Auto-inhibition in Calcium-dependent Protein Kinase CPK34 Results from a Combination of Competitive and Non-competitive Mechanisms
AdvisorHarper, Jeffrey F.
Biochemistry and Molecular Biology
AltmetricsView Usage Statistics
Calcium-dependent protein kinases (CPKs) are Ser/Thr protein kinases that are present in plants, ciliates and apicomplexan parasites. CPKs have a novel structure in which the kinase domain is connected to a calcium-sensing calmodulin-like domain (CaM-LD). The kinase and CaM-LD are connected via an auto-inhibitory junction sequence that can block the enzyme's active site. Like calmodulin, the CaM-LD for a typical CPK contains four calcium-binding EF-hands. The kinase is activated when Ca2+ binding to the CaM-LD changes the enzymes conformation. This conformational change is proposed to either directly disengage the auto-inhibitor from a pseudo-substrate competitive binding interaction with the active site (model-1), or disrupt a non-competitive inhibitory structure that alters access to, or conformation of, the active site (model-2). Here we engineered each of the four EF-hands with Ala substitutions designed to disrupt key Ca2+-binding ligand interactions at positions-1 and -12 within each EF hand. The mutant enzymes were expressed and purified from Escherichia. coli, and then used for in vitro kinase assays with Syntide-2 as a substrate. All four EF-hand mutants showed a 2-fold increase in the amount of Ca2+ required for activation, as well as a more than 2-fold decrease in their maximum activity. These results are consistent with a model in which each EF-hand mutation created a partially inhibited enzyme that was unable to fully disengage from their auto-inhibited conformation. Interestingly, peptide substrate KM determinations indicated that the partially autoinhibited conformations were not equivalent for all four EF-hand mutants. For EF 1 and -2 mutants, the peptide substrate KM for Syntide-2 was 3-fold weaker compared to WT. This is consistent with model-1 in which the auto-inhibitor can function as a competitive inhibitor with respect to a peptide substrate. In contrast, EF3 and 4 mutants showed KM's comparable to WT (only about 1.3-fold weaker), suggesting that these enzymes are still partially inhibited through a non-competitive mechansim. This study provides biochemical evidence that auto-inhibition in a CPK results from a combination of competitive and non-competitive mechanisms, with the N- and C-terminal lobes having kinetically distinct functions. Additional substitutions were made in EF2 to separately test the role of all five residues that have side chains that help coordinate calcium (residue positions-1, -3, -5, -9 and -12 in EF-loop). We found that an Ala substitution of each of the five residues significantly disrupted enzyme activation. Four of the five substitutions (all but position-3) resulted in a more than 2-fold decrease in Vmax, regardless of using high calcium concentrations. The enzymes' 50% calcium activation thresholds (K') were observed at [Ca2+] ≈ 1 μM in most cases except D420A mutant (position-3). The K' of D420A mutant was observed at [Ca2+] ≈ 83 μM, and its activity could be restored to near wild type levels at [Ca2+] ≈ 600 μM. This suggests that the primary contribution of position-3 in EF2 is to simply increase calcium affinity. In contrast, mutants with Ala substitutions at the other four positions displayed reduced activity that could not be further activated by elevated [Ca2+]. This indicates that positions-1, -5, -9 and -12 are not only important for overall calcium affinity, but that calcium binding to each position is required to promote a conformational change that disrupts auto-inhibition.Transgenes encoding CPK34 with selected EF-hand mutations were used to create stably transformed Arabidopsis plants, and tested for their ability to rescue a pollen growth and fertility defect associated with a loss of, cpk17/34 (-/-, -/-). Seed set analysis showed that Ala substitutions of position-12 in any of the four EF-hands resulted in cpk17/34 (-/-, -/-) plants with a partial restoration of seed set. In pollen transmission assays in which rescued mutant pollen were tested in competition with wild type pollen, the EF-hand mutants provided more than 10-fold weaker transmission efficiencies compared to a control wild type CPK34 transgene. In both seed set and pollen transmission assays, the most severely disrupted rescue potential was associated with EF2 mutations, and the least disrupted with EF3 mutations. These partial rescues are consistent with in vitro biochemical evidence indicating that mutant enzymes still retained partial activity, albeit with weaker calcium activation thresholds and reduced Vmax's,. Using the seed set rescue assay to compare the relative effects of Ala substitutions of EF2 positions-1, -3 and -12, the weakest rescue potential occurred with a position-3 substitution, D420A. This is interesting because CPK34-D420A can attain near wild type levels of activity at high [Ca2+]. However, at less than 1 μM calcium, the D420A mutant has less activity than Ala substitutions of position-1 or -12 (of all EF-hand mutants). This suggests that the D420A mutant can be used as an in vivo gauge for calcium concentrations at the plasma membrane druing pollen tube tip growth, and supports an estimate that CPK34 functions are normally being activated by Ca2+ signals with magnitudes of less than 1 μM.