BIOINFORMATICS<-->STRUCTURE
Jerusalem, Israel, November 17-21, 1996

Abstract


Dialog between sequence homology and 3D crystal structure defines interaction of functional domains in a multi-enzyme complex

Edward A. Bayer (1), Linda J.W. Shimon (2), Felix Frolow (2), Ely Morag (1), Raphael Lamed (3), Jose Tormo (4), Sima Yaron (5) and Yuval Shoham (5)

(1) Dept. Membrane Res. Biophys., Weizmann Institute of Science, Rehovot, Israel;
(2) Dept. Chem. Services, Weizmann Institute of Science, Rehovot, Israel;
(3) Dept. Molec. Microbiol. Biotechnol., Tel Aviv University, Ramat Aviv, Israel;
(4) Molec. Biophys. Biochem., Yale University, New Haven, CT, USA and
(5) Dept. Food Eng. Biotechnol., Technion-IIT, Haifa, Israel

bfbayer@wiccmail.weizmann.ac.il


Microbial degradation of cellulose is an immensely complex process which involves numerous cellulases and related enzymes which act synergistically to hydrolyze the recalcitrant substrate. Many cellulolytic microorganisms produce an intricate type of multienzyme complex - termed the cellulosome, which was first identified on the basis of combined biochemical, immunochemical and genetic techniques in the anaeorbic, thermophilic, cellulolytic bacterium, Clostridium thermocellum. The cellulosome is composed of a conglomerate of subunits, each of which comprises a set of interacting functional domains. Insight into the structural organization of the various cellulosomal enzymatic and the major non-enzymatic subunit was accomplished first by their cloning and sequencing. This approach has led to the identification of numerous functional domains, and supportive biochemical data have indicated their various functions.

The central role of the cellulosome structure is fulfilled by a multi-functional, scaffolding or integrating subunit (called scaffoldin) which is responsible for organizing the cellulolytic subunits into the multi-enzyme complex. This is accomplished by the interaction of two complementary classes of domain, located on two separate types of interacting subunits, i.e., a cohesin domain on scaffoldin and a dockerin domain on each enzymatic subunit. The cohesin-dockerin interaction apparently defines the structure of the cellulosome. The scaffoldin subunit in C. thermocellum also mediates adhesion of the bacterium to the substrate, by virtue of a cellulose-binding domain (CBD). Thus far, about 100 CBDs from a large variety of different cellulolytic bacteria and fungi have been sequenced and classified in about a dozen different families according to sequence homology. Over 20 different cohesin domains and a similar number of dockerin domains have also been described.

We have recently solved at 1.75 Å resolution the crystal structure of a recombinant form of the CBD from the scaffoldin subunit of the cellulosome of C. thermocellum. This particular CBD belongs to Family III, according to the classification scheme based on sequence homology. The protein forms a 9-stranded b-sandwich. The various faces of the CBD were subjected to a docking procedure with the known structure of cellulose. This procedure revealed a series of conserved surface-exposed amino acids which potentially interact with cellulose. Site-directed mutagenesis is currently being employed to systematically replace and analyze the contribution of the suspected binding residues.

Similarly, the cohesin-dockerin interaction is being analyzed on the structural level. A recombinant cohesin has now been crystallized and its structure has been solved (PDB ID code 1ANU) by multiple isomorphous replacement (with the aid of two heavy atom derivatives). Its fold is remarkably similar to that of the CBD, despite the complete lack of sequence identity. We are currently carrying out a similar analysis of the cohesin-dockerin complex. In this context, co-crystals of the complex have already been obtained.

The structures of the CBD and the cohesin domains have allowed us to refine their boundaries within the sequence of the scaffoldin subunit. This knowledge has important consequences concerning scaffoldins from other bacteria. Reexamination of previous efforts to define the limits of the various cellulosomal domains by sequence homology has indicated that past conclusions were erroneous and have led to mistakes in subsequent mutation analyses.


Back to the Abstract Index.