Analysis of proteins with the 'hot dog' fold: Prediction of function and identification of catalytic residues of hypothetical proteins

We have analyzed the structural features and sequences of proteins having the hot dog fold. This study reveals that though the basic architecture of the fold is well conserved in these proteins, significant differences exist in their sequence, nature of substrate and oligomerization. Segments with certain conserved sequence motifs seem to play crucial structural and functional roles in various classes of these proteins.The analysis led to predictions regarding the functional classification and identification of possible catalytic residues of a number of hot dog fold-containing hypothetical proteins whose structures were determined in high throughput structural genomics projects.The 'Hot dog' fold was first identified in the crystal structure of β-hydroxydecanoyl thiolester dehydratase (FabA) from E. coli [1]. The hot dog fold in this structure is formed by a sausage-like long central helix wrapped around by a highly curved six-stranded β-sheet resembling the bun (Fig. 1). Several protein structures determined since then were shown to contain this fold. These proteins take part in several functions including the dehydration step of fatty acid elongation, thioester hydrolysis, transcription regulation in fatty acid metabolism and degradation of phenylacetic acid and environmental pollutants.It appears that the hot dog fold is designed for coenzyme A (CoA) binding in cellular processes involving fatty acids and related molecules. The binding site in FabA has been identified as a well-formed deep tunnel present at the subunit interface [1]. Residues from both subunits contribute to the formation of the tunnel which is hydrophobic in nature except at the two catalytic residues His70 and Asp84. FabA, being a homodimer, consists of two tunnels related by a twofold symmetry. The active site is located slightly below the surface of the protein and the active residues at each site come from a different subunit.Dillon and Bateman [2] have unified the large superfamily of hot d