Regulation of Arabidopsis PI4P 5-kinases by phosphorylation
Final Report Abstract
Phosphoinositides (PIs) are minor but essential lipid constituents of eukaryotic membranes. The components of a PI signaling system in plants have long been known. However, only in recent years have PIs emerged as regulatory factors with roles in processes that are central foir plant function, such as the control of cell polarity in tip growing cells and of auxin distribution in vegetative tissues. A predominant PI is phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2), which can serve various alternative functions in the cell. PtdIns(4,5)P2 is formed from lipid precursors by PI4P 5-kinases. Research performed in this project was aimed at elucidating the unknown regulation of the important Arabidopsis PI4P 5-kinase isoform, PIP5K2. In order to test whether PIP5K2 was phosphorylated by Arabidopsis protein kinases, expression and purification protocols were developed that allowed for sufficient purified recombinant PIP5K2 protein for biochemical tests. The data establish that PIP5K2 can be phosphorylated not only be protein kinase A but also by protein kinase activities present in Arabidopsis leaf extracts. Phosphorylation by leaf extract occurs in different portions of the enzyme, as is evident from the analysis of N-terminally truncated variants of PIP5K2. Phosphorylation of PIP5K2 by the extracts results in an inhibition of PIP5K2 catalytic activity. Mass-spectrometry-based analysis of phosphopeptides after tryptic digestion of PIP5K2 protein that has been incubated with leaf extract revealed a number of phosphorylated serines and threonines. Further sites were suggested by computer-based prediction. Using sitedirected mutagenesis, variants of PIP5K2 were recombinantly generated, in which the determined phosphosites wer altered to either alanines (A) or aspartates (D), potentially mimicking the dephosphorylated and phosphorylated states of the parent protein. The biochemical analysis of these variants (e.g., T430, S500 or S370) indicates that some D variants display reduced catalytic activity, whereas the corresponding A variants are catalytically active. This pattern is consistent with an inhibition of catalytic activity by phosphorylation in these positions. A/D substitution in other sites did not display a discernable effect or resulted in overall reduced catalytic activity of the variant enzymes, suggesting these sited are important for enzyme architecture and not easily accessible to experimental analysis by this approach. Compared to the parental PIP5K2 enzyme, PIP5K2 T430A displayed reduced phosphorylation by leaf extract, suggesting this position as a relevant phosphorylation site. While catalytic activity was altered for some of the candidate A/D-variants, we observed no change in plasma membrane association of the variant enzymes, as was tested by monitoring the distribution of fluorescence-tagged enzymes during transient expression in tobacco pollen tubes. The incidence of morphological changes induced by the expression of PIP5K2 in pollen tubes was changed upon expression of some of the A/D-variants, indicating a relevant role for thecorresponding phosphorylation sites in vivo. Further effects on in vivo functionality were obtained from mutant complementation tests, expressing variant forms of PIP5K2 in a pip5k1 pip5k2 double mutant. In these tests, the catalytically active variant PIP5K2 T500A positively rescued the severe mutant phenotype, whereas the corresponding variant PIP5K2 T500D did not. Overall, the accumulated data provide proof that PI4P 5-kinases are phosphorylated in Arabidopsis and that the posttranslational modification has an effect on the function of the enzymes. The data are an important and necessary base for the identification of protein kinases involved in the control of PtdIns(4,5)P2-formation.
