Genomewide production of multipurpose alleles for
the functional analysis of the mouse genome
Frank Schnu ¨tgen*†, Silke De-Zolt*†, Petra Van Sloun*†, Melanie Hollatz†‡, Thomas Floss†‡, Jens Hansen†‡,
Joachim Altschmied*, Claudia Seisenberger†‡, Norbert B. Ghyselinck§, Patricia Ruiz†¶, Pierre Chambon§,
Wolfgang Wurst†‡, and Harald von Melchner*†?
*Department of Molecular Hematology, University of Frankfurt Medical School, 60590 Frankfurt am Main, Germany;‡Institute of Developmental Genetics,
GSF-National Research Center for Environment and Health, 85764 Neuherberg, Germany;§Institut Clinique de la Souris and Institut de Ge ´ne ´tique et de
Biologie Mole ´culaire et Cellulaire, CU de Strasbourg, 67404 Illkirch Cedex, France; and¶Center for Cardiovascular Research, Charite ´ Universita ¨tsmedizin, and
Department of Vertebrate Genomics, Max Planck Institute for Molecular Genetics, 14195 Berlin, Germany
Contributed by Pierre Chambon, March 22, 2005
A type of retroviral gene trap vectors has been developed that can
induce conditional mutations in most genes expressed in mouse
embryonic stem (ES) cells. The vectors rely on directional site-specific
recombination systems that can repair and reinduce gene trap mu-
into the mouse genome, genetic mutations can be produced at a
particular time and place in somatic cells. In addition to their condi-
tional features, the vectors create multipurpose alleles amenable to a
cell lines with conditional mutations in single genes yet assembled,
can be ordered from the German Gene Trap Consortium, are freely
available to the scientific community.
conditional mutagenesis ? ES cells ? gene trapping ? site-specific
recombination ? insertional mutagenesis
tional annotation of mammalian genes (1, 2). Among the various
is the most relevant for extrapolation to human genetic disease.
Although several model organisms have been used in a variety of
mutagenesis approaches, the mouse offers particular advantages
because its genome structure and organization are closely related
to those of the human genome. More importantly, mouse embry-
onic stem (ES) cells, which grow indefinitely in tissue culture, allow
the generation of mice with defined mutations in single genes for
functional analysis and studies of human disease.
Several mutagenesis strategies have been deployed in mice,
ranging from random chemical (N-ethyl-N-nitrosourea) mutagen-
esis coupled with phenotype-driven screens (3, 4) to sequence-
based approaches using ES cell technology, such as gene trapping
and gene targeting (5, 6).
Gene trapping is a high-throughput approach that is used to
introduce insertional mutations across the mouse genome. It is
performed with gene trap vectors where the principal element is a
gene trapping cassette consisting of a promoterless reporter gene
sequence (polyadenylation sequence; pA). When inserted into an
intron of an expressed gene, the gene trap cassette is transcribed
from the endogenous promoter in the form of a fusion transcript in
which the exon(s) upstream of the insertion site is spliced in frame
to the reporter?selectable marker gene. Because transcription is
terminated prematurely at the inserted polyadenylation site, the
processed fusion transcript encodes a truncated and nonfunctional
version of the cellular protein and the reporter?selectable marker
(7). Thus, gene traps simultaneously inactivate and report the
expression of the trapped gene at the insertion site and provide a
DNA tag (gene trap sequence tag, GTST) for the rapid identifi-
ith the complete sequencing of the human and mouse
genomes, attention has shifted toward comprehensive func-
cation of the disrupted gene. Because gene trap vectors insert
randomly across the genome, a large number of mutations can be
trap approaches have been used successfully in the past by both
academic and private organizations to create libraries of ES cell
lines harboring mutations in single genes (8–11). Collectively, the
existing resources cover ?66% of all protein-coding genes within
the mouse genome (12). However, the gene trap vectors that have
been used to generate the currently available resources induce only
null mutations; mouse mutants generated from these libraries can
the trapped gene. Therefore, for most of the mutant strains, the
significance of the trapped gene for human disease remains uncer-
tain, because most human disorders result from late-onset gene
dysfunction. In addition, between 20% and 30% of the genes
targeted in ES cells are required for development and cause
embryonic lethal phenotypes when transferred to the germ line,
precluding functional analysis in the adult (9, 13).
To circumvent the limitations posed by germ-line mutations,
to spatially and temporally restrict the mutation to somatic cells
(14). However, targeted mutagenesis in ES cells requires a detailed
knowledge of gene structure and organization to physically isolate
a gene in a targeting vector. Although the completed sequencing of
the mouse genome greatly assists targeted mutagenesis, the gen-
eration of mutant mouse strains by this procedure that can handle
only one target at a time is time consuming, labor intensive,
expensive, and relatively inefficient.
trap vectors based on a strategy for directional site-specific recom-
bination termed flip-excision (FlEx) (15). The vectors employ two
directional site-specific recombination systems that, when activated
in succession, invert the gene trap from its mutagenic orientation
on the sense, coding strand to a nonmutagenic orientation on the
antisense, noncoding strand. We show that mutations induced by
these vectors in ES cells can be both repaired and reinduced by
site-specific recombination and introduce a new resource of ES cell
lines primed for conditional mutagenesis.
Materials and Methods
Plasmids. pFlipROSA?geo was assembled in pBabeSrf, a modified
pBabepuro retroviral vector lacking the promoter and enhancer
elements from the 3? LTR (16). Pairs of the heterotypic frt?F3 and
Abbreviations: GTST, gene trap sequence tag; SA, splice acceptor; RTs, recombinase target
sequences; pA, polyadenylation sequence; FlEx, flip-excision; RBBP7, retinoblastoma bind-
ing protein 7; Glt28d1, glycosyltransferase 28 domain containing 1 gene.
Gene Trap Consortium.
?To whom correspondence should be addressed at: Department of Molecular Hematology,
University of Frankfurt Medical School, Theodor-Stern-Kai 7, 60590 Frankfurt am Main,
Germany. E-mail: firstname.lastname@example.org.
© 2005 by The National Academy of Sciences of the USA
www.pnas.org?cgi?doi?10.1073?pnas.0502273102 PNAS ?
May 17, 2005 ?
vol. 102 ?
no. 20 ?
lox511?loxP recombinase target sequences (RTs) were cloned in
the illustrated orientation (Fig. 1A) into the unique BamHI and
EcoRI sites of pBabeSrf, yielding the intermediate plasmid pBLF.
RTs were obtained by synthetic oligonucleotide annealing and
extension overlap PCR. To enable efficient recombination, 86- and
46-bp spacers were inserted between frt?F3 and loxP?lox511 sites,
respectively. To obtain pFlipRosa?geo, a SA?geopA cassette de-
rived from the gene trap vector ROSA?geo (17) was inserted into
the SnaBI site of pBLF between the inversely oriented RT pairs.
The final pFlipRosa?geo vector was verified by sequencing. The
pFlipRosaCeo vector was obtained from pFlipRosa?geo by replac-
ing the SA?geo cassette with the Ceo fusion gene derived from
pU3Ceo (16). The final pFlipRosaCeo plasmid was verified by
sequencing. Oligonucleotide and primer sequences used in the
various cloning steps are available on request.
The pCAGGS-FLPe expression plasmid was a gift from A.
Francis Stewart (18). The pCAGGS-Cre expression plasmid was
derived from pCAGGS-FLPe by replacing the FLPe cDNA with
the Cre cDNA of pSG5Cre (19).
ES Cell Cultures, Infections, and Electroporations. The [C57BL?6J ?
129S6?SvEvTac]F1 ES cell lines were grown on irradiated or
mitomycin C-treated MEF feeder layers in the presence of 1,000
U?ml leukemia inhibitory factor (LIF) (Esgro, Chemicon, Hof-
heim, Germany) as described in ref. 9.
Gene trap retrovirus was produced in Phoenix-Eco helper cells
by using the transient transfection strategy described previously
(20). ES cells were infected with the virus-containing supernatants
expressing ES cell lines were selected in 130 ?g?ml G418 (Invitro-
gen), manually picked, expanded, and stored frozen in liquid
Electroporations were carried out by using 1 ? 107ES cells, 10
in ref. 5. After incubating for 2 days in medium supplemented with
0.6 ?g?ml puromycin (Sigma–Aldrich), the cells were trypsinized
and seeded at low density (1,000 cells per dish) onto 60-mm Petri
dishes. Emerging clones were manually picked after 9 days and
Nucleic Acids and Protein Analyses.PCRswereperformedaccording
of reverse-transcribed total RNA in a total volume of 50 ?l. The
primer sequences used are available on request.
For Northern blotting, poly(A)?RNA was purified from total
RNA by using the Oligotex mRNA-minikit (Qiagen, Hilden,
Germany) according to the manufacturer’s instructions. The
mRNA (1–2 ?g) was fractionated on 1% formaldehyde?agarose
gels, blotted onto Hybond N? (Amersham Biosciences) nylon
membranes, and hybridized to32P-labeled cDNA probes (Hart-
mann Analytic, Braunschweig, Germany) in ULTRAhyb hybrid-
ization solution (Ambion, Austin, TX) according to manufac-
turer’s instructions. The Glt28d1-cDNA probe was obtained by
asymmetric RT-PCR (21) using an antisense primer comple-
mentary to exon 10 of the Glt28d1 gene.
Semiautomated 5?-RACE and sequencing were performed as
described in ref. 9. The sequences of the generic and vector-specific
primers used are available on request.
Western blotting was performed as described in ref. 22, by using
Biotechnology) primary antibodies.
GTST Analysis. GTSTs were analyzed as described in ref. 9 by using
the following databases: GenBank (release 144), UniGene (build
141), RefSeq (release 8) (all at www.ncbi.nlm.nih.gov), ENSEMBL
v26.33 (www.ensembl.org), MGI (www.informatics.jax.org) and
GeneOntology (December 2004 release) (www.geneontology.org).
Vector Design. Two gene trap vectors were designed for large-scale
contains a classic SA, ?-galactosidase?neomycin phosphotransfer-
ase fusion gene (?geo), pA cassette inserted into the backbone of
inverse transcriptional orientation relative to the virus (Fig. 1A)
(17). The second vector, FlipRosaCeo, is similar to FlipRosa?geo
except that SA?geo has been exchanged with Ceo, which is an
in-frame fusion between the human CD2 cell surface receptor and
the neomycin-resistance genes (16). Unlike ?geo, Ceo does not
Schematic representation of the retroviral gene trap vectors. LTR, long terminal
repeat; frt (yellow triangles) and F3 (green triangles), heterotypic target se-
quences for the FLPe recombinase; loxP (red triangles) and lox511 (purple trian-
gles), heterotypic target sequences for the Cre-recombinase; SA, splice acceptor;
?geo, ?-galactosidase?neomycin phosphotransferase fusion gene; pA, bovine
growth hormone polyadenylation sequence; TM, human CD2 receptor trans-
membrane domain. (B) Conditional gene inactivation by a SA?geopA cassette.
The SA?geopA cassette flanked by recombinase target sites (RTs) in a FlEx
Transcripts (shown as gray arrows) initiated at the endogenous promoter are
of the SA?geopA cassette. Thereby the ?geo reporter gene is expressed and the
endogenous transcript is captured and prematurely terminated at the cassette’s
pA causing a mutation. In step 1, FLPe inverts the SA?geopA cassette onto the
antisense, noncoding strand at either frt (shown) or F3 (not shown) RTs and
reinversion because the remaining frt and F3 RTs cannot recombine. This inver-
sion reactivates normal splicing between the endogenous splice sites, thereby
repairing the mutation. Cre-mediated inversion in steps 3 and 4 repositions the
SA?geopA cassette back onto the sense, coding strand and reinduces the muta-
tion. Note that the recombination products of steps 1 and 3 are transient and
transformed into the stable products of step 2 and 4, respectively (15).
Conditional gene trap vectors and mechanism of gene inactivation. (A)
www.pnas.org?cgi?doi?10.1073?pnas.0502273102Schnu ¨tgen et al.
a powerful cryptic 5? splice site close to its 5? end. Moreover, Ceo
i.e., secretory pathway genes (Fig. 1A) (16). Previous studies
involving the isolation of 3,620 ES cell lines with the retroviral gene
genes with ?80% efficiency. This is in contrast to the classic ?geo
vectors, of which only 19% insert into such genes. Thus, the classic
and the secretory pathway gene trap vectors are complementary
and, therefore, we equipped both with a conditional mechanism.
The mechanism relies on two site-specific recombination systems
(FLPe?frt and Cre?loxP) that enable gene trap cassette inversions
from the sense, coding strand of a trapped gene to the antisense,
noncoding strand and back. As a result, the gene trap vectors allow
(i) high-throughput selection of gene trap lines by using G418, (ii)
inactivation of gene trap mutations before ES cell line conversion
into mice by blastocyst injection, and (iii) reactivation of the
mutations at prespecified times and in selected tissues of the
We used an adaptation of a recently published site-specific
recombination strategy termed FlEx (15). FlEx uses pairs of
inversely oriented heterotypic RTs such as loxP and lox511 or frt
and F3. When inserted upstream and downstream of a gene trap
cassette, Cre or FLPe recombinases invert the cassette and place a
homotypic RT pair near to each other in a direct orientation.
Recombination between this pair of directly repeated RTs excises
one of the other heterotypic RTs, thereby locking the recombina-
flanking the gene trap cassettes of FlipRosa?geo and FlipRosaCeo
with pairs of heterotypic lox and frt sites (Fig. 1A), a successive
delivery of FLPe and Cre to a trapped ES cell line will induce two
directional inversions, thereby first repairing and then reinducing
the gene trap mutation as exemplified for the SA?geopA gene trap
cassette in Fig. 1B.
Gene Trap Cassette Inversions in ES Cells. To test for recombinase-
mediated inversions, several FlipRosa?geo-trapped ES cell lines
were selected for high levels of ?geo expression by using X-Gal
staining. X-Gal-positive (blue) cell lines were then transiently
versions in FlipRosa?geo-trapped ES cell lines. (A
and B) ES cells were infected with FlipRosa?geo
virus and selected in G418. X-Gal-positive sub-
lines (blue) were electroporated with FLPe (A) or
Cre (B) expression plasmids and stained with X-
Gal after incubating for 10 days. DNA extracted
from blue and white sublines was subjected to a
multiplex PCR to identify inversions. Primer po-
sitions within FlipRosa?geo are indicated by
large arrows; allele-specific amplification prod-
ucts are visualized on ethidium bromide-stained
gels to the right. (C) Sublines of the FS4B6 ES cell
line harboring Cre- or FLPe-inverted gene trap
insertions were electroporated with both FLPe-
and Cre-expression plasmids. The amplification
products obtained from the progeny lines by
bromide-stained gel to the right. t, trapped al-
lele; inv, inverted allele; re-inv, reinverted allele;
M, molecular weight marker (1 kb ? ladder,
Invitrogen). FS4B6 (1anes 1–3), parental
FlipRosa?geo-trapped ES cell line; FS4B6 C14
(lanes 4–6), Cre-inverted subline; and FS4B6 F14
(lanes 7–9), FLPe-inverted subline.
Site-specific recombinase-induced in-
Schnu ¨tgen et al.
May 17, 2005 ?
vol. 102 ?
no. 20 ?
transfected with FLPe or Cre expression plasmids, and emerging
subclones were stained with X-Gal. As shown in Fig. 2, exposure of
a mixture of X-Gal-positive (blue) and X-Gal-negative (white)
subclones, indicating that several cell lines have ceased to express
DNA from both the blue and the white sublines, and subjected it
to an allele-specific PCR. Fig. 2 A and B shows that, in each case,
the amplification products obtained from the blue and white clones
corresponded to a normal and to an inverted gene trap allele,
respectively. Taken together, the results indicate that both FLPe
and Cre can disrupt the gene trap expression by simply flipping it
to the antisense, noncoding strand.
To test whether the FLPe- or Cre-inverted cell lines would
reinvert after a second recombinase exposure, we reexpressed
FLPe and Cre in each of the cell lines and checked their progeny
readily reinverted the Cre-inverted subline FS4B6 C14 (lane 6) but
readily reinverted the FLPe-inverted subline FS4B6 F14 (lane 8)
but not the Cre-inverted subline FS4B6 C14 (lane 5). Taken
together, the results indicate that gene trap reinversions are induc-
ible only by the recombinase that was not involved in the original
inversion, suggesting that the recombination products obtained
with either recombinase are stable. Inversions induced by Cre and
FLPe in FlipRosaCeo-trapped ES cell lines were similarly stable
by long periods of exposure in culture, during development, or by
were unable to detect recombination between loxP and lox511 sites
background recombination does not significantly affect conditional
Reversibility of Gene Trap Mutations.Totestwhetherthemutations
induced by the conditional gene trap vectors are reversible, we
selected the Q017B06 and M117B08 gene trap lines for further
analysis. In Q017B06, the FlipRosa?geo gene trap vector dis-
rupted the retinoblastoma binding protein 7 (RBBP7) gene at
the level of the first intron. In M117B08, the FlipRosaCeo gene
trap vector disrupted the glycosyltransferase 28 domain contain-
ing 1 gene (Glt28d1) in the 10th intron. Both genes are located
on the X chromosome of a male-derived ES cell line, which
provided a haploid background for the mutational analysis. As
shown in Figs. 3 B and C and 4 B and C , the RBBP7 (Fig. 3) and
Glt28d1 (Fig. 4) genes were both expressed in the wild-type cells
as expected. However, expression was either blocked (RBBP7,
Fig. 3) or severely repressed (Glt28d1, Fig. 4) by the gene trap
insertions. Both trapped cell lines instead expressed fusion
transcripts as a result of splicing the upstream exons to the gene
trap cassettes (Figs. 3B and 4 B and C).
lines was whether endogenous gene expression would resume after
Cre- or FLPe-induced inversions. Toward this end, we expressed
Cre or FLPe in the Q017B06 and M117B08 cell lines, isolated
several sublines, and genotyped them by allele-specific PCR (Figs.
3A and 4A). Inverted sublines were then analyzed for RBBP7,
Glt28d1, and gene trap cassette expression by using RT-PCR in
combination with Northern and Western blotting. Figs. 3 B and C
and 4 B and C show that in both cell lines the endogenous gene
expression was restored to wild-type levels and the fusion tran-
scripts disappeared, indicating that the antisense gene trap inser-
tions do not interfere with gene expression. Finally, to test whether
coding strand would reinduce the mutation, we exposed inverted
in the RBBP7 gene (ENSEMBL ID: ENSMUSG00000031353). The Q017B06 gene
trap cell line (t) was transiently transfected with a FLPe expression plasmid,
staining and allele-specific PCR (inv). Inverted sublines were then electropo-
in G418 (re-inv). (A) X-Gal staining (Upper) and allele-specific PCR amplifica-
tion products (Lower) from the trapped RBBP7 locus in trapped (t), inverted
amplification of RBBP7 wild-type and trapped fusion transcripts expressed in
Q017B06 cells before and after exposure to FLPe and Cre recombinases. The
positions of the primers used are shown on top, wherein U19 ? 5?-GCT CTT
GAC TAG CGA GAG AGA AG-3?, B32 ? 5?-CAA GGC GAT TAA GTT GGG TAA
CG-3?, U34 ? 5?-CCA GAA GGA AAG GAT TAT GC-3?, and U35 ? 5?-ACA GAG
CAA ATG ACC CAA GG-3?. Amplification products are visualized below on
ethidium bromide-stained gels. Amplification of the RNA polymerase II (RNA
pol II) transcript serves as a positive control. wt, parental ES cells; t, trapped
Q017B06 cells; inv, inverted Q017B06 subline; re-inv, reinverted Q017B06
subline; endo, endogenous transcript; fus, fusion transcript. (C) Western blot
analysis of the RBBP7 protein expressed in Q017B06 cells. Crude cell lysates
using the anti-RbAp46 antibody. The anti-lamin A antibody served as a
Conditional mutation induced by a FlipRosa?geo gene trap insertion
www.pnas.org?cgi?doi?10.1073?pnas.0502273102Schnu ¨tgen et al.
subclones to FLPe or Cre. Figs. 3 and 4 show that the reinverted
sublines lost the endogenous gene expression and re-expressed the
that the FlipRosa?geo- and FlipRosaCeo-induced mutations can
be repaired and reinduced by the successive activation of the two
Large-Scale Conditional Mutagenesis in ES Cells. We isolated 4,525
ES cell lines with conditional gene trap insertions and recovered
4,138 GTSTs by 5?-RACE. Of these, 3,257 were derived from
FlipRosa?geo and 881 from FlipRosaCeo integrations. Ninety
percent of the FlipRosa?geo and 99% of the FlipRosaCeo GTSTs
belonged to RefSeq annotated genes (Table 1). The number of
annotated genes was nearly double that found in our previous
overall efficiency of trapping was similar to that observed in
previous studies, as was the number of preferred insertions sites
(i.e., hot spots) (Table 1). Insertions occurred in all chromosomes,
including one on the Y chromosome, and their number correlated
with the number of genes per chromosome (data not shown).
frt sites built into the gene trap vectors do not affect the efficiency
of trapping. Regardless of the vector, the vast majority of gene trap
insertions occurred into first and second introns, confirming the
reported preference of retroviral integrations near the 5? ends of
genes (Fig. 5) (25). As expected, the major difference between the
vectors was their ability to capture signal sequence genes. Whereas
?80% of the FlipRosaCeo insertions were in genes encoding
secreted or transmembrane proteins, only 21% of FlipRosa?geo
insertions captured secretory pathway genes according to Gene-
Ontology. Thus, like the nonconditional vectors, the two types of
conditional gene trap vectors complement each other in gene
site-specific recombination to develop an approach suitable for the
large-scale induction of conditional mutations in ES cells. The
strategy (FlEx) (15) that enables directional inversions of gene trap
cassettes at the insertion sites. By using gene trap integrations into
X-chromosomal genes, we have shown that gene trap vectors
can be used for generating mice with either null or conditional
mutations. For example, to obtain straight knock-outs the cell lines
can be converted directly into mice by blastocyst injection. How-
ever, to obtain conditional mutations, one would first repair the
mutation in ES cells, preferentially with FLPe to reserve the more
efficient Cre for in vivo recombination, and then proceed to mouse
production. Resulting mice would lack germ-line mutations but
would be vulnerable to somatic mutations inducible by Cre. De-
Table 1. Trapping efficiency with conditional gene trap vectors
Gene trap vector
annotated genes (%)
‘‘hot spots’’* (%)
*All genes with ?2 insertions were classified as hot spots.
Ceo gene trap insertion in the Glt28d1 gene (ENSEMBL
ID ENSMUST00000040338). The M117B08 gene trap
line was treated with recombinases and processed as
described for Q017B06 in the legend to Fig. 3, except
second. (A) Allele-specific PCR of the trapped Glt28d1
locus in trapped (t), inverted (inv), and reinverted (re-
inv) cell lines. (B) RT-PCR of Glt28d1wild type and
Glt28d1?gene trap fusion transcripts expressed in
M117B08 cells before and after exposure to Cre and
FLPe recombinases. The positions of the respective
primers within the trapped gene are shown on top,
wherein M117B8s ? 5?-GAG AGT GCT GGC CAG CTG
GAA C-3?, G01 ? 5?-CAA GTT GAT GTC CTG ACC CAA
G-3?, and M117B8as1 ? 5?-CCA CCA TAC TCC ACA CAC
TCT G-3?. Amplification products are visualized on
the glyceraldehyde-3-phosphate dehydrogenase
(GAPDH) transcript serves as a positive control. wt,
parental ES cells; t, trapped M117B08; inv, inverted
M117B08 subline; re-inv, reinverted M117B08 subline;
endo, endogenous transcript; fus, fusion transcript. (C)
Northern blot analysis of Glt28d1 transcripts expressed
in M117B08 cells. Samples (2 ?g) of polyadenylated
RNAs from wild-type (wt), Q017B06 (t), inverted
were fractionated on 1% formaldehyde?agarose gels
and hybridized to a32P-labeled Glt28d1 cDNA probe.
The Glt28d1 probe was obtained by asymmetric RT-PCR using a reverse primer in exon 10 to amplify sequences upstream of the insertions site. The loading of
each lane was then assessed by using a GAPDH probe. endo, endogenous transcript; fus, fusion transcript; Glt28d1, Glt28d1 transcript.
Conditional mutation induced by a FlipRosa-
Schnu ¨tgen et al.
May 17, 2005 ?
vol. 102 ?
no. 20 ?
pending on the type of Cre and the form of its delivery, the
mutations can be reactivated in prespecified tissues at prespecified
Because of the inherent recombinase target sites, the vector
insertions create multipurpose alleles enabling a large variety of
postinsertional modifications by recombinase-mediated cassette
exchange (26). Examples include replacing the gene trap cassettes
with Cre recombinase genes to expand the Cre-zoo, or with point
insertion of toxin genes for cell lineage-specific ablations.
from its position on the antisense, noncoding strand. In the two
examples described, the antisense insertions were innocuous, how-
ever, this will not always be the case. Factors likely to influence the
antisense neutrality include cryptic splice sites and transcriptional
termination signals. In line with this, we have shown that aberrant
splicing induced by an antisense gene trap insertion resulted in a
partial gene inactivation and an interesting phenotype (22). Thus,
the most likely outcome of antisense insertions that interfere with
gene expression are hypomorphic mutations, which have a merit of
their own. However, in silico analysis failed to identify sequences
of their insertions create bona fide conditional alleles.
By using the vectors in high-throughput screens, we have assem-
bled the largest library of ES cell lines with conditional mutations
of single protein-coding genes, including secretory pathway genes.
Presently it contains 1,000 potentially conditional alleles (Table 2,
which is published as supporting information on the PNAS web
site), which is about 10 times the number produced within the last
10 years by gene targeting. Considering that these gene trap lines
were isolated in less than a year, conditional gene trapping seems
significantly more efficient than conditional gene targeting. How-
ever, analysis of the existing gene trap resources indicates that gene
trapping is more efficient than gene targeting only up to ?50% of
all mouse genes, after which the mutation rate falls to a level
comparable to gene targeting (12). Moreover, effective gene trap-
28). Although gene trapping strategies have been described for
genes that are not expressed in ES cells, their performance is quite
inconsistent, making them unsuitable for high-throughput ap-
proaches (29–31). We believe that for a comprehensive mutagen-
esis of the mouse genome, a balance between gene trapping and
gene targeting, performed with generic gene trap cassettes inserted
into the targeting vectors, is likely to be the most efficient and
scientific community. Cell lines can be ordered directly from the
German Gene Trap Consortium.
We thank Julia Schmidt, Corinna Strolz, Silke Garkisch, Carsta Werner,
Beata Thalke, and Dorotha German for excellent technical assistance.
This work was supported by grants from the Deutsche Forschungsge-
Forschung (to the German Gene Trap Consortium).
1. Austin, C. P., Battey, J. F., Bradley, A., Bucan, M., Capecchi, M., Collins, F. S., Dove, W. F.,
Duyk, G., Dymecki, S., Eppig, J. T., et al. (2004) Nat. Genet. 36, 921–924.
A., Brown, S., Carmeliet, P., et al. (2004) Nat. Genet. 36, 925–927.
3. Cox, R. D. & Brown, S. D. (2003) Curr. Opin. Genet. Dev. 13, 278–283.
4. Brown, S. D. & Balling, R. (2001) Curr. Opin. Genet. Dev. 11, 268–273.
5. Floss, T. & Wurst, W. (2002) Methods Mol. Biol. 185, 347–379.
6. Mansouri, A. (2001) Methods Mol. Biol. 175, 397–413.
7. Stanford, W. L., Cohn, J. B. & Cordes, S. P. (2001) Nat. Rev. Genet. 2, 756–768.
8. Wiles, M. V., Vauti, F., Otte, J., Fu ¨chtbauer, E. M., Ruiz, P., Fu ¨chtbauer, A., Arnold, H. H.,
Lehrach, H., Metz, T., von Melchner, H., et al. (2000) Nat. Genet. 24, 13–14.
9. Hansen, J., Floss, T., Van Sloun, P., Fu ¨chtbauer, E. M., Vauti, F., Arnold, H. H., Schnu ¨tgen, F.,
Wurst, W., von Melchner, H. & Ruiz, P. (2003) Proc. Natl. Acad. Sci. USA 100, 9918–9922.
Lee, R. E., Yee, A., L’Italien, L., et al. (2003) Nucleic Acids Res. 31, 278–281.
11. Zambrowicz, B. P., Abuin, A., Ramirez-Solis, R., Richter, L. J., Piggott, J., BeltrandelRio,
H., Buxton, E. C., Edwards, J., Finch, R. A., Friddle, C. J., et al. (2003) Proc. Natl. Acad.
Sci. USA 100, 14109–14114.
12. Skarnes, W. C., von Melchner, H., Wurst, W., Hicks, G., Nord, A. S., Cox, T., Young, S. G.,
Ruiz, P., Soriano, P., Tessier-Lavigne, M., et al. (2004) Nat. Genet. 36, 543–544.
13. Mitchell, K. J., Pinson, K. I., Kelly, O. G., Brennan, J., Zupicich, J., Scherz, P., Leighton,
P. A., Goodrich, L. V., Lu, X., Avery, B. J., et al. (2001) Nat. Genet. 28, 241–249.
14. von Melchner, H. & Stewart, A. F. (2004) in Handbook of Stem Cells, eds. Lanza, R., Blau
H., Gearhart J., Hogan B., Melton D., Moore M., Pedersen, R., Thomas, E. D., Thomson
J., Verfaillie, C., et al. (Academic, Oxford), Vol. 1, pp. 609–622.
15. Schnu ¨tgen, F., Doerflinger, N., Calleja, C., Wendling, O., Chambon, P. & Ghyselinck, N. B.
(2003) Nat. Biotechnol. 21, 562–565.
16. Gebauer, M., von Melchner, H. & Beckers, T. (2001) Genome Res. 11, 1871–1877.
17. Friedrich, G. & Soriano, P. (1991) Genes Dev. 5, 1513–1523.
18. Rodriguez, C. I., Buchholz, F., Galloway, J., Sequerra, R., Kasper, J., Ayala, R., Stewart,
A. F. & Dymecki, S. M. (2000) Nat. Genet. 25, 139–140.
19. Feil, R., Wagner, J., Metzger, D. & Chambon, P. (1997) Biochem. Biophys. Res. Commun.
20. Nolan, G. P. & Shatzman, A. R. (1998) Curr. Opin. Biotechnol. 9, 447–450.
21. Buess, M., Moroni, C. & Hirsch, H. H. (1997) Nucleic Acids Res. 25, 2233–2235.
22. Sterner-Kock, A., Thorey, I. S., Koli, K., Wempe, F., Otte, J., Bangsow, T., Kuhlmeier, K.,
Kirchner, T., Jin, S., Keski-Oja, J. & von Melchner, H. (2002) Genes Dev. 16, 2264–2273.
23. Kolb, A. F. (2001) Anal. Biochem. 290, 260–271.
24. Lauth, M., Moerl, K., Barski, J. J. & Meyer, M. (2000) Genesis 27, 153–158.
25. Bushman, F. D. (2003) Cell 115, 135–138.
26. Baer, A. & Bode, J. (2001) Curr. Opin. Biotechnol. 12, 473–480.
27. Ramalho-Santos, M., Yoon, S., Matsuzaki, Y., Mulligan, R. C. & Melton, D. A. (2002)
Science 298, 597–600.
28. Ivanova, N. B., Dimos, J. T., Schaniel, C., Hackney, J. A., Moore, K. A. & Lemischka, I. R.
(2002) Science 298, 601–604.
29. Zambrowicz, B. P., Friedrich, G. A., Buxton, E. C., Lilleberg, S. L., Person, C. & Sands, A. T.
(1998) Nature 392, 608–611.
30. Horie, K., Yusa, K., Yae, K., Odajima, J., Fischer, S. E., Keng, V. W., Hayakawa, T., Mizuno,
S., Kondoh, G., Ijiri, T., et al. (2003) Mol. Cell. Biol. 23, 9189–9207.
31. Osipovich, A. B., Singh, A. & Ruley, H. E. (2005) Genome. Res. 15, 428–435.
the trapped intron within genes. The data are based on National Center for
Biotechnology Information (NCBI) mouse genome build 33 and RefSeq
Distribution of gene trap insertions according to the position of
www.pnas.org?cgi?doi?10.1073?pnas.0502273102Schnu ¨tgen et al.