Protein Design

Jean-LucJestin and Frederic Pecorari 9.1

Introduction

Most natural proteins are not adapted to our needs for therapeutic or even bio-technological uses. The challenge of protein design is to define efficient and comprehensive ways for the identification of new or improved proteins with the required properties. The design of new proteins relies on sequence-structure-function relationships and is limited by our partial knowledge of the protein folding mechanism. But as we will see in this chapter, creating new proteins or proteins with desired properties is clearly possible.

A striking characteristic of natural proteins is that they are able to adopt a folded state despite their very high sequence diversity. About one thousand different folds are known. Natural proteins constitute the richest family of folded molecules. When targeting specific functions and applications, non-natural folded proteins are of obvious interest. They can be created by protein engineering thanks to their plasticity using diverse strategies including mutagenesis (random mutagenesis, recombination or site-directed mutagenesis making use of structures and models), domain fusions and sequence-activity relationships deriving from natural and directed evolution.

This widely applied strategy aims to explore the potential benefit of natural proteins. There is an advantage in knowing the protein's three-dimensional structure, as modifications can be introduced rationally into the protein while trying not to destroy its foldability. This predictive approach can be used to change li-gand or reaction specificities, or to increase stability, for example. New proteins can be also designed by combining several natural proteins or their functional domains in one polypeptide chain. Although many successes have been reported with rational approaches to protein design, it remains very difficult in many cases to make drastic changes to a protein. Thus, with the development of powerful molecular biology tools, the combinatorial approach has progressed remarkably during the last 15 years. The principle is to introduce mutations in the gene of interest and create large collections of variants of the encoded protein. The result-

Table 9.1 Comparison of the properties of selection versus screening.

SELECTION in vivo or in vitro SCREENING

1 Analysis in parallel in series

2 Number of variants analyzed 10 6 to 1013 103 to 108

3 Substrate quantities required mg to mg g

4 Isolated variants selected population one or a few ing library is then screened or selected to identify clones with the desired property. Several rounds of random mutagenesis and of screening or selection can be conducted in order to obtain variants with better properties (Table 9.1). Such processes represent an accelerated mimicry of the natural evolution of proteins, since there is a generator of variability associated to a selection pressure (Fig. 9.1). These approaches make general use of the link between the genotype (the gene which can be amplified and sequenced) and the phenotype (the protein's function) (Fig. 9.2). The most efficient way to achieve the expected result is to combine the predictive and combinatorial approaches in order to target randomizations in only a part of the protein, for example the active or binding site. The idea is to minimize perturbation of the folding and stability of the protein. It has therefore been possible to obtain new proteins with desired properties and mutations that would have been impossible to predict.

Directed protein evolution

Amplification Transcription Translation

Amplification Transcription Translation

Protein library 10« to 10"

Selected proteins

Protein library 10« to 10"

Selected proteins

Screening or selection

Screening or selection

Sequencing of corresponding genes Biochemical characterisation

Characterised proteins

Fig. 9.1 The evolution of protein libraries of more than 106 proteins can be directed experimentally by screening or by selection. To enrich the population of variants in proteins of interest, several cycles of selection of proteins and of amplification of their genes are necessary. An evaluation of the number of cycles required can be estimated by measuring enrichment factors [80]. Amplifications may also be mutagenic, thereby generating new libraries from a selected population of proteins of interest.

0 0

Post a comment