PHYLOGENETIC AND FUNCTIONAL CLASSIFICATION OF MITOGEN-ACTIVATED AND STRESS-ACTIVATED PROTEIN-KINASES

All currently sequenced stress-activated protein kinases (SAPKs), extr acellular signal-regulated kinases (ERKs), and other mitogen-activated protein kinases (MAPKs) were analyzed by sequence alignment, phylogen etic tree construction, and three-dimensional structure modeling in or der to classify members of the MAPK family. Based on this analysis the MAPK family was divided into three subgroups (SAPKs, ERKs, and MAPK3) that consist of at least nine subfamilies. Members of a given subfami ly were exclusively from animals, plants, or yeast/fungi. A single sig nature sequence, [LIVM][TS]XX[LIVM]XT[RK][WY]YRXPX[LIVM][LIVM], was id entified that is characteristic for all MAPKs and sufficient to distin guish MAPKs from other members of the protein kinase superfamily. This signature sequence contains the phosphorylation site and is located o n loop 12 of the three-dimensional structure of MAPKs. I also identifi ed signature sequences that are characteristic for each of the nine su bfamilies of MAPKs. By modeling the three-dimensional structure of thr ee proteins for each MAPK subfamily based on the resolved atomic struc tures of rat ERK2 and murine p38, it is demonstrated that amino acids conserved in all MAPKs are located primarily in the center of the prot ein around the catalytic cleft. I conclude that these residues are imp ortant for maintaining proper folding into the gross structure common to all MAPKs. On the other hand, amino acids conserved in a given subf amily are located mainly in the periphery of MAPKs, indicating their p ossible importance for defining interactions with substrates, activato rs, and inhibitors. Within these subfamily-specific regions. amino aci ds were identified that represent unique residues occurring in only a single subfamily and their location was mapped in three-dimensional st ructure models. These unique residues are likely to be crucial for sub family-specific interactions of MAPKs with substrates, inhibitors, or activators and, therefore, represent excellent targets for site-direct ed mutagenesis experiments.