucial energy parameters and showed acceptable Ramachandran statistics when compared to the experimentally resolved NMR solution structures and/or X-Ray diffracted crystal structures (templates). The superimposition of the functional WRKY domains from SlWRKY3 and SlWRKY4 revealed remarkable structural similarity. The sequence specific DNA binding for two WRKYs was explored through DNA-protein interaction using Hex Docking server. The interaction studies found that SlWRKY4 binds with the W-box DNA through WRKYGQK with Tyr⁴⁰⁸, Arg⁴⁰⁹, and Lys⁴¹⁹ with the initial flanking sequences also get involved in binding. In contrast, the SlWRKY3 made interaction with RKYGQK along with the residues from zinc finger motifs. Protein-protein interactions studies were done using STRING version 10.0 to explore all the possible protein partners involved in associative functional interaction networks. The Gene ontology enrichment analysis revealed the functional dimension and characterized the identified WRKYs based on their functional annotation.

Keywords: DNA binding domain; DNA-protein docking; homology modeling; monophyletic origin; transcription factors.

Figure 1

A generalized model showing functional mechanism and role of PcWRKY1 (Petroselinum crispum) in PAMP induced WRKY gene regulation through direct binding and auto regulatory negative feedback loop.

Figure 2

(A) Sequence alignment of the functional domain region from SlWRKY3 and SlWRKY4. The substituted amino acid residues have been shown in three letter codes (B) Phylogenetic tree showing the evolutionary origin of SlWRKY3 (C) Phylogenetic tree for evolutionary origin of SlWRKY4.

Figure 3

(A) The sequence alignment of the conserved functional WRKY domain region from N-terminal end in SlWRKY3, (B) N-terminal end of WRKY4, (C) C-terminal end of SlWRKY3, and (D) C-terminal end of SlWRKY4.

Figure 4

(A) The motif scan analysis represented along with phylogenetic tree showing the distribution and presence or absence of common and uncommon motifs found in Solanum lycopersicum, Solanum pennellii, Solanum tuberosum, and Nicotiana sylvestris discovered through MEME and MAST results. (B) The BLOCK diagram showing the sequence of the discovered motifs for SlWRKY3. The red arrows indicates the presence of uncommon motifs (motif 18, 19, 20, and 21) found in Nicotiana sylvestris and absent from other members, the red square indicate the presence of common motif (motif 15) that is absent from N. sylvestris. (C) The sequential logo of the motif 1 showing consensus WRKY sequences and present in all the representative members of tomato family.

Figure 5

(A) The motif scan analysis represented along with phylogenetic tree showing the distribution and presence or absence of common and uncommon motifs found in Solanum lycopersicum, Solanum pennellii, Solanum tuberosum, and Nicotiana sylvestris discovered through MEME and MAST results. (B) The BLOCK diagram of the discovered motifs for WRKY4 protein in tomato. The red arrows indicates the uncommon motifs found exclusively in Nicotiana sylvestris and showing the distribution of motif 19 absent from Solanum lycopersicum and Solanum penellii. (C) The sequential logo of the motif 2 showing consensus WRKY sequences and present in all the representative members of tomato family.

Figure 6

Functional interactive associative network of tomato WRKY3 with other protein family members as found on STRING server datasets at medium confidence leveland represented in multifaceted way where the color nodes describe query proteins and first shell of interactors whereas white nodes are second shell of interactors. The large node size represent characterized proteins and smaller nodes for uncharacterized proteins.

Figure 7

Comparative analysis of the tomato WRKY genes (SlWRKY3 and SlWRKY4) with the others members of family Solanaceae along with model Arabidopsis thaliana to investigate the similarities and differences. The circular map generated based on percentage similarity matrices (obtained through phylogenetic clustering using ClustalW) and have been visualized through Circos software.

Figure 8

(A) Predicted structure of the C-terminal WRKY3 domain visualized through the Discovery Studio 3.0. (B) Structure of the C-terminal WRKY4 domain (C) structure of the W-Box (TTTGACCA)DNA sequence.

Figure 9

Superimposition results represented with their respective global and local RMSD-values showing the structural conservation of WRKY domain. Superimposition revealed that both SlWRKY3 and SlWRKY4 were found to be structurally more closer to their template 2AYD rather than other template 1WJ2 as evident from fluctuations in their RMSD values. Sequence alignment between SlWRKY3 and SlWRKY4 predict the synonymous (deep blue) and non-synonymous substitutions (light blue) color. We have labeled the residues only from tomato WRKY3 to show different residues with respect to their topology and their sitewise probable occurrence in protein secondary structure.

Figure 10

Qualitative evaluation of the predicted model using PROCHECK and ProSA analysis. (A) The stereo chemical spatial arrangement of amino acid residues in the predicted models (SlWRKY3 and SlWRKY4) as computed with the PROCHECK server and were compared with experimentally resolved protein structures (1WJ2 and2YD). Most favored regions are colored red, additional allowed, generously allowed, and disallowed regions are indicated as yellow, light yellow, and white fields, respectively. (B) Qualitative evaluation using ProSA webserver, which generates a plot measuring the structural error at each residues in the protein and calculate the overall score for quality measurement. The ProSA score for SlWRKY3 and SlWRKY4 were found closer to the native structures.

Figure 11

(A) Comparative evaluation of the docked complexes with experimentally solved structures. The NMR resolved solution structure of complex of the C-terminal WRKY domain of AtWRKY4 with W-box DNA. The residues LRWRKYGQK made interaction with W-box DNA and compared with our predicted complexes where (RKYGQK) along with residues forming zinc finger got involved in interaction (SlWRKY3) and similar residues KWRKYGQK with initial flanking sequences (SlWRKY4) participated in interaction with W-box. (B) Structure of the docked complex (SlWRKY3 with DNA) as visualized by Discovery Studio 3.0 (C) Structure of the docked complex (SlWRKY4 with DNA).

Figure 12

Gene ontology enrichment analysis using ReviGO web server. The functional and significant GO terms were shown on scattered plot digramme using hypergeometric test distribution in terms of their controlled functional vocabularies (A, biological process; B, molecular function; and C, cellular processes involved).

Figure 13

The prediction of functional gene annotation along with subcellular localization using CELLO2GO web server. The functional vocabolaries are represented in pie chart digramme evaluating the significant terms in form of their percentage contribution.