Flp recombinase regulated lacZ expression at the ROSA26 locus.

Manipulation of gene expression patterns through gene targeting in mouse embryonic stem (ES) cells (Thomas and Capecchi, 1987) and the more recent advent of conditional mutagenesis has revolutionized mouse genetics. The recombinase system most commonly used for conditional alleles in mouse is Cre/loxP derived from bacteriophage P1 (Sauer and Henderson, 1988). While less utilized, FLP/FRT derived from S. cerevisiae (O’Gorman et al., 1991) has been elegantly used by Martin and colleagues in combination with Cre to create a series of alleles for Fgf8 of varying strength (Meyers et al., 1998). The utility of Flp as a robust tool for mammalian genetics has been markedly improved by Stewart and colleagues who cleverly used cycling mutagenesis to select for variant alleles of Flp that carry out recombination at high efficiency in mammalian cells (Buchholz et al., 1998). Experimental strategies to sequentially inactivate and reactivate a given gene, or to test epistatic relationships between genes in somatic cells, are now possible using a combination of Cre with the enhanced Flpe. As Cre and Flp transgenes used to regulate conditional alleles may not recapitulate the expected pattern of expression (e.g., Tallquist and Soriano, 2000), it is essential to characterize the activity of these recombinase alleles to correctly interpret the phenotypes of conditionally mutant animals. Whereas several reporter lines for Cre have been described, we have found the allele created by Soriano in which Cre regulates lacZ expression at the ROSA26 locus to be especially useful (Soriano, 1999). ROSA26 insertions are ubiquitously expressed and show no overt phenotype when homozygous (Zambrowicz et al., 1997). We therefore created an analogous ROSA26 allele, which we term R26FR, to monitor Flp activity in vivo. The targeting vector contains an FRT-flanked neomycin resistance gene with four transcriptional stop signals placed upstream of a -galactosidase-neo fusion ( -geo; Friedrich and Soriano, 1991) all inserted into ROSA26 genomic sequence (Fig. 1a). The vector was transfected into V6.5 ES cells (Eggan et al., 2001), 48 clones were analyzed by Southern blotting, and 24 correctly targeted clones were identified (Fig. 1b). To demonstrate that the R26FR clones could activate – geo, a CMV-FLPe expression plasmid (Buchholz et al., 1998) was introduced by transient lipofection. One hundred ninety-two clones were picked and aliquots analyzed for – galactosidase activity by X– gal staining. Both Flpe-recombined (Fig. 2a) and parental clones (Fig. 2d) were identified, and absence or presence of the stop cassette was confirmed by Southern blotting (Fig. 1b). To create the reporter mouse strain, completely ESderived mice (Eggan et al., 2001; Nagy et al., 1990) were made by injecting targeted clones into tetraploid blastocysts. Four viable pups were produced, and two were used to derive tail fibroblast cell lines. To demonstrate functionality of the reporter in somatic cells, Flpe was cloned into an MSCV-IRES-GFP retroviral vector (Van Parijs et al., 1999) and used to infect the transformed reporter cell lines. Cells sorted for and against GFP expression were enriched by FACS and compared with those infected with an antisense Flpe retroviral control vector. Sorted cells were stained with X-gal and analyzed by Southern blotting, which demonstrated conditional recombination (Fig. 1b) and -galactosidase activity (Fig. 2b) dependent upon Flpe. To confirm activity of the reporter allele in vivo, R26FR mice were crossed to a -actin-Flpe transgenic mouse (Rodriguez et al., 2000). Embryos dissected at day 10.5 pc were stained with X-gal and showed ubiquitous Flp-mediated activation (Fig. 2c). The purpose of this FLP reporter allele is to specify the extent of FLP activity with a convenient and widely used enzymatic activity, -galactosidase. Development of FLP/ FRT tools such as our R26FR reporter allele and the recently described alkaline phosphatase reporter (Awatramani et al., 2001) improve the ability to characterize Flpe alleles in vivo and therefore expand the opportunities for using precisely regulated dual recombinase experimental strategies.