Genome-wide identification and analysis of WRKY gene family in maize provide insights into regulatory network in response to abiotic stresses

Abstract

Background: The WRKY transcription factor family plays significant roles in biotic and abiotic stress responses, which has been associated with various biological processes in higher plants. However, very little is known regarding the structure and function of WRKY genes in maize.

Results: In this study, a total of 140 ZmWRKY proteins encoded by 125 ZmWRKY genes were eventually identified in maize. On the basis of features of molecular structure and a comparison of phylogenetic relationships of WRKY transcription factor families from Arabidopsis, rice and maize, all 140 ZmWRKY proteins in maize were divided into three main groups (Groups I, II and III) and the Group II was further classified into five subgroups. The characteristics of exon-intron structure of these putative ZmWRKY genes and conserved protein motifs of their encoded ZmWRKY proteins were also presented respectively, which was in accordance with the group classification results. Promoter analysis suggested that ZmWRKY genes shared many abiotic stress-related elements and hormone-related elements. Gene duplication analysis revealed that the segmental duplication and purifying selection might play a significant role during the evolution of the WRKY gene family in maize. Using RNA-seq data, transcriptome analysis indicated that most of ZmWRKY genes displayed differential expression patterns at different developmental stages of maize. Further, by quantitative real-time PCR analysis, twenty-one ZmWRKY genes were confirmed to respond to two different abiotic stress treatments, suggesting their potential roles in various abiotic stress responses. In addition, RNA-seq dataset was used to conduct weighted gene co-expression network analysis (WGCNA) in order to recognize gene subsets possessing similar expression patterns and highly correlated with each other within different metabolic networks. Further, subcellular localization prediction, functional annotation and interaction analysis of ZmWRKY proteins were also performed to predict their interactions and associations involved in potential regulatory network.

Conclusions: Taken together, the present study will serve to present an important theoretical basis for further exploring function and regulatory mechanism of ZmWRKY genes in the growth, development, and adaptation to abiotic stresses in maize.

Keywords: Abiotic stress; Expression patterns; Gene duplication; Regulatory network analysis; WRKY transcription factor; Zea mays L.

Fig. 1

Phylogenetic analysis among WRKY domains of WRKY proteins of maize, rice and Arabidopsis. The Unrooted neighbor-joining (NJ) phylogenetic tree was constructed with WRKY domains of WRKY proteins from maize, rice and Arabidopsis using MEGA7.0 with a bootstrap of 1000. Gray circles of different sizes represent the level of support. Eight main clades are marked: IN, IC, IIa, IIb, IIc, IId, IIe, and III with different colored ranges. Protein sequences were downloaded from National Center for Biotechnology Information (NCBI) and Maize genome database. The respective WRKY domain sequences of candidate WRKY proteins were downloaded from the Smart database

Fig. 2

Chromosomal location of WRKY family genes in maize. Distribution of WRKY genes in maize chromosomes. A total of 125 ZmWRKY genes were mapped on the ten maize chromosomes with an uneven distribution. The chromosome numbers are indicated at the top of each vertical gray bar. The gene names on the both side of each chromosome correspond to the approximate locations of each ZmWRKY genes. The scale on the left is in megabases

Fig. 3

Synteny analysis of ZmWRKY genes in maize genome. Duplicated blocks in maize chromosomes were revealed. The circular image retrieved from PLAZA database show inter-chromosome homologous regions connected by bands in different colors. The chromosome numbers and ZmWRKY genes are indicated outside. Gray lines mean all syntenic regions in whole maize genome and red lines mean ZmWRKY gene pairs with segmental duplication

Fig. 4

Unrooted phylogenetic tree, gene structures and conserved motifs of the ZmWRKY family in maize. The evolutionary tree from the ZmWRKY protein sequences was constructed using the MEGA 7.0 program with the Neighbor-joining method. Different colors represented various groups. The conserved motifs of ZmWRKY proteins were analyzed using MEME. Boxes with different colors denote different motifs and their position in each ZmWRKY protein. All 125 genes’ structures were obtained using GSDS 2.0. CDS, intron and 5′/3′ UTR are denoted by orange rectangles, single lines and green rectangles, respectively. (Color figure online)

Fig. 5

Hierarchical clustering of expression profiles of ZmWRKY gene family in 10 tissues. Ten tissues from different developmental stages including endosperm, husk, tassel, leaf, silks, shoot tip, stem, seedling, root and seed were investigated .The expression values were shown as log2 of the RPKM values. The scale bar is shown in the middle and higher expression levels are presented as red color

Fig. 6

Expression profiles of 21 ZmWRKY genes under slat stress treatments in two different tissues. QRT-PCR data was normalized using maize Actin gene. X-axes represent various treatments (CK, normal condition; S1, S2, S4 and S8 indicate hours of salt treatment.) and different genes and y-axes are scales of relative expression level. Error bars result from three biological replicates. Asterisks on top of the error bars represent the significance levels. *Significantly different at P < 0.05; ** significantly different at P < 0.01

Fig. 7

Expression profiles of 21 ZmWRKY genes under drought stress treatments in two different tissues. QRT-PCR data was normalized using maize Actin gene. X-axes represent various treatments (CK, normal condition; D1, D2, D4 and D8 indicate hours of drought treatment in leaves. D2 and D4 indicate hours of drought treatment in roots.). And different genes and y-axes are scales of relative expression level. Error bars result from three biological replicates. Asterisks on top of the error bars represent the significance levels. *Significantly different at P < 0.05; ** significantly different at P < 0.01

Fig. 8

Co-expression network analysis of ZmWRKY genes. (a) Elgengene adjacency heatmap. A high degree of correlation between modules is indicated by red. (b) Module sample correlation. Each row corresponds to a module. Each column corresponds to specific tissue. The color of each cell at row-column intersection indicates the correlation coefficient between the module and the tissue. A high degree of correlation between a specific module and tissue type is indicated by red

Fig. 9

Co-expression network analysis of modules of ZmWRKY genes. Heatmaps showing genes in each module that were expressed in tissues. Right correlation networks in the module corresponding to heatmap. Network is visualized in Cytoscape. Green and red color Spheres (nodes) represent ZmWRKY genes, and the thick lines (edges) represent the high correlation

Fig. 10

GO analysis of ZmWRKY proteins. The results are grouped into three main categories: biological process, cellular component and molecular function. The right y-axis indicates the number of genes

Fig. 11

Protein Protein Interaction network of specific ZmWRKY proteins. Black and red color characters represent Arabidopsis and maize, and the thick lines represent the high interaction