Comparing zinc finger nucleases and transcription activator-like effector nucleases for gene targeting in Drosophila

Abstract

Zinc-finger nucleases have proven to be successful as reagents for targeted genome manipulation in Drosophila melanogaster and many other organisms. Their utility has been limited, however, by the significant failure rate of new designs, reflecting the complexity of DNA recognition by zinc fingers. Transcription activator-like effector (TALE) DNA-binding domains depend on a simple, one-module-to-one-base-pair recognition code, and they have been very productively incorporated into nucleases (TALENs) for genome engineering. In this report we describe the design of TALENs for a number of different genes in Drosophila, and we explore several parameters of TALEN design. The rate of success with TALENs was substantially greater than for zinc-finger nucleases , and the frequency of mutagenesis was comparable. Knockout mutations were isolated in several genes in which such alleles were not previously available. TALENs are an effective tool for targeted genome manipulation in Drosophila.

Figure 1

TALEN targets in the Drosophila ry and y genes. Each gene is diagrammed approximately to scale, with rectangles denoting exons and coding sequences as shaded rectangles. The locations of the TALEN targets are shown with black vertical lines, and the corresponding sequences are illustrated. TALEN binding sites are in capital letters, spacers in lower case. The ZFN binding sites that overlap the ryT2, ryT3, and yT2 sites are underlined.

Figure 2

The most common single mutations found at overlapping ZFN and TALEN targets in ry exon 3 (ryAB ZFNs, ryT3 TALENs). Gray rectangles denote the binding sites for the DNA-binding modules; the spacer sequences are written out. The most common ZFN product is an apparent fill-in and blunt join of the 4-nt 5′ overhang created by cleavage: the duplicated 4 bp are underlined and in bold. The most common TALEN product is a 7-bp deletion supported by a 2-bp microhomology (underlined).

Figure 3

Distribution of deletion sizes for ZFNs and TALENs. The median deletion size was 2 bp for ZFNs and 8 bp for TALENs. The data for ZFNs reflect results for the y and ry targets (Bibikova et al. 2002; Beumer et al. 2006, 2008). The TALEN data include all sites presented in this study, with more examples from the y and ry targets than from others.

Figure 4

Sequences of the junctions of the 2.5-kb deletions created by coinjection of mRNAs for the ryT1 and ryT3 TALENs. The target sequences for the individual TALEN pairs are shown at the top; the positions of these sequences relative to the start of ry transcription are given. The sequences of the deletions, which have been given arbitrary numerical designations, are aligned to them below. As in Figure 1, TALEN binding sites are shown in capital letters, spacers in lower case. An apparent single-base substitution in deletion 106A is underlined.

Figure 5

Histograms of the percent of mutant progeny scoring positive for homologous recombination achieved with oligonucleotide donors in conjunction with the ryT3 TALENs. Data are from experiments in a wild type strain (Canton S) and in a strain that lacks DNA ligase IV (lig4⁻).