PiggyBac transposon-based polyadenylation-signal trap for genome-wide mutagenesis in mice

Abstract

We designed a new type of polyadenylation-signal (PAS) trap vector system in living mice, the piggyBac (PB) (PAS-trapping (EGFP)) gene trapping vector, which takes advantage of the efficient transposition ability of PB and efficient gene trap and insertional mutagenesis of PAS-trapping. The reporter gene of PB(PAS-trapping (EGFP)) is an EGFP gene with its own promoter, but lacking a poly(A) signal. Transgenic mouse lines carrying PB(PAS-trapping (EGFP)) and protamine 1 (Prm1) promoter-driven PB transposase transgenes (Prm1-PBase) were generated by microinjection. Male mice doubly positive for PB(PAS-trapping (EGFP)) and Prm1-PBase were crossed with WT females, generating offspring with various insertion mutations. We found that 44.8% (26/58) of pups were transposon-positive progenies. New transposon integrations comprised 26.9% (7/26) of the transposon-positive progenies. We found that 100% (5/5) of the EGFP fluorescence-positive mice had new trap insertions mediated by a PB transposon in transcriptional units. The direction of the EGFP gene in the vector was consistent with the direction of the endogenous gene reading frame. Furthermore, mice that were EGFP-PCR positive, but EGFP fluorescent negative, did not show successful gene trapping. Thus, the novel PB(PAS-trapping (EGFP)) system is an efficient genome-wide gene-trap mutagenesis in mice.

Figure 1. Schematic diagram of PB(PAS-trapping (EGFP)) gene trapping vector and transposase constructs of the piggyBac Transposition System for Mice.

(A) PB donor constructs, named as PB(PAS-trapping (EGFP)). The reporter gene of PB(PAS-trapping (EGFP)) is an EGFP gene driven by the human CMV promoter. An unstable mRNA signal sequence (ARE) to the 3′ end of EGFP was added. The IRES, three ICs and SD were added between EGFP and ARE. There is the option to remove the IRES using Cre, thereby removing the truncated protein (but not GFP). A SA was added in front of EGFP reporter gene. The PAS-trapping (EGFP) cassette was placed between a pair of PB terminal domains (PBL and PBR, black arrows). (B) PB transposase helper constructs. CMV-PBase was constructed by inserting CMV promoters in front of PBase gene, followed by BGH pA. The piggyBac transposase gene (PBase) driven by Prm1 promoters (Prm1-PBase) were followed by SV40 late poly (A) signal. (SA: splicing acceptor; CMV promoter: cytomegalovirus immediate early promoter; TC: termination codon; IRES: internal ribosome entry site; IC: initial codon; SD: splicing donor; ARE: an RNA instability element; PBL: PB repeat left termini; PBR: PB repeat right termini; pA: poly (A); BGH pA: bovine growth hormone poly (A).; Prm1: Protamine 1).

Figure 2. Illustration depicting PB(PAS-trapping (EGFP)) gene trapping strategies and its mechanism of action.

(A) The lack of a dedicated poly (A) signal for the EGFP leads to no EGFP expression after transposition outside a gene. (B) Schematic diagram of gene trap and insertional mutation by PB(PAS-trapping (EGFP)) gene trap vector when transposition into an exon of a gene with the same direction of the endogenous gene reading frame. (C) Schematic diagram of gene trap and insertional mutation by PB(PAS-trapping (EGFP)) gene trap vector when transposition into an intron of a gene with the same direction of the endogenous gene reading frame.

Figure 3. Overview of the breeding scheme used to generate mutant mice in the mouse germline.

GFP was used as a marker to monitor transposition events. (A) Obtaining the male mice doubly positive for PB(PAS-trapping (EGFP)) and Prm1-PBase. Transgenic mouse lines carrying PB(PAS-trapping (EGFP)) and protamine 1 (Prm1) promoter-driven PB transposase transgenes (Prm1-PBase) were generated by microinjection. The male mice doubly positive for PB(PAS-trapping (EGFP)) and Prm1-PBase were obtained by mating the PB(PAS-trapping (EGFP)) and Prm1-PBase transgenic lines. (B) Generating mutant mice in the mouse germline. The mice doubly positive for PB(PAS-trapping (EGFP)) and Prm1-PBase were crossed with WT females to generating offspring with various insertion mutations.

Figure 4. Screening for mutant mice.

(A,B) Screening for mutant mice performed by GFP expression. Newborn mice were examined by fluorescence microscopy. (C,D,E) Molecular analysis of flanking genomic DNA sequences at sites of new transposon insertion. (C) Schematic diagram of reverse PCR for identification of insertion sites, after digested with restriction enzyme Sau3A I. (D) The PCR products were amplified from ligation products of the SauAI-digested genomic DNA with T4 ligase. (E) Clone selection of the positive clones, who had the PCR product (obtained in Fig. 4D) by PCR. After electrophoresis of the PCR products in 1.2% agarose gels, the desired nucleic acid fragments were purified and imported into pMD18-T Simple vector.

Figure 5. Five independent insertion sites were mapped in EGFP-expressing mice.

(A) Dedd (inserted in 1th intron) was inserted by PB (PAS-trapping (EGFP)) trap vector. (B) Dgkb (inserted in 19th intron) was inserted by PB (PAS-trapping (EGFP)) trap vector. (C) Gm9733 (inserted in 1th intron) was inserted by PB (PAS-trapping (EGFP)) trap vector. (D) Ift80 (inserted in 11th intron) was inserted by PB (PAS-trapping (EGFP)) trap vector. (E) Fam96a (inserted in 2th intron) was inserted by PB (PAS-trapping (EGFP)) trap vector. Dgkb: Diacylglycerol kinase beta, Dedd :death effector domain-containing, Ift80:intraflagellar transport 80 homolog (Chlamydomonas), Gm9733 and Fam96a are putative genes.