While the use of enzymes as biocatalysts to assist in the industrial manufacture of fine chemicals and pharmaceuticals has enormous potential, application is frequently limited by evolution-led catalyst traits. The advent of designer biocatalysts, produced by informed selection and mutation through recombinant DNA technology, enables production of process-compatible enzymes. However, to fully realize the potential of designer enzymes in industrial applications, it will be necessary to tailor catalyst properties so that they are optimal not only for a given reaction but also in the context of the industrial process in which the enzyme is applied.The past two decades have led to major advances in our understanding of the subtleties of protein structure-function interrelationships. Mechanisms of protein stability in aqueous and nonaqueous environments 1,2 , the links between conformational mobility, structural integrity and activity 3,4 , and the complexities of substrate specificity have all succumbed to the onslaught of advanced molecular methods, including crystallography 5-7 , site-specific mutagenesis, gene shuffling, and protein evolution [8][9][10][11] . Scientists are now in a position to visualize, if not design, catalytic systems that approach the functional "ideal".The "ideal" catalyst is typically considered by the biochemist in terms of turnover number (k cat ) or, for a given process, in terms of maximum specificity constant (A cat /KM). However, from a bioprocess viewpoint, each bioprocess is constrained by a set of conditions dictated by the specific properties of the substrates, products, and the bioconversion reaction. Thus, while for any given bioprocess it is clearly possible to specify a set of catalyst properties that would constitute the ideal for that process, only broad generalizations for ideality may otherwise be identified.In this review, we discuss molecular properties of enzymes from a bioprocess viewpoint. A paradigm for design based on the ideal process determining the desired features of the catalyst is presented. "Ideal" characteristics (generic and process-specific) and methodologies for seeking or engineering these are discussed. We then review the extent to which significant changes in protein functional properties have been achieved by application of these methods, together with issues that require more research and/or development.
Ideal processesBioprocess engineers develop processes not only against fixed capital and operating expenditure, but also rapid, and often, very aggressive timelines. This is particularly the case for pharmaceutical products where patent expiry on the product demands a robust and scalable process with short development times 12 . Conventionally, once a synthetic route is fixed and the biocatalytic step defined, bioprocess design and operation are centered on the properties of the