Abscisic acid (ABA) and cytokinin are key hormones controlling plant development. How ABA and cytokinin interplay to control the transition from a dry seed into a young seedling remains elusive. Here we undertook a gain-of-function genetic screen to identify ABA-insensitive mutants during seed germination in Arabidopsis using an estradiol-inducible approach. In the presence of estradiol, one of these mutants gim1 (germination insensitive to ABA mutant 1) exhibited an elevated level of cytokinin that was attributed to the estradiol-induced expression of AtIPT8 that encodes an isopentenyltransferase for the biosynthesis of cytokinins. Our data on OE-2 and Com-1 transgenic plants carrying the ectopically expressing AtIPT8 gene indicated that the elevation of cytokinin level was responsible for the ABA-insensitivity of gim1 seed germination. Further analyses on alterations of gene transcriptomes in the gim1 mutant demonstrated that the expression of some ABA-inducible genes, including ABI5, was reduced, and could not be restored by exogenous ABA treatment. Moreover, we also failed to observe the ABA-mediated repression of a family of cytokinin signal transducers and transcription repressors called type-A ARR4, ARR5 and ARR6 in the gim1 seedlings. Further analysis demonstrated that type-A ARR4, ARR5 and ARR6 could negatively regulate ABI5 expression, and the physical interaction of ABI5 and type-A ARR4, ARR5 and ARR6 proteins was detected. In summary, our study suggests that the interaction of ABA and cytokinin during seed germination and seedling growth can be mediated by the interplay of transcriptional regulators in Arabidopsis

Supporting Information

Figure S1. Cloning and examining AtIPT8 gene in gim1. (a) The diagram shows XVE-T-DNA insertion site in AtIPT8 gene of gim1 mutant (not in scale). Arrows indicate the transcriptional orientation and T-DNA insertion site in promoter region of AtIPT8 gene. The gray boxes stand for two exons, and a line stands for one intron. (b) The expression level of AtIPT8 gene was examined by semi-quantitative RT-PCR in seedlings of Col and gim1 without (−E.) or with (+E.) induction of estradiol (10 μM). Expression level of ACTIN2 represents the loading control.

Figure S2. Quantifications of relative-contents of cytokinins. Cytokinin contents were quantified with 2-week-old seedlings of Col, gim1, OE-2 and ipt8-1-ko. Seedlings of gim1 and OE-2 were induced with 10 μM estradiol for 24 h. For controls, seedlings of Col and ipt8-1-ko were treated with DMSO. The data represent the means ±SE of two reproducible experiments (FW, fresh weight).

Figure S3. Analysis of gene expression in Col and gim1 plants. The gene expression levels of representative downstream targets in ABA signaling which were identified in our microarray assay were validated by qRT-PCR analysis in Col and in gim1 seedlings. Seedlings were first treated by 10 μM estradiol for 24 h (−ABA) (white bars) and then treated with ABA (50 μM) for 3 h (+ABA) (black bars). The data are the mean values of three reproducible experiments (±SD).

Table S1. Representative ABA- and/or stress-responsive genes identified in microarray analysis were compared in gim1 versus Col.

Table S2. Representative cytokinin-related genes identified in microarray analysis were compared in gim1 versus Col.