Despite the existence of structures of substrate-bound and substrate-free forms of two highly specific members of the family of phosphagen (guanidino) kinases, namely arginine and creatine kinase, the full determination of the structural correlates of substrate specificity within this family of enzymes has remained elusive. Such has prompted the expansion of this research to elucidate the structural correlates of specificity of each phosphagen kinase for its preferred substrate to lombricine kinase (LK), the largest and least specific of these enzymes. The culmination of this work has firmly established a paradigm for substrate specificity in multi-substrate enzyme reactions.
As the least specific of the phosphagen (guanidino) kinases, the study of lombricine kinase complements that of the highly specific arginine and creatine kinases. The primary goal of this research was a more advanced structural understanding of the basis for substrate specificity within this family of enzymes, to elucidate how lombricine kinase catalyzes reactions with a wider variety of substrates, and specifically to test the hypothesis that its activity on this wider variety of substrates is attributable to its ability to accurately pre-align these substrates in a way that arginine and creatine kinases can not.
Here we report the largely identical structures of two substrate-free forms of lombricine kinase crystallized at two different pHs and analyze them within the context of their differing space groups, non-crystallographic symmetry, and structural differences upon superimposition with other phosphagen kinase structures. A predicted structural homology model for the closed lombricine kinase transition-state analogue conformation has then been generated from a sequence alignment and quasi-rigid sub-domain motions known to exist in AK, as described by Yousef, 2003. The robustness of the predicted model of the closed LK TSAC structure was supported by CHARMM-based targeted molecular dynamics (TMD) methods, using the open substrate-free lombricine kinase structure as the starting model and the AK TSAC structure as the target. Interpretation of the model along with evidence from multiple sequence alignments, comparisons of substrate structures and structural superimposition of the predicted LK model with the closed arginine kinase transition-state structure indicates that the structural correlates of the lombricine specificity are likely to include regions of sequence that are conserved in both known LKs and physically close to the beta carbon of the substrate-analog arginine in the hybrid model. Lysine 83, that is distinct from AKs and strictly conserved in both known LK sequences, appears to position its side chain nitrogen toward the beta carbon of arginine. Further analysis of this model shows a pair of histidines (H187 and H313) and a pair of cysteines (C58 and C270), one on each side of the substrate, are in the general vicinity and could mediate lombricine specificity through salt bridges that would stabilize the negative charge on the phosphate moiety between the beta and gamma carbons of the substrate lombricine, the only position where lombricine differs from the structure of the substrate arginine (Figure 1). The results show conclusive evidence that Lysine 83 of lombricine kinase is clearly involved in mediating the specificity of LK toward the substrate lombricine.
