Genomics, evolution, and crystal structure of a new family of bacterial spore kinases

Abstract

Bacterial spore formation is a complex process of fundamental relevance to biology and human disease. The spore coat structure is complex and poorly understood, and the roles of many of the protein components remain unclear. We describe a new family of spore coat proteins, the bacterial spore kinases (BSKs), and the first crystal structure of a BSK, YtaA (CotI) from Bacillus subtilis. BSKs are widely distributed in spore-forming Bacillus and Clostridium species, and have a dynamic evolutionary history. Sequence and structure analyses indicate that the BSKs are CAKs, a prevalent group of small molecule kinases in bacteria that is distantly related to the eukaryotic protein kinases. YtaA has substantial structural similarity to CAKs, but also displays distinctive features that broaden our understanding of the CAK group. Evolutionary constraint analysis of the protein surfaces indicates that members of the BSK family have distinct clade-conserved patterns in the substrate binding region, and probably bind and phosphorylate distinct targets. Several classes of BSKs have apparently independently lost catalytic activity to become pseudokinases, indicating that the family also has a major noncatalytic function.

Figure 1

Conserved motifs in BSKs, arranged by proposed phylogeny. Logos show the relative conservation at selected positions within each of the major family members, with HSK2 as an outgroup. Key residue numbers in YtaA are labeled on top, and consensus CAK motifs and structural elements are labeled on the bottom (o = any hydrophobic residue; x = any residue; lowercase = partially conserved). Skull and crossbones indicate predicted pseudokinases; some CotS orthologs may also be pseudokinases. Structural residues such as H148 and D269 are highly conserved, whereas catalytic residues are lost in pseudokinases and many other positions are conserved but distinct between classes. The hydrophobic linker and putative substrate binding motifs are structural motifs that are discontinuous in the primary sequence, and are shown with the intervening sequence removed. The tree is a schematic, inferred from phylogenetic analysis (Supporting Information Figure S3) and species taxonomy (Supporting Information Table S1). The major taxonomic ranges of each BSK are shown on the right.

Figure 2

Conservation patterns near the ytaA/cotS loci. Despite considerable rearrangements, the ytaA/cotS genes are consistently colocated and coregulated with glycosyl transferases (ytcC and cotSA) and frequently with nucleotide sugar metabolizing genes (ytdA and ytcA-B). Genes are color-coded by orthology; gray represents genes neither conserved in the cluster nor spore-associated. Gene lengths are not to scale.

Figure 3

Crystal structure of YtaA, colored by secondary structure. CAK-specific elements are labeled with an underline.

Figure 4

Overview comparison of YtaA with other CAKs and PKA. Common secondary structure elements shared by all structures are shown in gray, and labeled in the PKA structure. Distinctive structural elements specific to the CAKs are shown in blue and yellow, with the analogous (but structurally distinct) regions of the PKA structure shown in identical colors. The unique helix in YtaA, α2bi, is shown in red. The CAK-specific elements are labeled on all three CAK structures in underline. A: PKA; B: APH; C: ChoK; and D: YtaA.

Figure 5

Comparison of the adenosine binding pocket in YtaA and other CAKs. Structures are presented in identical orientation, based on structural alignment and superposition with DaliLite. This presentation highlights changes in location and orientation of the adenosine molecule in each structure. Key adenosine-interacting residues in each structure are shown (side chains are omitted for residues that contribute only backbone atoms). H-bonds are shown as dotted lines and metal atoms as blue spheres. Portions of the structures are omitted to improve clarity. A: APH; B: ChoK; and C: YtaA.

Figure 6

Comparison of active sites of APH and YtaA. Structures are presented in identical orientation, based on structural alignment and superposition with DaliLite. Conventions are as in Figure 5, except that metal atoms are shown as transparent spheres. A: APH and B: YtaA.

Figure 7

The hydrophobic linker motif conserved only in putatively active BSKs. A: Cutaway view of motif residue interactions. The YtaA structure is in approximately the same orientation as in Figure 4, with secondary structure elements identically colored. Motif residues are rendered with a yellow space-filling shell. H148, a residue highly conserved in almost all PKL kinases, is shown with a space-filling shell in white. Unconserved residues interacting with the motif are in ball-and-stick view. B: Logo of motif, showing selective conservation only in putatively active BSKs. The motif is discontinuous in sequence, and shown with intervening sequence removed.

Figure 8

Predicted substrate binding region of the BSK family, mapped onto the YtaA surface. The orange and blue surface show the region generally conserved throughout the family. The blue region is specifically highly conserved in YtaA (Table IV). The key active site residue D239^YtaA is shown with a red space-filling shell. The unknown ligand in the YtaA structure (green) and the adenosine are shown in ball-and-stick.