Development of Zinc Finger Domains for Recognition of the 5′-CNN-3′ Family DNA Sequences and Their Use in the Construction of Artificial Transcription Factors

Birgit Dreier,Roberta P Fuller,David J Segal,Caren V Lund,Pilar Blancafort,Adrian Huber,Beate Koksch,Carlos F Barbas

doi:10.1074/jbc.m506654200

Abstract

Considerable progress has been made in recent years in the design of transcription factors for the directed regulation of endogenous genes. Although many strategies involve selection methods that must be applied for each new target sequence, we have developed an approach based on linkage of predefined zinc finger domains that each recognize a three-base pair DNA sequence to construct artificial transcription factors that bind to a desired sequence. These domains can be assembled to recognize unique 18-base pair DNA sequences with high specificity. Here we report the development and characterization of zinc finger domains that bind to 15 of the 16 5'-CNN-3' subsites. These domains were created through a combination of phage display selection, site-directed mutagenesis, and de novo design. Furthermore, these domains were used to generate a highly specific six-finger protein targeting the ERBB-2 promoter. When fused to regulatory domains, this protein was capable of up- and down-regulating the expression of the endogenous ERBB-2 gene. With the addition of this collection of predefined zinc finger domains, most 5'-CNN-3'-, 5'-GNN-3'-, and 5'-ANN-3'-containing sequences can now be rapidly targeted for directed gene regulation and nuclease cleavage.

Highlights

Summary of the most optimal zinc finger domain sequences identified in this study for binding to 5؅-CNN-3؅ subsites
The helices that bound to the 5Ј-CNN-3Ј target sites are listed
The zinc finger domains described here that recognize 5Ј-CNN-3Ј DNA subsites are suitable for the rapid construction of artificial transcription factors. These 5Ј-CNN-3Ј zinc finger domains augment the 5Ј-GNN-3Ј, 14 5Ј-ANN-3Ј, and 2 5Ј-TNN-3Јdomains we have previously developed and together provide for the rapid construction of more than 10 billion proteins that bind 5Ј-[(G/A/C)NN]63Ј sites

Summary

MATERIALS AND METHODS

Construction of Zinc Finger Library and Selection via Phage Display—Construction of the zinc finger library was based on the earlier described C7 protein (29). Multitarget Specificity Enzyme-linked Immunosorbant Assay (ELISA)—The zinc finger-coding sequence was subcloned from pComb3H (30, 31) into a modified bacterial expression vector pMal-c2 (New England Biolabs). Freeze/thaw extracts of these bacterial cultures or purified proteins were applied in 1:2 dilutions to 96-well plates coated with streptavidin (Pierce) and were tested for DNA binding specificity against each of the 16 5Ј-GAT CNN GCG-3Ј target sites using the hairpin oligonucleotide described above. Gel Mobility Shift and DNase I Footprint Analysis—The coding sequence of pE2S was subcloned into a modified pMAL-c2 (New England Biolabs) bacterial expression vector and transformed into Escherichia coli strain XL1-Blue (Stratagene). Purified protein was used to perform DNase I footprints and gel mobility shift assays to determine the DNA binding site and affinity. Hydrogen bonds were considered plausible when the distance between the heavy atoms was 3 (Ϯ 0.3) Å and the angle formed by the heavy atoms and hydrogen was 1200 or greater

RESULTS

DISCUSSION

TABLE ONE