Development of synthetic selfish elements based on modular nucleases in Drosophila melanogaster

Abstract

Selfish genes are DNA elements that increase their rate of genetic transmission at the expense of other genes in the genome and can therefore quickly spread within a population. It has been suggested that selfish elements could be exploited to modify the genome of entire populations for medical and ecological applications. Here we report that transcription activator-like effector nuclease (TALEN) and zinc finger nuclease (ZFN) can be engineered into site-specific synthetic selfish elements (SSEs) and demonstrate their transmission of up to 70% in the Drosophila germline. We show here that SSEs can spread via DNA break-induced homologous recombination, a process known as 'homing' similar to that observed for homing endonuclease genes (HEGs), despite their fundamentally different modes of DNA binding and cleavage. We observed that TALEN and ZFN have a reduced capability of secondary homing compared to HEG as their repetitive structure had a negative effect on their genetic stability. The modular architecture of ZFNs and TALENs allows for the rapid design of novel SSEs against specific genomic sequences making them potentially suitable for the genetic engineering of wild-type populations of animals and plants, in applications such as gene replacement or population suppression of pest species.

Figure 1.

Model of HEG-mediated homing and of ZFNs and TALENs. (A) Mode of action of HEGs. When a HEG is expressed in a trans-heterozygous cell (top), it recognizes its target site on the homologous chromosome generating a DSB (middle). The broken chromosome activates the repair machinery of the cell that can employ the non-homologous end-joining pathway (bottom left) or homologous recombination (bottom right), using the HEG-containing chromosome as a repair template. A HEG heterozygous is converted into a HEG homozygous, a process called ‘homing’. (B) Schematic of ZFNs and (C) TALENs. They both act as a dimer in which each monomer binds on complementary DNA strands (binding sequences are underlined and in capital letter) and the FoKI nuclease directs its activity to the spacer between the recognition sequences generating a 5’-overhang. Each ZFN module recognizes a DNA triplet and four fingers are linked together to selectively recognize a 12 bp target sequence. Individual TALEN modules bind to single nucleotide and 17 modules are assembled together to form a functional nuclease.

Figure 2.

Schematic representation of the genetic markers used to follow the structure of the progeny. (A) Donor and target constructs were inserted by ΦC31site-specific integration in the same attP2 docking line (on chromosome 3L, position 11 063 638 bp). Donor nuclease sequences are inserted as cassette (top) within their corresponding target site interrupting the green fluorescent protein (GFP) coding sequence. Three different SSEs were constructed encompassing the following elements: the male germline promoter Rcd-1r (white triangle), a TALEN or a ZFN-pair nuclease (blue shape) and the ß56D-tubulin 3’-UTR (black bar). The TALELAT construct carries a RFP marker gene driven by the eye-specific promoter 3xP3 (black triangle). The left and the right ZFNs are separated by a Furin-2A self-cleavage ribosomal stuttering peptide (yellow). The donor constructs are adjacent to a functional mini-white gene (orange box) that restores the red pigmentation in the fly's eyes as phenotypical marker and the recessive marker curled (cu), on chromosome 3R (position 7 023 314). The recipient (target) chromosome carries the nuclease target sequence (as shown. The nuclease recognition sequences are underlined and in capital letters) in-frame with a functional GFP gene (green box), driven by the eye-specific promoter 3xP3. The mini-white marker in the target line was inactivated by a frame-shift mutation (marked by an arrow head). (B) Schematic of the homing assay. Donor/target trans-heterozygous flies are crossed to attP2 cu flies (the genetic background). When the nucleases are expressed in the germline, the target site is cleaved and the chromosome repair mechanisms can lead to different progeny outcomes, which can be discriminated by fluorescent and phenotypical markers, as indicated. Donor and target chromosome (marked in grey and white, respectively) can be discriminated by the presence of dominant mini-white (w) and recessive curled (cu) markers. We defined ‘homing’ as any recombination event which leads to the conversion of the target chromosome into a nuclease expressing chromosome. (C) PCR reactions were performed on w- GFP-negative flies with different sets of primers to distinguish HR-dependent repair events from imprecise NHEJ. The combination P1–P2 generates a PCR amplicon only in the case of homing of the locus while the combination P1–P3 is diagnostic for NHEJ events. Lanes 1–10 were loaded with PCR reactions generated from GFP-negative flies derived from TH males crossed to wt females.

Figure 3.

Phenotypical analysis of SSEs activity in vivo. (A) Phenotypic analysis of progeny originating from crosses of donor/target trans-heterozygous (TH) and wild-type flies according to the markers described in Figure 2. ‘Male’ and ‘female’ denote the gender of the trans-heterozygous parent. The bars indicate the fraction of offspring carrying the donor chromosome (red; this is directly inherited from parent to offspring and is not the result of nuclease activity), homed target chromosome (striped red; fraction of homing), unmodified target (green) and out-of-frame NHEJ (grey). The inset on the right shows the phenotypic analysis of progeny from target-chromosome only (w-) as reported in Table 1. Error bars indicate SEM between independent crosses. (B) Phenotypic analysis of GFP expression in F2 progeny from F1 homed flies individually crossed to fresh target flies. Each dot indicates the GFP fraction of an individual cross outcome. The crosses in which the GFP-positive fraction statistically differs from 0.5 (Chi-square, P < 0.05) are shown in purple.

Figure 4.

Molecular analysis of F2 dysfunctional homing products. PCR analysis from homed flies that carry dysfunctional TALEN SSE product. (A) Schematic of the TALELAT donor construct and the expected size of PCR amplicons (grey areas). (B) Primers c–d are expected to generate an amplicon encompassing the TALEN sequence of 3264 bp in size (marked with a star at the right as in the positive control, +). Different sizes of PCR products or failure to amplify were observed from genomic DNA extracted from all non-functional F2 donor-expressing flies. Note the ‘laddering’ effect in the positive control as a PCR artefact of TALEN repeats amplification. (C) PCR from all homed flies (except lane 9) gives a positive product of the expected size with primers a–b. (D) Primers c–d generate the expected PCR product from TALELAT donor stock flies (3264 bp, star). L: Hyperladder I. + and – indicate positive and negative PCR controls, respectively.

Figure 5.

Characterization of NHEJ events originating from TALEN and ZFN activity. Sequencing characterization of imprecise NHEJ events originating from TALEN and ZFN activity, as indicated. The first line shows the GFP coding sequence that includes the nuclease target site (the nucleases binding sequences are underlined). The majority of repair events following ZFN cleavage leave microdeletions in proximity of the cleavage site whereas in the case of TALEN, the repaired chromosome exhibits bigger deletion (up to 300 bp) mainly at the 3’ of the cleavage site. In few cases, partial HR resulted in segmented of donor cassette being inserted in the target site, from either side of the DSB (RFP or Rcd-1r sequence). Insertions are highlighted. The numbers of identical repair events are indicated in squared brackets on the right.

Figure 6.

Population invasion experiments of ZFN and TALEN SSEs. Fraction of GFP-negative alleles in populations of flies monitored for up to 15 generations after a single initial release of 20–25% of SSE trans-heterozygous males in a GFP target population. Each data set refers to individual populations of 400 and 100 flies each for (A) ZFN-AAVS1 and (B) TALEN, respectively. The experimental points from (A) five and (B) six populations (black lines) were compared to 30 independent iterations of a stochastic model (grey line) based on the experimental data of cleavage and homing reported in Table 1. The average of the 30 stochastic simulations is also plotted (dotted blue lines). As a control trans-heterozygous males were introduced into a population of 400 refractory (non-cleavable) target flies at an initial frequency of 25% (black dotted line).