| One of the earliest responses of legumes to symbiotic signalling is the oscillation of calcium concentration in the nucleoplasm of root epidermal cells. The integration and decoding of the calcium-spiking signal involves a calcium-and calmodulin-dependent protein kinase (CCaMK) and its phosphorylation substrate CYCLOPS. Multiple phosphorylation sites are predicted in both proteins, pointing to a complex regulatory mechanism, however, their relevance during symbiosis is unclear.Here the Lotus japonicus supll mutant was selected as a presumed suppressor of the Lotus japonicus harl-1hypernodulation phenotype through ethyl methanesulfonate (EMS). Using Map-based cloning and Next Generation Sequencing, two linked mutations were found and called ccamk-14and nph3-1, respectively. sup11was backcrossed to wild-type L. japonicus Gifu in order to identify the corresponding single mutant individuals. Genotyping of har1-1, ccamk-14and nph3-1loci in the resulting504F2progeny allowed for the selection of individuals that were homozygous wild-type for two loci but remained heterozygous at a third locus by Derived Cleaved Amplified Polymorphic Sequence Markers (dCAPS). These individuals were allowed to self and subsequent genotyping of F3progenies led to the selection of single, har1-1, ccamk-14and nph3-1mutant lines.nph3-1had no discernible negative impact on nodulation and mycorrhization, although shoot growth of the mutant plant was somewhat diminished in comparison with wild-type Gifu. Consistent with these observations, the wild-type L. japonicus CCaMK gene was able to restore normal mycorrhization in transgenic hairy roots induced on sup11shoots. The defective nodule infection and AM root colonization phenotypes of the ccamk-14single mutant were also restored by the CCaMK gene. Thus, ccamk-14was defined as the causative mutation, which was further confirmed through detailed analyses of the ccamk-14single mutant.The ccamk-14mutant is defective for both rhizobial and AM fungal infection but able to initiate nodule organogenesis. The frequency of the epidermal infection events were enhanced in ccamk-14, while the ccamk-14mutant was severely defective in the cortical infection by rhizobia and AM fungi, which suggested that ccamk-14points to a cell-type specific regulation of bacterial infection. The ccamk-14mutation causes a serine to asparagine substitution at position337located within the calmodulin binding site. The wild-type CCaMK was subjected to autophosphorylation in the presence of Ca2+by an in vitro kinase assay and the resulting phosphorylated sites were identified by HPLC-MS/MS. This analysis revealed that the S337residue is subjected to phosphorylation. Given that S337is located within the CaM-binding domain and that phosphorylation of discrete amino-acid residues within the CaM binding domain of the multifunctional Ca2+/calmodulin-dependent protein kinase II (CaMKII) in mammals has been shown to prevent Ca2+/CaM binding, we first tested whether the S337residue is pertinent to this process. In addition to CCaMKS337N (corresponding to the ccamk-14mutation), a phosphorylation-mimetic CCaMK S337D was constructed by site-directed mutagenesis and used in CaM binding assays. Approximately three times more CCaMKS337N was retained in comparison with CCaMK, showing that the substitution of S337to N significantly enhances this binding. In contrast, CCaMKS337D showed a strongly reduced binding to CaM, indicating interference by the phosphomimetic replacement.The biological relevance of the S337to D substitution was subsequently tested by expressing CCaMKS337D from the L. japonicus CCaMK promoter in transgenic hairy roots induced on non-transgenic ccamk-13mutant shoots. Unlike wild-type CCaMK, which when expressed from the endogenous promoter, restored normal nodulation and arbuscular mycorrhization to the ccamk-13null mutant background, no complementation was observed with CCaMKS337D, showing the impairment in functionality of this mutant protein. Like CCaMKS337N, CCaMKS337D did not confer spontaneous nodule formation in transgenic hairy roots that were grown in the absence of M. loti.Given the influence of S337modification on CaM binding, we also tested whether CCaMK, CCaMKS337N and CCaMKS337D exhibit an alteration in the interaction with CYCLOPS, a known interactor and an in vitro phosphorylation substrate of CCaMK. Quantitative yeast two-hybrid (Y2H) analysis revealed that a kinase-dead version, CCaMKG30E was strongly impaired in its interaction with CYCLOPS, while the interaction of both CCaMKS337N and CCaMKS337D with CYCLOPS was at a similar level as wild-type CCaMK.As Ca2+/CaM binding was altered in ccamk-14, we subsequently tested the relevance of the S337site for the CCaMK kinase activity in vitro with respect to both, autophosphorylation (auto-P) and substrate phosphorylation. A commonly used artificial phosphorylation substrate, myelin basic protein (MBP) and CYCLOPS, the known in vivo binding partner and an in vitro CCaMK phosphorylation substrate, were analyzed in parallel to compare wild-type CCaMK and the two mutant proteins, CCaMKS337N and CCaMKS337D. Our data showed:CCaMKS337D remained almost totally unresponsive to addition of Ca2+or Ca2+/CaM. In contrast, substrate-dependent responses were observed for CCaMK and CCaMKS337N In the presence of CYCLOPS, the basal, wild-type CCaMK auto-P activity approximately doubled upon addition of Ca2+and returned to the basal level in the presence of Ca2+/CaM, which indicated the opposite effects of Ca2+and Ca2+/CaM on in vitro autophosphorylation of CCaMK. This pattern was completely different in both CCaMKS337N and CCaMKS337D mutant versions. Strikingly, in contrast to MBP, CYCLOPS phosphorylation by CCaMKS337N was much more sensitive to stimulation by Ca2+/CaM and reached a level3-fold higher than that observed for the wild-type CCaMK. In contrast, CCaMKS337D CYCLOPS phosphorylation activity remained unchanged upon Ca2+addition, and showed only a slight increase, when Ca2+/CaM were added to the reaction.Taken together, the substitution of S337in the ccamk-14mutant participates in the binding of Ca2-/CaM and the negative regulation of the CCaMK activity. As inferred by CCaMKS337D, the phosphorylation of S337likely limits further activation of CCaMK by preventing its responsiveness to Ca2+and restricting Ca2+/CaM binding, the regulatory capacity that has been lost in CCaMKS337N. Future experiments will have to establish whether the phosphorylation of T265within the CCaMK kinase domain triggers a phosphorylative burst that targets S337within the CaM binding domain. |
PDF Full Text Request