Modification of maltose binding protein for generation of nanomaterials and for improving crystallization

Protein crystallization can be considered as a self-assembly process, since protein molecules in solution form a well-ordered lattice in the crystal. Self-assembly processes are a common route to generating nanomaterials. Nanomaterials can be generated from protein crystals by crosslinking the adjacent molecules in the protein crystal. By engineering cysteine mutations on the surface of the protein of interest, such that crystallization brings the cysteine residues close enough to form disulfide bonds, adjacent molecules in the crystal can be covalently crosslinked. Based on the position and number of crosslinks, different nanoarchitectures such as protein fibers and sheets can be generated. This strategy was applied to maltose binding protein (MBP) and crosslinked crystals of MBP were obtained. Fibers of MBP were obtained on dissociation of the crystals.; The second part of this work deals with improving protein crystallization. Iyer et al. have calculated potentials for interactions of residues at the interfaces in protein crystals, based on a statistical analysis of 223 protein crystal structures. These potentials (non pairwise) indicate that lysines and hydrophobic residues tend to be excluded from crystal contacts, while arginines and glutamines are favored. So replacement of lysines, and hydrophobic residues on the protein surface with more favorable residues should be a viable strategy for improving crystallization. The potentials can also be used to determine unfavorable residue interactions in a known crystal structure (pairwise potentials). Mutations based on both pairwise and non pairwise potentials were made in a model protein, MBP. The effects of the mutations on the crystallization properties of MBP were studied by sparse matrix screening of mutants to see if mutants provided an increased number of hits. In one template, all K to R mutants indicated slight improvement in crystallizability, while mutation of a residue predicted to be favorable (R354S) drastically reduced the crystallizability of the protein. Crystal quality of mutants and control were also compared by X-ray diffraction analysis. The conclusions from these studies indicate that the potentials calculated do have some predictive value for generating mutants with improved crystallizability.