list of worm proteins not in yeast from Chervitz et al.

In Response To:

S A Chervitz, L Aravind, G Sherlock, C A Ball, E V Koonin, S S Dwight, M A Harris, K Dolinski, S Mohr, T Smith, S Weng, J M Cherry & D Botstein (1998). "Comparison of the Complete Protein Sets of Worm and Yeast: Orthology and Divergence," Science 282: 2022-2028.

Unpublished Summary Text

A recent review by Chervitz et al. (cited above) compared the yeast genome to the newly finished worm genome. Based on sequence similarity, Chervitz et al. partitioned the ~19,000 proteins in the worm into those shared and not shared with yeast, with the latter group comprising 3557 "worm-only" proteins. It is revealing to look at these shared and not shared proteins from the point of view of protein structure, and in this vein, I have briefly reanalyzed Chervitz et al.'s results, using the scop classification of protein folds and methods described elsewhere (1,2,3). Only subsets of the yeast and worm proteins are amenable to this sort of analysis (e.g. only 2472 worm proteins have a match in the PDB structure database). One finds that the number of protein folds shared between yeast and worm is fairly high: Of the 157 known folds in yeast and the 190 in worm, 139 are shared. More to the point, Chervitz et al.'s 3557 worm-only proteins, which have no obvious sequence similarity to any yeast protein, contain 63 folds shared with yeast. This illustrates dramatically how much more structure is shared than sequence.

The common TIM-barrel fold provides an excellent example of this fold sharing between worm and yeast. It occurs often in each genome (>35 times). Some of these TIM-barrels are associated with functions unique to the worm, representing cases of a common scaffold acquiring worm-specific functions (e.g. 2K1058, a probable methylmalonyl-coA mutase precursor). In contrast, other TIM-barrels in Chervitz et al.'s "worm-only" set roughly correspond to yeast TIM-barrels carrying out similar functions, suggesting they represent cases of marginal sequence similarity that could perhaps be detected by more sensitive comparison programs (e.g. worm C50B6.7 and yeast YBR299W or F01F1.12 and YKL060C).

The 45 folds that are present in the worm and not in yeast likely comprise protein shapes that are essential for carrying out worm-specific functions. 31 of these are also not present in E. coli or any of the first ten microbial genomes to be sequenced (3), indicating that they are probably "metazoan-only" folds. These include the 4-bladed beta-propeller, cystine-knot, and EGF/Laminin folds. The latter was described as a worm-only "domain" by Chervitz et al., but it is interesting to see its uniqueness extend to the level of protein structure. Further information regarding these comparisons is available from bioinfo.mbb.yale.edu/genome/worm_yeast.

(1) A G Murzin, SE Brenner, T Hubbard & C Chothia. (1995). "scop: a structural classification of proteins database for the investigation of sequences and structures" Journal of Molecular Biology 247:536-540.

(2) M Gerstein & M Levitt (1997). "A Structural Census of the Current Population of Protein Sequences," Proceedings of the National Academy of Sciences USA 94: 11911-11916.

(3) M Gerstein (1998). "Patterns of Protein-Fold Usage in Eight Microbial Genomes: A Comprehensive Structural Census," Proteins 33: 518-534.