The consensus concept for thermostability engineering of proteins

Previously, sequence comparisons between a mesophilic enzyme and a more thermostable homologue were shown to be a feasible approach to successfully predict thermostabilizing amino acid substitutions. The 'consensus approach' described in the present paper shows that even a set of amino acid sequences of homologous, mesophilic enzymes contains sufficient information to allow rapid design of a thermostabilized, fully functional variant of this family of enzymes. A sequence alignment of homologous fungal phytases was used to calculate a consensus phytase amino acid sequence. Upon construction of the synthetic gene, recombinant expression and purification, the first phytase obtained, termed consensus phytase-1, displayed an unfolding temperature (T-m) of 78.0 degreesC which is 15-22 degreesC higher than the T-m values of all parent phytases used in its design. Refinement of the approach, combined with site-directed mutagenesis experiments, yielded optimized consensus phytases with T-m values of up to 90.4 degreesC. These increases in T-m are due to the combination of multiple amino acid exchanges which are distributed over the entire sequence of the protein and mainly affect surface-exposed residues; each individual substitution has a rather small thermostabilizing effect only. Remarkably, in spite of the pronounced increase in thermostability, catalytic activity at 37 degreesC is not compromised. Thus, the design of consensus proteins is a potentially powerful and novel alternative to directed evolution and to a series of rational approaches for thermostability engineering of enzymes and other proteins. (C) 2000 Elsevier Science B.V. All rights reserved.