Transcription factors relevant to auxin signalling coordinate broad-spectrum metabolic shifts including sulphur metabolism

Abstract

A systems approach has previously been used to follow the response behaviour of Arabidopsis thaliana plants upon sulphur limitation. A response network was reconstructed from a time series of transcript and metabolite profiles, integrating complex metabolic and transcript data in order to investigate a potential causal relationship. The resulting scale-free network allowed potential transcriptional regulators of sulphur metabolism to be identified. Here, three sulphur-starvation responsive transcription factors, IAA13, IAA28, and ARF-2 (ARF1-Binding Protein), all of which are related to auxin signalling, were selected for further investigation. IAA28 overexpressing and knock-down lines showed no major morphological changes, whereas IAA13- and ARF1-BP-overexpressing plants grew more slowly than the wild type. Steady-state metabolite levels and expression of pathway-relevant genes were monitored under normal and sulphate-depleted conditions. For all lines, changes in transcript and metabolite levels were observed, yet none of these changes could exclusively be linked to sulphur stress. Instead, up- or down-regulation of the transcription factors caused metabolic changes which in turn affected sulphur metabolism. Auxin-relevant transcription factors are thus part of a complex response pattern to nutrient starvation that serve as coordinators of the metabolic shifts driving sulphur homeostasis rather then as direct effectors of the sulphate assimilation pathway. This study provides the first evidence ever presented that correlates auxin-related transcriptional regulators with primary plant metabolism.

Keywords: Auxin; sulphur metabolism; systems biology; transcription factors.

Fig. 1.

Expression analysis of IAA28, IAA13, and ARF1-BP, respectively, in overexpressing lines from 28-d-old soil-grown plants with northern blot hybridization and q-RT-PCR, respectively. (A) Wild-type and empty-vector plants served as controls. RNA from 10 plants per treatment was extracted and 10 μg total RNA was subjected to northern blot analysis. Gene-specific probes were radioactively labelled and hybridized as described in Materials and methods. Equal loading in each track is demonstrated by comparing the amount of the 28S rRNA. Transcript sizes are given in brackets. (B) Quantification of gene expression in transgenic plants overexpressing IAA28, IAA13, and ARF1-BP, respectively. Gene expression was assessed by q-RT-PCR as described in Materials and methods. The data represent the average relative to the gene expression in control plants.

Fig. 2.

Growth phenotypes of overexpressing lines of IAA28, IAA13, and ARF1-BP, respectively. Wild-type plants (left in each picture) and three independent transgenic lines per construct and three plants per line were grown simultaneously in soil for 10 weeks. (A–C) Lines expressing IAA28, (D–F) lines expressing IAA13, and (G–J) line expressing ARF1-BP, respectively. The wild-type plants were grown for 6 weeks in soil and are shown for comparison of regular and mutant shoot phenotypes. (J) Close-up of leaf morphology of lines overexpressing ARF1-BP in comparison with leaves of wild-type plants (10 weeks old). Two wild-type leaves are shown in the left part of the picture (10 weeks old).

Fig. 3.

Phenotype of mature IAA28 knock-down plants, a sketch depicting the location of the T-DNA insertions, and quantification of gene expression in mutant plants. (A) Wild-type plants (left in each picture) and three mutant plants (right) at 10 weeks. Each of three individually selected homozygote mutant plants is presented. The plants were grown at the same time. (B) Identification of IAA28 T-DNA insertion. Boxes in black represent exons, lines represent non-coding regions, and boxes in grey indicate T-DNA insertions. (C) q-RT-PCR analysis of plants derived from the IAA28 knock-down screen. RNA was extracted from 28-d-old soil-grown plants. Ratios to controls are shown.

Fig. 4.

Contents of cysteine (upper row), γ-glutamylcysteine (GEC; middle row), and glutathione (GSH; lower row) are shown for Arabidopsis plants overexpressing IAA28, IAA13, and ARF1-BP, respectively, or down-regulated with respect to IAA28. Plants were grown for 10 weeks on soil before thiol extraction. IAA28 knock-downs are represented by cross-hatched columns, overexpressing lines by white columns, and wild-type (WT) and empty-vector control lines (EV) by black columns. Values are the mean ±SD of three independent experiments. Asterisks indicate that the difference between the wild-type plants and the manipulated transgenic plants was significant using t-tests (P ≤0.05).

Fig. 5.

Heat map generated from amino acid measurements reflecting log base 2-transformed and normalized amino acid levels and its similarity among themselves and the genotypes. The top colour bar indicates the relative log base 2-fold changes ranging between reduced relative (red) and increased relative (blue) contents of amino acids with respect to the wild-type.

Fig. 6.

Heat-map visualization and cluster tree representations of amino acid contents and genotypes. Data were obtained from experiments where plants were starved of sulphate for 10 d. The heat-map was generated by using log base 2-transformed fold changes. The given data represent the ratio of the determined amino acids for control and starved plants. Each amino acid is represented by a single column and each genotype by a single row. Red indicates decreased relative metabolite content whereas blue indicates increased relative contents of amino acids compared with the wild-type. Separated heat-map visualization of amino acid contents in control and mutant plants are presented in Fig. S3 in Supplementary data available at JXB online and the respective diagrams in Fig. S2.