Deep sequencing of large library selections allows computational discovery of diverse sets of zinc fingers that bind common targets

Abstract

The Cys2His2 zinc finger (ZF) is the most frequently found sequence-specific DNA-binding domain in eukaryotic proteins. The ZF’s modular protein–DNA interface has also served as a platform for genome engineering applications. Despite decades of intense study, a predictive understanding of the DNA-binding specificities of either natural or engineered ZF domains remains elusive. To help fill this gap, we developed an integrated experimental-computational approach to enrich and recover distinct groups of ZFs that bind common targets. To showcase the power of our approach, we built several large ZF libraries and demonstrated their excellent diversity. As proof of principle, we used one of these ZF libraries to select and recover thousands of ZFs that bind several 3-nt targets of interest. We were then able to computationally cluster these recovered ZFs to reveal several distinct classes of proteins, all recovered from a single selection, to bind the same target. Finally, for each target studied, we confirmed that one or more representative ZFs yield the desired specificity. In sum, the described approach enables comprehensive large-scale selection and characterization of ZF specificities and should be a great aid in furthering our understanding of the ZF domain.

Diversification of Transcription Factor Paralogs via Noncanonical Modularity in C2H2 Zinc Finger DNA Binding

“…For protein engineers, the appeal of this family has been the ability to model and manipulate DNA binding specificity by altering a small set of canonical residues. While it is widely appreciated that residues outside of the canonical recognition positions affect DNA binding, the mechanisms by which these residues alter binding remain unclear (Lam et al, 2011; Persikov et al, 2014; Persikov and Singh, 2011; Ramirez et al, 2008). Here we present two novel mechanisms for altering ZF DNA binding specificity that operate via stabilization of alternate binding modes.…”

“…Second, the ability to switch between different binding modes and bind different DNA sites could lead to off-target binding in a ZF design experiment - one might optimize residues to bind select target sites only to find that an alternate mode permitted binding to undesirable sites. Multiple binding modes further motivates the utility of selection assays and the inability to simply ‘stich-together’ ZF domains of characterized specificity (Beerli et al, 1998; Choo and Klug, 1997; Enuameh et al, 2013; Klug, 2010; Persikov et al, 2014; Wolfe et al, 2000) and suggests that studies aimed at finding ways to inhibit alternate binding modes may minimize off-target binding of synthetic ZF proteins.…”

“…C2H2 ZF proteins (hereafter referred to as “ZF proteins”) bind DNA using arrays of ZF domains, each containing an alpha helix and two beta strands (Klug, 2010) (Figure 1D). Extensive experimental (Beerli et al, 1998; Choo and Klug, 1997; Enuameh et al, 2013; Persikov et al, 2014; Wolfe et al, 2000) and computational (Benos et al, 2002; Kaplan et al, 2005; Mandel-Gutfreund and Margalit, 1998; Persikov and Singh, 2011; Siggers and Honig, 2007) analyses have established a ZF DNA recognition code in which amino acids at 4 canonical ‘recognition’ positions (− 1, 2, 3, and 6, as in (Elrod-Erickson et al, 1996)) in each ZF domain mediate DNA base contacts (Figure 1, Figure S1A). This stereotyped binding mode has made ZFs an object of intense research for the design of artificial TFs and custom ZF nucleases for site-specific genome editing (Klug, 2010).…”

Engineering synthetic TALE and CRISPR/Cas9 transcription factors for regulating gene expression

“…Despite many notable successes with this technology (1, 3, 5, 26), engineering new ZFPs with high activity and specificity remains technically challenging for most researchers. Since predictions of ZFP DNA-binding specificity and affinity are complex (27, 28), it is typically necessary to build and screen many rationally designed proteins or use high-throughput selections to find functional proteins within large libraries (24, 29, 30). Consequently, academic laboratories are adopting the newer TALE and CRISPR/Cas9 platforms that have straightforward DNA-recognition properties (31, 32).…”

Protein Engineering of Cas9 for Enhanced Function

“…Screening systems can effectively cover libraries of sizes up to ~10 6 , which is roughly the amount of E. coli that can conveniently be sorted 10× by flow cytometer in an hour. On the other hand, selection systems, which rely on repression/activation of a toxic/essential gene for growth, can screen libraries of random protein variants of up to ~10 9 in size (Persikov, Rowland, Oakes, Singh, & Noyes, 2014). …”