doi:10.1016/S0076-6879(06)20007-7
Copyright © 2006 Elsevier Inc. All rights reserved.
Engineering Embryonic Stem Cells with Recombinase Systems
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Frank Schnütgen, A. Francis Stewart, Harald von Melchner and Konstantinos Anastassiadis
Available online 8 December 2006.
Abstract
The combined use of site-specific recombination and gene targeting or trapping in embryonic stem cells (ESCs) has resulted in the emergence of technologies that enable the induction of mouse mutations in a prespecified temporal and spatially restricted manner. Their large-scale implementation by several international mouse mutagenesis programs will lead to the assembly of a library of ES cell lines harboring conditional mutations in every single gene of the mouse genome. In anticipation of this unprecedented resource, this chapter will focus on site-specific recombination strategies and issues pertinent to ESCs and mice. The upcoming ESC resource and the increasing sophistication of site-specific recombination technologies will greatly assist the functional annotation of the human genome and the animal modeling of human disease.
 |
 |
Fig. 1. Characteristics of Cre and FLP site-specific recombination substrates. (A) Cre or FLP recombinations targets (RTs) known as loxP or FRT, respectively, are depicted along with several useful variations. The 13-bp inverted binding sites are illustrated by inverted arrows flanking the spacer sequences, which contain the region of recombination. The recombination region is denoted by arrowheads above and below the sequence at each end of the spacer. For the variant RTs, only the sequence differences are shown. (B) Excision recombination between two directly repeated RTs. The excised DNA is released as a covalent circle. Reintegration is possible but disfavored. (C) Inversions occur when RTs are arranged as indirect repeats. (D) Recombination between RTs on different molecules results in a reciprocal translocation. The orientation of the translocated fragments is dictated by the orientation of the RTs. White pentagons indicate RTs and their orientation. (E) Idealized kinetics for excision, inversion, and translocation. Note that all three reactions approach 50% with equivalent kinetics. Thereafter, excisions approach 100%, whereas inversions and translocations equilibrate at 50%. See Logie and Stewart, 1995, for supporting data. (F) Site-specific recombination frequency as a product of the distance between two RTs, as determined for naked DNA (Ringrose et al., 1999). Note the maximum recombination efficiency at 400 bps.
 |
Fig. 2. Directional recombination using Cre and FLP. (A) Inversion between mutant lox66 and lox71 generates a wild-type loxP site and a double mutant lox66/lox71 site that cannot recombine. (B) Recombinase-mediated cassette exchange (RMCE). Intermolecular recombination occurs between two different (heterotypic) RTs, here shown as wild-type (white pentagons) and mutant (black pentagons). Intramolecular recombination between heterotypic RTs is not possible. The RTs flank the exchange cassettes on a linear recipient molecule (in gray) and on a circular donor molecule (in black). The insertion of the donor cassette is followed by the excision of the recipient's casette (gray) in 50% of cases. (C) Flip-excision (FLEx). Inversion of the RT flanked fragment occurs either at wild-type (white pentagons) or mutant (black pentagons) RTs. After either inversion, pairs of homotypic RTs in direct orientation flank a heterotypic RT. Excision between the directly repeated homotypic RTs excises the heterotypic RT, thus locking the recombination product. The final product is flanked by heterotypic RTs that cannot recombine.
Fig. 3. Products of ΦC31 recombination. The ΦC31-specific RTs, attP and attB are shown in excision (A) or inversion (B) orientations. After recombination, the RTs are changed to attL and attR and cannot be recombined by ΦC31 in the reverse reaction.
Fig. 4. A common strategy for creating conditional alleles in the mouse. In the targeting vector, an exon (here exon 2) of the target gene is flanked by loxP sites in direct (excision) orientation (white pentagons). The targeting vector contains a selection cassette (here neomycin; neo), which is flanked by FRT sites in direct orientation (white polygons). After introducing the vector into ESCs and selecting for homologous recombinants, the neomycin cassette is removed using FLPe, either in ESCs, mice, or oocytes. Homozygous mice for the conditional allele are crossed to mice expressing Cre in a spatially and/or temporally restricted manner. This deletes the loxP-flanked exon 2 from the target gene and causes a mutation.
 |
Fig. 5. Conditional gene inactivation by gene trap vectors. (A) Single cassette strategy. A flexed SAßgeopA gene trap cassette is illustrated after integration into an intron of an expressed gene. White polygons, FRT sites; gray polygons, F3 sites; white pentagons, loxP sites; black pentagons, lox5171 sites; ßgeo, ß-galactosidase/neomycin phosphotransferase fusion gene. Transcripts (shown as gray arrows) initiated at the endogenous promoter are spliced from the splice donor (SD) of an endogenous exon (here exon 1) to the splice acceptor (SA) of the SAßgeopA cassette. Thereby, the ßgeo reporter gene is expressed, and the endogenous transcript is captured and prematurely terminated at the cassette's polyadenylation sequence (pA), causing a mutation. FLPe inverts the SAßgeopA cassette onto the anti-sense, noncoding strand at either FRT (shown) or F3 (not shown) RTs and positions FRT and F3 sites between direct repeats of F3 and FRT RTs, respectively. By simultaneously excising the heterotypic RTs, the cassette is locked against reinversion, because the remaining FRT and F3 RTs cannot recombine. This reactivates normal splicing between the endogenous splice sites, thereby repairing the mutation. Cre-mediated inversion repositions the SAßgeopA cassette back onto the sense coding strand and reinduces the mutation (Schnütgen et al., 2005). (B) Double cassette strategy. Two gene trap cassettes in opposite orientation relative to each other are illustrated after integration into an intron of an expressed gene. Neo, neomycin phosphotransferase; ßgal, ß-galactosidase. The sense cassette (SAneopA) is flanked by wild-type FRT sites in direct orientation and is amenable to excision by FLPe. The antisense cassette (SAßgalpA) is flexed by heterotypic lox sites, hence invertable by Cre. Transcripts initiated at the endogenous promoter are spliced from the splice donor of the endogenous exon to the splice acceptor site of the SAneopA cassette, causing a mutation (see earlier). FLPe repairs the mutation by excising the SAneopA cassette and reactivating normal splicing between the endogenous splice sites. Cre-mediated inversion places the flexed SAßgalpA cassette onto the sense coding strand and reinduces the mutation.
 |
Fig. 6. Gene expression switch. (A) A selectable marker (5′marker-pA), flanked by loxP sites (white pentagons) in direct orientations, is expressed from a constitutively active promoter (P) and simultaneously blocks the expression of a downstream gene (3′marker-pA) by premature polyadenylation (pA). Cre deletes the upstream cassette, which activates the downstream gene. (B) Use of the gene expression switch in a genetic screen for inducible genes. A switch cassette is integrated into ESCs by selection for the upstream selection marker, which is flanked by loxP sites (white pentagons). The downstream marker is not expressed because of premature polyadenylation (pA). These cells are transduced with a Cre gene trap vector. Vector integrations into active genes express Cre so that these are eliminated by continued selection for the upstream marker. Hence, surviving cells do not express Cre, which must be integrated into silent genomic sites (1. OFF). Then, cells are treated with an inducer, which may be a cytokine, a hormone, or a differentiation protocol. If this change activates a gene trap insertion site to express Cre, recombination ensues, and selection applied for the activation of the downstream marker can be used to isolate these cells (2. ON). Regardless of whether Cre expression stays on or is subsequently turned off, the expression of the downstream marker continues (3. ON or OFF), hence facilitating the identification of transient sites of activation of Cre expression.
Fig. 7. Recombination in ESCs. (A) The expression vector pCAGGs-Flpe-IRES-puro-pA that is used for transfection in ESCs. (B) Schematic presentation of an imaginary gene in all three possible conformations: wt-allele, targeted allele, and recombined allele. Depicted are only exons 1 and 2 (gray boxes). The selection marker (S.M.) flanked by FRTsites (white pentagons) has been integrated in the first intron. Vertical arrows show positions of the hypothetical restriction site (x). Horizontal arrows represent the PCR primers used for testing recombination efficiency. Dashed horizontal lines that connect the restriction sites (x) represent the different fragments (a, b, c) that can be detected after hybridization with probe 1 (black bar). (C) Schematic presentation of Southern blot and the resulting fragments using probe 1. (D) Schematic presentation of PCR using different primer combinations.
Fig. 8. Multiplex PCR strategy for the verification of gene trap cassette inversions in ESCs and mice. Positions of primers (A) and expected amplification products (B) from the three possible alleles inducible by recombination of the FlipRosaßgeo gene trap vector.