The present invention relates to a novel class of gene trap vectors for efficiently identifying silent or weakly expressed target genes in mammalian genomes, methods of their production and methods for identifying and mutating target genes. The gene trap vectors of the present invention can also be used for inducing the expression of silent genes and enhancing the expression of weakly expressed genes. The use of the enhanced gene trap vectors for creating transgenic organisms to study gene function and to validate pharmaceutical compounds prior to clinical applications is a further aspect of the present invention. The novel gene trap vectors as referred to herein are termed as "enhanced gene trap vectors" (eGTV).

Background of the Invention

In the past few years, a variety of gene trap vectors have been shown as being useful tools for the identification and analysis of permanently or transiently expressed genes. Standard gene trap vectors are DNA or retroviral vectors that insert a promoterless reporter gene into a large number of chromosomal sites. A classic gene trap vector integrates into introns, which are the non-expressed regions of a gene. Introns are flanked by exons, which are the expressed regions of a gene. Transcription of a trapped mammalian gene yields a primary messenger RNA consisting of exon, intron and vector sequences. Primary mRNA processing removes the intron sequences and splices the exons together at specific sites (splice sites) located at the 5' and 3' ends of each exon. As a result, the gene trap vector sequences encoding for the reporter gene become associated with the upstream exons in a processed fusion transcript from which a truncated cellular protein is translated together with the reporter protein.

With the completion of sequencing of the human and mouse genomes, the interest in tools suitable for performing genome-wide mutagenesis has significantly increased. Large scale insertional mutagenesis in mammalian cells has been most effectively induced with conventional gene trap vectors (Hansen, J. et al., Proc. Natl. Acad. Sci. USA 100:9918-22 (2003); Skarnes, W. C. et al., Nat. Genet. 36:543-4 (2004); Wiles, M.V. et al., Nat. Genet. 24: 13-4 (2000); Zambrowicz, B. P.

et al., Proc. Natl. Acad. Sci. USA 100: 14109-14 (2003)). When selecting genes by means of their expression, recombinants will be obtained in which the reporter gene is fused to the regulatory elements of an endogenous gene. Transcripts generated by these gene fusion faithfully reflect the activity of individual cellular genes and serve as molecular tags to identify and/or clone any genes linked to specific functions. Thus, gene trap vectors simultaneously mutate and report on the expression of an endogenous gene at the site of insertion and provide a DNA tag for a rapid identification of the disrupted gene. The application of this technique in a genome-wide manner should allow for the identification of most, if not all, active transcripts in a genome and is thus an important tool for genome annotation. More importantly, gene trapping in mouse embryonic stem (ES) cells enables the establishment of ES cell libraries with mutations in a substantial fraction of genes in the mouse genome, which can be used to produce transgenic mice ²⁴ . Thus, the gene trapping methodology enables the analysis of gene function in the context of an entire organism.

For some years targeted mutagenesis in pluripotent mouse embryonic stem (ES) cells has been used to inactivate genes for which cloned sequences were available (Capecchi, M. R., Trends Genet. 5:70-6 (1989)). Since ES cells can pass mutations induced in vitro to transgenic offspring in vivo, it is possible to analyze the consequences of gene disruptions in the context of entire organisms. As a result, numerous mouse strains with functionally inactivated genes ("knock out mice") have been created by this technology. However, targeted mutagenesis requires detailed knowledge of gene structure and organization as well as its physical isolation in a cloning vector. Overall, the generation of mutant mouse strains by this procedure is still time consuming, labor intensive, expensive and inefficient because it can handle only one gene at the time.

The principal element of a standard gene trap vector is a gene disruption and selection cassette (GDSC) consisting of a promoterless reporter gene and/or selectable marker gene flanked by an upstream 3' splice site (splice acceptor; SA) and a downstream transcriptional termination sequence (polyadenylation sequence; polyA; see Figure 1). The GDSC is inserted into an intron of a target gene and transcription takes place from the upstream target gene promoter. Since the 3' end of the exon upstream of the vector insertion is flanked by a splice donor (SD) site, it is spliced to the GDSC resulting in a fusion transcript in which the upstream

exons of the trapped gene are fused in frame to the reporter and/or selectable marker gene. Due to the presence of a polyA sequence in the GDSC, transcription is terminated prematurely, and, as a result, any exon(s) downstream of the GDSC are not transcribed anymore. Consequently, the processed fusion transcript encodes a truncated form of the target gene, consisting of the upstream exon(s), and the reporter/selectable marker gene.

From the above it becomes apparent that standard gene trap vectors can only disrupt genes that are actively transcribed in the target cell. Genes that are not expressed or expressed only too weakly for detection, i.e. at low expression levels, cannot be recovered by standard gene trapping. This poses a significant problem for genome-wide mutagenesis programs seeking a large scale and cost-effective functional analysis of the ~30,000 mammalian genes. In mouse embryonic stem (ES) cells, for example, only about one half of all genes are expressed, leaving ~ 15,000 genes inaccessible to standard gene trapping. The overall impact of a gene trap resource for elucidating gene function in vivo will thus rest on the fraction of the genome that is accessible with the standard gene trapping technology.

In order to trap genes that are not accessible to standard trapping, gene trap vectors that can be activated independently of gene expression have been developed previously. These vectors are based on a selectable marker gene flanked upstream by a constitutive promoter and downstream by a 5' splice site (splice donor, SD) (Zambrowicz, B. P. et al., Nature 392:608-11 (1998)). These elements are inserted downstream of a standard GDSC such as described above.

An insertion of these standard vectors into an intron of a gene induces splicing of the selectable marker gene, which, in turn, becomes associated with the downstream exon(s) of that gene. As a result, the cells express a fusion transcript initiating at the constitutive promoter and terminating at the polyA site of the trapped gene (= polyA trap). Since the selectable marker gene is expressed independently of the trapped gene's expression, poly-A traps should, at least in principle, enable the recovery of mutations in any gene.

However, there are some major drawbacks with these gene trap vectors and gene trapping methods. Several large scale screening efforts in ES cells with this technology have shown that polyA-containing gene trap vectors generate a high number of false positive recombinants and, more importantly, are not considered to be highly mutagenic (Zambrowicz, B. P. et al., Proc. Natl. Acad. Sci. USA

100: 14109-14 (2003)). So far two main reasons have been cited for their poor performance: (i) the vectors frequently acquire cryptic polyA sites on the non- coding strands of genes, and (ii) selection is biased for gene trap insertions close to the 3' ends of genes, which are frequently non-mutagenic.

From the above it follows that there exists a need for gene trap vectors and gene trapping methods that overcome the above drawbacks, and which are efficient in the identification and mutation of cellular genes that are either not expressed or expressed too weakly to be detected by standard detection methodology. Thus, the provision of a gene trap strategy making most, if not all, genes of a genome accessible to effective trapping in a target cell would be highly desirable.

The problem underlying the present invention can thus be regarded as the provision of a gene trap vector and a gene targeting cassette that allows for the identification of gene products that are normally not expressed or expressed at non-detectable expression levels in a mammalian target cells.The solution provided by the present invention thus concerns a gene trap vector (eGTV) as defined in independent claim 1.

Summary of the Invention

The present invention relates to a novel class of gene trap vectors that are capable of targeting genes independently of their expression.

In a first aspect of the invention, the gene trap vector of the invention comprises a gene disruption and selection cassette (GDSC) and at least one cell type-specific enhancer element that can be placed at any site within the vector, preferably upstream of the GDSC.

In a further aspect of the invention ubiquitous enhancer elements may be used. Such ubiquitous enhancer elements are e.g. the aldolase A enhancer (Moch C. et al., Gene Expr. 6: 1-14 (1996)), the ployoma virus enhancer (Tanimoto K., et al., Nucleic Acids Res. 27:3130-3137 (1999)), the Oct-1 enhancer (Kemler I. et al., Nucleic Acids Res. 19:237-242 (1991)), and the murine adenosine desamidase enhancer (Winston J. H. et al., Somat. Cell MoI. Genet. 22:261-278 (1996)).

In a further aspect the gene trap vector of the invention comprises more than one enhancer element upstream and/or downstream of the GDSC. The GDSC of the

invention comprises from 5' to 3': a splice acceptor sequence, a reporter gene and/or selectable marker gene and a transcription termination site. Preferably the transcription termination site comprises a polyA stretch consisting of adenylic acid (poly A) repeats.

In another aspect the enhancer elements of the present invention are located downstream of a target gene promoter and contain binding sites for transcription activating factors. Preferred enhancer elements of the invention are of the class of responsive elements containing repeat units of specific recognition sites for the corresponding enhancer element-binding partners.

In a further preferred embodiment, the enhancer elements are stretches of nucleic acid sequences of natural or synthetic, viral or non-viral origin that bind transcription activating factors in a sequence-specific manner. Examples are enhancer elements comprising hormone responsive elements, transcription factor binding elements, viral enhancer elements. Enhancer elements of the invention may comprise transcription factor binding sites for AP-I, AP-2, CRE, SRE, NF-kB, SRF, SPl, Octl, Oct2, Oct3, Oct4 transcription factor binding sites. Preferably, the transcription factor binding sites are arranged as tandem repeats.

In a further aspect the enhanced gene trap vector of the invention comprises recombinase recognition elements for introducing GDSC inversions by site-specific recombinases. Examples of such recognition elements are FRT and loxP recombination target sequences.

In a preferred embodiment, the gene trap vector of the invention contains one ore more Oct-4 responsive enhancer elements that are inserted between two homotypic or heterotypic site-specific recombination targets (RTs).

In a preferred embodiment the reporter gene of the GDSC is β-galactosidase and the selectable marker gene is the neomycinphosphotransferase spliced together in a fusion gene.

In a preferred embodiment, the gene disruption and selection cassette (GDSC) and the enhancer element(s) are integrated in a retrovirus or a plasmid.

Preferred embodiments of retroviral gene trap vectors are FlipROSAβgeo, eFlip3ROSAβgeo and eFlipβROSAβgeo comprising a puromycin resistance gene inserted downstream of the GDSC allowing for the quantification of gene trap insertions (WO 01/29208).

In another aspect, the present invention relates to a method for generating a transgenic non-human organism comprising

(i) incorporation of a gene trapping construct of the present invention into a cell of said non-human organism; and

(ii) selection of cells in which said gene trapping construct is incorporated in a gene.

In a preferred embodiment of said method, it is suitable for identifying and/or isolating of a target gene in a non-human organism and comprises the steps:

(i) incorporation of a gene trapping construct according to the present invention in a vertebrate cell;

(ii) selection of cells in which the gene trapping construct is incorporated in a gene; and optionally

(iii) identification and/or isolation of the gene in which said gene trapping construct is incorporated.

In yet another aspect of the present invention, the gene trap vector can be used for mutating a target gene in a mammalian cell. The method for mutating a target gene in a mammalian cell, comprises

(i) transfection/infection of said cell with a gene trapping construct according to the present invention;

(ii) incorporation of the gene trapping construct into the target gene, wherein the incorporation results in a truncated non-functional expression product.

It thus follows that the gene trap vectors of the present invention can be used for detecting, identifying or mutating a functional gene in a cell.

In another aspect, the gene trap vectors of the invention can be used for the generation of a gene trap library comprising gene trap insertions identified by the gene trapping methods of the present invention.

In a further aspect, the gene trap vectors of the invention can be used for targeted gene trapping in combination with homologous recombination.

In another aspect, the gene trap vectors of the invention can be used to create mouse mutant strains which are, among others, useful as models for genetic human disease and for validation of pharmaceutical compounds by monitoring in vivo effects of said compounds.

Description of Figures

Figure 1: Mechanism of eGTV activation. Insertion of enhancer elements binding a cell type specific transcription factor along with a gene disruption and selection cassette (GDSC) into an intron of a silent gene leads to its activation. SA = splice acceptor, pA = polyadenylation sequence.

Figure 2: Enhanced target-directed gene trapping (targeted trapping). A GDSC is inserted together with enhancer elements binding a specific transcription factor into the intron of a silent target gene by homologous recombination. The eGTV activates gene expression and enables drug selection of homologous recombinants. SA = splice acceptor, pA = polyadenylation sequence, HA = homology arms.

Figure 3: The Oct-4 intronic enhancer of the osteopontin gene. (A.) Position of the Oct-4 enhancer in the osteopontin gene. (B.) Sequence of the Oct-4 responsive element (ORE) used in the eGTV vectors. The Sox2 (S) binding sequence has been inactivated by mutagenesis. O = Oct-4 binding sequence, E = engrailed binding sequence.

Figure 4: Conditional gene trap vectors. (Top) classic; (Bottom) enhanced. LTR = long terminal repeat, SA = splice acceptor, βgeo = β-galactosidase/neomycin- phosphotransferase fusion gene, pA = polyadenylation sequence, ORE = Oct-4 responsive elements. Triangles represent the homotypic and heterotypic site specific recombinase recognition targets for the FLPe (frt, F3) and Cre (loxP, Iox511).

Figure 5: Comparative rates of trapping exhibited by enhanced and non-enhanced gene trap vectors. Unique genes were identified by blasting (BLASTN) the GTSTs against the RefSeq database. The number of novel genes among accumulating insertions was dertemined chronologically.

Definitions

"Upstream" refers to nucleotide sequences that precede the codons specifying the mRNA or that precede (are on the 5' side of) the protein coding sequence.

"Downstream" refers to nucleotide sequences that follow the codons specifying the mRNA or that follow (are on the 3' side of) the protein coding sequence.

"Target Gene" defines a specific locus on a chromosome consisting of exons and introns to be trapped by a gene trap vector.

"Transcription activating factor" refers to a compound, e.g. a protein, polypeptide or peptide that has the ability of inducing gene expression by binding to a specific nucleic acid sequence (DNA, RNA).

"Transcription factor" refers to a compound, such as a protein, polypeptide or peptide that binds to one or more transcription factor recognition sites in the proximity of a gene promoter resulting in an activation of mRNA synthesis of said gene.

"Silent promoter" refers to a promoter that is not activated and consequently does not express the gene under control of this promoter.

"Silent gene" refers to a gene that is not expressed.

"Weakly expressed gene" refers to a gene, which is expressed at low expression level, wherein said expression levels are lower than that of a normally expressed gene. Also encompassed by the terms "weakly expressed" are a number of copies of the expression product that is too low to be detected by conventional standard detection methods.

"Gene disruption and selection cassette (GDSC)" refers to genetic elements comprising from 5' to 3' a splice acceptor sequence, a reporter and/or selection gene and a transcription termination site (e.g. poly A repeats).

"Gene trapping" refers to a random mutagenesis approach in functional genomics and is based on the random integration of a gene disruption and selection cassette into a genome.

"Targeted trapping" refers to a gene specific mutagenesis approach in functional genomics and is based on the insertion of a GDSC into the genome by homologous recombination.

"Gene trap vector" refers to a promoterless gene trapping construct consisting of nucleic acid, wherein the gene trapping construct is capable of generating a fusion transcript with the targeted endogenous gene. The presence of splice acceptor elements in the gene trap vector results in the generation of a fusion protein after its insertion into introns.

"Enhanced gene trap vector" (eGTV) refers to a gene trap vector, which contains one or more enhancer elements in addition to a gene disruption and selection cassette (GDSC).

"Reporter gene" refers to a gene encoding for a gene product (e.g. CAT, β galactosidase, βgeo, GFP, EGFP, alkaline phosphatase) that can be readily detected by standard biochemical assays.

"Selectable marker gene" refers to a gene whose expression in a transfected cell allows for the isolation of gene trap vector-expressing cells in drug-containing media (e.g. neomycin, puromycin, diphtheria toxin).

"PoIyA" (A = adenylic acid) refers to a nucleic acid sequence that comprises the AAUAAA consensus sequence, which enables polyadenylation of a processed transcript. In a gene disruption or selection cassette (GDSC), the polyA sequence is located downstream to the reporter and/or selectable marker gene and signals the end of the transcript to the RNA-polymerase.

"Splicing" refers to the process by which non-coding regions (introns) are removed from primary RNA transcripts to produce mature messenger RNA (mRNA) containing only exons.

"5 ¹ splice site" (splice donor SD)" and "3 ¹ splice site" (splice acceptor SA) refer to intron flanking consensus sequences that mark the sites of splicing.

"Enhancer element" or "enhancer" refers to a nucleic acid sequence, which can increase the levels of transcription of a gene into messenger RNA. Typically, an enhancer element functions in either orientation and at various distances from a cellular promoter.

"Responsive enhancer element" refers to an enhancer element, which is specific for a particular transcription factor. When bound by the specific transcription factor, the levels of expression of the gene are enhanced. An example is the Oct-4 responsive element (ORE), which in combination with the transcription factor Oct-4 increases the gene ^' s expression level.

"GDSC inversion" refers to an aberration in which a GDSC segment is deleted, reinserted and turned by 180 degrees from its original orientation, so that the gene sequence for the segment is reversed with respect to that of the rest of the chromosome. Said inversions can by accomplished by using recombinase enzymes (e.g. Cre, FLPe, ΦC31).

"Tandem repeats" refers to copies of genetic elements repeated one after another along a genomic or vector site.

"Homotypic" means being of the same type or form. "Heterotypic" means being of different type or form.

"ROSA" (Reverse-Orientation-Splice-Acceptor) refers to a gene trap cassette inserted into a retroviral backbone in reverse transcriptional orientation relative to the retrovirus (Friedrich, G., Soriano, P., Genes Dev. 5: 1513-1523 (1991)).

Sequence Listing

SEQ ID NO • free text

1 FlipROSAβGeo(int)

2 pBABErfl

3 FlipROSAβGeo

4 eFlip3ROSAβGeo

5 eFlip6ROSAβGeo

6 FlipROSAβGeoPuro

7 eFlip3ROSAβGeoPuro

8 eFlip6ROSAβGeoPuro

9 - 16 primer 11, 13, 12, 14, 16, 18, 15 and 17

17 -18 oligonucleotides P5 and P6

19 - 20 primer P7and P8

21 - 24 primer

25 - 32 oligonucleotides 116, SRl, 115, ISR2, 114, iPCRu3, 113 and iPCRu4

Detailed Description of the Invention

The present invention relates to a novel class of gene trap vectors with the ability to trap silent or weakly expressed genes in a vertebrate genome. More specifically, the gene trap vectors and gene trapping constructs of the invention allow for trapping genes that are expressed at a low expression level, i.e. genes that usually escape detection by standard detection methods because the copy number of their expressed products is too low to be detectable by standard detection methodology (e.g. by utilization of antibodies in Western blotting or ELISA (enzyme-linked immunosorbent assay). This novel class of gene trap vectors will be in the following referred to as "enhanced gene trap vectors" (eGTV).

The invention is based on the finding that the expression of a silent or weakly expressed gene can be induced or enhanced by using a gene trap vector construct

that contains one or more enhancer elements capable of activating the trapped gene ^' s promoter. This, in turn, activates gene trap expression, which enables further analysis of the gene. By using enhancer elements that are responsive to cell type specific transcription factors in gene trap vectors, it is possible to disrupt genes that are not accessible to standard methodology.

Gene trap vectors have been designed containing a standard GDSC in combination with at least one enhancer element that can insert either randomly (gene trapping) or specifically (targeted gene trapping) throughout the genome. It is preferred that the gene trap vectors of the invention integrate in non-expressed sites of the genome (introns), i.e. between the expressed regions (exons) of a gene. Following transcriptional activation of the trapped gene by means of the inserted enhancer elements, a fusion transcript is generated between the upstream exons and the GDSC resulting in a selectable mutation. To determine the identity of the trapped gene, the fusion transcript can be reverse transcribed, amplified by PCR and subsequently sequenced.

Conventional gene trap vectors and classic gene targeting methods require gene expression for the successful mutagenesis of target genes. However, genes that are not actively transcribed cannot be disrupted by these methods. Depending on the cell type, between 50 - 80% of all genes are not transcribed under natural conditions because the cell does either not require the gene product for survival and proliferation, or because the cell has ceased to differentiate any further.

We therefore sought to provide a gene trap vector with the ability to either induce gene expression or to enhance gene expression of a target gene. To achieve this, the gene trap vectors of the invention contain at least one enhancer element serving as a binding site for transcription activating molecules that, when bound, are able to turn on the transcription of the trapped gene.

The gene trap vectors of the invention are equipped with cell type specific enhancer elements that are placed into the target genes upon vector insertion. It is preferred that one, two or more enhancer elements are arranged within the gene trap vector depending on the degree of activation and/or class of genes to be analyzed. The enhancer elements can be placed at various distances from the promoter of the gene of interest, preferably they are arranged downstream of the promoter region. In preferred embodiments, the enhancer elements are of natural or synthetic, viral or non-viral origin and bind transcriptional activators. In yet another preferred

embodiment of the invention, the enhancer elements comprise responsive elements such as hormone responsive elements, transcription factor binding elements, and viral enhancer elements. The specific selection of the enhancer elements will depend on the cell type and class of genes to be analyzed. In a further preferred embodiment, the enhancer elements comprise binding sites of the following transcription factors: AP-I, AP-2, CRE, SRE, NF-kB, SRF, SPl, Octl, Oct2, Oct3, Oct4 (Nakabeppu, Y. et al., Cell 55:907-15 (1988); Bosher, J. M. et al., Oncogene 13: 1707-7 (1996); Gotquin, V. et al., Genes Dev. 12 2073-90 (1998); Scholer, H. R. et al., Nature 344:435-9 (1990)). Preferably, the transcription factor binding sites are arranged as tandem repeats.

The binding of a transcription factor to its cognate enhancer element induces the expression of silent genes and enhances the expression of weakly expressed genes. By activating gene expression, the enhancer-bound transcription factors also activate the inserted enhanced gene trap vector (eGTV) of the invention in much the same way than standard gene trap cassettes are activated, however, with the advantages referred to below. Thus, by using the enhanced gene trap vectors of the invention, it is possible to identify and select genes that would normally escape such a selection (see Figure 1).

The enhanced gene trap vectors of the present invention are not only suitable for random mutagenesis but can also be used for targeted gene trapping involving the introduction of a GDSC along with one or more specific enhancer elements into a silent gene by homologous recombination (see Figure 2). The activation of a target gene by specifically binding a transcription factor simultaneously activates the GDSC, and thus enables the recovery of homologous recombinants by drug selection. The number of genes accessible to trapping in a particular target cell is thereby increased above the number of genes accessible to standard trapping (see examples).

In a preferred embodiment, these enhancer elements include intronic Oct-4 enhancer elements derived from the osteopontin gene (see Figure 3). Oct-4 is a transcription factor, which is highly and specifically expressed in undifferentiated embryonic stem (ES) cells (Scholer, H. R. et al., Nature 344:435-9 (1990)). The Oct-4 transcription factor specifically recognizes and binds to the conserved octamer motif - ATGCAAAT-, which is present in Oct-4 enhancers (see Figure 3). In ES cells, insertions of a gene trap vector of the invention comprising Oct-4

responsive enhancer elements (OREs) into the introns of silent or weakly expressed genes will activate these genes by binding the Oct-4 transcription factor (see examples). This, in turn, induces GDSC expression, which enables selection for gene trap events.

It is preferred that the reporter gene of the invention is β-galactosidase, βgeo or a gene encoding for a luminescent gene, e.g. a fluorescent gene such as green fluorescent protein (GFP) or firefly luciferase. It is further preferred that the selectable marker gene encodes an anti-antibiotic such as neomycin- and hygromycin-phosphotransferases or puromycin-acetyl- transferase.

In another aspect of the invention, the gene trap vectors comprise target sequences for site-specific recombinases to enable postinsertional GDSC inversions and/or excisions. Thus, in an embodiment the enhanced gene trap vectors of this aspect of the invention contains: (i) a GDSC consisting of a β-galactosidase- neomycinphosphotransferase fusion gene (βgeo) flanked by SA- and polyA- sequences for gene trapping, (ii) elements enabling GDSC inversions by site specific recombinases for conditional mutagenesis, and (iii) one or more Oct-4 responsive enhancer elements (ORE), preferably inserted between two heterotypic site-specific recombination targets (RTs) (see Figure 3).

Any suitable recombinase can be used for inversion, such as Cre recombinase, FIp recombinase or ΦC31 recombinase.

It is preferred that the gene trap vectors of the invention are contained in a retrovirus or a plasmid. These gene trapping constructs contain in addition to a GDSC at least one enhancer element for transcriptional activation of an upstream or downstream gene promoter. Retroviruses insert a single copy of the reporter and/or selectable marker cassette per locus, with no rearrangement of flanking sequences. In particular, retroviruses have a preference for insertions at the 5' ends of genes.

The gene trap vectors of the invention are preferably retroviral gene trap vectors selected from the group consisting of FlipROSAβgeo, eFlip3ROSAβgeo and eFlipβROSAβgeo (WO 01/29208). These vectors preferably comprise a puromycin resistance gene inserted downstream of the GDSC allowing for the quantification of most if not all gene trap insertions.

The above specified retroviral or plasmid based gene trap vectors are suitable for disrupting both silent and expressed genes across any mammalian genome (e.g. human, mouse). For example, by using the gene trap vectors of the invention in a genome wide manner, a large collection of embryonic stem (ES) cell lines harboring gene trap insertions in single genes can be assembled and used to make mutant mice. In particular, for pharmaceutical research seeking to validate the utility of specific genes and their products as targets for drug development, mutant mice are excellent genetic tools.

In another aspect, the invention relates to a method for the identification of a functional yet silent gene in mammalian cells. The method comprises the transduction of cells with an enhanced gene trap vector as described herein and the incorporation of the reporter gene and/or selectable marker cassette into genomic sites. If the vector inserts into the intron of silent genes, gene trap vector-induced promoter activation results in GDSC expression, which in turn enables cell selection. Preferably, the disrupted gene is identified by RT-PCR (RACE) or PCR (PCR = polymerase chain reaction; RT = reverse transcription) (Hansen, J. et al., Proc. Natl. Acad. Sci. USA 100:9918-22 (2003); von Melchner et al., Proc. Natl. Acad. Sci. USA 87:3733-7 (1990)).

In a preferred embodiment, the method of the present invention is adapted to reach a cost-effective saturation of the genome with insertional mutations in the fastest possible way. The steps involved in this method are: (i) transducing a large number of embryonic stem (ES) cells with enhanced gene trap vectors (eGTV) of the invention, (ii) selecting eGTV-expressing clones and establishing cell lines from them, (iii) creating an ES cell bank containing ES cell lines with mutations in single genes, (iv) amplifying genomic sequences adjacent to the eGTV insertion by PCR and/or sequences appended to eGTV transcripts by RT-PCR (RACE) from the ES cell lines, (v) sequencing the amplification products to obtain cell line specific gene trap sequence tags (GTSTs, "flank bank"), (vi) identifying and cataloguing the disrupted genes by GTST homology searches in the public databases, (vii) making mutant strains of mice using ES cells from the ES cell bank.

The invention further encompasses a method for mutating a functional gene within the genome. The mutation is introduced by incorporating a gene trap vector of the invention into intronic sites of a gene. As a result of the splicing process (see supra), a fusion product will be obtained comprising one or more exons of the gene

and the reporter/selectable marker gene cassette. Due to the fact that transcription terminates at the polyA site, the downstream exons of the gene will not be part of the fusion product. Therefore, the mRNA of the gene is not complete but truncated since every exon that follows the gene trap vector will not be expressed.

A further aspect of the invention relates to a method for producing both "null" and

"conditional" mutations in genes of an organism regardless of whether the genes are expressed or not. In a preferred embodiment the method comprises the following steps:

(i) incorporation of a gene trapping vector construct in a suitable cell;

(ii) selection of cells having the vector construct incorporated in a gene by expression analysis of the selectable marker; (iii) identification and/or isolation of the gene in which the construct is incorporated by PCR or Reverse Transcriptase (RT)-PCR.

The gene trap vectors of the invention and their uses allow both random or targeted mutagenesis in mammalian cells. For a targeted-insertional mutagenesis (= targeted gene trapping), sequence information of the gene of interest is required for specific integration, whereas in a random insertional mutagenesis the integration occurs at non-specific sites. A method for targeted-insertional mutagenesis using enhanced targeted gene trapping constructs of the present invention comprises the following steps: (i) design of a gene trapping construct comprising the enhancer elements and the GDSC from an enhanced gene trap vector flanked by sequences homologous to the an intron preferably flanking a 5' exon of the specific target gene (= homology arms; see Figure 2); (ii) transduction of the gene trapping construct into a suitable cell; (iii) isolation of homologous recombinants by selecting for GDSC expression; (iv) verification of homologous recombination in the selected cells by a suitable detection method, preferably 5' RACE, genomic PCR or Southern blot analysis.

Taken together the enhanced gene trap vectors (eGTV) of the invention and the methods employing them are suitable to solve the drawbacks of conventional gene trap- and gene targeting vectors currently used in the field. In particular, it shows that the eGTV have the capability to

(i) induce mutations in genes regardless of their expression, including genes for which cloned sequences are not available; (ii) increase the number of genes accessible to trapping and targeted trapping in comparison to the number presently accessible by standard gene trap and targeted trapping vectors; (iii) increase the gene trapping rate by reducing redundancy, and thereby saving costs.

In its application to pharmaceutical research, the eGTV technology greatly assists the creation of a library of ES cell clones, which includes all genes of the mouse genome (~30,000), modified by the insertion of an eGTV vector. Sequence analysis from the modified alleles allows, as described above, to identify the genes modified in individual clones, resulting in a database. This collection is the resource required to quickly generate a large number of mutant mouse strains for pharmaceutical research. Such mutant strains are an optimal tool to study the function of genes in the mouse as a mammalian model organism for human disease. The evaluation of gene function, in particular in mouse strains prone to develop a disease, allows to validate the utility of an individual gene for pharmaceutical drug development. For example, a given number of genes could be thought to be involved into insulin receptor signaling as a therapeutic target for Diabetes treatment. The generation and physiological analysis of mouse mutants for these genes would identify which gene products play a role in insulin signaling. Subsequently, drug development could be focused only on such "validated targets". Thus, within pharmaceutical drug development, mutants serve as a validation instrument to identify useful target molecules.

In addition, a subset of mutants could develop diseases, which mimic known human disorders and serve as a model for their treatment.

Once validated targets are available, the mutants can be used in combination with gene expression profiling to determine "on" and "off" target effects of candidate drugs.

Since the use of the gene trap technology is not restricted to murine ES cells and mouse mutants its application can be extended to any other vertebrate or invertebrate model organism (e.g. rats, zebra fish, Drosophila) to characterize the biological function of selected genes. As described above for the mouse, such

mutants could be used for the validation of target genes for pharmaceutical drug development.

In another application the gene trap technology can be used to validate the utility of selected genes of plants for agricultural purposes. Plant mutants can be used to identify valuable target genes for herbicide development as well as to identify genes involved into the fertility of economically useful species.

As further exemplified in the examples below, the eGTV vectors of the invention do not only allow the identification of expressed genes in a cell but also the identification of yet unidentified, silent or poorly expressed genes. In summary, eGTVs provide superior tools for the field of genomics and functional genetical analysis.

In the following examples, material and methods of the present invention are provided. It should be understood that these examples are for illustrative purpose only and are not to be construed as limiting this invention in any manner. All publications, patents, and patent applications cited herein are hereby incorporated by reference in their entity for all purposes.

Examples

Example 1

To test the trapping efficiency of enhanced gene trap vectors, the retroviral gene trap vectors FlipROSAβgeo, eFlip3ROSAβgeo and eFlipβROSAβgeo (Figure 4) were equipped with a puromycin resistance gene, which is independently transcribed from a pgk promoter. This element, inserted downstream of a β-galactosidase (LacZ) -neomycinphosphotransferase fusion gene (βgeo) flanked by SA- and polyA- sequences enables the quantification of all gene trap insertions across the genome of a target cell. Enhanced gene trap vectors contain in addition to a GDSC either 3 or 6 Oct-4 responsive elements (OREs) arranged in tandem repeats (see Figure 4).

A. Construction of the gene trap vectors FlipROSAβqeo, eFlip3ROSAβqeo and eFlip6ROSAβqeo.

Construction of the gene trap vector FlipROSAβgeo: For the construction of

FlipROSAβgeo an overlap extension PCR strategy was chosen using the oligonucleotides Il (5'- CGC CTC CTC TTC CTC CAT CC-3'; SEQ ID NO:9) and 13

(5'- ACT CTT CCG CTT CCT CGC TCC ACC GCG GCT TCG AGA CCG T-3'; SEQ ID

NO: 10) for amplification of the 5' f3-frt recombinase target (RT) sequences from FlipROSAβgeo(int) (see SEQ ID NO: 1) and the oligonucleotides 12 (5'- GGG CCT CTT CGC TAT TAC GC-3'; SEQ ID NO: 11) and 14 (5'- ACG GTC TCG AAG CCG CGG TGG AGC GAG GAA GCG GAA GAG T-3'; SEQ ID NO: 12) for amplification of the 5' lox511-loxP RTs from pFIEx; Schnutgen, F. et al., Nat. Biotechnol. (2003)). The two PCR products were purified, annealed and re-amplified using the oligonucleotides Il and 12. The resulting PCR product was cloned into the BamHI site of the retroviral vector pBABESrfl (modified from pBABE puro; see SEQ ID NO:2) (Morgenstern, J. P., Land, H., Nucl. Acids. Res. 18:3587-3596 (1990)) to obtain -pBLF5'-. Similarly, the 3' RTs were amplified with the oligonucleotides 16 (5'-GCT CCT CGC CCT TGC TCA CC-3'; SEQ ID NO: 13) and 18 (5'-TAG AAG CGG TTT TCG GGA GAA TAC GAC TCA CTA TAG GGC G-3'; SEQ ID NO: 14) for loxP-lox511 and with the oligonucleotides 15 (5'-TGC TGG CCT TTT GCT CAC AT-3'; SEQ ID NO: 15) and 17 (5'-CGC CCT ATA GTG AGT CGT ATT CTC CCG AAA ACC GCT TCT A-3'; SEQ ID NO: 16) for the frt-f3. The two products were purified, annealed and re-amplified with the oligonucleotides 16 and 17. The resulting PCR product containing all four RTs was cloned into the EcoRI restriction site of pBLF5' to obtain pBLF. The GDSC (SA-βgeo-pA) was recovered as an Xhol restriction fragment from ROSAβgeo (Friedrich, G., Soriano, P., Genes. Dev. 5: 1513-1523 (1991)) and cloned into the SnaBI restriction site of pBLF to obtain the final gene trap vector -FlipROSAβgeo-. The final vector was verified by DNA sequencing (see SEQ ID NO:3). Construction of the gene trap vector eFlip3ROSAβgeo: Oct 4 responsive elements (ORE) from the osteopontin gene (Botquin, V. et al., Genes Dev. 12:2073-90 (1998) were obtained by annealing the oligonucleotides P5 (5'-GAT CCT GCA CTG ACC TTT CAG CTT TGT ATA ATG TAA GTT AAA ATC ACA TTT GAA ATG CAA ATG GAA AAG CA-3'; SEQ ID NO: 17) and P6 (5'-GAT CTG CTT TTC CAT TTG CAT TTC AAA TGT GAT TTT AAC TTA CAT TAT ACA AAG CTG AAA GGT CAG TGC AG-3'; SEQ ID NO: 18) and cloning into the BamHI / BgIII sites of pEGFP-Nl resulting in pEGFP- 01. pEGFP-02 to pEGFP-06 was obtained by subsequent ligation of additional to pEGFP-01. 3-6 OREs obtained from pEGFP-03 and pEGFP-06 were cloned as BamHI/Bglll fragments into the BgIII site located in the spacer between the frt and the f3 sites of FlipRosaβgeo to obtain eFlip3ROSAβgeo and eFlip6ROSAβgeo. The primary structure of the final vectors was confirmed by DNA sequencing (SEQ ID NOs:4 and 5).

B. Construction of the gene trap vectors FlipROSAβqeoPuro, eFlip3ROSA-βqeoPuro and eFlip6ROSAβqeoPuro

Construction of the gene trap vector FlipROSAβgeoPuro: The puromycin cDNA was obtained from pBabePuro (Morgenstern, J. P., Land, H., Nucl. Acids. Res. 18:3587- 3596 (1990)) by PCR using the primers P7 (5'- GGG GGC TGC AGA CTT ACA GCG GAT CCC CTC AGG CAC CGG GCT TGC-3'; SEQ ID NO: 19) and P8 (5'-GGG GGC TGC AGC CAA TAT GAC CGA GTA CAA GCC CAC-3'; SEQ ID NO: 20). The puromycin cDNA was then used to replace the neomycin resistance gene of expression plasmid pD383 (Zazopoulos, E. et al., Nature 390:311-5 (1997)), obtain pD383/puro. The pgk promoter-puro-polyA cassette of pD383/puro was cloned as a Xhol restriction fragment into a unique Sad site of FlipROSAβgeo, eFlip3ROSAbgeo and eFlip6ROSAbgeo to obtain of FlipROSAβgeoPuro, eFlip3ROSAβgeoPuro and eFlip6ROSAβgeoPuro, respectively (SEQ ID NOs:6-8).

C. Isolation of puromycin resistant clones and quantification of gene trap events by LacZ staining.

Generation of retroviral particles: Virus producer cells (4 x 10 ⁵ ) (Kinsella, T.M., Nolan, G. P., Hum. Gene Ther. 7: 1405-13 (1996)) were seeded onto P90 dishes and grown in DMEM (high glucose) supplemented with 10 % FCS, 20 mM glutamine, 1 x non-essential amino acids, and 0.1 mM β-mercaptoethanol. After incubating for 3 days, the cell were transfected with the retroviral constructs using Lipofectamin 2000 (Invitrogen) and the manufacturers instructions. 48 hours later, cells were overlaid with 5 ml or ES cell medium (see below) and virus particle containing supernatants were harvested after incubating for 4.5 hours. Supernatants were filtered through a 0.45 μm Millipore filter, supplemented with 5 μg/ml Poybrene and stored at -80 ⁰ C until use.

ES cell cultures: 129/Sv/C57BL6 Fl-ES-cells were grown in DMEM (high glucose) supplemented with 15 % FCS, 2 mM glutamine, 20 mM HEPES, 1 mM sodium pyruvate, 1 x non-essential amino acids, 0.1 mM mercapto-ethanol, and 1500 U/ml leukemia inhibitory factor (LIF) (Chemicon).

Infection of mouse embryonic stem cells: 1 x 10 ⁵ ES cells were seeded onto gelatinized P60 Petri dishes and allowed to attach overnight. Cells were then exposed to 2 ml virus containing supernatant for 4.5 h. After adding 2 ml of fresh ES cell medium and incubating overnight, cells were put in selection for 12 days using ES cell medium containing either 0.6 μg/ml puromycin.

X-GaI staining: After washing in PBS, cells were fixed in 3 % formaldehyde and incubated overnight in lacZ staining buffer (5 mM K ₃ [Fe(CN ₆ )], 5 mM K ₄ [Fe(CN) ₆ ], 2 mM MgCI ₂ in PBS) and 1.25 ml X-GaI solution (40 mg/ml in dimethylformamide).

D. Results: Since retroviruses integrate mostly randomly throughout the genome, only a small fraction of the vectors will produce a gene trap event by inserting into an expressed gene. Gene trap events induce β-galactosidase (LacZ) expression, which can be visualized by staining with X-GaI. Consistent with previous observations, less than 4% of all FlipROSAβgeoPuro insertions (= number of puromycin resistant clones) generated a gene trap event (= number of LacZ+ clones) (Table 1). In contrast, over 30% of the eFlipβROSAβgeoPuro insertions generated a gene trap event, suggesting an ORE/Oct-4 mediated induction of gene expression at the insertion site (Table 1). Since activation can involve both expressed and non-expressed genes, the vectors of the invention will (i) improve the efficiency of trapping by decreasing the threshold for reporter protein detection, and (ii) increase the fraction of genes accessible to trapping by including silent genes.

Table 1: Induction of gene expression by enhanced gene trap vec :tors*

*F1-ES cells were infected with retrovirus particle containing supernatants from FNXEco producer cells after transiently transfecting the gene trap plasmids. After

selecting in 0.8 μg/ml puromycin for 10 days, resistant clone were stained with X- GaI and counted. Results are from duplicate plates of 3 independent experiments are shown.

Example 2:

To determine whether enhanced gene trap vectors trap genes more efficiently, we produced several hundreds of ES cell clones with each of the conditional gene trap vectors - FlipRosaβgeo and eFlipβRosaβgeo - (Figure 4). After isolating the GTSTs by 5' RACE, we determined the number of genes trapped by each vector that had not been trapped before with the collection of standard gene trap vectors used by the German Gene Trap Consortium (GGTC).

A. Isolation of mutant ES cell lines expressing FlipRosaβqeo and eFlip6ROSAβqeo: eFlip6ROSAbgeo gene trap virus was produced as described in Example IB for the FlipROSAβgeoPuro class of vectors. ES cells were infected with the virus containing supernatants at an M.O.I. < 0.5 as described in Example IB. Gene trap expressing ES-cell lines were selected in 130 μg/ml G418 (Invitrogen), manually picked, expanded, and stored frozen in liquid nitrogen until use.

B. Recovery of GTSTs by 5' RACE and database analysis: cDNAs were prepared from the polyadenylated RNA using a RoboAmp robotic device (MWG Biotech, Ebersberg, Germany) with a processing capacity of 96 samples/day. Samples of 2xlO ⁵ cells were lysed in 1 ml of lysis buffer containing 100 mM Tris/HCI pH 8.0, 500 mM LiCI, 10 mM EDTA, 1% LiDS and 5 mM DTT. Polyadenylated RNA was captured from the lysates by biotin-labeled oligo-d(T)-primers according to the manufacturers instructions (Roche Diagnostics Corp., Indianapolis, IN, USA) and placed on streptavidin-coated 96-well plates (AB Gene, Surrey, UK). After washing, solid-phase cDNA synthesis was performed in-situ using random hexamers and Superscript II RT (Invitrogen, Karlsruhe, Germany). To remove excess primers the cDNAs were filtered through Multiscreen PCR plates (Millipore Corp. Bedford, MA, USA). The 5 ^' ends of the purified cDNAs were tailed with dCTPs using terminal transferase -TdT- (Invitrogen, Karlsruhe, Germany) following the manufacturer's instructions.

For PCR amplification of GTSTs, the following vector-specific primers were used: 5 ^' -CTA CTA CTA CTA GGC CAC GCG TCG ACT AGT ACG GGI IGG GII GGG IIG-3 ^'

(SEQ ID NO:21) and 5 ^' -GCC AGG GTT TTC CCA GTC ACG A-3 ^' (SEQ ID NO:22); 5 ^' -CTA CTA CTA CTA GGC CAC GCG TCG ACT AGT AC-3 ^' (SEQ ID NO: 23) and 5 ^' - TGT AAA ACG ACG GCC AGT GTG AAG GCT GTG CGA GGC CG-3 ^' (SEQ ID NO:24) (nested). Amplification products were directly sequenced using AB377 or ABI3700 sequencing machines (Applied Biosystems ABI, Foster City, USA). After filtering sequences against repeats and removing all vector sequences from the GTSTs, a PHRED score was assigned to each individual nucleotide. GTSTs qualified as informative if they were at least 50 nucleotides long and exhibited a minimum mean PHRED score of 20 (Figure 1, supplementary information). Homology searches were performed using the publicly available sequence databases and the BlastN algorithm. Databases included GenBank, UniGene, OMIM (all at http://www.ncbi.nlm.nih.gov), ENSEMBL (http://www.ensembl.org), RIKEN (http://www.rarf.riken.go.jp) and GeneOntology (http://www.geneontology.org).

C. Results: GGTCs library is presently the largest public library of ES cell lines with gene trap insertion in single genes. It presently contains 13,616 GTSTs (www.genetrap.de) corresponding to 3,349 unique genes. Table 2 shows, that 15% of the genes trapped by eFlipRosaβgeo were not present in the database, i.e. had not been trapped before, which is almost twice as much the number of novel genes trapped by standard FlipRosaβgeo. This suggests that in ES cells the genomic target accessible to enhanced gene trap vectors is significantly larger.

Table 2: Trapping efficiency of novel genes by enhanced gene trap vectors as estimated by 5 ¹ RACE*

*GTSTs recovered from FlipRosaβgeo and eFlip6Rosaβgeo expressing ES cell clones were aligned to the GGTC-GTST database using the BlastN algorithm

This conclusion was re-enforced by determining the number of sentinel genes trapped by the two vectors that were not present in any gene trap resource, including the International Gene Trap Consortium's (IGTC) database with 27,000 GTSTs and Lexicon Genetics' Omnibank with 200,000 GTSTs (all available at NCBI). "Sentinel" genes are fully genome annotated genes, which are presently at 7,984

(Skarnes, W.C. et al., Nat. Genet. 24: 13-4 (2000)). Table 3 shows that eFlipRosaβgeo trapped sentinel genes not present in both resources about 40% more efficiently than FlipRosaβgeo.

Table 3: Trapping efficiency of novel sentinel genes by enhanced gene trap vectors*

* GTSTs recovered from FlipRosaβgeo and eFlip6Rosaβgeo expressing ES cell clones corresponding to sentinel genes were aligned to the IGTCs and Lexicon's databases using the BlastN algorithm.

Example 3:

The large scale recovery of GTSTs from mutant ES cell lines produced with conventional gene trap vectors relies on the PCR amplification of fusion transcripts using semiautomatic 5' RACE. The method, while generally robust, is dependent on transcript levels, which if too low preclude PCR amplification. In contrast to conventional gene trap expressing clones for which over 80% of RT-PCR amplifications were successful, less than 50% of the eGTV expressing clones gave meaningful 5' RACE amplification products, suggesting that the eGTV insertions occurred into either silent or weakly expressed genes. To test this hypothesis, we subjected 51 ES cell lines that failed to generate amplification products to genomic (inverse) PCR.

A. Recovery of GTSTs by inverse PCR: Genomic DNA was isolated using the DNeasy kit of Quiagen according to the manufacturers protocol. DNA was eluted in 150 μl. Approximately 3 μg genomic DNA were digested in 100 μl with 20 u Nspl at 37°C overnight. Digested DNA was purified using the Qiaquick kit according to the manufacturers protocol, ligated in 300 μl at 16°C overnight and again purified using the Qiaquick kit.

5' inverse PCRs were carried out using the oligonucleotides 116 (5'-CGA GCC CCA GCT GGT TCT TTC-3'; SEQ ID NO:25) and SRl (5'-GCT AGC TTG CCA AAC CTA CAG GTG G-3'; SEQ ID NO:26). Nested PCR was carried out using the oligonucleotides

115 (5'-GTC TCA GAA GCC ATA GAG CCC-3'; SEQ ID NO:27) and SR2 (5'-GCC AAA CCT ACA GGT GGG GTC TTT-3'; SEQ ID NO:28). 3' inverse PCR was carried out using the oligonucleotides 114 (5'-ACT ATC CCG ACC GCC TTA CTG C-3'; SEQ ID NO:29) and iPCRu3 (5'-CCT CCG ATT GAC TGA GTC GCC C-3'; SEQ ID NO:30). Nested PCR was carried out using the oligonucleotides 113 (5'-TGT TTT GAC CGC TGG GAT CTG C-3'; SEQ ID NO:31) and iPCRu4 (5'-TAC CCG TGT ATC CAA TAA ACC C-3'; SEQ ID NO:32).

B. Results: Sequencing of the amplification products showed that of 25 eGTV insertions in annotated genes, 15 (60%) were novel and not present in GGTCs database, suggesting that the novel genes were either silent or poorly transcribed prior to insertion.

Table 3: eFlipβRosabgeo insertions into annotated genes identified by inverse PCR

*refers to genes not present in the GGTC database.

Example 4

This example describes the comparative rate of trapping of the FlipROSAβgeo and eFlipβROSAβgeo gene trap vectors. Figure 5 shows that at average eFlipβROSAβgeo traps a novel gene in with every 1.4 insertions. In contrast, the non-enhanced FlipROSAβgeo requires 2.8 insertions to trap a novel gene, clearly indicating that the enhancer elements in the eFlipβROSAβgeo gene trap vector significantly improve the efficiency of trapping.

Example 5

This example describes another enhanced gene trap vector of the invention containing the cytomegalo-virus (CMV) immediate early enhancer which is essentially ubiquitous. This enhancer has been shown to activate gene expression in embryonic stem cells (Chung, S. et al., Stem Cells 20: 139-45 (2002)) and in all mouse tissues in either orientation and up to a distance of several kbp from the promoter (Dorsch-Hasler, K. et al., Proc. Natl. Acad. Sci. USA 82:8325-9 (1985)). The CMV enhancer has been used in combination with the chicken β-actin promoter to drive the expression of transgenes in the mouse (Rodriguez, C.I. et al., Nat. Genet. 25: 139-40 (2000); Zong H. et al., Cell 121:479-492 (2005); Okabe M. et al., FEBS Lett. 407:313-319 (1997)).

Example 6

This example describes another enhanced gene trap vector of the invention containing the enhancer from the mouse embryonic stem cell virus. This virus is a synthetic retrovirus derived from a mutant myeloproliferative sarcoma virus (PCMV). The enhancer element of this virus is demonstrated activity in embryonic carcinoma cells as well as embryonic stem cells ²¹ .

Example 7

This example described the use of trapped ES cell lines for making mutant mice. ES-cell derived chimeras were generated by injecting C57BI/6 blastocysts with ES cells from the following trapped lines as obtained in Example 2: P015F03 P016F03, P023A01, P023F01, Q001D04, and Q016D06. Male chimeras were obtained with each clone and were bred to C57BI/6 females. Litters were analyzed for germline transmission using the agouti coat color marker and tail blotting. So far, the clones P015F03 and P016F03 generated transmitted the mutation to the Fl generation. Fl mice were intercrossed to obtain homozygous (mutant) F2 offspring for phenotype analysis.