Fig 1.
BLI protein localizes in Golgi and nucleus.
A-R, Subcellular localization of BLI in Arabidopsis bli-11 mutant plants. BLI:BLI-GFP was co-expressed with the Golgi marker SYP32-RFP, the trans-Golgi/early endosome marker VTI12-RFP, the ER marker WAK2-RFP, or the nucleus marker H2B-RFP. Root cells in the elongation zone (A-C, G-I, M-R) or meristematic zone (D-F, J-L) were observed under a confocal microscope and merged. Close-up pictures are shown in A-L. Scale bars are 20 μm.
Fig 2.
bZIP60 mutation suppresses UPR gene expression but not shoot growth phenotype of BLI mutant plants.
A-B, Phenotypic analysis. T-DNA mutant bli (bli-1) was crossed to either bZIP28 mutant (bzip28) or bZIP60 mutant (bzip60) to generate the respective double mutant plants. 2-week-old plant seedlings were photographed (A) and siliques lengths were measured later at reproductive stage (B). Error bars represent SD (n = 10). Letters above the bars in (B) indicate significant differences as determined by LSD test following ANOVA analysis (p<0.05). Bar = 5 mm. C, UPR gene expression analysis. Total RNA was exacted from 2-week-old plants for qRT-PCR. Fold change is the gene expression level in the mutants normalized to that in the WT, both of which were normalized to the expression of ACTIN. Asterisks indicate significance levels when comparing to the WT control in t-test. (*, p<0.05; **, p<0.01; n.s., not significant at p<0.05).
Fig 3.
IRE1 mutations suppress the shoot growth phenotype of BLI mutant plants.
A-D, Phenotypic analysis. T-DNA mutant bli (bli-1) was crossed to either IRE1A mutant (ire1a) or IRE1B mutant (ire1b) to generate the respective double mutant plants. 2-weeks-old plants were photographed (A), weighted (B). Representative siliques were photographed (C) and siliques lengths were measurement at reproductive stage (D). Error bars represent SE (n = 3) in (B) and represent SD (n = 12) in (D). Letters above the bars indicate significant differences as determined by LSD test following ANOVA analysis (p< 0.05). Bar = 5 mm. E-F, Cell death and ROS accumulation analysis. 2-weeks-old plants were stained with trypan blue (E) or DAB (F). Arrows and arrow heads point to sites of cell death and ROS accumulation, respectively. Bar = 1 mm.
Fig 4.
IRE1A mutation suppresses both canonical and non-canonical UPR gene expression in BLI mutant plants.
A-B, Gene expression analysis. T-DNA mutant bli (bli-1) was crossed to either IRE1A mutant (ire1a) or IRE1B mutant (ire1b) to generate the respective double mutant plants. Total RNA was exacted for RNA-Seq analysis (A) and qRT-PCR (B). Up-regulated or down-regulated genes in the mutants relative to the wild-type plants were used to draw the Venn diagrams. Fold change is the gene expression level in the mutants normalized to that in the WT, both of which were normalized to the expression of ACTIN. Error bars represent SE (n = 3). Asterisks indicate significance levels when comparing to the WT control in t-test. (*, p<0.05; **, p<0.01; n.s., not significant at p<0.05). bZIP60S, spliced bZIP60 transcript. Note that bZIP60-S, BiP3, NSF, DERD3A, SARA1A and NAC103 are up-regulated while MRN1 and LTP4 are not up-regulated by canonical ER stress inducer tunicamycin.
Fig 5.
Residues in the protein kinase domain of IRE1A is required for the shoot growth phenotype in BLI mutant plant background.
A-B, Site-specific mutagenesis analysis of the IRE1A auto-phosphorylation. Domain structure of IRE1A is shown in (A) with the mutated amino acids highlighted. The wild-type form and various mutated forms of IRE1A were expressed in E. coli and purified proteins were treated with CIAP for de-phosphorylation assays. Western blotting analysis was done to detect the molecular weight shifts. C-D, Genetic complementation analysis. T-DNA mutant bli (bli-1) was crossed to either IRE1A mutant (ire1a) or IRE1B mutant (ire1b) to generate the respective double mutant plants. The wild-type form and various mutated forms of IRE1A were expressed the bli ire1a double mutant background and 2-weeks-old T3 transgenic plants were photographed (C) and fresh weight were measured (D). Error bars represent SE (n = 3). There were at least 10 plants for each of the three biological replicates. Letters above the bars indicate significant differences as determined by LSD test following ANOVA analysis (p<0.05). Bar = 5 mm.
Fig 6.
Residues in the protein kinase domain of IRE1A is required for non-canonical UPR gene expression in the BLI mutant background.
T-DNA mutant bli (bli-1) was crossed to either IRE1A mutant (ire1a) or IRE1B mutant (ire1b) to generate the respective double mutant plants. The wild-type form and various mutated forms of IRE1A were expressed in the bli ire1a double mutant background and 2-weeks-old T2 transgenic plants were collected for gene expression analysis. Fold change is the gene expression level in the mutants or the genetically complemented materials normalized to that in the WT, both of which were normalized to the expression of ACTIN. Error bars represent SE (n = 3). Asterisks indicate significance levels when comparing to the WT control in t-test. (*, p<0.05; **, p<0.01; n.s., not significant at p<0.05).
Fig 7.
A working model for IRE1-mediated growth regulation in BLI mutant plants.
In the presence of BLISTER (BLI) in wildtype (WT) plants, BLI may directly or indirectly inhibit IRE1A activation and clustering. In the absence of BLI in bli mutant plants, the inhibitor is removed and IRE1A is activated. The RNase activity of IRE1A for bZIP60 mRNA splicing and downstream canonical UPR gene expression is dependent on N780 in the RNase domain and D570/K572 in the ATP-binding pocket of IRE1A. In contrast, the auto-phosphorylation activity of IRE1A is dependent on D570/K572 and D590 of IRE1A, which is important for non-canonical UPR gene expression and the shoot growth retardation phenotype.