Generalized lacZ expression with the ROSA26 Cre reporter strain
Philippe Soriano
Program in Developmental Biology, Division of Basic Sciences, Fred
Hutchinson Cancer Research Center, 1100 Fairview Avenue North
, Seattle, Washington 98109,
USA. psoriano@fhcrc.org
Mouse strains expressing the site-specific recombinase Cre (or Flp) facilitate
conditional ablation of gene function when one or several exons of the gene
of interest are flanked by loxP (or FRT) sites1.
Cre expression achieved by classic transgenesis or targeting to an appropriate
locus might be tissue specific, temporally restricted or inducible2,
3.
In such experimental outlines, it is necessary to monitor Cre activity at
desired time points as well as to verify that Cre was not active previously
during development. Other investigators have generated transgenic4,
5
or knock-in6 lines in which lacZ expression is conditional
on the removal of an intervening segment. However, such lines are most useful
if lacZ can be expressed in all cell types and hence is driven off
a constitutively active promoter in the mouse.
We have previously described a gene-trap strain, ROSA geo 26, in
which expression of the geo reporter appears to be constitutive during
embryonic development7,
8. I report here successful targeting
at the ROSA26 locus and the derivation of a reporter line for monitoring Cre
expression. To target the locus, a 5-kb genomic fragment was subcloned in
a plasmid vector along with a diphtheria toxin (DTA) expression cassette for
negative selection to produce the vector pROSA26-1. A splice acceptor sequence
(SA) identical to the one used in the original gene-trap allele, a neo expression
cassette flanked by loxP sites, a lacZ gene and a polyadenylation
(bpA) sequence were inserted at a unique XbaI site approximately 300-bp 5´
of the original gene-trap integration site (Fig. 1a).
A triple polyadenylation sequence9 was added to the 3´
end of the neo expression cassette to prevent transcriptional read-through.
This ROSA26 reporter (R26R) construct was linearized with KpnI and
electroporated into AK7 embryonic stem (ES) cells. Following G418 selection,
8 of 23 G418r colonies were found to have correctly undergone
homologous recombination by PCR and were further verified by Southern-blot
analysis (Fig. 1b). Three clones were used to
derive germline chimaeras. Heterozygous mice did not display an overt phenotype,
and were bred to obtain viable and fertile homozygous mutant progeny.
a, Top, restriction map of the locus. PCR primers from ROSA26
flanking (5´-CCTAAAGAAGAGGCTGTGCTTTGG-3´) and splice acceptor
(5´-CATCAAGGAAACCCTGGACTACTG-3´) sequences were used to amplify
an approximately 1.2-kb diagnostic fragment (grey arrowheads). The probe used
for Southern-blot analysis is shown as a shaded box. LoxP sites are
indicated by black arrowheads. Only EcoRV sites are indicated for pROSA26-R.
b, Left, Southern-blot analysis of targeted clone (1-8) and wild-type
(WT) DNA digested with EcoRV; 1C and 2C are populations of ES clones
1 and 2 transiently transfected with PGKCrebpA (a gift of M. Komada) showing
the expected shorter targeted fragment due to deletion of the neo segment.
Right, only the shorter EcoRV fragment is seen in offspring also expressing
Cre (lanes 6,7) in contrast with the reporter allele alone (lane 2). ROSA26-R
embryos could be genotyped by PCR (approximately 500-bp wild-type and (approximately
250-bp mutant fragments) using three oligonucleotides: 5´-AAAGTCGCTCTGAGTTGTTAT-3´, 5´-GCGAAGAGTTTGTCCTCAACC-3´
and 5´-GGAGCGGGAGAAATGGATATG-3´.
The recombination efficiency will depend on the level of Cre expression
and thus vary between different mouse strains. Heterozygous R26R mice were
bred with R26Cre mice, a general deletor mouse line made by targeting Cre
to the ROSA26 locus, and embryos were collected at various stages between
embryonic day (E) 8 and E16 and stained with X-Gal for lacZ activity.
Embryos heterozygous for both R26Cre and R26R alleles displayed ubiquitous
blue staining, whereas wild-type or heterozygous R26R embryos did not show
any staining (Fig. 2). These results indicate that the
reporter line functions as planned and that following recombination at preimplantation,
at which time the ROSA26 promoter is activated7, lacZ
can be expressed in all cells of the embryo. Crossing R26R mice with other
Cre-expressing strains resulted in different or more restricted lacZ
expression patterns (data not shown). The R26R mouse strain should be of wide
use for monitoring Cre expression, as well as for analysing cell lineages
during development, and is available from the Induced Mutant Resource of the
Jackson Laboratory (stock numbers 3309 and 3310). Another reporter mouse line
has been generated at the BT5 gene-trap locus using a similar approach (S.K.
Michael and E.J. Robertson, pers. comm.). Although the ROSA26 promoter has
been shown to have activity on its own and might lead to broad gene expression
in transgenic mice8, the high rate of homologous recombination
and generalized lacZ expression observed here suggest that targeting
of genes to the ROSA 26 locus may be a desirable method to achieve ubiquitous
expression during development or in the adult.
Figure 2.LacZ expression following Cre recombination
a, Whole-mount X-Gal staining of E9 R26R heterozygous (left) and
R26R/R26Cre (right) compound heterozygous embryos. b, Sagittal section
of an E9 R26R/R26Cre compound heterozygous embryo. c, Higher magnification
(200) sagittal section showing lacZ expression in all cells
in the somites and the underlying mesenchyme; scale bar, 50 m. d,
Cross-section through a somite (1000); scale bar 10m.
geo 26, in
which expression of the 
200) sagittal section showing lacZ expression in all cells
in the somites and the underlying mesenchyme; scale bar, 50 
m. d,
Cross-section through a somite (
