Exploring the functional potential of de novo proteins

Abstract

Nature has provided many exquisite functional proteins, each built from a different sequence of the twenty naturally-occurring amino acids. The natural proteins we observe today are the result of the sampling and shaping of a small subset of possible sequences. Scientists have the ability to create completely novel proteins, ones which do not exist in nature and are not homologous to naturally occurring proteins, from the same set of amino acid building blocks. But will they behave as natural proteins do? To investigate this, our lab has created several libraries of de novo proteins designed to fold into four-helix bundles. Proteins from these libraries have been shown to bind small molecules, catalyze reactions at low-levels, and rescue E. coli with conditionally lethal gene deletions.

The research described in this thesis explores the interplay between de novo proteins and living systems. Specifically, the effects of de novo proteins expressed by E. coli in lethal situations, imposed by either internal factors (deletion of fes, a critical gene, Chapter 3) or external factors (exposure to toxins, Chapter 4), was investigated. This work led to two important discoveries: a novel enzyme (Syn-F4) that could catalyze the stereospecific hydrolysis of ferric enterobactin, as well as a protein (MlnH7) that could increase the resistance of E. coli cells to certain antibiotics.

In addition, the role of evolution in shaping the specificity and efficiency of de novo protein function is detailed in Chapters 2 and 3. This work began with a bifunctional protein (Syn-IF), capable of rescuing both Δfes and ΔilvA cells. Over several rounds of directed evolution towards faster rescue of either Δfes or ΔilvA cells, two monofunctional specialist proteins (Syn-F4 and Syn-I3) resulted, each now rescuing only the strain in which they were evolved.