Chemistry Reference
In-Depth Information
mRNA-protein fusions [75] are only some of the techniques devised. In coupling gen-
otype and phenotype they enable in vitro selection of proteins with interesting binding
or catalytic properties from vast combinatorial libraries. Genetic selection in vivo re-
presents yet another powerful tool for enzyme (re)design that effectively harnesses the
power of evolution [76].
Recent advances in computation, combined with the methods of directed evolution,
should make future enzyme engineering much easier. Conversion of a ribose-binding
protein into a triose phosphate isomerase mimic represents an exciting first step [77].
If the results of this study can be generalized, computation could become an effective
alternative to the immune system for efficiently exploring sequence space. Such an
approach would not be limited to a single scaffold, but binding pockets could be
matched to the requirements of the reaction being catalyzed. It may even be possible
to design new protein folds with tailored active sites de novo. Since reaction mechan-
ism and transition state properties can be considered explicitly during the computa-
tional design process, the chances of obtaining a good experimental starting point for
further development are likely to be considerably improved. As illustrated by the arti-
ficial triose phosphate isomerase example [77], random mutagenesis coupled with
high-throughput screening or mutagenesis allows subsequent fine-tuning.
4.9
Outlook
Antibody catalysts are arguably the most successful enzyme mimics so far described.
They can be prepared by a general strategy that takes advantage of our knowledge of
reaction mechanism and physical organic chemistry. Although the rates achieved as
yet are not “enzyme-like“, enzymes - with millions of years of evolutionary optimiza-
tion - may set an unrealistically high standard to judge synthetic catalysts that can be
evolved in the immune system over a few weeks or months. This approach has already
provided valuable model systems for exploring how protein binding energy can be
exploited to accelerate numerous chemical transformations, including those lacking
biological counterparts. Even if catalytic antibodies never achieve the broad applicabil-
ity originally anticipated, the strategies employed in their productions and the lessons
learned can be expected to contribute significantly to the success of future efforts to
create tailored proteins with truly enzyme-like properties.
Acknowledgments
The author is indebted to the ETH Z¨rich, the Swiss National Science Foundation and
Novartis Pharma for generous support.
