Genome-wide Identification, Expression Profiling and Evolutionary Analysis of Auxin Response Factor Gene Family in Potato (Solanum tuberosum Group Phureja)

Abstract

Auxin response factors (ARFs) play central roles in conferring auxin-mediated responses through selection of target genes in plants. Despite their physiological importance, systematic analysis of ARF genes in potato have not been investigated yet. Our genome-wide analysis identified 20 StARF (Solanum tuberosum ARF) genes from potato and found that they are unevenly distributed in all the potato chromosomes except chromosome X. Sequence alignment and conserved motif analysis suggested the presence of all typical domains in all but StARF18c that lacks B3 DNA-binding domain. Phylogenetic analysis indicated that potato ARF could be clustered into 3 distinct subgroups, a result supported by exon-intron structure, consensus motifs, and domain architecture. In silico expression analysis and quantitative real-time PCR experiments revealed that several StARFs were expressed in tissue-specific, biotic/abiotic stress-responsive or hormone-inducible manners, which reflected their potential roles in plant growth, development or under various stress adaptions. Strikingly, most StARFs were identified as highly abiotic stress responsive, indicating that auxin signaling might be implicated in mediating environmental stress-adaptation responses. Taken together, this analysis provides molecular insights into StARF gene family, which paves the way to functional analysis of StARF members and will facilitate potato breeding programs.

Figure 1

Genomic distribution of StARF genes on DM1-3 chromosomes. The chromosome numbers and size are indicated at the top and bottom of each bar, respectively. The arrows next to gene names show the transcription directions. The number on the right side of the bars designated the approximate physical position of the first exon of corresponding StARF genes on potato genome.

Figure 2

Analysis of conserved domains in StARF porteins. (A) Schematic organization of conserved domains in StARF proteins. The B3 DNA-binding domain, Aux_Resp domain, and PB1 domain are shown in blue, red and green, respectively. (B) Amino acid composition of MR domains in StARF proteins. Bars represent the percentage of different amino acid residues in MR domains of StARFs, and each color represents one kind of amino acid.

Figure 3

Classification of S. tuberosome ARF proteins. (A) Neighbor-joining tree were generated using MEGA 7.0 to determine the phylogenetic relationship between S. tuberosome ARF proteins (left). According to classification proposed by Finet et al. (2013), StARFs were divided into three subgroups: (A–C) Shadow colors were used to distinguish different StARF subgroups. (B) The intron-exon organization of StARF genes was plotted using Gene Structure Display Server (Version 2.0). Blue boxes represent exons, and grey lines represent introns.

Figure 4

Phylogenetic analysis of ARF proteins in liverwort, the moss, and flowering plants. Neighbor-joining tree was constructed based on the alignment of ARF protein sequences from Marchantia polymorpha, Physcomitrella patens, Oryza sativa, S. lycoperiscum, S. tubersomem and Arabidopsis thaliana. The position of the root was determined from an outgroup consisting of an ARF-like sequence from green algae Chlamydomonas reinhardtii. The percent bootstrap support for 500 replicates is shown on each branch with >50% support. Some nodes were designated with letters in bold and italic. Sequences used in phylogenetic analysis were provided in Supplementary Data.

Figure 5

Expression profiles of StARF genes with hierarchical clustering in different tissues. The Illumina RNA-Seq data were obtained from PGSC database, and the FPKM value of representative transcripts of StARF genes were log₂ transformed for further analysis. The normalized expression data was used to generate heatmap with hierarchical clustering based on the Manhattan correlation with average linkage using MeV software package. Color scale below heatmap shows the expression level; red indicates high transcript abundance while green indicates low abundance.

Figure 6

Heatmap representation and hierarchical clustering of StARF genes under abiotic stresses (A), biotic stresses (B), and phytohormone treatments (C). The Illumina RNA-Seq data were obtained from PGSC database, and the relative expression of StARF genes was calculated with respect to control samples using FPKM values of representative transcripts corresponding to StARF genes. Fold changes of StARF expression were log₂ transformed, and the normalized expression data was used to generate heatmap with MeV software package using the same parameters in Fig. 4. Color scale below heatmap shows the expression level; red indicates high transcript abundance while green indicates low abundance.

Figure 7

qRT-PCR analysis of StARF genes in different tissues (A) and phytohormones (B) StARF transcript levels measured by real-time qRT-PCR from the various tissues or under phytohormone treatments at indicated time points. Data are means of three biological replicates (8 pooled plants each), and error bars denote SE. Potato elongation factor StEF-1α gene was used as an internal control. Stars above the error bars indicate significant differences between treatments and controls (according to student’s t-test). qRT-PCR primers for StARF and StEF-1α genes were provided in Supplementary Table S1.

Figure 8

Predicted protein-protein interaction network of StARFs and StIAAs. Each node represents either one StARF protein (blue label) or StIAA (black label) protein. The disconnected nodes in the network were hidden due to the lack of supporting information. Edges between nodes represent protein-protein associations predicted by experimentally determined (pink line) or from curated databases (blue line).