Protein products obtained by site-preferred partial crosslinking in protein crystals and "liberated" by redissolution.

The use of protein crystals as a source of nanoscale biotemplates has attracted growing interest in recent years owing to their inherent internal order. As these crystals are vulnerable to environmental changes, potential applications require their stabilization by chemical crosslinking. We have previously shown that such intermolecular chemical crosslinking reactions occurring within protein crystals are not random events, but start at preferred crosslinking sites imposed by the alignment of protein molecules and their packing within the crystalline lattice. Here we propose a new working hypothesis and demonstrate its feasibility in enabling us to extricate homogeneous populations of single protein molecules that display chemical point mutations or of dimers that show homogeneous chemical crosslinking, and that have the potential for isolation of higher structures. Characterization of the crosslinking mechanism and its end products opens the way to the potential retrieval of such specific modified/intermolecular crosslinked products simply by effecting partial crosslinking at identified preferred sites, followed by time-controlled arrest of the crosslinking reaction and dissolution of the crystals by medium exchange complemented by chromatographic purification.

Noncellulosomal cohesin from the hyperthermophilic archaeon Archaeoglobus fulgidus.

The increasing numbers of published genomes has enabled extensive survey of protein sequences in nature. During the course of our studies on cellulolytic bacteria that produce multienzyme cellulosome complexes designed for efficient degradation of cellulosic substrates, we have investigated the intermodular cohesin-dockerin interaction, which provides the molecular basis for cellulosome assembly. An early search of the genome databases yielded the surprising existence of a dockerin-like sequence and two cohesin-like sequences in the hyperthermophilic noncellulolytic archaeon, Archaeoglobus fulgidus, which clearly contradicts the cellulosome paradigm. Here, we report a biochemical and biophysical analysis, which revealed particularly strong- and specific-binding interactions between these two cohesins and the single dockerin. The crystal structure of one of the recombinant cohesin modules was determined and found to resemble closely the type-I cohesin structure from the cellulosome of Clostridium thermocellum, with certain distinctive features: two of the loops in the archaeal cohesin structure are shorter than those of the C. thermocellum structure, and a large insertion of 27-amino acid residues, unique to the archaeal cohesin, appears to be largely disordered. Interestingly, the cohesin module undergoes reversible dimer and tetramer formation in solution, a property, which has not been observed previously for other cohesins. This is the first description of cohesin and dockerin interactions in a noncellulolytic archaeon and the first structure of an archaeal cohesin. This finding supports the notion that interactions based on the cohesin-dockerin paradigm are of more general occurrence and are not unique to the cellulosome system.

Structure of a family 3b' carbohydrate-binding module from the Cel9V glycoside hydrolase from Clostridium thermocellum: structural diversity and implications for carbohydrate binding.

Family 3 carbohydrate-binding modules (CBM3s) are associated with both cellulosomal scaffoldins and family 9 glycoside hydrolases (GH9s), which are multi-modular enzymes that act on cellulosic substrates. CBM3s bind cellulose. X-ray crystal structures of these modules have established an accepted cellulose-binding mechanism based on stacking interactions between the sugar rings of cellulose and a planar array of aromatic residues located on the CBM3 surface. These planar-strip residues are generally highly conserved, although some CBM3 sequences lack one or more of these residues. In particular, CBM3b' from Clostridium thermocellum Cel9V exhibits such sequence changes and fails to bind cellulosic substrates. A crystallographic investigation of CBM3b' has been initiated in order to understand the structural reason(s) for this inability. CBM3b' crystallized in space group C222(1) (diffraction was obtained to 2.0 A resolution in-house) with three independent molecules in the asymmetric unit and in space group P4(1)2(1)2 (diffraction was obtained to 1.79 A resolution in-house and to 1.30 A resolution at a synchrotron) with one molecule in the asymmetric unit. The molecular structure of Cel9V CBM3b' revealed that in addition to the loss of several cellulose-binding residues in the planar strip, changes in the backbone create a surface 'hump' which could interfere with the formation of cellulose-protein surface interactions and thus prevent binding to crystalline cellulose.

Crystal structure of a type-II cohesin module from the Bacteroides cellulosolvens cellulosome reveals novel and distinctive secondary structural elements.

The incorporation of enzymes into the multi-enzyme cellulosome complex and its anchoring to the bacterial cell surface are dictated by a set of binding interactions between two complementary protein modules: the cohesin and the dockerin. In this work, the X-ray crystal structure of a type-II cohesin from scaffoldin A of Bacteroides cellulosolvens has been determined to a resolution of 1.6 angstroms using molecular replacement. The type-II B. cellulosolvens cohesin (Bc-cohesin-II) is the first detailed description of a crystal structure for a type-II cohesin, and its features were compared with the known type-I cohesins from Clostridium thermocellum and Clostridium cellulolyticum (Ct-cohesin-I and Cc-cohesin-I, respectively). The overall jelly-roll topology of the type-II Bc-cohesin is very similar to that observed for the type-I cohesins with three additional secondary structures: an alpha-helix and two "beta-flaps" that disrupt the normal course of a beta-strand. In addition, beta-strand 5 is elevated by approximately 4 angstroms on the surface of the molecule, relative to the type-I Ct and Cc-cohesins. Like its type-I analogue, the hydrophobic/aromatic core of Bc-cohesin-II comprises an upper and lower core, but an additional aromatic patch and conserved tryptophan at the crown of the molecule serves to stabilize the alpha-helix of the type-II cohesin. Comparison of Bc-cohesin-II with the known type-I cohesin-dockerin heterodimer suggests that each of the additional secondary structural elements assumes a flanking position relative to the putative dockerin-binding surface. The raised ridge formed by beta-strand 5 confers additional distinctive topographic features to the proposed binding interface that collectively distinguish between the type-II and type-I cohesins.

Preliminary X-ray characterization and phasing of a type II cohesin domain from the cellulosome of Acetivibrio cellulolyticus.

The N-terminal type II cohesin from the cellulosomal ScaB subunit of Acetivibrio cellulolyticus was crystallized in two different crystal systems: orthorhombic (space group P2(1)2(1)2(1)), with unit-cell parameters a = 37.455, b = 55.780, c = 87.912 A, and trigonal (space group P3(1)21), with unit-cell parameters a = 55.088, b = 55.088, c = 112.553 A. The two crystals diffracted to 1.2 and 1.9 A, respectively. A selenomethionine derivative was also crystallized and exhibited trigonal symmetry (space group P3(1)21), with unit-cell parameters a = 55.281, b = 55.281, c = 112.449 A and a diffraction limit of 1.97 A. Initial phasing of the trigonal crystals was successfully performed by the SIRAS method using Cu Kalpha radiation with the selenomethionine derivative as a heavy-atom derivative. The structure of the orthorhombic crystal form was solved by molecular replacement using the coordinates of the trigonal form.