Stable and efficient cassette exchange under non-selectable conditions by combined use of two site-specific recombinases

Abstract

Work of the last decade has proven the 'one gene- one product-one function' hypothesis an oversimplification. To further unravel the emerging 'one gene-multiple products-even more functions' concept, new methods (such as subtle knock-in and tightly regulated conditional mutations) for the analysis of gene function in health and disease are required. Another class of improvements (such as tetraploid fusion and cassette exchange) addresses the efficiency with which targeted mutant strains can be generated. Recombinase-mediated cassette exchange (RMCE), which in theory is well suited for the rapid generation of multiple alleles of a given locus, is hampered by its low efficiency in the absence of selection and, especially in vivo, by the promiscuity of the participating recombinase recognition sites. Here we present a novel approach which circumvents this problem by the use of two independent recombinase systems. The strategy, which uses loxP on one and FRT on the other side of the cassette together with a Cre/Flpe expression vector, prevents excisive events and results in higher rates of cassette integration without selection than previously described. This method has a huge potential for the generation of allelic series in embryonic stem cells and, importantly, in pre-implantation embryos in vivo.

Figure 1

(A) The proposed reaction mechanism of Froxing is a two-step process. First, Cre inserts the replacement construct via loxP (green triangles) into the genome, leading to a co-integrate structure with two loxP and two FRT (orange triangles) sites. Second, depending on the enzyme acting on the co-integrate, either the original state is re-established (upon Cre recombination) or, upon Flp recombination, the final exchange product is generated. The plasmid product is lost during successive cell divisions. The potentially re-established original state is immediately available for another exchange round. Once a full recombination round has taken place, it is virtually impossible that the former genomic sequence is re-integrated because the replacement plasmid is introduced in great excess compared to the single genomic copy. (B) Schematic drawing of the plasmids used. White triangles represent loxP sites, black triangles FRT sites. Flp recombinase expression is driven by the strong CAGGS promoter, followed by an internal ribosomal entry site (IRES) and the puromycin resistance gene. Cre expression is under the control of the Herpes simplex thymidine kinase gene promoter. Hygro, hygromycin B phosphotransferase; Neo, neomycin phosphotransferase; EGFP, enhanced green fluorescent protein; PGK, phosphoglycerate kinase promoter; PGKpA, poly(A) of the PGK gene; Sv40pA, Sv40 poly(A).

Figure 2

Analysis of site-specific exchange reactions in ES cells with and without selection. Only locus A is depicted. Primers used for amplification of junction regions are shown as arrows together with the respective primer number. (A) PGK-FroxNeo ‘platform’ transgene. The PCR-amplified junction regions and the corresponding primer pairs are shown on the right. Digestion of genomic DNA with BamHI and probing with a Neo probe yields a 12 kb fragment on the Southern blot. (B) PGK-Frox-Hygro transgene obtained after Frox exchange of Neo by Hygro (selection applied). The characteristics of correctly recombined clones are depicted on the right: presence of Hygro and the corresponding loxP and FRT junctions, absence of Neo. (C) PGK-Frox-EGFP transgene obtained after Frox exchange of Neo by EGFP (no selection applied). The characteristics of correct EGFP integration are shown on the upper right: presence of EGFP and the corresponding loxP and FRT junctions, absence of Neo. The lower right panel shows froxed ES cell clones on embryonic feeder cells after EGFP exchange. One clone (indicated by the arrowhead in the brightfield picture) is clearly positive for EGFP fluorescence.

Figure 3

Flow cytometric quantification of site-specific EGFP integration in ES cells. (Left) Non-transfected PGK-FroxNeo ES cells. (Middle) PGK-FroxNeo cells transfected with pCre/Flp and pFrox-EGFP (one experiment shown as an example). Counts were gated on living cells. The numbers in the diagrams indicate the percentage of GFP-positive cells. Shown on the right are the percentages achieved in several independent experiments with and without a short puromycin treatment.