Figure 1.
Failure to detect PtdIns5P by HPLC.
HPLC analysis of deacylated lipid products isolated from rosette leaves of in vivo radiolabeled whole plants with orthophosphate (300 µCi/ml) for 16 hours at room temperature. Peak identification: 1, phospholipids with positive charge; 2, phosphatidylinositol; 3, phosphatidic acid; 4, free inorganic phosphate; 5, PtdIns3P; 6, PtdIns4P; 7, PtdIns5P; 8, PtdIns(4,5)P2.
Figure 2.
PtdIns5P in Arabidopsis tissues determined by the PIP4Kα assay.
Thin layer chromatographic separation of the products of the in vitro phosphorylation reaction between PIP4Kάandendogenous phosphoinositides isolated from leaves (lane 2), stems (lane 3), flowers (lane 4) and siliques (lane 5). A PtdIns(4,5)P2 standard is shown (lane 1) (panel a). HPLC analysis of deacylated products of the in vitro phosphorylation reaction between endogenous phosphoinositides (from leaves) and PIP4Kα. Deacylated PtdIns(3,4)P2 and PtdIns(4,5)P2 used as standards (panel b). The products of the in vitro phosphorylation reaction between endogenous phosphoinositides isolated from rosette leaves and PIP4Kα were incubated in the absence (lane 1) or presence (lane 2) of yeast YNK-5 phosphatase. The reaction products were analyzed by thin layer chromatography. The position of a PtdIns4P standard is shown (lane 3) (panel c).
Figure 3.
Time-course of water loss from detached leaves.
Water-loss in detached rosette leaves after exposure to ambient air and a temperature of 20°C determined as percentage of residual tissue mass taken as 100%. Data are from four independent experiments; bars are s.d. (panel a). Lipids extracted from leaves after 30 minutes or 120 minutes of air-exposure were processed for PtdIns5P content and expressed as the percentage of the fresh sample (indicated as 100%). Data are from six independent experiments; bars are s.d. (panel b). Detached (column 1), air-exposed (column 2) and water-submerged leaves (column 3) were treated in parallel for 120 minutes. Cellular PtdIns5P is indicated as pmoles PtdIns5P/mg initial fresh tissue mass. Bars are s.d. (panel c).
Figure 4.
PtdIns5P levels and ATX1 activity in AtMTM1 mutant cells.
PtdIns5P levels in fresh and dehydrated wild type leaves, in OX-AtMTM1 leaves, and in leaves from homozygous mtm1 plants (panel a). Relative WRKY70 transcripts in non-stressed and in stressed wild-type, OX-AtMTM1, and mtm1 mutant cells, quantified by real-time PCR (panel b). Quantitative PCR of ChIP assays of H3K4me3 methylation in wild-type, OX-ATMTM1 and mtm1 mutant chromatins in fresh leaves and in leaves after dehydration-stress (panel c). Relative enrichment in WRKY70-specific fragments vs input DNA.
Figure 5.
ATX1 activity under drought conditions and different backgrounds.
Relative WRKY70 transcripts in freshly harvested leaves and in leaves after 2 h air-exposure in wild type, OX-AtMTM1, and mtm1 backgrounds quantified by real-time PCR normalised versus actin (panel a). Quantitative PCR of ChIP assays of WRKY70 H3K4me3 levels in wild type control and dehydration-stressed tissue and in respective samples in the OX-AtMTM1 and mtm1 backgrounds. The y-axis represents the relative enrichment of recovered DNA versus the input (panel b). Quantitative PCR ChIP assay with anti-ATX1 antibodies at the WRKY70 locus before and after dehydration-stress in wild type, OX-AtMTM1, and mtm1 mutant chromatins; relative enrichment in WRKY70-specific fragments vs input DNA (panel c). Relative ATX1 transcript levels in freshly harvested leaves and in leaves after 2 h air-exposure in wild type, OX-AtMTM1, and mtm1 backgrounds quantified by real-time PCR normalised versus actin (panel a). Bars are s.d.
Figure 6.
FLC and AG transcripts are not regulated by AtMTM activity.
Relative FLC transcripts determined in leaves of 12 day old seedlings in non-stressed and in stressed wild-type, atx1, OX-AtMTM, and mtm mutants, quantitated by real-time PCR and normalized against actin (panel a). Relative AG transcripts in inflorescences of wild-type, atx1, OX-AtMTM, and mtm mutants, under non-stressed and stressed conditions (panel b). Bars are s.d.
Figure 7.
Sub-cellular distribution of ATX1 co-expressed with AtMTM1.
ATX1 and its derivatives were expressed as ATX1-GFP fusion proteins (green signal) while AtMTM1 was expressed as an RFP fusion protein (red signal). a) all images in this column illustrate cells expressing ATX1 or derivatives alone; b) all images in this column illustrate cells co-expressing ATX1 (or ATX1 derivatives) with AtMTM1. Arrows point to nuclei in cells expressing a GFP-fusion protein. ATX-N is the N-terminal ATX1 portion, ATX1-C is the C-terminal ATX1 portion containing the ePHD and the SET domains (SF7). Arrows point to nuclei; Bars are 50 µm.
Figure 8.
Cells showing nuclear GFP-signal associated with ATX1 protein or its derivatives.
The number of cells showing nuclear localization of transiently expressed ATX1 alone is set at 100%. ATX+M represents the population of cells displaying nuclear signal when co-expressing the entire ATX1 and AtMTM; ATX-N+M represents cells co-expressing the entire N-terminal portion of ATX1 and AtMTM; ATX-C+M are cells co-expressing the ATX1-C-terminal portion and AtMTM, respectively; bars are s.d. from four independent experiments.
Figure 9.
Structure of representative constructs used in the study.
The FYRN and FYRC domains are called DAST, for Domain Associated with SET in Trithorax; the ePHD belongs to a distinct, extended PHD, family of proteins. The putative AtMTM domains are labeled according to their homology to the human MTMR2.
Figure 10.
PtdIns5P in the cytoplasmic retention of ePHD is dependent on AtMTM1 catalytic efficiency.
Nuclear localisation of ePHD-GFP expressed alone. Arrows in all panels point to nuclei (panel a). ePHD-GFP co-expressed with wild-type AtMTM1-RFP causes depletion of the green nuclear signal (panel b). The consensus amino acids in the active AtMTM1 phosphatase site and the amino acid substitutions in mut1 and mut2 (panel c). Phosphoinositide 3′-phosphatase activity of recombinant wild type, mut1 and mut2 phosphatases. The Vmax values towards PtdIns(3,5)P2 as a substrate are shown as the percentage of the wild-type (WT) ATMTM1 activity. Bars are s.d. from three measurements (panel d). Distribution of ePHD-GFP co-expressed with mut1AtMTM1-RFP show nuclear localizations of the green signal (panel e). Co-expression of ePHD-GFP with mut2AtMTM1-RFP. The green signal is largely distributed in the cytoplasm (panel f). Bars are 50 Pm.