Random AA

In vitro transcription/


mRNA-puromycin fusion

mRNA-puromycin fusion

Fig. 35.18. Experimental outline for the selection of ATP-binding proteins from a 1012-member combinatorial protein library.

Combinatorial Protein Library (6 X1012 members)

functional protein from first principles is not yet an achievable goal. As an alternative, researchers have begun producing combinatorial libraries of proteins with varying levels of randomization, and then screening or selecting from these libraries for valued function.

Given the size of typical protein domains, fully randomized protein combinatorial libraries are, for practical purposes, infinitely large. For a 100-amino-acid protein, for example, there are 20100 (10130) possible sequences, and it is thus impossible to examine a substantial portion of this diversity. To evaluate as large a fraction as possible, researchers have turned to in vitro selection methods (discussed in Section 35.3.2) such as mRNA-protein fusions. Using the same mRNA display system outlined in Section 35.3.2, but selecting for ''protein'' behavior instead of''peptide'' behavior, Keefe and Szostak recently identified four ATP binders starting from a library of 6 x 1012 proteins in which 80 contiguous amino acids were randomized (Fig. 35.18) [147]. One of these variants showed very high affinity (KD = 5 nm) and specificity for ATP, and was capable of being further truncated to a protein of only 45 residues. Like most in vitro selection strategies, the ATP-binding proteins were identified based on their ability to elicit binding responses, but it is possible to select proteins on the basis of their catalytic activity [148-150], which may also allow the directed evolution of novel enzymatic activities.

Fully randomized combinatorial protein libraries are of limited practical utility, since the frequency of foldable, much less functional, sequences in them is expected to be very low. As an alternative, partially randomized protein combinatorial libraries can provide the same benefits of randomization but in a sequence that is predicted to fold into a stable structure. For example, the utilization of basic structural information, such as the sequence preferences of helices and sheets, termed

Fig. 35.19. (A) Structure of the AroQ-scaffold chorismate mutase. (B) Using a two-stage selection procedure, active chorismate mutase variants were selected that had a dramatically reduced amino acid palette



51 84 88 95



Fig. 35.19. (A) Structure of the AroQ-scaffold chorismate mutase. (B) Using a two-stage selection procedure, active chorismate mutase variants were selected that had a dramatically reduced amino acid palette

(X = phenylalanine, isoleucine, leucine, or methionine; Z = aspartate, glutamate, asparagine, or lysine) in the helical regions of the protein.

binary patterning [151-153], has been used to design "scaffolded" combinatorial libraries from which soluble proteins were selected [154-156]. To further probe the utility of such libraries, not only for folding structures but also for functional catalysts, we have randomized up to 80% of the total sequence of a six-helix bundle chorismate mutase, using a binary pattern and only eight different amino acids in the randomized regions (Fig. 35.19) [157]. Using in vivo genetic selection for chorismate mutase catalytic activity [158], we identified proteins that upon further analysis had physical properties characteristic of natural enzymes. Our results suggest that such patterned combinatorial protein libraries could be useful for protein design projects.

Analogous to the way that some small-molecule chemotypes make attractive starting scaffolds for combinatorial chemistry libraries, some protein structures have been suggested as being very suitable scaffolds into which any number of possible activities could be built by randomizing or rationally mutagenizing specific regions of the scaffold. In particular, the (a/b)8 barrel, which is perhaps the most commonly occurring protein fold [159], has been suggested as the best starting point from which to produce combinatorial libraries to screen for new functions [160, 161]. One impressive example of the combinatorial redesign of the (a/b)8 barrel is the conversion of one enzyme on the tryptophan biosynthesis pathway, indole-3-glycerol-phosphate synthase (IGPS), into another (phosphoribosylanthranilate iso-merase, PRAI) with very high catalytic activity (Fig. 35.20) [162]. Rational redesign was used to remodel the IGPS structure by elimination of an a-helix and the re-

A Gin Glu PRPP

Was this article helpful?

0 0

Post a comment