The present invention relates to a model for tumourigenesis. In particular, the invention relates to an animal model for leukaemia which is based on a chromosomal translocation which creates a fusion from the Ews and ERG genes or the MLL and AF4 genes.

Introduction

The aetiology of tumours often involves chromosomal translocations in haematopoietic malignancies, sarcomas and epithelial tumours and these can either activate proto-oncogenes or create novel fusion genes which function in tumourigenesis (Rabbitts, 1994). These translocation products act at various levels in afflicted cells, although the translocation genes always encode intracellular proteins located in either the cytoplasm or nuclei. The recurrent chromosomal translocations in haematopoietic malignancies and in sarcomas generally display a tropism in respect of the cell type in which they are found, for instance the MLL gene fusion sub-types being restricted to either lymphoid or myeloid malignancies (Downing and Shannon, 2002; Look, 1997; Rowley, 1998). The MLL gene is a frequent site of chromosomal translocations in human leukaemias (Corral et al., 1993; Djabali et al., 1992; Domer et al., 1993; Gu et al., 1992; Thirman et al., 1993; Tkachuk et al., 1992), affecting over 30 different chromosomes and resulting in many leukaemia-associated MLL-gene fusions (Ayton and Cleary, 2001; Collins and Rabbitts, 2002). Different fusions distinguish distinct leukaemias, for instance the MLL-AF9 fusion is almost exclusively in myeloid leukaemias while MLL-AF4 is confined to B-lymphoid leukaemias (Gu et al., 1992; Rowley, 1998). Others, such as MLL-ENL can be found in myeloid and lymphoid leukaemias (Ayton and Cleary, 2001).

Mouse models which mimic the effects of chromosomal translocations, or which recapitulate these events, are essential to understand the determinants of frank malignancy and those dictating the restriction of translocations to specific lineages. In addition, with advancement of efforts to devise new reagents which can interfere with protein function, mouse models of human cancer are crucial to develop and test these rationally devised therapeutic approaches, prior to application in patients. Finally, mouse models of human cancer will be needed for in vivo testing of new compounds, which may have promising biochemical properties, in order to accelerate the translation of these into clinical use.

Mouse cancer models involving chromosomal translocation-genes have been either gain of function transgenic models (Adams et al., 1999; Pandolfi, 1998), retroviral transduction models or translocation mimics in which a gene fusion is created by homologous recombination into an endogenous target gene using embryonic stem cells (ES cells) (Corral et al., 1996). For instance, the consequences of MLL-associated translocations have been modelled in mice using retroviral transduction of bone marrow progenitors (Lavau et al., 1997; Schulte et al., 2002; Slany et al., 1998) or by homologous recombination knock-in of fusion genes (Corral et al., 1996; Dobson et al., 1999; Dobson et al., 2000). In the knock-in approach, homologous recombination was used to introduce fusion oncogenes into mice, via embryonic stem cells. In this way, a mouse model of the human MLL-AF9 gene fusion was made in which knock-in mice were found to develop an acute myeloid malignancy similar to that found in human patients with the chromosomal translocation t(9; 11) (Corral et al., 1996; Dobson et al., 1999). While the knock-in approach is a close mimic of natural chromosomal translocations, it is limited by the fact that the homologous recombination event generates a knock-in of one allele which is subsequently transferred to all cells of the mouse. Thus the fusion gene will be expressed in all sites where the targeted gene is expressed, for instance as found for the MII-AF9 knock-in fusion (Corral et al., 1996). hi addition, dominant, embryonic lethal effects of knock- in alleles have been observed (Okuda et al., 1998; Yergeau et al., 1997). These can be overcome by creating conditional knock-in alleles such as the Amll-Eto example, which circumvents the embryonic lethal effects of the fusion, and causes AML to arise in mice (Higuchi et al., 2002). The most ideal concept is the de novo creation of chromosomal translocations which could potentially be achieved by in vivo recombination systems, such as the Cτe-loxP system (Smith et al., 1995; van Deursen et al., 1995).

SSRs of the integrase class such as FLP, from the 2μ episome of S. cereviseae, Cre, from the E. coli Pl phage, and R from Z. rouxii are able to mediate precise and conservative recombination between their binding sites. These proteins share overall similarity in their primary sequence and attract attention because they have been shown to function efficiently in a broad range of organisms from E.coli to mice. Their binding sites are 13bp inverted repeats which are sequence specifically recognised by a monomer per half site. Other SSRs having homologous structures are known.

The use of SSRs is potentially the nearest that could be achieved of a direct recapitulation of the de novo occurrence of human cancer-associated chromosomal translocations. The Cτe-loxP recombination system has been shown to facilitate de novo chromosomal translocations in mice (Buchholz et al., 2000; Collins et al., 2000) but no tumour incidence was reported in these initial translocator mice.

Human cancer originates in somatic cells acquiring mutations, either point mutations or more gross changes manifest by chromosomal translocations, inversions or deletions i. All tumors have some or all of these changes. Chromosomal translocations are amongst the best studied since they are usually visible by microscopy, facilitating the molecular characterisation of the breakpoint sequences and of the genes (oncogenes) involved 2,3, Differences occur in the nature of the consistent chromosomal translocations in haematopoietic (leukemias and lymphomas) or mesenchymal tumors (sarcomas) and epithelial tumors (carcinomas) since the former two tend to have reciprocal translocations whereas the latter often have non-reciprocal translocations 4. The main outcomes of the reciprocal chromosomal translocations are either forced oncogene expression as found in lymphoid malignancies or gene fusion found in both leukemias and sarcomas 2,3. Gene fusion results when the break and re¬ joining during the translocation occurs within the introns of genes, such that the two translocated chromosomes have new exon organization, leading to the formation of chimaeric mRNA species and in turn, chimaeric proteins.

Different tumor phenotypes can result from specific versions of related fusion proteins, for instance the Philadelphia t(9;22) translocation (which yields the BCR-ABL fusion) occurs in both myeloid- and lymphoid-lineage tumors, dictated largely by the position of the translocation junction within the 5' BCR gene 5. The variability of gene fusions is even greater with some leukemia genes, in particular the MLL gene for which, more than 30 different fusion partners are known 6,7. This situation is not confined to haematopoietic malignancies, as the EWS-FUS family, typically involved in sarcomas, exhibits similar characteristics g-π. The EWS gene was identified in Ewing's sarcoma by its association with FLIl and ATFl 12,13 and subsequently found in subsets of Ewing's with either ERG or ETVl 14,15. Further, the EWS gene is involved in other sarcomas and with yet different partners, such as WTl in desmoplastic small round cell sarcoma 16 or CHN in myxoid chondrosarcoma π. A further ramification of this infidelity, is that the FUS gene, first identified by its fusion with the CHOP gene in t(12;16) of malignant myxoid liposarcoma and related to the EWS gene 18,19, has also been found in Ewing's sarcoma or acute myeloid leukemia fused with the ERG gene 20- 22. Finally the EWS gene can also be involved with the CHOP gene by chromosomal translocation in malignant myxoid liposarcoma 23, analogous to FUS-CHOP. These observations suggest that the EWS/FUS fusion proteins are effective in mesenchymal cells or in haematopoietic cells and their presence may be dictated simply by the cell type in which the relevant chromosomal translocation took place. Conditional mouse models of the EWS and Frø-associated chromosomal translocations should give insights into whether these fusions influence the malignant phenotype.

Conditional mouse models which mimic chromosomal translocations are important to provide a biology basis to study how oncogenes influence tumorigemcity. The present inventors have previously developed two mouse models mimicking chromosomal translocations using homologous recombination in ES cells to create mutant alleles and transferring these alleles into the mouse germline. The first approach was the knock-in model 24 which involved fusing the MLL gene partner AF9 into the mouse Mil gene. This model resulted in acute myeloid leukemias in mice, preceded by a myeloproliferation in the bone marrow 24,25. A drawback of this knock-in mouse model is that the fusion is effectively a germ-line mutation and thus cell specific effects cannot be achieved.

This deficiency was obviated in the translocator 26,27 in which de novo chromosomal translocations are created in mice 28,29, resulting in cell specific chromosomal translocations and leukemias 30. The translocator mouse model involves the use of loxP sites, introduced by homologous recombination into two target chromosomes, to facilitate site-specific recombination mediated by Cre recombinase 28-30.

In an Mll-Enl model which was also developed by the present inventors, myeloid leukemias arose quickly and with high penetrance when Cre was expressed under the control of the haematopoietic stem cell gene Lmo2 30. The translocator approach could, in principle, be used to model any human chromosomal translocation but if the gene orientation on the mouse chromosomes is incompatible with transcription following CτQ-loxP recombination, a conditional version of the knock-in chromosomal translocation mimic strategy must be invoked. A conditional version of the original knock-in approach, using a loxP flanked (floxed) transcription stop, has been used to circumvent embryonic lethality 31 but even this approach cannot always prevent low levels of read-through transcription, to avoid the dominant lethal effect of a fusion protein (MM, AF, RP and THR, unpublished results).

Thus there remains a need in the art to provide a mouse model of human chromosomal rearrangements which overcomes the problems described above.

SUMMARY OF THE INVENTION.

To avoid the problems of the prior art described above, the present inventors have developed a novel version of the knock-in translocation mimic, named the invertor model. In this approach, a floxed cassette carrying a cDNA sequence with an intron segment and acceptor splice site is knocked-in to the intron of a target gene but in the reverse transcription orientation to the latter. Consequently and importantly, no RNA splicing can take place between the inserted cassette until it is inverted by expression of Cre recombinase. The method of the invention ensures that RNA splicing and thus transcription only occurrs in conjunction with the expression of Cre recombinase. This important step therefore ensures that read-through transcription does not ocurr and thus the problems associated with the expression of embryonically lethal fusion proteins is avoided.

Thus in a first aspect the present invention provides a method for generating a non- human animal model of a chromosomal rearrangement, which method comprises creating a transgenic non-human mammal wherein an invertor cassette (comprising a first nucleic acid sequence and one or more nitrons comprising one or more acceptor splice sites from a gene of interest, wherein the first nucleic acid sequence and the intronic sequence/s are flanked by site/s which recognise a site specific recombinase) is knocked into a target gene of interest with that animal so that the first nucleic acid sequence is in a reverse transcription orientation with respect to that gene of interest and wherein that animal expresses a site-specific recombinase under the control of a cell type-specific promoter, such that a chromosomal inversion is catalysed by tiie recombinase upon its expression.

The method referred to herein and described above has been named the invertor method.

According to the method of the invention, the first nucleic acid sequence is comprised within a nucleic acid cassette (invertor cassette). Preferably, the nucleic acid is cDNA. Advantageously, the nucleic acid cassette comprises one or more of the following components in addition to a first nucleic acid sequence: a segment of intronic sequence including an acceptor splice site from the gene of interest, a poly A site and a segment for the selection of homologous recombinant clones, preferably a MCl-neo-pA seqment. These regions are flanked by one or more site/s which recognise and bind the recombinase. Advantageously, the recognition sites are loxPsites and the recombinase is the Cre recombinase. Further according to the above aspect of the invention, the nucleic acid cassette (invertor cassette) for introduction of the first nucleic acid sequence into the gene of interest may be introduced into the gene of interest, preferably the mouse gene of interest using any suitable method known to those skilled in the art. Advantageously, it is introduced using gene targetting in suitable cells, for example embryonic stem cells (ES cells). According to the method of the invention, advantageously, the nucleic acid invertor cassette according to the invention is introduced into the gene of interest downstream from a donor splice site. By introducing the invertor cassette 3' to 5' with respect to the orientation of the gene of interest, then a splice from the gene of interest (comprising a splice donor site) to the acceptor site of the invertor cassette is not possible. Thus read-through transcription is prevented using the method of the invention.

Thus in a further aspect the present invention provides an invertor cassette comprising a first nucleic acid sequence and one or more introns comprising one or more acceptor splice sites from a gene of interest, wherein the first nucleic acid sequence and the intronic sequence/s are flanked by site/s which recognise a site specific recombinase.

According to the above aspect of the invention, preferably the nucleic acid is cDNA. Advantageously, the invertor cassette comprises one or more of the following components in addition to those listed above: a poly A site and a segment for the selection of homologous recombinant clones, preferably a MCl-neo-pA seqment. These regions are flanked by one or more site/s which recognise and bind the site- specific recombinase. Advantageously, the recognition sites are loxPsites and the recombinase is the Cre recombinase.

According to the method of the invention, once suitable cells (for example ES) have been targetted so that they contain the invertor cassette within the gene of interest, then chimeras may be generated by injecting such cells into blastocysts or other suitable cells. Heterozygous carrier mammals which possess the Ews-ERG invertor allele may then be generated from these chimeras. Details are provided in the detailed description of the invention.

When the method of the invention is performed, chromosomal inversion and subsequent transcription may be performed. Those skilled in the art will appreciate that an essential feature of the method of the invention is that transcription of the fusion oncogene cannot occurr until inversion has been performed. In turn inversion step is dependent upon the expression of the specific recombinase. Thus, fusion gene transcription is dependent upon the process of homologous recombination ocurring. This method therefore prevents read through transcription with the possible generation of lethal fusion proteins.

Using the invertor method described above, oncogene fusions were created in several mouse models, in particular the inventors created a mouse model of the Ews-ERG gene fusion which is commonly found in human cancers and which gene fusion has been found by the present inventors to produce lethal fusion proteins when it is attempted to generate such a mouse model by other frequently used techniques such as the translocator technique or the loxP-stop method. Details of the generation of such mouse model according to the method of the invention are provided in Example 1.

.In addition an MLL-AF4 fusion has been created by the present inventors using the invertor model. This mutation has been commonly found in leukemias. Details of me generation of such a mouse model of leukemia using the invertor model described herein are given in Example 2.

Thus, in a further aspect the present invention provides the use of the invertor method according to the invention in the preparation of a non-human model of a chromosomal rearrangement

In a further aspect still, the present invention provides an Ews-ERG mouse model or an MLL- AF4 mouse model of a human chromosomal gene rearrangement. In a preferred embodiment of the above aspect of the invention, the Ews-ERG mouse model and/or the MLL- AF4 mouse model is obtainable, preferably obtained, using the invertor method described herein.

According to the above aspects of the invention, the Ews-ERG advantageously represents a mouse model of human sarcoma and/or leukemia. Moreover, the MLL- AF4 model advantageously represents a mouse model of adult and/or infant leukemia.

The present inventors have found that the Ews-ERG fusion protein and the MLL- AF4 fusion protein when expressed in lymphoid tissue in invertor mice prepared according to the methods of the invention plays a role in the onset and or maintenance of leukemogenesis. Specifically, the present inventors found that 25/29 mice which carried both the Ews-ERG and Ragl-|Cre genes developed leukemia associated with thymoma, whereas those displaying Ews-ERG only did not display leukemias. Morever, the inventors consider that the presence of the Ews-ERG fusion protein is crucial for malignancies in the invertor mice since such malignancies only arose in mice with both the Ews-ERG and ragl-Cre genes.

In a final aspect the invention provides the use of a non-human animal tumour model according to the present invention in the testing of potential anti-tumour compounds.

Advantageously, according to the above aspect of the invention, the non-human animal tumour model is the MLL-AF4 mouse model described herein, hi an alternative embodiment of the above aspect of the invention the non-human animal tumour model is the Ews-ERG mouse model described herein.

The present inventors tested the invertor approach by modelling the Ews-ERG fusion found in human sarcomas and in addition sought to determine if an Ews-ERG fusion can be oncogenic in a cell type in which it is not generally associated in human tumors (specifically in lymphoid cells). The inventors found that clonal T cell leukemias arose in the invertor model when Cre expression was controlled by the lymphocyte-specific Ragl gene, showing that Ews-ERG can invoke leukemia in lymphoid cells. The inventors results show the efficacy of the invertor translocation mimic model as a general method for creating fully conditional mouse models and suggest that in the example studied, that EWS-ERG is promiscuous in its ability to direct the development of malignancies.

According to the above aspect of the invention, the chromosomal rearrangement may be any rearrangement, for example an insertion, deletion, inversion or translocation; advantageously, the chromosomal translocation is a reciprocal translocation.

The chromosomal rearrangement is preferably tumourigenic, that is, gives rise to tumours. The invention thus provides non-human animal models of tumour conditions associated with chromosomal rearrangements.

Preferably, the cell type-specific promoter is the Ragl promoter. The Ragl gene is known in the art.

Recombinases are known in the art and include Cre, FIp and R recombinases. The sequences of these polypeptides are known: for example, the sequence of Cre recombinase is given in GenBank GI:5690440 and GI: 17016300; FIp recombinase is given in GI: 173333 and GI: 173335; R recombinase, from Zygosaccharomyces rouxii, is given in GenBank GI:5260.

The cognate sites for these recombinases are lox (or loxP) sites for Cre, frt sites for FIp and Rs sites for R. Any recombinase system may be used in the invention; advantageously, it is selected from CTQ-IOX, Flp-frt and R-Rs.

Leukaemias which can be modelled by the method of the invention include all those which involve a chromosomal rearrangement; such rearrangements are very common in leukaemias, hi particular, the human Ews gene is involved in a large number of leukaemias, and Ews-ERG fusions are found in human sarcomas and leukaemias. Moreover, MLL translocations ocurr in 5-10% of acute leukemias in children and adults. Further, MLL-AF4 is the most common abheration in infant leulaemia and is associated with poor treatment outcome.

Sites recognised by the recombinase are inserted into the genome of the animal in positions such that recombination between the sites will give rise to the desired rearrangement. Rearrangements include deletions, inversions, duplications and both reciprocal and non-reciprocal translocations. Advantageously, the rearrangement is a translocation, preferably a reciprocal translocation, in which parts of non-homologous chromosomes exchange positions in the rearrangement.

The recombinase, which is advantageously Cre, FIp or R as described above, is placed under the control of a cell type-specific promoter which is advantageously the Rag promoter.

Advantageously, the reciprocal translocation gives rise to a Ews-ERG fusion.or an M11-AF4 fusion protein as herein described.

The animal model according to the invention advantageously has leukaemia (or another tumour). It has surprisingly been found that the Ews-ERG fusion protein is crucial for malignancies in 'invertor mice' as malignancies only arose in mice with both the Ews-ERG and Ragl-Cre genes.

Animal models in accordance with the invention are useful in testing anti-tumour therapeutics, or compounds with a potential anti-tumour effect. Compounds may be tested for effectiveness against any type of tumour, including leukeamias and sarcomas but also other tumours which involve a chromosomal rearrangement.

In a further aspect, therefore, the invention provides a method for testing anti-tumour compounds, comprising exposing an non-human animal model in accordance with the present invention to said compounds and assessing the incidence of tumours in said model.

According to the above aspect of the invention, advantageously the non-human animal model is either an MLL-AF4 mouse model obtained using the inverter method described herein or the Ews-ERG mouse model obtained using the inverter model described herein.

hi a further aspect the present invention provides a method for the testing of an anti- tumour agent which method comprises the steps of: (a) providing a non-human animal tumour model according to the invention, (b) administering to that model one or more potential anti-tumour agents; and (c) testing for a change in one or more characteristics of tumour/s comprising the model according to step (a).

In a preferred embodiment of the above aspect of the invention step (c) (testing for a change in one or more characteristics of tumour/s comprising the model according to step (a)) involving the testing for a change in any one or more characteristics of one or more tumours in the group consisting of the following: size, morphology, presence and/or levels of one or more tumour markers and cytological assays.

DEFINITIONS.

In general, an 'animal model' is one in which pathological processes (for example, cancer) can be investigated, and in which the phenomenon in one or more respects resembles the same phenomenon in humans or other species of animals. An animal model, as referred to herein, is a model based on a whole organism or part thereof such as a tissue which possesses cellular organisation which replicates that seen in the whole organism. Thus, "animal model" does not include cell-based assays performed on single cells or cultures of cells. Preferably, "animal model" refers to complete animals. The animals are non-human animals, and may be mammals, birds, repltiles, fish etc. Mammals are preferred. Rodent models are convenient and within the scope of the present invention; mice and rats are preferred rodent models. Other models include canine, feline, monkey and primate models.

A "chromosomal rearrangement" is a change to the structure of a chromosome which leads to a genetic change in the subject. Rearrangements include deletions, where DNA is lost from the chromosome, insertions, inversions, duplications and translocations. Deletions can be located within a chromosome (interstitial) or can remove the end of a chromosome (terminal). Deletions can be small (intragenic), affecting only one gene, or can span multiple genes (multigenic). A deletion is a rearrangement that results in an increase in copy number of a particular chromosomal region. In tandem duplications, the duplicated regions lie right next to one another, either in the same order or in reverse order, hi non-tandem duplications, the repeated regions lie far apart on the same chromosome or on different chromosomes. An inversion is a rearrangement in which a chromosomal segment is rotated 180 degrees. Inversions in which the rotated segment includes the centromere are called pericentric inversions; those in which the rotated segment is located completely on one chromosomal arm and do not include the centromere are called paracentric inversions. A translocation is a chromosomal rearrangement in which part of one chromosome becomes attached to a non-homologous chromosome (non-reciprocal), or in which parts of two non-homologous chromosomes exchange places (reciprocal).

A "site-specific recombinase" is an enzyme that catalyses site-specific recombination of nucleic acid between paired sets of cognate recognition sequence sites. Examples are set forth above. Cre Recombinase is a Type I topoisomerase from bacteriophage Pl that catalyses the site-specific recombination of DNA between loxP sites. Recombination products depend on the location and relative orientation of the lox (or loxP) sites. Two DNA species containing single lox sites will be fused. DNA between directly repeated lox sites will be excised in circular form while DNA between opposing lox sites will be inverted with respect to external sequences. Members of the Cre family cleave DNA substrates by a series of staggered cuts, during which the protein becomes covalently linked to the DNA through a catalytic tyrosine residue at the carboxy end of the recombinase motif. A "cell type-specific promoter" is a promoter which directs expression of a nucleic acid in a cell type-specific manner; i.e. such that the said expression essentially occurs in certain tissues only. Many cell type-specific promoters are known in the art, and haematopoietic cell type-specific promoters are also known. The Ragl promoter is preferred in the present invention. Cell type-specific promoters may "leak", that is give rise to gene expression in cells other than the target cell type. This is acceptable, as long as sufficient levels of recombinase are achieved in the desired cell type to direct the rearrangement event therein to a sufficient level for the biological effect of the rearrangement to manifest itself.

Sites recognised by a recombinase are the cognate recognition sites. For example, Cre recognises lox or loxP sites. The loxP recognition element is a 34 bp sequence comprised of two 13 bp inverted repeats flanking an 8 bp spacer region which confers directionality. Other recognition sites, such asfrt and Rs sites, are known in the art.

A "tumour", or neoplasm, is a growth of cells resulting from abnormal regulation of cell growth leading to excessive mitotic division. Tumours may be benign, that is slow-growing and usually harmless, or malignant, that is fast growing and with a tendency to metastasise to other tissues. A "tumourigenic" event is an event that leads to or enhances tumour formation or growth.

A 'leukaemia', broadly speaking, is a malignant proliferation of haematopoietic cells, characterised by replacement of bone marrow by neoplastic cells. The leukaemic cells usually are present in peripheral blood, and may infiltrate other organs of the reticuloendothelial system, such as liver, spleen and lymph nodes. Leukaemia is broadly classified into acute and chronic leukaemia, with multiple distinct clinicopathologic entities subclassified in each category. Leukaemias are divided into four main types: acute myelogenous (or myeloid) leukaemia (AML), chronic myelogenous (CML), acute lymphocytic (ALL), and chronic lymphocytic (CLL).

BRIEF DESCRIPTION OF THE TABLES AND FIGURES. Table 1: Frequency of targeted clones and chimaera generation with Ews knock- in ES clones. Ews knock-in targeted clones were made using homologous recombination in ES (CCB) cells and successful targeting established by growth on G418 (due to expression of the neo resistance gene) and gancyclovir (homologous events deleting thz-Hsv-tk gene from the targeting vector). IA: The number of targeted clones identified from the indicated number of colonies analysed. Ews cassette control represents the knock-in of pC2A-neo (Figure 5 online), lacking additional cDNA sequences, into Ews exon 7. Ews-AF9, FHl, Chop and ATF represent respectively knock-ins of fusion cDNAs as shown in Supplementary Figure 2. IB: The number of chimaeras (identified by coat colour) resulting from injection of Evustargeted ES clones into blastocysts following transfer to recipient females.

Table 2: Tumor characteristics of Ews-ERG invertor mice expressing Cre recombinase from the Ragl gene. A cohort of 29 Ews-ERG; Ragl-Cre mice was established and monitored for ill health. 25 of these mice developed leukemia. The main post-mortem characteristics (including the age in days to disease manifestation) are shown. The status of the ERG cassette was determined by filter hybridization to ascertain that the transcription orientation was 5'->3' (inv) with respect to the Ews gene in the thymoma DNA; + = ERG cassette inversion found in thymoma DNA and - = ERG cassette inversion not found in thymoma DNA. T cell receptor β gene (TCRJβ236) and IgH (heavy chain intron μ enhancer probe 37) rearrangements were determined by filter hybridization of thymoma DNA and the status of each allele is indicated as R = rearranged; G = germ-line. The cell surface protein expression profile of the cells present in the thymomas was determined using flow cytometry. Pairs of markers were used (Thyl and B220 or CD4 and CD8) and the data summarised for the latter.

Table 3 : Sequences of rearranged TCR Jβ loci in Ews-ERG thymoma DNA. Thymoma DNAs were prepared from Ews-ERG; Ragl -Cre mice and the V-D- Jβ2 rearrangements were analysed (Fig. 4 and summary Table 2). The rearranged V-D- Jβ2 segments were amplified by PCR using primer pairs previously described 38,39 and the pools of Vβ primers (as indicated) with the single Jβ2 primer. Amplification was assayed using agarose gel electrophoresis (not shown) and the PCR products were directly sequenced using the Jβ2N primer 39. Assignment of sequences was accomplished by comparison with those in the IGMT database 40.

V, D or J segments are underlined and nucleotides between represent N-region additional nucleotides. P = productive V-D-J join; NP = non-productive V-D-J join resulting in out of frame joint.

Table 4: shows the Tumour characterisitcs of the Ews-ERG ήivertor mice expressing Cre Recombinase from the Ragl gene..

Figure 1: Strategy for generating Ews-ERG invertor gene by homologous recombination. A. A partial restriction map of the mouse Ews gene including the Xbal-EcoRI fragment (indicated above the line) cloned from a mouse 129 genomic library and the invertor cassette comprising loxP-ERG-polyA-MClneopA-loxP (using human ERG cDNA sequence, Figure 6) was cloned into the Xbal site just to the 3 'side of Ews exon 7, to generate a targeting vector. The transcription orientation of the ERG cDNA cassette was opposite from that of the endogenous Ews gene. B. A partial restriction map of the targeted Ews-ERG invertor allele. Homologous recombination in ES cells using the targeting vector (indicated above the line) resulted in insertion of the inverted ERG cassette just 3' of Ews exon 7 on one allele. Two probes for detecting homologous recombinants were located 5' and 3' of the Ews targeting fragment, d = donor RNA splice site (from Ews exon 7); a = acceptor RNA splice site (from the ERG invertor cassette, origin Af4 intron 4; see Supplementary information). In the targeted clone, the transcription orientation for these donor and acceptor sites is 5'>3' and 3'>5' respectively. . C. Expression of Cre recombinase causes inversion of the invertor cassette to bring the transcription orientation of ERG to be the same as Ews and the acceptor splice of the invertor cassette is thus upstream of ERG cDNA, allowing post-transcriptional RNA splicing from the donor site of Ews exon 7 to this acceptor site associated with ERG. R = EcoRI; K = Kpnl; H = Hindlll; RV = EcoRv; X = Xbal; S = Sad; B = BamHI; Bg =BglII; d = donor splice site; a = acceptor splice site D. Genomic sequence adjacent to the Ews exon 7 and the ERG invertor cassette (the derived amino acid sequence is shown in the single letter code) following Cre- mediated inversion obtained from DNA of ES cells and of thymus after Cre expression. The boxed sequence (TCTAG/CGAT) corresponds to the ligation of filled- in Xbal (Ews genomic) and CIaI (ERG invertor cassette) sites. The boxed Xbal site (TCTAGA) corresponds to the position of cloning of the mouse A/4 intronic sequence (shaded) in the ERG invertor cassette. On the 3' side of the fused Xbal-Clal sites is a loxP site of the ERG invertor cassette (highlighted in yellow; the second site resides at the 3' end of the ERG cassette; see Figure 7) followed by a partial sequence (shaded) of the Af4 intron 4 (note this sequence is noncontiguous and the dots represent the gap; the full sequence appears in Figure 7) upstream of the ERG cDNA sequence, and up to the Af4 acceptor splice site. This is followed by the ERG cDNA sequence from the pC2A transfer vector (see Methods and Figure 7). The Notl site derived from the pC2A vector is boxed. A represents splice donor of Ews exon 7. E: Nucleotide sequence, and derived protein sequence, of the junction between Ews and ERG sequences in the fusion rnRNA obtained by RT-PCR of mRNA made from Ews-ERG ES cells transiently transfected with a Cre expression vector or from, spleen RNA of Ews-ERG; Ragl-Cre mice. The underlined sequence comes from mouse Ews, the highlighted sequence from the cassette used to construct the intron-Ei?G-polyA sequence (see Figs. 5 and 8) and the plain sequence is the human ERG cDNA fused to Ews. The Notl site derived from the pC2A vector is boxed.

Figure 2: Leukemia incidence and characteristics of Ews-ERG invertor mice. Cohorts of 29 Ews-ERG;Ragl-Cre mice and 20 Ews-ERG control mice were established and monitored over a period of -17 months. Individual mice were culled and post-mortem conducted when signs of ill health were observed. Leukemia/lymphoma was established by various criteria (see Methods). Top left panel. Survival curve of Ews-ERG; Ragl-Cre mice as a function of time (days). Top right panel. Histology of blood and bone marrow from leukemic mice (M21 and M20 respectively). Blood smears were stained with MGG and photographed at 400X magnification, whilst bone marrow was photographed at 10OX. Bottom panel. Histological sections of tissues from mice with thymoma, illustrating three levels of severity of leukemic infiltration. Tissues were fixed in 10% formalin and wax embedded, followed by sectioning and staining with haematoxylin and eosin. Mouse M9 had the most severe phenotype, with thymus and spleen heavily disrupted by tumor cells and marked peri-vascular deposits of leukocytes in liver and kidney. Mouse M21 had severe alterations to thymus and spleen but little involvement of liver or kidney. M6 has the least severe phenotype, with only its thymus being involved with an abnormal structure. Sections with 0.4μm and photographed as 400X times.

Figure 3: Cell surface analysis of thymomas from Ews-ERG invertor mice. Thymus tissue was resected from mice with thymoma and FACS analysis performed to determine T cell phenotype (summarised in Table 2). The data in the figure shows representative flow diagrams of three tumor-bearing mice (M2, M3 and M4) compared with a wild type C57BL/6 control, using anti-Thyl (y axis) plus anti-B220 (x-axis) antibodies or using anti-CD4 (y-axis) plus anti-CD8 (x-axis) antibodies. The three thymomas show a range of CD4 and CD8 co-expression phenotypes characterizing the thymomas generated in the Ews-ERG; Ragl-Cre invertor mice.

Figure 4: Ews-ERG;Ragl-Cre invertor mice have clonal T cell Ieukemias. Genomic DNA was prepared from various tissues of invertor mice with thymomas. Genomic analysis was carried out using filter hybridization to assess the presence of the inverted ERG cassette in the tumor cells (A), and to assess whether the lymphocytes involved in the Ieukemias were clonal T (B) or B cells (C). A. Inversion hybridization autoradiograph: DNA was prepared from the thymoma and other tissues of Ml 8, cleaved with EcoRI and hybridised to the 5' Ews probe (which detects a 5Kb targeted ERG cassette fragment or a 6.5Kb Cre-inverted fragment). If the ERG cassette is inverted by Cre activity, the size of the EcoRI fragment increases from the initial targeted gene size, as indicated in the maps below the figure. The data shown are for DNA made from spleen (spl), thymus (thy), liver (liv), kidney (kid) or tail of invertor mouse Ml 8 (ES cell DNA is used as a control). The Cre-mediated inverted band (~6.5Kb) is evident in thymus DNA (thymoma) but not the other tissues. The summary of the data from the cohort of invertor mice is in Table 2. Restriction fragment sizes are represented as GL = germline; inv = inverted allele of ERG cassette; Tgt; = initial targeted Ews allele. The organisation of the targeted Ews allele is indicated underneath the hybridization autoradiograph showing the location of Ews exon 7 and the initial targeted orientation (bottom) or inverted orientation of ERG invertor cassette after Cre-mediated recombination (top). a = acceptor splice site B. Autoradiograph showing rearrangement of T cell receptor β locus: A TCR Jβ2 probe 36 (diagrammatically shown below the autoradiograph) detects a 5Kb germline HindIII Jβ2 band whilst V-D- Jβ2 joins in T cells result in new HindIII sized bands depending on the nature of the rearrangement. Twelve thymoma DNAs were compared and each showed one or two Jβ2 alleles rearranged, signifying that these tumors were clonal T cells. In some thymoma samples, there is almost complete absence of the germline band indicating that the thymuses of these mice is solely comprised of malignant clonal T cells. DNA sequence analysis of the V-D-J junctions of mice M2, M5, M13, M18 showed functional VDJ joins (see Table 3). C. Autoradiograph showing rearrangement status of immunoglobulin heavy chain genes: A Cμ intron probe 37 (diagrammatically shown below the autoradiograph) was used to hybridize a set of thymoma DNAs for the presence of Igh rearranged bands. Only two samples showed rearrangements. D. Detection of Ews-Erg fusion protein in thymoma cells. Single cell suspensions were made of T cells from a normal thymus (wt) or from the thymoma of Ml 8. 18μg protein was solubilised in SDS-loading buffer, fractionated on 4-20% acrylamide gel, transferred to nitrocellulose membrane and incubated with anti-Ews or anti-ERG antibody. Bound antibody was detected with anti-rabbit or anti-goat HRP respectively and exposed to X-ray film by ECL (Amersham).

Figure 5: Targeting cassettes for knock-in or invertor cDNA cloning A: Partial restriction maps of pC2-neo, pC2A-neo and pC2A-frt-neo-frt transfer vectors for inserting cDNA sequences for knock-in. These vectors were made from ρBSpt-C2 (B) or pBSpt-C2A (C). For knock-in constructs, cDNA was PCR amplified using primers with Notl sites added to the ends and cloned into the unique Notl of the vectors, giving the reading frame appropriate for the site of homologous recombination in the gene of interest. The resulting plasmids can be digested with either Sfil (pC2) or BgIII (pC2A), for insertion into an appropriate targeting vector made from 129 mouse DNA for homologous recombination in ES cells. Each transfer cassette includes the SV40 polyA sequence (shown in D) cloned between Notl and BarnHI sites of pBSpt- C2 or pBSpt-C2A, MCl-neo-pA or PGK-neo-pA fragments, for selection of ES clones in G418, cloned as BamHI fragments. A cassette similar to pC2A-neo is available with the PGK-hygromycin gene sequence as a BgIII fragment cloned into the BamHI site as above (thereby destroying both BamHI and BgIII sites but leaving the BgIII sites at the end for knock-in cloning) instead of the MCl-neo-pA cassette (pC2A-hygro) pC2Afrt-neo-frt also has a PGK-neo fragment flanked frt sites cloned into the BamHI site as above instead of the MCl-neo-pA cassette (pC2-frt-neo-frt). This allows deletion of the neomycin selectable marker gene using the FLP recombinase after homologous recombinants have been identified in ES cells. A similar cassette has loxP sites flanking the MCl -neomycin cassette.B: Polylinker of intermediate vector pBSpt- C2. The clone was constructed by cloning an oligonucleotide pair into the Notl-BamHI sites of pBSpt (and at the same time, destroying the original Notl site in pBSpt) to create an Sfil-Notl polylinker to give pBSpt-C2. C: Polylinker of intermediate vector pBSpt-CA2: The clone was made as above but using the inserted oligoncucleotide sequence shown with BgIII sites on either side of the Notl-BamHI sites. D: The SV40 polyA signal was PCR amplified from SV40 DNA and cloned into the NotIBamHI sites of either cassette pBSpt-C2 or pBSpt-C2A and the MCl-neo-polyA clone 1 cloned into the BamHI site. The sequence of the PCR fragment made from SV40 DNA is shown comprising the dual polyadenylation signals (highlighted in yellow). The primers used for amplification are also highlighted and included Notl or BamHI sites for cloning into pBSpt-C2 or pBSpt-C2A vectors.

Figure 6: Ews knock-in targeting clones cDNA sequences were cloned as Notl PCR fragments into pC2A-neo (Fig. 1) for generation of targeted knock-in Ews genes in ES cells. FIi 1 cDNA was prepared from MEL RNA (A), AF9 cDNA was prepared from HEL RNA (B), Chop from 3T3L1 RNA (C), ATF from Jurkatt RNA (D) and ERG from COLO320 RNA (E) (latter was used for the invertor cassette). The Notl restriction sites (lower case) at each end of the PCR fragments were added as part of the PCR primers, for cloning into the Notl site of pC2A-neo. F. The fusion junction of Ews exon 7 BamHI site and the BamHI/Bglll junction of Ews and FHI (as an example) All the derived protein sequences are shown as single letter code.

Figure 7: Strategy for construction of the invertor cDNA cassette. A. Organisation of the floxed ERG cDNA inversion cassette. The ERG invertor cDNA cassette is depicted (middle line) as it would be in the genome following homologous recombination and Cre-mediated inversion. From the 5' end, there is a loxP site (sequence shown in D), the Af4 intron (sequence shown in E), ERG cDNA (sequence shown in Fig S2E), the SV40 pA (sequence shown in Figure 5D online), MCl-neo-pA 1 and a 3' loxP site (sequence shown in F), The top line of A illustrates the orientation in which the cassette is targeted after homologous recombination, adjacent to Ews exon 7 (which is indicated with the Ews ORF and the donor splice site location). The bottom line indicates the relationship between the Ews exon 7 ORF and donor splice sequences and the ERG invertor cDNA cassette ORF and acceptor splice sequences after Cre-dependent inversion. For general use, any invertor cDNA cassette may be constructed in either pC2 or pC2A vector. B. Sequence of pMGl vector polylinker. This vector was adapted from pBSpt by replacing the polylinker sequence as indicated to include a single site for the rare cutter Seel restriction endonuclease (TAGGGATAA/CAGGGTAAT). C. Sequence of pMG2 vector polylinker. This vector was adapted from pBSpt by replacing the polylinker sequence as indicated to include two sites for the Seel restriction endonuclease. D. Sequence of the 5' loxP site of the cDNA invertor cassette. The loxP site (highlighted in yellow) was cloned from pPGK-loxP-hygro 2 as a 120bp EcoRI fragment (rilled in ends) into the filled-in Notl site of pMG2, thereby removing both sites. E. Sequence of the mouse Af 4 gene intron 4. The indicated Af4 gene segment was amplified from 129 DNA using the primers (underlined) including Xbal and BamHI restriction sites for cloning into the loxP vector indicated in D. F. Sequence of the 3' loxP site of the cDNA invertor cassette. The loxP site (highlighted in yellow) was cloned from pPGK-loxP-hygro 2 as a 120bp EcoRI fragment (filled in ends) into the EcoRV site of the pMG2-derived vector with loxP- Af4 intron (D, E). This gave the pLOXKI clone (depicted in G). The ERG cDNA transfer cassette (in pC2A and comprising ERG cDNA, SV40 p A and MCl-neo sequences as a BgIII fragment, see Figs. 5 and 6 ) was cloned into the unique BamHI site of the pLOXKI invertor cassette. This final ERG invertor cassette was cleaved with Seel and CIaI, end-filled and cloned into the Xbal (end-filled) of Ews genomic fragment downstream of exon 7 for homologous recombination (see Fig. 1).

Figure 8: Cassettes for knock-in of the Cre recombinase gene into targeting vectors for gene specific expression. Three vectors are available for the knock-in of Cre (or of FLP) recombinase genes into target genes of interest (illustrated in D). A: In each case, the Cre cDNA sequence was cloned as a Notl flanked PCR fragment into the Notl of pC2A vectors giving the ORF shown (the single letter code is used for protein and those residues in upper case correspond to the N-terminus of Cre, including the added nuclear localisation signal). B: To create a Cre knock-in into the coding sequence of any gene of interest, a mutation would normally be required just 3' of the initiation codon of the target gene for insertion of a pC2A-Cre clone in frame. In the example here, an in-frame knock-in of Cre into Ragl 3 was achieved by mutagenesis of Ragl to include a BamHI site adjacent to the Ragl initiator converting the sequence as shown. C: Cre was cloned into the pC2A-Cre clone digested with Notl and the cassette cleaved with BgIII and cloned into the Ragl mutated BamHI site. This yielded in- frame knock-in of Cre into Ragl which produces a fusion protein comprising two amino-acids of Ragl and the whole Cre sequence, including the nuclear localisation signal. D: Three Cre transfer vectors are available as shown, hi the case of pC2A-Cre-frt-neo- frt, the frt-neo-frt portion (frt-PGK-neo-frt) was cloned as (Xhol-Sall) from pfrt-neo- frt (a gift from Dr A. McKenzie) into the filled in BamHI site of pC2A (note: a single BgIII site in pfrt-neofrt was removed by linearization with BgIII, repairing the cohesive ends followed by religation to form circular plasmid). Cre was cloned as the Notl PCR fragment. NOTE : If required FLP recombinase can be cloned into pC2A- lox-neo-lox using PCR to clone the relevant coding region of FLP, facilitating knock- in of FLP into genes of interest.

Fig 9: The invertor strategy- An MLL- AF4 fusion gene model.

Fig 10: Leukemia incidence in MLL-AF4 invertor using different Cre expressing genes.

Fig 11: Cre dependent inversion of MLL-AF4 gives a functional MLL- AF4 fusion mRNA.

Fig 12: MLL-AF4 have a myeloid phenotype but have Ig gene rearrangements.

Fig. 13: Mouse models are useful for pre-clinical testing of drugs prior to use in patients.

Fig 14: Strategy for generating Ews-ERG invertor gene by homologous recombination. (A) The method for making invertor mice is described in detail elsewhere [35]. In summary, an inverter cassette comprising a short intronic region with an acceptor splice site, human ERG cDNA sequence, a polyA site and the MClneopA gene all flanked by loxP sites (depicted by black triangles) was knocked in, using homologous recombination, into Ews intron 8. The transcription orientation of the targeted ERG cDNA invertor cassette was opposite from that of the endogenous Ews gene after initial homologous recombination. Germline carrier mice of this targeted allele were crossed with Cre-expressing mouse strains [36] and, as indicated in the right hand side of the diagram, the invertor cassette is turned around to create a transcription orientation identical with that of Ews. Thus after transcription, a pre-mRNA is made with the donor splice site of Ews exon 7 adjacent to the acceptor site of the invertor cassette allowing post-transcriptional fusion of Ews with ERG in an analogous format to that found in human sarcomas with t(21 ;22). (B) Genomic sequence adjacent to the Ews exon 7 and the ERG invertor cassette (the derived amino acid sequence is shown in the single letter code) obtained from DNA of ES cells and of thymus after Cre-mediated inversion. The Ews intron 7/8 donor splice site is indicated by an arrow. The boxed sequence (TCTAG/CGAT) corresponds to the ligation of filled-in Xbal (Ews genomic) and CIaI (ERG invertor cassette). The boxed Xbal site (TCTAGA) corresponds to the position of cloning of the mouse Af 4 intronic sequence (shaded) in the ERG invertor cassette. On the 3' side of the fused Xbal-Clal sites, there is a loxP site originating from the ERG invertor cassette followed by the Af4 intron 4 (shaded) upstream of the ERG cDNA sequence. The Af4 acceptor splice site is indicated by an arrow, and is followed by Hie ERG cDNA sequence. A Notl site used for cloning the amplified ERG sequence is boxed (note this sequence is non¬ contiguous and the dots represent a gap;

Figure 15. Incidence of haematological malignancy and characteristics of Ew$- ERG invertor mice. Cohorts of 29 Ews-ERG;Ragl-Cre mice and 20 Ews-ERG control mice were monitored over a period of ~17 months. Mice were culled and post-mortem conducted when signs of ill health were observed. Leukaemia/lymphoma was established by various criteria (see Methods). Top panel. Survival curve of Ews-ERG; Ragl-Cre (+Cre) or Ews-ERG (-Cre) mice as a function of time (days). Bottom panel. Histology of blood and bone marrow from leukaemic mice (M21 and M20 respectively). Blood smears were stained with MGG and photographed at 400X magnification, whilst bone marrow was photographed at 100OX.

Figure 16. Analysis of cell surface antigens of thymomas from Ews-ERG invertor mice using flow cytometry. Thymus tissue was resected from mice with thymoma and FACS analysis performed to determine T cell phenotype. The data in the figure shows representative flow diagrams of three tumour-bearing mice (M2, M3 and M4) compared with a wild type C57BL/6 control, using anti-Thyl (y axis) plus anti-B220 (x axis) antibodies or using anti-CD8 (y axis) plus anti-CD4 (x axis) antibodies. The three thymomas show a range of CD4 and CD8 co-expression phenotypes characterizing the thymomas generated in the Ews-ERG; Ragl-Cre invertor mice.

Figure 17. Clonality of T cell neoplasias in Ews-ERG invertor mice Genomic DNA was prepared from various tissues of invertor mice with thymomas. Genomic analysis was carried out using filter hybridization to assess the presence of the inverted ERG cassette in the tumour cells (A), and to assess whether the lymphocytes involved in the tumours were clonal T (B) or B cells (C). (A) Inversion hybridization autoradiograph: DNA was prepared from the thymoma and other tissues of Ml 8, cleaved with EcoRI and hybridized to the 5' Ews probe (which detects a 5Kb targeted ERG cassette fragment or a 6.5Kb Cre-inverted fragment). If the ERG cassette is inverted by Cre activity, the size of the ΕcoRI fragment increases from the initial targeted gene size, as indicated in the maps below the figure. The data shown are for DNA extracted from spleen (spl), thymus (thy), liver (liv), kidney (kid) or tail (ES cell DNA is used as a control). The Cre-mediated inverted band (~6.5Kb) is evident in thymus DNA (thymoma).. Note that despite extensive infiltration of spleen detected by histology, we cannot see the inverted band by Southern blot; this presumably reflects regional clustering of neoplastic cells in spleen. Restriction fragment sizes are represented as GL = germline; inv = inverted allele of ERG cassette; Tgt = initial targeted Ews allele. The organization of the targeted Ews allele is indicated underneath. The hybridization autoradiograph showing the location of Ews exon 7 and the initial targeted orientation (bottom) or inverted orientation of ERG invertor cassette after Cre-mediated recombination (top), a = acceptor splice site (B) Autoradiograph showing rearrangement of T cell receptor β locus: A TCR Jβ2 probe (diagrammatically shown below the autoradiograph) detects a 5Kb germline HindIII Jβ2 band whilst V-D- Jβ2 joins in T cells result in new HindIII sized bands depending on the nature of the rearrangement. Twelve thymoma DNAs were compared and each showed one or two Jβ2 alleles rearranged, signifying that these tumours were clonal T cells. In some thymoma samples, there is almost complete absence of the germline band indicating that the thymuses of these mice are solely comprised of malignant clonal T cells. DNA sequence analysis of the V-D-J junctions of mice M2, M5, Ml 3, Ml 8 showed functional VDJ joins.. (C) Autoradiograph showing rearrangement status of immunoglobulin heavy chain genes: A Cμ intron probe, (diagrammatically shown below the autoradiograph) was used to hybridize a set of thymoma DNAs for the presence of Igh rearranged bands. Only two samples showed rearrangements. (D) Detection of Ews-Erg fusion protein in thymoma cells. Single cell suspensions were made of T cells from a normal thymus (wt) or from the thymoma of M6, protein fractionated on 4-20% acrylamide gel, transferred to nylon membranes. Specific proteins were detected with anti-Ews or anti-ERG antibody.

Figure 18. B and T cells express the Ews-ERG fusion RNA. A 96 day old mouse with both Ews-ERG and Ragl-Cre alleles was used as a source of spleen and thymus cells. Single cell suspensions of spleen cells were labelled with anti-B220 or with anti-thyl.2 and cells purified using a MoFIo preparative flow cytometer. Estimated purities were achieved of >95%. cDNA was prepared from RNA extracted from sorted cells or from aliquots of unsorted populations and RT-PCR (approximately 3400 B220+ or 6400 thyl.2+ cell equivalents per PCR reaction) carried out with specific Pax5 (A), CDS (B) or Ews-ERG (C) primers. PCR reaction products were fractionated on 1% agarose gels and either stained with ethidium bromide and photographed (A, B) or the gel blotted and hybridised with an Ews-ERG probe (C).

Figure 19. Invertor strategy for generating MII-AF4 mice. Invertor MII-AF4 was made employing die strategy described previously (Codrington et al., 2005; Forster et al., 2005) with a human AF4 cDNA sequence. In outline, a floxed invertor cassette was constructed comprising a short stretch of intron, an acceptor splice site, the AF4 eDNA followed by a polyadenylation site (pA) and a neomycin gene for selecting homologous recombination events. This cassette was inserted into an Mil targeting vector at a BgIII site and transfected into ES cells followed by identification of clones in which the invertor cassette had entered the genomic DNA in a 3'->5' orientation with respect to the endogenous Mil gene. ES clones were injected into blastocysts, and germline carrier mice made. The floxed invertor cassette was made such that Cre recombinase recognizes the loxP sites and inverts the cassette to make the transcription orientation 5 '-3', tins allowing transcription of the MII-AF4 fusion gene. (A) Diagram of part of the mouse Mil gene and of the AF4 invertor cassette. A partial restriction map of mouse Mil around exons 8-10 is shown in the top line, indicating the BgIII site into which the AF4 invertor cassettes were introduced. Two probes (used for detecting targeted alleles) were located 5' and 3' of the respective Mil fragment (Mil 5' probe, Mil 3 'probe). The invertor cassette is depicted below the Mil map and two forms were used, one flanked by wild type loxP sequences (loxP) and a second generation by the mutant loxP sites (Iox66 and lox 71 (Albert et aL, 1995)). These sequences are shown below the AF4 invertor cassette. (B) Organization of the targeted MII-AF4 invertor allele before Crc mediated inversion of the cassette. The floxed cassette carrying the AF4 cDNA sequence was knocked-in to the Mil intron 10 but in the opposite orientation for transcription of the Mil gene (as depicted by the arrow beneath the map). RNA splicing from the Mil exon 10 donor splice site to the acceptor site (a) in front of the AF 4 cDNA insert cannot occur until the cassette is inverted by expression of Cre recombinase. (C) Organization of the targeted MII-AF4 invertor allele after Cre mediated inversion of the cassette. Cre recombinase mediated inversion of the floxed cassette allows post-transcriptional RNA splicing from the donor site of Mil exon 10 to the acceptor site upstream of AF 4 cDNA, forming a fusion transcript representative of the chimeric mRNA found in human MLL-AF4 positive leukemia. Primers located in Mil exon 10 (MEF), Af 4 intron 4 (AIR, fluorescence labeled), and AF4 exon 5 (AER) are indicated by red arrows. The genomic fusion sequence was amplified using primers MEF and AIR. Primer AIR was labeled with 5' fluorescein to quantify the inversion efficiency and Crc specificity by competitive PCR. Primers MEF and AER were used for RT-PCR to amplify and sequence the expressed fusion mRNA. (D) MII-AF4 fusion transcript and derived protein sequence (written in single letter code) obtained from spleen RNA from MII-AF4 lox; Cre mice. (E) Fusion transcript and derived protein sequences of MII-AF4 Iox66/71; Cre invertor. Note that the exact sequence at the junction of Mil and AF4 in the two forms of invertor mice differ slightly due to splice variation. Sp, Sphl; Xb, Xbal; R, EcoRI; H, Hindlll; K, Kpnl; Bg5 BgIII; B, BamHI. chr. = chromosome

Figure 20. MIU-AF4 invertor mice develop hematopoietic malignancy Survival curves of MII-AF4 invertor mice. Cohorts of Mll-AF4lox; Cre mice and controls lacking either the invertor or Cre alleles were established and post-mortem examinations carried out at signs of ill health. Latency and frequency of disease is plotted for invertor mice with Cre recombinase expression under the control of either the Rag-], Lck and CDl 9 promoters (McCormack et al., 2003). Incidence of lymphoid tumors was 31/36 mice (86%) for MII-AF4 loxP; Rag-Cre; 25/29 (86%) for MII-AF4 1 ox66/71; Rag-Cre with a median latency of 466 and .317 days, respectively. Incidence was 11/12 mice (92%, mean latency 416 days) for MII-AF4 loxP; Lck-Cre and 17/18 (94%, mean latency 472 days) for MII-AF4 Iox66/71; Lck-Cre. Incidence was 6/7 mice (86%, mean latency 475 days) for MII-AF4 loxP; GD19-Cre mice and 10/14 (71%, mean latency 460 days) MII-AF4 Iox66/71; CDI9-Cre mice. Black lines represent MII-AF4 Iox66/71 strains, grey lines MU-AF4 loxP strains and dashed lines Mil wt; Cre strains.

Figure 21. Tumor cells infiltrating MII-AF4 invertor mouse tissues have gene inversion and fusion transcripts. (A, C) Filter hybridization analysis of Sphl digested genomic DNA from the indicated tissues using the Mil 5 'probe is shown for one representative individual for each Cre expressor strain ofMll-AF4 loxP (panel A) and MII-AF4 tox66/71 (panel C) invertor mice. The 9.7 kb fragment (inv) indicates the inversion of the invertor cassette. Signal intensity depends on the degree of tumor infiltration in the respective organ. gl = germline allele; tg = targeted allele, inv = inverted allele; Mes. LN = Mesenteric lymph node; Axill LN. = Axillary lymph node (B, D) Detection of MII-AF4 RNA using RT-PCR with primers MEF and AER confirms the presence of MU-AF4 fusion transcripts in affected tissues of MII-AF4 loxP (B) and MII-AF4 Iox66/71 mice (D), respectively. The integrity of mRNA is confirmed by β-actin amplification products. (E) Cell type specific inversion by Rag-Cre, Lck-Cre, and GD19-Cre expression Fluorescence labeled PCR products of the genomic fragment following inversion (primers MEF and AIR) and exon 2 of the single copy gene Lmo2 as a control were quantified and compared in unsorted, Macl+, B220+, or Thyl.2+ sorted cell populations. Columns show mean and SEM of triplicate PCR reactions from representative DNA preparations.

Figure 22. MII-AF4 invertor mice develop diffuse large B cell lymphomas. Fixed tissues from MII-AF4; Cre neoplastic mice were analyzed by histology or by immunohistochemistry. (A) Hematoxylin and eosin (H&E) stained paraffin sections (10Ox magnification) show disruption of spleen and liver architecture due to tumor infiltration in a representative MIIAF4 tox66/71; Rag-Cre mouse compared to a CS7BL/6 mouse (wt). The bottom right panel shows lymphoma cells stained with the B-cell marker B220 (40Ox magnification). (B) In the top row showing high magnification (100Ox), infiltrates can be identified as diffuse large cell lymphoma in each of the three different Cre expressor strains. H&E staining shows large vesicular nuclei, clumping chromatin, prominent nucleoli, and mitoses in lymphoma cells. Immunohistochemistry demonstrates the B cell phenotype (B220+) of the lymphoma. Reactive T cells (CD3+) with small regular nuclei are also present within the infiltrated areas. No myeloid cells (as judged by lack of expression of the marker F4/80) are present in the lymphoma, whereas myeloid cells in remaining red pulp of spleens stain with this marker (inset).

Figure 23. Lymphomas of MII-AF4 inverters have clonal immunoglobulin gene rearrangements. Genomic DNA from lymphoma infiltrated spleens or mesenteric lymph nodes were analyzed by filter hybridization using either Igh (A), Tcr Cβl (B) or Tcr Jβ2 (C) probes. Examples of 2 mice of each strain are presented. Positions of the respective germline (gl) fragments are indicated by arrows. Partial restriction maps of the JH or Cβl -Jβ2 genomic regions and location of the probe fragments are shown between the respective southern blots filter hybridizations. (D) Results of 53 mice are summarized in the table. Each box shows the numbers of tumors with Igh or Tcr rearranged/total number in group. H, Hindlll; X, Xbal; R, EcoRI

Figure 24. MII-AF4 inverter tumors are transplantable in Rag-I null mice. A single cell suspension was made from the mesenteric lymph node of an MII-AF4 1 ox66/71; CD19-Cre tumor mouse and 5x 106 cells injected into the tail vein of a Rag- 1 -/- mouse. A massive lymphoma infiltration of spleen and mesenteric lymph nodes as well as liver and kidneys of the first recipient mouse occurred within 56 days and the spleen of this mouse was used as a source of cells for flow cytometric analysis, DNA preparation and re-transplantation into four secondary Rag-1 -/- recipients. The latter developed tumors after 17-23 days and the spleens from these mice were taken for analysis. (A) Flow cytometry using anti-B220, anti-Thyl.2 or anti-Macl FITC coupled antibodies of mesenteric lymph node cells from the MII-AF4 Iox66/71; CDI9-Cre mouse (top row) or spleen cells from the first Rag-/- (middle row) or one of the 4 secondary Rag-/- recipients (bottom row). (B) DNA was prepared from the indicted tissues of the primary MII-AF4 Iox66/7I; CD19-Cre mouse and from the spleens of the Rag-1 -/- transplant recipient mice. Filter hybridization was carried out using an Mil probe to detect inversion of the AF4 cassette (top panel), an Igh probe (middle) or a Tcr probe (bottom). Mes. LN = mesenteric lymph node; inv = inverted allele; gl = germline allele; tg = targeted Mil allele; rearr. = rearranged Igh allele

DETAILED DESCRIPTION OF THE INVENTION.

General Techniques

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art (e.g., in cell culture, molecular genetics, nucleic acid chemistry, hybridisation techniques and biochemistry). Standard techniques are used for molecular, genetic and biochemical methods (see generally, Sambrook et al., Molecular Cloning: A Laboratory Manual, 2d ed. (1989) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N. Y. and Ausubel et al., Short Protocols in Molecular Biology (1999) 4th Ed, John Wiley & Sons, Inc. which are incorporated herein by'\zz reference) and chemical methods. In addition Harlow & Lane., A Laboratory Manual Cold Spring Harbor, N. Y, is referred to for standard Immunological Techniques.

Generation of transgenic animals

Methods for the generation of transgenic animals are generally known to those skilled in the art and are discussed in detail in First N, Haseltine FP., Transgenic animals: Stoneham, UK: Butterworth-Heinmann, 1991; and Grosveld F, Kollias G., Transgenic animals: San Diego, California: Academic Press, 1992. Commonly used methods include the ES cell method and pronuclear microinjection. In the context of the present invention, ES cell techniques are preferred. In this technique, embryonic stem cells (ES cells) are harvested from the inner cell mass (ICM) of animal blastocysts. They can be grown in culture and retain their full potential to produce all the cells of the mature animal, including its gametes. The cultured cells are exposed to the DNA which is intended to be inserted into the cells so that some cells will incorporate it. The cells are then tested for successful incorporation of the DNA, usually by selecting for the presence of a neo marker, and injected into the inner cell mass of an animal blastocyst. A proportion of the cells of the embryo resulting from the blastocyst will incorporate the desired DNA.

A pseudopregnant animal is prepared and the embryos transferred into the uterus. A proportion of the embryos will develop into animals carrying the inserted DNA.

Targeted gene insertion requires some additional steps. The vector used to transfer the DNA typically includes: the desired gene neor, a gene that encodes an enzyme that inactivates the antibiotic neomycin (and its relatives) tk, a gene that encodes thymidine kinase, an enzyme that phosphorylates the nucleoside analogue gancyclovir. DNA polymerase fails to discriminate against the resulting ■ nucleotide and inserts this non-functional nucleotide into freshly-replicating DNA.

The culture of ES cells is treated with a preparation of vector DNA: Most cells fail to take up the vector; these cells will be killed if exposed to neomycin. In a few cells: the vector is inserted randomly in the genome. In random insertion, the entire vector, including the tk gene, is inserted into host DNA. These cells are resistant to neomycin but killed by gancyclovir.

In still fewer cells: homologous recombination occurs. Stretches of DNA sequence in the vector find the homologous sequences in the host genome and the region between these homologous sequences replaces the equivalent region in the host DNA. Positive selection of transformed cells is achieved by culturing all cells in neomycin. The cells (the majority) that fail to take up the vector are killed. Negative selection of all cells (tk+) in which the vector was inserted randomly by culturing cells surviving the neomycin selection is performed by culturing the cells in gancyclovir. This step leaves a population of cells transformed by homologous recombination (enriched several thousand fold). These cells are injected into the inner cell mass of animal blastocysts.

Methods for introduction of recombinase specific sites into the genome of a transgenic animal are described in Collins et al, (2000) EMBO Reports 1:127-132.

The insertion of genes encoding a site-specific recombinase under the control of a cell type-specific promoter has been described in the art. For example, Cre recombinase has been placed under the control of promoters for:

The gene encoding the recombinase may be inserted in the same procedure as that used to insert the cognate sites, or by a separate procedure. It can be introduced into the same animal, or a separate animal. For example, a transgenic animal carrying the recombinase gene, introduced by ES cell transformation or pronuclear microinjection, can be crossed with an animal carrying the cognate sites to produce an animal of the desired genotype.

Site-Specific Recombination

Site-specific recombinase enzymes include the hit recombinase of bacteriophage (with or without Xis) (Weisberg, R. et al, in Lambda II, (Hendrix, R., et al, Eds.), Cold Spring Harbor Press, Cold Spring Harbor, NY, pp. 211-50 (1983); Tpnl and the -lactamase transposons (Mercier, et al, J. Bacteriol, 172:3745-57 (1990)); the Tn3 resolvase (Flanagan & Fennewald J. Molec. Biol, 206:295-304 (1989); Stark, et al, Cell, 58:779-90 (1989)); the yeast recombinases (Matsuzaki, et al, J. Bacteriol ., 172:610-18 (1990)); the B. subtilis SpoIVC recombinase (Sato, et al, J. Bacteriol. 172:1092-98 (199Oj); the FIp recombinase (Schwartz & Sadowski, J Molec. Biol, 205:647-658 (1989); Parsons, et al, J. Biol. Chem., 265:4527-33 (1990); Golic & Lindquist, Cell, 59:499-509 (1989); Amin, et al, J. Molec. Biol, 2\A:55-12 (1990)); the Hin recombinase (Glasgow, et al, J. Biol. Chem., 264:10072-82 (1989)); immunoglobulin recombinases (Malynn, et al, Cell, 54:453-460 (1988)); and the Cin recombinase (Haffter & Bickle, EMBO J., 7:3991-3996 (1988); Hubner, et al, J. Molec. Biol, 205:493-500 (1989)). Such systems are discussed by Echols (J. Biol. Chem. 265:14697-14700 (1990)); de Villartay {Nature, 335:170-74 (1988)); Craig, {Ann. Rev. Genet, 22:77-105 (1988)); Poyart-Salmeron, et al, {EMBO J. 8:2425-33 (1989)); Hunger-Bertling, et al.,{Mol Cell. Biochem., 92:107-16 (1990)); and Gregg & Madden {Mol Gen. Genet, 219:320-23 (1989)).

Cre recombinase has been purified to homogeneity, and its reaction with the loxP site has been extensively characterised (Abremski & Hess J. MoI Biol. 259:1509-14 (1984); Gopaul and van Duyne, (1999) Curr Opin Struct Biol 9:14-20). Cre protein has a molecular weight of 35,000 and can be obtained commercially from New England Nuclear/DuPont. The cre gene (which encodes the Cre protein) has been cloned and expressed (Abremski, et al, Cell 32:1301-11 (1983)). The Cre protein mediates recombination between two loxP sequences (Sternberg, et al, Cold Spring Harbor Symp. Quant. Biol. 45:297-309 (1981)), which may be present on the same or different DNA molecule. Because the internal spacer sequence of the loxP site is asymmetrical, two loxP sites can exhibit directionality relative to one another (Hoess & Abremski Proc. Natl. Acad. ScL U.S.A. 81:1026-29 (1984)). Thus, when two sites on the same DNA molecule are in a directly repeated orientation, Cre will excise the DNA between the sites (Abremski, et al, Cell 32:1301-11 (1983)). However, if the sites are inverted with respect to each other, the DNA between them is not excised after recombina-tion but is simply inverted. Thus, a circular DNA molecule having two loxP sites in direct orientation will recombine to produce two smaller circles, whereas circular molecules having two loxP sites in an inverted orientation simply invert the DNA sequences flanked by the loxP sites. In addition, recombinase action can result in reciprocal exchange of regions distal to the target site when targets are present on separate DNA molecules.

Modelling of chromosomal rearrangements

Chromosomal rearrangements are commonplace in cancers and are efficiently modelled by the method of the invention. The use of a recombinase-catalysed rearrangement at recombinase-specific sites in the absence of any read-through transcription permits the generation of realistic models of chromosomal rearrangements. Moreover, the method of the invention permits the generation of chromosomal rearrangements commonly found in cancers which cannot be generated using prior art methods such as the translocator method devised by the present inventors and the LoxP-STOP method due to the problem of read-through transcription using these prior art methods and the consequent production of embryonically lethal fusion proteins.

Many chromosomal rearrangements are known in the art, which can be modelled in accordance with the present invention. These include:

1. Chronic Myeloproliferative Diseases (MPD)

Agnogenic myeloid metaplasia Atypical Chronic Myeloid Leukaemia (aCML) Chronic Myelomonocytic Leukaemia (CMML) Chronic myelogenous leukaemia (CML) Essential thrombocythemia Idiopathic myelofibrosis Idiopathic thrombocythemia Juvenile Chronic Myelogenous Leukaemia (JCML) Polycytemia vera del(13q) in myeloid malignancies del(16)(q22) del(20q) in myeloid malignancies del(5q) in myeloid malignancies del(9q) solely i(l 7q) in myeloid malignancies idic(X)(ql3) ins(3;3)(q26;q21q26) inv(16)(pl3q22) inv(3)(q21q26) t(12;13)(pl2;ql2-14) t(12;22)(pl3;ql l-12) t(16;16)(pl3;q22) t(l;18)(qlθ;qlθ) t(l;3)(p36;P21) t(l;3)(p36;q21) t(l;7)(qlθ;plθ) t(3;3)(q21;q26) t(5;10)(q33;q21) t(5;l l)(q31;q23) t(5;12)(q33;pl3) t(7;ll)(pl5;pl5) t(8;14)(ql l;q32) t(8;22)(pll;ql l) t(9;12)(p24;pl3) t(9;12)(q34;pl3) t(9;22)(q34;ql 1) in CML t(Y;l)(ql2;ql2)

2. Myelodysplasia Syndromes (MDS)

1 Iq23 rearrangements in leukaemia 12p abnormalities in myeloid malignancies f W I ✓ WW t-ww / W W \J W ^

38

Childhood myelodysplastic syndromes Chronic Myelomonocytic Leukaemia (CMML) Congenital leukaemias Infant leukaemias Neonatal leukaemias del(l 3q) in myeloid malignancies del(16)(q22) del(17p) in myeloid malignancies del(20q) in myeloid malignancies del(5 q) in myeloid malignancies del(9q) solely i(17q) in myeloid malignancies idic(X)(ql3) ins(3;3)(q26;q21q26) inv(16)(pl3q22) inv(3)(q21q26) t(ll;16)(q23;pl3) t(l l;17)(q23;q25) t(12;13)(pl2;ql2-14) t(12;22)(pl3;ql l-12) t(16;16)(pl3;q22) t(16;21)(q24;q22) t(l;16)(ql l;ql l) t(l;18)(qlθ;qlθ) t(l;19)(pl3;pl3.1) t(l;3)(p36;p21) t(l;3)(p36;q21) t(l;7)(qlθ;plθ) t(2;ll)(p21;q23) t(3;12)(q26;P13) t(3;21)(q26;q22) t(3;3)(q21;q26) t(5;l l)(q31;q23) t(5;12)(q33;pl3) t(5;7)(q33;ql l) t(6;8)(q27;pl2) t(6;9)(p23;q34) t(7;12)(q36;pl3) t(8;13)(pl2;ql2) t(9;12)(q22;pl2) t(Y;l)(ql2;ql2)

3. Treatment Related Leukaemias (t- ANLL)

11 q23 rearrangements in leukaemia 11 q23 rearrangements in therapy related leukaemias 12p abnormalities in myeloid malignancies Biphenotypic Acute Leukaemia (BAL) Childhood myelodysplastic syndromes del(16)(q22) del(l 7p) in myeloid malignancies del(5 q) in myeloid malignancies inv(16)(pl3q22) t(10;l l)(pl2;q23) t(l l;12)(pl5;ql3) t(ll;15)(q23;ql4) t(ll;16)(q23;pl3) t(ll;17)(q23;pl3) t(l l;17)(q23;q25) t(ll;19)(q23;pl3.1) t(ll;19)(q23;pl3.3) t(l l;22)(q23;ql3) t(15;17)(q22;q21) t(16;16)(pl3;q22) t(16;21)(q24;q22) t(17;21)(ql l.2;q22) t(18;21)(q21;q22) t(19;21)(ql3.4;q22) t(l;21)(p36;q22) t(l;3)(p36;p21) t(l;3)(p36;q21) t(l;7)(p36;q34) t(l;7)(qlθ;plθ) t(3;l l)(p21;q23) t(3;l l)(q25;q23) t(3;l l)(q28;q23) t(3;21)(q26;q22) t(4;l l)(q21;q23) t(4;12)(ql l-q21;pl3) t(5;l l)(q31;q23) t(6;l l)(q27;q23) t(6;9)(p23;q34) t(8;16)(pl l;pl3) t(8;21)(q22;q22) t(9;ll)(p22;q23)

4. Acute Non Lymphocytic Leukaemias (ANLL)

11 q23 rearrangements in leukaemia 12p abnormalities in myeloid malignancies Acute Erythroid leukaemias Acute basophilic leukaemia Biphenotypic Acute Leukaemia (BAL) Congenital leukaemias Infant leukaemias MO acute non lymphocytic leukaemia (MO-ANLL) Neonatal leukaemias Systemic mast cell disease (SMCD) del(13q) in myeloid malignancies del(16)(q22) del( 17p) in myeloid malignancies del(20q) in myeloid malignancies del(5q) in myeloid malignancies del(9q) solely idic(X)(ql3) ins(3;3)(q26;q21q26) inv(14)(ql l;q32.1) inv(16)(pl3q22) inv(3)(q21q26) t(10;ll)(pll.2;q23) t(10;l l)(pl2;q23) t(10;ll)(pl3;q21) t(ll;16)(q23;pl3) t(l l;17)(ql3;q21) t(l l;17)(q23;pl3) t(l l;17)(q23;ql2) t(l l;17)(q23;q21) t(l l;17)(q23;q25) t(l l;19)(q23;pl3.1) t(ll;19)(q23;pl3.3) t(ll;22)(q23;ql l.2) t(12;13)(pl2;ql2-14) t(12;22)(pl3;qll-12) t(14;14)(qll;q32.1) t(15;17)(q22;q21) t(16;16)(pl3;q22) t(16;21)(pl l;q22) t(16;21)(q24;q22) t(17;21)(qll.2;q22) t(l;ll)(q21;q23) t(l;12)(q25;pl3) t(l;18)(q25;q23) t(l;19)(pl3;pl3.1) t(l;19)(q23;pl3) t(l;22)(pl3;ql3) t(l;3)(p36;P21) t(l;3)(p36;q21) t(l;7)(p36;q34) t(l;7)(qlθ;plθ) t(2;ll)(p21;q23) t(3;12)(q26;pl3) t(3;21)(q26;q22) t(3;3)(q21;q26) t(3;5)(q25;q34) t(4;l l)(q21;q23) t(4;12)(ql l-q21;pl3) t(5;l l)(q31;q23) t(5;l l)(q35;pl5.5) t(5;14)(q33;q32) t(5;17)(q35;q21) t(6;l l)(q27;q23) t(6;8)(q27;pl2) t(6;9)(p23;q34) t(7;l l)(pl5;pl5) t(7;12)(q36;pl3) t(8;16)(pll;pl3) t(8;21)(q22;q22) t(9;ll)(p22;q23) t(9;l l)(q34;q23) t(9;12)(q34;pl3) t(9;22)(q34;qll) in ANLL t(X;l l)(ql3;q23) t(X;6)(pl l;q23)

5. B-CeIl Acute Lymphocytic Leukaemias (B-ALL)

1 Iq23 rearrangements in leukaemia 12p rearrangements in ALL 9p Rearrangements in ALL Biphenotypic Acute Leukaemia (BAL) Congenital leukaemias High hyperdiploid acute lymphoblastic leukaemia Infant leukaemias Near haploid acute lymphoblastic leukaemia Neonatal leukaemias Severe hypodiploid acute lymphoblastic leukaemia del(6q) abnormalities in lymphoid malignancies dic(9;12)(pll-13;pll-12)

ins(5;ll)(q31;ql3q23) inv(14)(qll;q32.1) t(10;ll)(pl2;q23) t(10;ll)(pl3;q21) t(l l;17)(q23;pl3) t(l l;17)(q23;q25) t(ll;19)(q23;pl3.3) t(12;13)(pl2;ql2-14) t(12;21)(pl2;q22) t(14;14)(ql l;q32.1) t(14;18)(q32;q21) t(17;19Xq22;pl3) t(18;22)(q21;qll) t(l;18)(q25;q23) t(l;19)(q23;pl3) t(l;3)(p36;p21) t(2;ll)(p21;q23) t(2;14)(pl3;q32) t(2;18)(pll;q21) t(2;8)(pl2;q24) t(4;ll)(q21;q23) t(4;12)(qll-q21;pl3) t(5;ll)(q31;q23) t(5;14)(q31;q32) t(5;15)(pl5;qll-13) t(8;14) (q24;ql l) t(8;14)(qll;q32) t(8;14)(q24;q32) t(8;22)(q24;qll) t(9;12)(p24;pl3) t(9;12)(q34;pl3) t(9;22)(q34;ql I) Ui ALL

6. T-CeIl Acute Lymphocytic Leukaemias (T-ALL)

1 Iq23 rearrangements in leukaemia 12p rearrangements in ALL Ip32 rearrangements 9p Rearrangements in ALL Biphenotypic Acute Leukaemia (BAL) del(6q) abnormalities in lymphoid malignancies del(9q) solely dic(9;12)(pll-13;pll-12) inv(14)(ql l;q32.1) t(10;l l)(pl3;q21) t(10;14)(q24;ql l) t(ll;14)(pl3;ql l) t(l l;14)(pl5;ql l) t(l l;19)(q23;pl3.3) t(12;13)(pl2;ql2-14) t(14;14)(ql l;q32.1) t(l;14)(p32;ql l) t(l;19)(q23;pl3) t(l;3)(p32;p21) t(l;5)(p32;q31) t(l;7)(p32;q34) t(2;14)(pl3;q32) t(4;ll)(q21;pl5) t(4;ll)(q21;q23) t(5;14)(q35;q32) t(5;17)(ql3;q21) t(6;l l)(q27;q23) t(7;10)(q34;q24) t(7;l l)(q35;pl3) t(7;9)(q34;q32) t(8;14) (q24;ql l) t(9;12)(p24;pl3) t(9;22)(q34;ql 1) in ALL t(X;ll)(ql3;q23)

7. Non Hodgkin Lymphomas (NHL)

3q27 rearrangements in non Hodgkin lymphoma Anaplasic large cell lymphoma (ALCL) Angioimmunoblastic T-cell lymphoma . Burkitt's lymphoma (BL) Classification of T-CeIl disorders Diffuse large cell lymphoma Follicular lymphoma (FL) Hodgkin's disease Lymphoepithelioid lymphoma Lymphoplasmacytic lymphoma Mantle cell lymphoma (incomplete) Marginal Zone B-cell lymphoma Small lymphocytic lymphoma T-cell large granular lymphocyte leukaemia Waldenstrom's macroglobulinemia (WM) del(l Iq) in non-Hodgkin's lymphoma (NHL) del(13q) in non-Hodgkin's lymphoma del(17p) in non-Hodgkin's lymphoma (NHL) del(6q) abnormalities in lymphoid malignancies del(7q) in non-Hodgkin's lymphoma (NHL) dic(9;12)(pll-13;pll-12) inv(14)(qll;q32.1) inv(2)(p23q35) t(10;14)(q24;ql l) t(l l;14Xql3;q32) t(l l;18)(q21;q21) t(14;14)(ql l;q32.1) t(14;18)(q32;q21) t(14;19)(q32;ql3) t(18;22)(q21;ql l) t(l;13)(q32;ql4) t(l;14)(p22;q32) in non Hodgkin's lymphoma (NHL)

t(l;19)(q23;pl3) t(l;2)(q25;p23) t(l;3)(p36;p21) t(l;7Xq21;q22) t(2;14)(pl3;q32) t(2;18)(pll;q21) t(2;22)(p23;qll.2) t(2;3)(pl2;q27) t(2;3)(p23;q21) t(2;5)(p23;q35) t(2;8)(pl2;q24) t(3;13)(q27;ql4) t(3;14)(q21;q32) t(3;14)(q27;q32) t(3;22)(q27;ql l) t(3;4)(q27;pl3) t(3;Var)(q27;Var) in non Hodgkin lymphoma t(6;8)(ql l;qll) t(6;8)(q27;pl2) t(7;10)(q34;q24) t(8;13)(pl2;ql2) t(8;14) (q24;qll) t(8;14)(q24;q32) t(8;22)(q24;qll) t(9;14)(pl3;q32) t(X;2)(qll;p23)

8. Chronic Lymphoproliferative Diseases (CLD)

Iq rearrangements in multiple myeloma B-cell prolymphocyte leukaemia (B-PLL) Chronic lymphocytic leukaemia (CLL) Fibrogenesis imperfecta ossium Hairy Cell Leukaemia (HCL) and Hairy Cell Leukaemia Variant (HCL-V) Multiple myeloma Plasma cell leukaemia (PCL) Splenic lymphoma with villous lymphocytes T-cell prolymphocytic leukaemia (T-PLL) del (13q) in chronic lymphoproliferative diseases del (6q) in Multiple Myeloma del(l 3q) in multiple myeloma del(6q) abnormalities in lymphoid malignancies inv(14)(ql l;q32.1) t(ll;14)(pll;q32) t(ll,14)(ql3;q32) t(ll;14)(ql3;q32) in multiple myeloma t(14;14)(ql l;q32.1) t(14;18)(q32;q21) t(14;19)(q32;ql3) t(18;22)(q21;qll) t(l;3)(p36;q21) t(2;14)(pl3;q32) t(2;18)(pll;q21) t(4;14)(pl6;q32) t(9;14)(pl3;q32)

See also Mitelman's Catalog of Chromosome Aberrations in Cancer '98, John Wiley & Sons, NY5 USA; ISBN 0-471-17603-6.

The present inventors consider that the inverter model could be also used as a conditional transgenic approach for making 'normal' (non-fusion) transgenes which are conditional on inversion.Advantageously, the disease modelled by the methods of the invention is leukaemia. Preferably, it is a myeloid or lymphoid leukaemia which produces an Ews-ERG fusion protein. Most advantageously, the leukemia which produces Ews-ERG fusion protein will manifest itself in the form of clonal T-cell malignancies. Further details are provided in Example 2. Sequence from the rearrangement sites is obtained from publicly-available sources, such as GenBank, and used to construct homologous recombination vectors which insert recombinase specific sites, such as lox sites, into the regions of interest.

Disease models of chromosomal translocations according to the invention replicate the aetiology of human chromosomal translocations with surprising fidelity. Preferably, the translocation according to the inventions results in an Ews-ERG fusion, which is associated with leukaemia in humans. As in humans, the chromosomal translocation between mouse Ews and Erg results in leukaemia. The data from the inventors study shows that mice with the compound genotype comprising an invertor cassette according to the invention and also Ragl-Cre develop leukaemia only if Cre recombinase is expressed. Further the inventors have shown that the Ews-ERG chromosomal translocation results in leukemia in 25/29 mice tested. Thus, the inventors have shown that the Ews-ERG fusion protein is crucial for malignancies in invertor mice as malignancies only arise in mice with both the Ews-ERG and Ragl- Cre genes. (Fig 2).

The presence of thymomas in all afflicted Ews-ERG; Ragl-Cre mice suggested that the hematological malignancy in these mice comprised T cells. This was confirmed by FACS analysis of T cell surface markers and by determination of T cell or B cell receptor gene rearrangement status. The majority of normal mouse thymic T cells express both the CD4 and CD8 surface markers (double positive or DP cells), as well as the pan-T cell marker Thyl (CD90) (wt, Fig. 3). The situation with the thymomas of invertor mice varied from mainly CD4+CD8+ DP cells (e.g. M2, Fig. 3), mainly CD8+ SP cells (e.g. M3, Fig. 3) or mainly DN cells (e.g. M4, Fig. 3). Thirteen of the cohort of (25) leukemic invertor mice were analysed for their thymic T cell profile by FACS (summarised in Table 2). While we found either CD8+ SP or CD4+, CD8+ DP thymomas, none showed a CD4+ SP phenotype. Two of the thymomas had a mixed phenotype comprising mainly CD8+ SP cells, with a subpopulation of DP cells (M3 and M6, Table 2).

Detection of tumour formation The presence of tumours in animals according to the invention is possible by a variety of methods. These include genetic testing, testing for tumour markers, physical observation, cytological assays and the like.

Tumour markers are substances that can be detected in higher than normal amounts in the blood, urine, or body tissues of animals with certain types of cancer. A tumour marker may be produced by the tumour itself, or by the body in response to a cancer presence. When diagnosing cancer, blood and tumour tissue biopsies may be tested. For example, the following markers are known to associated with tumours:

Tests for tumour markers are available commercially; for example, immunoassays for particular polypeptides may be used to assess their presence and/or amounts in biological samples.

Cytological assays are frequently used in the detection of leukaemia, for example by analysis of various tissues post mortem or analysis of bone marrow tissue obtained by biopsy. Another technology useful for detecting leukaemia, multiparameter immunological detection by flow cytometry, is claimed to be capable of detecting one leukaemic cell among 10,000 normal cells. The principle behind this assay is that leukaemia cells display certain surface, cytoplasmic and nuclear leukocyte antigens which normal cells do not. A fluorescent dye is tagged to an antigen-specific monoclonal antibody, incubated with a sample of bone marrow aspirate from a leukaemic animal, and the cells monitored by flow cytometry for the attachment of fluorescent dyes to the cells. If the cells fluoresce, they possess leukaemia antigens (see Coustan-Smith etal, Lancet 1998; 351: 550-4).

EWS-ERG is a promiscuous fusion protein able to mediate myeloid, lymphoid or mesenchymal tumorigenesis.

Human cancers of mesenchymal origin involves the EWS gene with various partner genes in different cell settings, such as in the chromosomal translocation t(21;22)(q22;q21) where EWS fuses with ERG to encode a novel EWS-ERG fusion protein (14). Thus, the present inventors considered that EWS-fusion proteins can be involved in tumor formation of various cell types and that restrictions in humans are dictated largely by the cells in which the specific chromosomal translocation occurs. The present inventors tested this hypothesis, and at the same time validated the invertor chromosomal translocation mimic approach, by assessing the possible universality of EWS-ERG as an oncogene. The Ews-ERG invertor allele was activated in lymphocytes, using the Ragl-Cre knock-in mice (35), to determine if lymphoid malignancies would arise.

It was found that invertor Ews-ERG mice develop clonal T cell malignancies. The T cells were of varying phenotypes but the majority of cases expressed CD8 on the cell surface and had productive V-D- Jβ rearrangements, indicative of mature T cells. Since Cre-mediated inversion through loxP sites recreates the loxP sequence, there is potential for the cassette to flip back and forth with continued Cre expression. This had no obvious deleterious effect on turnorigenesis in the Ews-ERG invertors as judged by various criteria (tumor penetrance, consistent T cell phenotype, pathology). The potential Cre-mediated flip-flop inversion of the ERG cassette may partly explain why predominantly mature T cells in the overt tumors were observed (Table 2 and 3) since CD 8+ or CD4+ T cells express lower levels of Ragl-Cre 35. Thus the invertor allele of transformed cells may be fixed in the correct orientation for synthesis of Ews-ERG mRNA when Rag-Cre is no longer effective. On the other hand, the lack of CD4+ (SP) thymomas is difficult to encompass with this explanation, suggesting that other biological mechanisms may be involved in this bias.

The Ragl-Cre mouse line used constitutively expresses Cre in lymphocytic cells (35) causing T cell malignancies to arise. This implies that the Ews-ERG fusion in lymphoid cells causes T lymphocyte tumors and that the EWS/FUS fusion proteins could be effective in a spectrum of cell types, outside those normally seen in human malignancies (i.e. sarcomas and myeloid leukemias). But in man, the cell in which an EWS/FUS associated chromosomal translocation occurs could be crucial in deteπnining if a tumor arises or not. In some cells, it may be that EWS/FUS fusions are lethal and those cells acquiring a translocation would die; in others the fusion protein may be tolerated and thus may become tumors, hi this respect, the absence of B cell tumors in the Ews-ERG invertors is of interest as both B and T cells should undergo inversion of the Ews-ERG cassette, since Ragl-Cre is expressed in both cell types (35).

The absence of B cell tumors may reflect toxicity of the fusion protein for B cells. Alternatively, it may mean that Ews-ERG does not function in B cells or that the development of B cell tumors is inhibited by the T cell malignancies which may arise earlier and then dominate the lymphoid compartment.

MLL-AF4 fusion proteins are commonly found in patients suffering from Leukemia.

The recurrent chromosomal translocation t(4;l l)(q21;q23) is the most common abnormality in acute leukemias of patients less than 1 year old, and overall the most common MLL fusion protein, giving rise to B lineage tumors invoking a poor prognosis. New methods to combat these leukemias are needed and a mouse model of MLL-AF4 tumorigenesis is needed. Modeling MLL-AF4 effects in mice has been hampered by embryonic lethality due to expression of MII-AF4 knock-in alleles, even using /oxP-STOP conditional alleles. The present inventors have obviated this problem using the invertor mouse conditional technology in which MII-AF4 fusion cannot take place until Cre expression occurs, giving many options for cell-specific generation of MU-AF4 fusion. The results show that the MII-AF4 invertor mice develop B cell tumors with high penetrance providing an ideal model for studying MLL- AF4 biology.

The invertor technology has allowed the first generation of B cell specific models for MLL-AF4. This is a valuable addition to the repertoire of MLL-models allowing studies of secondary mutations co-operating with MLL-fusions, lineage specific target genes and an important element to development of new therapies to target MLL- leukemia. This last point is of particular clinical relevance as the prognosis of MLL- AF4 associated leukemia is very poor, especially if it occurs before 1 year of age (Pui et al., 2002) and new drugs are needed to treat this leukemia.

More details of this model are provided in Example 5 herein. Uses of transgenic animals according to the invention

The armoury of approaches to chromosomal translocation mimics includes the knock- in model, the conditional knock-in model, the translocator model (28-30) and the invertor model described here. The new invertor mouse model adds an approach to generation of gene fusions, mimicking the consequence of chromosomal translocations, which is fully dependent of expression of Cre recombinase and inversion of the floxed intron-cDNA cassette. No fusion mRNA can be produced until inversion has taken place as the acceptor splice site is incorrectly orientated prior to inversion. This approach is applicable to any fusion gene using the transfer cassettes described herein. The invertor approach overcomes an inability to generate 'standard' knock-in Ews and MLL fusion alleles in ES cells and precludes any leakiness due to read through transcription, which can occur with /oxP-STOP conditional knock-ins, since the invertor approach gives an absolute conditional allele, dependent on Cre recombinase.

Recently, de novo translocations have been made in mice using the translocator model which gave rise to leukemias (30). This model is ideal as the abnormal karyotype of the translocator mice is created somatically by Cτe-loxP recombination. Translocator mice can only be made when the two genes into which loxP sites are targeted are orientated on the relevant mouse chromosomes in the same way to generate 5 '-3' exon fusion following translocation. The expanding sequence data available on the mouse genome has revealed that the transcription orientation of some genes equivalent to those that fuse in human tumor associated chromosomal translocations, differ in mouse and this would result in dicentric chromosomes in the translocator model, hi such situations, the invertor mice can be used as an alternative. The invertor methodology is simple and reliable since no fusion mRNA can be generated until Cre-mediated inversion. Possible effects of re-inversion could be further minimised in future by using inducible, cell-specific Cre expression, such as might be achieved with the Cre- oestrogen receptor-tamoxifen induction methods (42). Alternatively, use of additional recombination sites such as frt sites for FLP recombinase activity, could be used to delete one loxP site and render the invertor cassette stable in one orientation.

The de novo chromosomal translocations in mice are the closest to natural, human chromosomal translocations of any model thus far. The finding that inter- chromosomal translocations between Ews and ERG and MLL and AF4 respectively in mice leads to leukaemia suggests that the invertor method of the present invention can be employed, with appropriately specific Cre expressing mice, to any pair of genes as long as the orientation to the centromere does not result in dicentric aberrant chromosomes. Moreover, the advantage of the invertor method described herein is that read-through transcription is prevented in the absence of gene translocation mediated by Cre expression which is induced in adult mice, hi this way the production of potentially embryonically lethal fusion proteins generated by read-through transcription of an inserted gene cassette into a gene of interest is avoided in gene- targetted cells such as ES cells is avoided.

The time course of tumorigenicity in the invertor mice suggests that Ews-ERG is necessary but not sufficient for cancer in this model, and thus that 2nd mutational event(s) may have to occur to fix the overt tumor. As a consequence the Ews fusion protein would no longer be needed for tumor viability. It will thus be valuable to evaluate whether the function of the EWS-ERG fusion protein (or for that matter, any of the EWS or FUS fusions) remains essential for the tumor at the time of overt tumor presentation and therefore whether it is a suitable target for cancer therapy.

Engineering mouse models of chromosomal translocations i.e. knock-in, invertor and translocator mice, should be useful as a pre-clinical setting to test new therapeutic approaches, hi particular, it is noteworthy that chromosomal translocation products are invariably intracellular (reviewed in 43), anti-cancer reagents must therefore work in the intracellular milieu and drug targeting is an important problem. Thus one motivation for generating mouse models of chromosomal translocations is for potential use as a pre-clinical setting, prior to then: translation into use with patients. Several new therapeutic regimes are available but each is at a developmental stage which requires careful experimental analysis in vivo to assess not only efficacy but also possible side effects (toxicity). In addition, there is the all-important issue of delivering reagents, such as for siRNA or intracellular antibodies, to the tumor cells (44). An approach to these crucial points can be made in mouse models, as a precursor to use in humans.

Thus in a further aspect the present invention provides a method for the testing of an anti-tumour agent which method comprises the steps of: (a) providing a non-human animal tumour model according to the invention, (b) administering to that model one or more potential anti-tumour agents; and (c) testing for a change in one or more characteristics of tumour/s comprising the model according to step (a).

In a preferred embodiment of the above aspect of the invention step (c) (testing for a change in one or more characteristics of tumour/s comprising the model according to step (a)) involving the testing for a change in any one or more characteristics of one or more tumours in the group consisting of the following: size, morphology, presence and/or levels of one or more tumour markers and cytological assays.

Drugs used routinely for the treatment of cancer currently include the foloowing: •Classic cytotoxics: these cover alkylating agents, antibiotics including topoisomerase inhibitors, plant alkaloids, taxanes, and antimetabolites; and • Sensitizers: ie chemosensitisers, which are agents that increase the sensitivity of tumour cells to chemotherapeutic agents (radiosensitisers will not be covered in this article).

Classic cytotoxics Alkylating agents Alkylating agents act by binding to guanine on the DNA strand and damaging the genetic material of the cancer cell, often by creating cross-links between the guanine and its partner base, cytosine, on the complementary DNA strand. Such cross-linking prevents separation of the DNA strands, an event which must occur for cell division. Outgrowth of resistant cancer cells is common, and is probably due to the upregulation of enzymes, such as 0(6)-alkylguanine DNA transferase, which break the alkyl bond. Hence the development of therapeutic approaches which seek to downregulate the expression of these resistance-associated enzymes during the course of chemotherapy.

There are over 30 marketed alkylating agents, including drugs such as carmustine, dacarbazine, fomustine, temozolomide, lomustine, streptozotin and procarbazine, as well as the well-known agent cisplatin.

Anti-metabolites

These drugs affect dividing cells; they may be directly incorporated into DNA, thus disrupting the cell's capacity to function, or they may interfere with metabolic pathways necessary for DNA synthesis. Marketed antimetabolites include the 5-FU derivatives doxifluridine and capecitabine, and methotrexate and gemcitabine.

Antimetabolites which fall into the former group include the pyrimidine analogues and the purine analogues (pyrimidines and purines are the two types of base found in the DNA strand). Analogues of pyrimidines include doxifluridine and the well-known 5- fiuorouracil, which replace thymine and cytosine bases. Analogues of purines include thioguanine and mercaptopurine, which replace adenine and guanine bases. Antimetabolites which fall into the latter group include the antifolates, which inhibit folic acid metabolism thereby inhibiting DNA synthesis. These include inhibitors of dihydrofolate reductase and methionine transmethylase, such as methotrexate. There are over a dozen marketed antimetabolites.

Plant alkaloids

These are represented by the vinca alkaloids, which bind to microtubules (sub-cellular structures which must function correctly in order for cell division to occur). Interference with microtubules interferes with cell division and therefore interrupts tumour growth. Marketed alkaloids include vincristine, vinblastine, vindesine and vinorelbine; other microtubule poisoning agents are under development.

Taxanes

Like the vinca alkaloids, taxanes are plant-derived molecules which target the microtubules. There are only two marketed taxanes, paclitaxel and docetaxel. Rather than preventing appropriate microtubule formation, the taxanes promote it, but then prevent microtubule dissociation. This in turn prevents cell division, leading to a kind of paralysis of the cancer cell, rather than to cell death.

The best known of these agents is paclitaxel, which facilitates the assembly of intracellular microtubules which in turn inhibit cell division and therefore prevent the proliferation of tumour cells and growth of tumours. Paclitaxel appears to be effective in ovarian lung and breast tumours. Side effects appear to be relatively minor, being mainly limited to hypersensitivity reactions which can be largely prevented by premedication. Other more rare side effects are peripheral neuropathy and neutropenia.

Anti-angiogenic agents

Tumours cannot grow to any significant size without forming their own blood supply. Agents which prevent the formation of new blood vessels therefore would be expected to prevent the progression of metastatic disease. Many agents of anti-angiogenic potential are proteins or peptides and therefore fall outside the remit of this article. However, there are some small molecules being developed which may inhibit new vessel formation. Combretastatin is a small molecule that is said to target the tumour vasculature; it appears to directly kill angiogenic tumour cells by the induction of apoptosis.

Antibiotics

Antibiotics Some antibiotics have a degree of cancer cell-selectivity in that they either inhibit enzymes involved in the synthesis of DNA or else they directly interfere with DNA. Anthracycline antibiotics, which include daunorubicin and doxorubicin, are antibiotics which are have the ability to inhibit topoisomerases. Topoisomerases are enzymes which unwind supercoiled DNA in a step that is an essential precursor for DNA translation, and replication / cell division. Topoisomerase inhibitors interfere with this step, for example by preventing the DNA re-associating after the unwinding event (eg etoposide) and hence have a direct action on dividing cells. There are over 10 marketed topoisomerase inhibitors including actinomycin D and mitomycin C.

Chemosensitisers

Many of these agents are aimed at decreasing the capability of tumour cells to develop or maintain resistance to anticancer drugs. Such resistance often is mediated by a single gene (MDR-I), and therefore may in theory be inhibited by any drug which targets the product of that gene. Some drugs which have been used to combat the MDR-I phenotype to date include cyclosporin and verapamil; several others are in development.

Some drugs have been developed which are sensitive to light of particular wavelengths. These can be administered systemically; subsequent exposure of a localised area of tissue to light results in localised conversion of the drug into a short¬ lived cytotoxic form (eg by generation of free radicals). This may be a successful route for the treatment of some inoperable solid tumours; however, it is unlikely that photodynamic therapy will be converted into a treatment for disseminated disease.

The testing of these and other agents in cancer treatment may be performed using the cancer models of the invention.

The invention is further described, fur the purpose of illustration, in the following examples:

EXAMPLES. Example 1.

Methods Generation of targeting constructs and homologous recombination. Ews knock-ins An Ews genomic clone was isolated from a 129 λ phage library 45 and a targeting vector made extending from the Xbal site on the 5' side of exon 7 to the EcoRI site on its 3' side (Fig. IA). This initial Ews targeting vector was made in two steps in pBSpt. First, the 5' Xbal-BamHI 2Kb fragment was cloned into Sall-BamHI vector sites with Sail and Xbal sites filled-in to eliminate the Xbal site. Second, the 3 ' BamHI-EcoRI Ews 4.6Kb fragment was cloned into the BamHI-Clal sites of pBSpt with CIaI and EcoRI sites filled in to eliminate the EcoRI site. MCl-tk 46,47 was cloned into the EcoRV site for negative selection of the homologous recombinants in the ES cells. The knock-in clones to generate Ews- fusions with AF9, FUl, CHOP and A TF were made using a transfer vector (clone pC2A-neo, Figure 5) into which respective cDNA sequences were cloned as Notlfianked PCR fragments. The mRNA sources for RT- PCR amplification were from cell lines; the mouse erythroleukemia MEL (FHl), the human erythroleukemia HEL (AF9), mouse 3T3L1 fibroblasts (Chop) and the human T cell leukemia Jurkatt (ATF). The sequences of the cDNA cassettes are given in Figure 6. Each transfer cassette comprised the corresponding fusion part of the EWS partners, a polyA signal from SV40 and a selectable marker (neo) for selection of homologous recombinants and the transfer cassettes were cloned into the BamHI site in Ews exon 7 (Fig.l).

Knock-in was carried out by homologous recombination as described (45). Two probes used for detecting homologous recombinants in filter hybridizations of ES cell genomic DNA were cloned into pBSpt (Fig. 1); the 5' probe was a 1.4Kb Kpnl fragment and the 3' probe was a 0.25Kb EcoRI-BamHI fragment. These were used to detect homologous recombinants in EcoRI or BgIII digested genomic DNA for 5' or 3' probes respectively. The targeting vectors also had the HSV tk gene for negative selection of non- homologous integrations (46,47). The cDNA transfer cassettes were cloned into the Ews exon 7 BamHI site within the XbaIEcoRI fragment indicated in Figure IA. These were subsequently transfected into the ES cell line CCB as described, G418-gancyclovir-resistant colonies grown and DNA was prepared and homologous recombination determined using filter hybridization 48. Targeted clones for Ews-AF9 or the control targeting with the transfer vector only (pC2-neo, Figure 5) were expanded, verified by hybridizing fresh DNA preparations. These were injected into C57BL/6 blastocysts and chimaeras assessed by coat colour. High percentage male chimaeras were bred with C57BL/6 females and the genotypes of pups analysed to determine germline transmission.

Invertor knock-in The ERG inversion knock-in cassette was prepared using the pC2A-neo transfer cassette (Figure 5A). RT-PCR was carried with cDNA made from human colon carcinoma COLO320 mRNA, and primers allowing cloning into the unique Notl site of pC2A-neo. The sequence of the ERG segment, corresponding to the part of ERG found in EWS-ERG fusion (15), was determined (see Figure 6E) and the resulting p- ERG-C2A-neo transfer cassette was cloned as a BgIII fragment into the BamHI site of pLOXKI (Figure 7 online) prepared in the pMG2 vector (Figure 7B and C; pMGl and pMG2 was prepared by mutagenesis of pBSpt to incorporate respectively one or two Seel sites; this restriction enzyme recognizes an 18bp site). An EcoRI 120bp loxP- containing fragment, from pPGK-loxP-hygro 29, was blunt end cloned into the filled- in Notl site of pMG2 (Figure 7D; both EcoRI and Notl sites are eliminated), followed by a 530bp Xbal-BamHI fragment of intron 4 from the mouse Af 4 gene, to include the distal region of the intron up to and including the acceptor splice site (this fragment was PCR amplified as an Xbal-BamHI fragment, from mouse 129 genomic DNA, Figure 7E). A further loxP fragment was cloned as blunt EcoRI fragment into the EcoRV site of the vector (Figure 7F online), in an orientation compatible with Cre recombinase mediated inversion of the cassette. This yielded the final pLOXKI cloning vector (Figure 7G) for receipt of the ERG transfer cassette which was cloned as a BgIII fragment into the unique BamHI site of pLOXKI. The overall structure of the ERG inversion cassette is shown in Figures IA and Figure 8 A, and this was cloned into the unique Xbal site of Ews (XbaI)-EcoRI targeting vector (described above, Fig. IA) as a blunt ended Scel-Clal fragment, leaving an Xhol site for linearization prior to transfection into ES cells. The orientation of the ERG inversion cassette was 3'-> 5' with respect to Ews and this was determined by restriction site mapping and genomic DNA sequencing (Figure 7A depicts the organisation of the ERG inversion cassette before and after Cre-mediated inversion in the Ews targeting vector). Homologous recombinant knock-in alleles were detected and isolated as above. These ES cells were injected into C57BL/6 blastocysts, chimaeric mice produced and germline transmission of the inversion knock-in allele obtained. These mice were crossed with the line of mice expressing Cre from a Ragl knock-in 35 and specimen spleen DNA prepared and a PCR product made to determine sequence of the Cre-mediated inversion junction (Fig. ID). The previously published Rag-Cre knock-in allele (35) was made as outlined in Figure 8A-C, which also shows relevant junction sequences of three Cre cassettes suitable for knock-in of Cre into any gene of interest (Figure 8D).

Molecular analyses Genomic DNA was prepared from tissues using proteinase K digestion and phenolchloroform extraction and filter hybridizations was carried out with radiolabelled probes as described (48,49). The probes used to detect antigen receptor gene rearrangements were a heavy chain Ig μ intron enhancer probe (37) and a T cell receptor Jβ2 probe (36). T cell receptor V-D-Jβ junction sequences were obtained as described (38,39) using pools of Vβ primers combined with a single Jβ2 primer to amplify products covering the junction of the joints between Vβ, Dβ and Jβ2, followed by direct sequencing of the PCR product.

Analysis of leukemia/lymphoma A cohort of mice was established by inter-breeding Ews-ERG invertor line with a Ragl-Cre line (35) to establish littermates with Ews-ERG + Ragl-Cre or Ews-ERG genotypes. Genotypes were determined using filter hybridization of genomic DNA from a small tail biopsy. The health status of these mice were monitored and if signs of ill-health appeared, mice were sacrificed and a post-mortem was carried out. Tissue samples were removed for single cell preparation for determination of surface protein expression phenotype, for nucleic acid preparation or for fixation in 10% formalin for histology. Blood smears were mice prepared on glass slides and stained with May- Grϋnwald-Giemsa (MGG). For histology,, fixed tissues were embedded in wax and 0.4μ sections made, stained with haematoxylin and eosin (H&E) after mounting on slides. FACS analysis was conducted using a FACSCalibur with fluorescent antibodies purchased from BD Biosciences. Data was analysed with CellQuest software. Western protein detection was carried out as described 50 using 18μg protein per lane and proteins fractionated on 4-20% SDS-PAGE. The separated proteins were electrotransferred to Protran membranes (Schleicher & Schuell) and specific proteins detecting using anti-Ews antibody (Santa Cruz; detecting the amino terminus of Ews) or anti-ERG antibody (Santa Cruz; detecting the carboxy terminus of ERG). Antibody bound to filter was detected using horse radish peroxidase (HRP) detection and ECL as described by the Manufacturer (Amersham).

Example 2.

Conditional inversion of the ERG gene and fusion with Ews The use of homologous recombination to generate oncogene fusions was established with the creation of the MII-AF9 mouse model 24. We employed a similar technology to make Ews fusions of tibte type found in specific human sarcomas and leukemias. Transfer cassettes were made into which RT-PCR fragments of relevant isW-fusion partners were cloned into an Ews genomic fragment for selecting homologous recombinants in embryonic stem cells (ES cells) (the details of tihree transfer cassettes and targeting clones are given in Figs. 5 and 6;). Ews fusions with Chop, ATF and FUI were made and an Ews-AF9 fusion (not having been observed as a human EWS fusion genes) made as a control (sequences of these fusions cDNAs are shown in Figure 6). While we were able to obtain homologous recombination knock-in clones with either a vector carrying the transfer cassette only or the Ews-AF9 control fusion, no clones were obtained with any of the Ews-fusions equivalent to those naturally found in human cancers (Table IA). Furthermore, chimaeras could be made with the Ews cassette and the Ews-AF9 knock-in (Table IB) and the Ews-AF9 chimaeras were bred to give heterozygous carriers of the knock-in allele (data not shown). We conclude that Ews-fusions of the type naturally associated with human cancers are lethal in ES cells and therefore that a conditional knock-in model was required for this system.

An alternative approach developed for knock-in technology is the conditional transcription /OJCP-STOP method, developed to overcome embryonic lethality of fusion proteins 31. We attempted this approach with Ml gene fusion knock-ins, but found that even a low level of transcription read-through of the knock-in allele was sufficient for lethality (unpublished data). This finding necessitated a more robust condition form of knock-in, in which transcription of the fusion gene is impossible until a conditional activation step is invoked. LoxP-Cre recombination systems are known to facilitate inversion of sequences added into the genome 32-34, suggesting that knocking-in an inverted cassette flanked by loxP sites to create an absolute conditional allele in which Cre mediated inversion was a mandatory step for protein fusion. This has the additional advantage that cell-specific Cre expression will allow spatial timing to the generation of the fusion allele.

We tested this approach by knocking-in an ERG invertor cassette in the Ews gene. The invertor cassette was constructed to include a short segment of intronic sequence (from intron 4 of the mouse Af4 gene) including the acceptor splice site, the ERG cDNA made by RT-PCR of the relevant part of ERG usually involved in human translocations, a polyA site and the MCl-neo-pA segment for selection of homologous recombinant clones, all flanked by loxP sites with an orientation compatible for inversional recombination by Cre (Fig. IA and detailed sequences of the components of the inversion cassette are shown in Figs. 5 and 7). Gene targeting in ES cells was used to introduce the invertor cassette, cloned in an Ews genomic fragment, into the mouse Ews gene downstream of the donor splice site of exon 7 (Fig. IA). The selected targeted Ews allele (identified using the external 5' probe and verified with the external 3' probe, Fig. IA) had the ERG invertor cassette orientated 3' to 5' with respect to Ews, thus making a splice from Ews exon 7 donor site to the acceptor site of the invertor cassette impossible. After Cre-mediated inversion to give the correct transcription orientation, the correct splice occurred (illustrated in Fig. 1C) as confirmed by transiently transfecting targeted ES cells with a plasmid encoding Cre recombinase and amplifying the genomic junction of the the Ews-ERG allele (sequence shown in Fig. ID). By sequencing the RT-PCR product from these cells, we also confirmed an in-frame sequence of Ews fused to ERG (Fig. IE). The targeted ES cells were injected into blastocysts and these yielded normal numbers of chimaeras which gave rise to heterozygous carriers of the Ews-ERG invertor allele (invertor mice).

Example 3: The Ews-ERG fusion protein causes T cell leukemia in invertor mice.

We investigated the ability of Ews-ERG fusion protein to contribute to leukemogenesis by causing the inversion of the Ei?G-containing cassette in lymphoid cells using a Ragl-Cre knock-in mice 35 (Figure 8 illustrates the plasmid pC2A-neo used for knocking-in the Cre gene into the Ragl gene; Figure 8C shows the junction sequence of Ragl and Cre; Figure 8D shows other transfer cassettes suitable for knock-in homologous recombination of Cre recombinase into genes of interest). 25 out of the cohort of 29 mice (86%) mice carrying both Ews-ERG and Ragl-Cre genes developed leukemia associated with thymoma, whereas the Ews-ERG-oxήy cohort (20 mice) did not display leukemias (Fig. 2). Blood smears at the time of sacrifice typically contained elevated numbers of leukocytes, with lymphoid morphology of different maturities including lymphoblasts (Fig. 2). In addition, bone marrow sections showed high levels of infiltrating lymphoblasts (Fig. 2). The diseased mice all had large thymomas, often associated with splenomegaly (Table 2), and the histology of spleens showed variously either complete loss of normal architecture (loss of demarcated white and red pulp), partial loss or normal architecture (Fig. 2C, M9, M21 and M6 respectively). Similarly, the level of infiltration of leukemic lymphocytes into liver and kidney varied, with some having large amounts of perivascular deposits and others marginal levels. All mice had thymomas with homogeneous structures of leukemic cells.

The Εws-ΕRG fusion appears crucial for malignancies in the invertor mice as these only arose in mice with both the Ews-ERG and Ragl-Cre genes (Fig. 2). The presence of Ews-ERG mRNA was confirmed in spleens of these mice identical with that in ES cells (Fig. IE and data not shown). In addition, the steady state orientation of the loxP- flanked ERG cassette was studied using filter hybridization and we found that the ERG sequence had become inverted into the 5' to 3 Orientation with respect to the Ews gene, in each case of thymoma (Fig. 4A and Table 2). The other tissues examined have little evidence of an inverted band. Finally, the presence of the Ews-ERG fusion protein was shown by Western blotting of thymoma proteins using antibodies binding to either Ews or ERG which detected respectively the normal mouse Ews or Erg proteins and the Ews-ERG fusion molecule in the thymoma T cells of an Ews-ERG; Ragl-Cre mouse (Fig. 4D).

The presence of thymomas in all afflicted Ews-ERG; Ragl-Cre mice suggested that the hematological malignancy in these mice comprised T cells. This was confirmed by FACS analysis of T cell surface markers and by determination of T cell or B cell receptor gene rearrangement status. The majority of normal mouse thymic T cells express both the CD4 and CD8 surface markers (double positive or DP cells), as well as the pan-T cell marker Thyl (CD90) (wt, Fig. 3). The situation with the thymomas of invertor mice varied from mainly CD4+CD8+ DP cells (e.g. M2, Fig. 3), mainly CD8+ SP cells (e.g. M3, Fig. 3) or mainly DN cells (e.g. M4, Fig. 3). Thirteen of the cohort of (25) leukemic invertor mice wereanalysed for their thymic T cell profile by FACS (summarised in Table 2). While we found either CD8+ SP or CD4+, CD8+ DP thymomas, none showed a CD4+ SP phenotype. Two of the thymomas had a mixed phenotype comprising mainly CD8+ SP cells, with a subpopulation of DP cells (M3 and M6, Table 2).

The clonality of these T cell tumors was assessed by filter hybridization of genomic thymoma DNA with a T cell receptor β probe (Tcrb) from the Jβ2 region 36 (diagrammatically shown in Fig. 4B). The probe detects a 5Kb band in non-lymphoid DNA corresponding to the intact Tcrb gene; if V-D-J or D-J joins have occurred, new restriction fragments are created giving rise to 'rearranged' bands on the hybridization autoradiograph. Figure 4B shows the rearrangement status of 12 of the Ews-ERG thymomas. All except one (M9) showed at least one rearranged allele and several have two rearranged alleles (e.g. Ml 8). Variable amounts of residual germline allele were 03337

67

present. The Tcrb2 rearrangement status of the cohort of Ews-ERG leukemic mice is summarised in Table 2. In addition, the status of the immunoglobulin locus was examined with a probe from the region between JH and Cμ segments 37 (Fig. 4C). Single Igh allele rearrangements were observed in about half of the cohort (Table 2) including ones in which FACS analysis and Tcrb clonality hybridizations showed the tumors to be clonal T cell malignancies. The Igh rearrangements are therefore likely to be abortive rearrangements.

Confirmation that the Tcrb rearrangements observed in the clonality hybridizations were due to productive V-D-J join, was made by sequence analysis of the Tcrb genes in four of the cases. PCR amplification from thymoma DNA was carried out with pools of Vβ primers and a Jβ2 reverse primer (primer sequences from 38,39) and the product fractionated on agarose to determine the presence of a single amplified band, hi turn, each band was eluted and the sequence obtained using the Jβ2N primer 39 and the sequences identified by comparison with the IGMT database 40,41. Table 3 shows the V-D-J junctional sequences indicating the presence of productive T cell receptor β genes. In all cases, N-region diversity is present between the V-D and D-J junctions. In the thymoma DNA from M2, a nonproductive join has also occurred, in addition to the productive one, resulting in an out-offrame joint.

The invertor model of chromosomal translocations has successfully been applied with the creation of an Ews-ERG invertor allele in mice. Our data show that the strategy is effective in producing both cell-specific fusion protein and allowing for the manifestation of the oncogenic phenotype of the protein (in this exemplification clonal T cell malignancies).

This chromosomal translocation mimic is applicable to any fusion gene of interest using the set of vectors described herein.

Example 4: The Ews-ERG Fusion Protein can initiate neoplasia from lineage- committed haematopoietic cells. To assess whether EWS-ERG can be a tumour initiator in cells other than mesenchymal cells, we have analysed Ews-ERG function in a cellular environment not typical of that found in human cancers, namely committed lymphoid cells. We have used Ews-ERG invertor mice having an inverted ERG cDNA cassette flanked by loxP sites knocked in the Ews intron 8, and were crossed with mice expressing Cre recombinase under the control of the Ragl gene to give conditional, lymphoid-specific expression of the fusion protein. Clonal T cell neoplasias arose in these mice. This first conditional Ews gene fusion model of tumourigenesis shows that Ews-ERG can cause haematopoietic tumours and the precursor cells are committed cells. Thus Ews-ERG can function in cells that do not have to be pluripotent progenitors nor mesenchymal cells.

Materials and Methods Generation of targeting constructs and homologous recombination. The Ews genomic targeting clone was constructed from λ phage DNA isolated from mouse 129 DNA. The knock-in clones for potential Ews- fusions with ^1FP, FHl, Chop and ATFl were made using a transfer vector pC2A-neo. Knock-in was achieved by homologous recombination. The ERG inversion knock-in cassette was prepared using cDNA made from human colon carcinoma COLO320 mRNA (the sequence of the ERG segment, corresponding to the part of ERG found in EWS-ERG fusion. A diagram of overall structure of the ERG inversion cassette is shown in Figure 14 A, indicating the initial orientation of the ERG inversion cassette (3'-> 5' with respect to Ews) and this was determined by restriction site mapping and genomic DNA sequencing. After identification of targeted ES cells, these were injected into C57BL/6 blastocysts, chimaeric mice produced and germline transmission of the inversion knock-in allele obtained. These mice were bred with mice expressing Cre from a Ragl knock-in. Specificity of expression of the Ragl -Cre allele has previously been described and was sustained using the ROSA26-R reporter mice.

Molecular analyses. Genomic DNA was prepared from tissues using proteinase K digestion and phenol-chloroform extraction and filter hybridizations was carried out with radiolabeled probes. The probes used to detect antigen receptor gene rearrangements were a heavy chain Ig μ nitron enhancer probe and a T cell receptor Jβ2 probe. T cell receptor V-D- Jβ junction sequences were obtained by PCR amplification from thymoma DNA with pools of Vβ primers and a Jβ2 reverse primer (primer sequences from and the product fractionated on agarose to determine the presence of a single amplified band. In turn, each band was eluted and the sequence obtained using the Jβ2N primer and identified by comparison with the IGMT database

RT-PCR was carried out on cDNA as described.. RNA was prepared using RNeasy (Quiagen) from cells which were labelled with FITC-conjugated antibodies and isolated by flow cytometry. Sequences have been described for Pax5 and CD3 RT- PCR primers and the sequences of the Ews-ERG primers were 5'- CCACAGGATGGTAACAAGCCTGC-3' (Ews) and 5'- CGAACTTGTAGGCGTAGC-3' (ERG). The hybridisation probe was a 303bp fragment of Ews-ERG residing within the RT-PCR product.

Analysis of leukaemia/lymphoma. A cohort of mice was established by inter-breeding Ews-ERG invertor line with a Ragl-Cre line to generate littermates with Ews-ERG + Ragl-Cre or Ews-ERG genotypes. Genotypes were determined using filter hybridization of genomic DNA from a small tail biopsy. The health status of these mice was monitored and if signs of ill-health appeared, mice were sacrificed and a post-mortem was carried out. Tissue samples were removed for single cell preparation for determination of surface protein expression phenotype, for nucleic acid preparation or for fixation in 10% formalin for histology. Blood smears were prepared and stained with May-Griinwald-Giemsa (MGG). For histology, fixed tissues were embedded in wax and 0.4μμ sections made, stained with haematoxylin and eosin (H&E) after mounting on slides. FACS analysis was conducted using a FACSCalibur with fluorescent antibodies purchased from BD Biosciences. Data were analyzed with CellQuest software. Western protein detection was carried out as described using 18μg protein per lane and proteins fractionated on 4-20% SDS-PAGE. The separated proteins were electro-transferred to PVDF nylon membranes (Millipore) and specific proteins detected using anti-Ews antibody (Santa Cruz; raised against the amino terminus of Ews) or anti-ERG antibody (Santa Cruz; raised against the carboxy terminus of ERG). Antibody bound to filter was detected using secondary antibodies and ECL as described by the manufacturer (Amersham).

Results: The Ews-ERG fusion protein causes T cell neoplasia in invertor mice.

We investigated the ability of Ews-ERG fusion protein to contribute to Leukaemogenesis by causing the inversion of the Ei?G-containing cassette in lymphoid cells using Ragl-Cre knock-in mice. Twenty five out of the cohort of 29 mice (86%) mice carrying both Ews-ERG and Ragl-Cre genes developed leukaemia associated with thymoma within 500 days, whereas the Ews-ERG-oτάy cohort (20 mice) did not display neoplasia. Blood smears at the time of sacrifice of Ews-ERG; Ragl-Cre mice typically showed elevated numbers of large leukocytes, with lymphoid morphology of different maturities including lymphoblasts. hi addition, bone marrow sections showed high levels of infiltrating lymphoblasts. The diseased mice all had large thymomas, usually associated with splenomegaly and lymphadenopathy, and the histology of spleens showed variously either complete loss of normal architecture (loss of demarcated white and red pulp), partial loss or normal architecture. Similarly, the level of infiltration of leukaemic lymphocytes into liver and kidney varied, with some having large amounts of peri-vascular deposits and others marginal levels. AU mice had thymomas with homogeneous presence of large leukaemic cells. These characteristics, together with evidence of clonal TCRB gene rearrangements and expression of TCR-associated CD8 and CD4, permits diagnosis of large cell anaplastic T cell lymphoma according to the Bethesda proposals for lymphoid tumours in mice. The Ews-ERG fusion appears crucial for malignancies in the invertor mice as disease only arose in mice with both the Ews-ERG and Ragl-Cre genes. The presence of Ews-ERG mRNA was confirmed in spleens of these mice, identical with that in Ews-ERG ES cells transfected with Cre expression plasmids. In addition, the steady state orientation of the /øxP-flanked ERG cassette was studied using filter hybridization and we found that the ERG sequence had become inverted into the 5' to 3 'orientation with respect to the Ews gene, in each case of thymoma. The other tissues examined have little evidence of an inverted band. Finally, the presence of the Ews- ERG fusion protein was shown by Western blotting of thymoma proteins using antibodies binding to either Ews or ERG, which detected respectively the normal mouse Ews or Erg proteins and the Ews-ERG fusion molecule in thymoma T cells of an Ews-ERG; Ragl-Cre mouse.. Studies of progenitor gene expression have indicated the promiscuous expression of lymphoid markers precedes lineage commitment and thus, the Ragl-Cre allele might be expressing in cells destined to become non-lymphoid. The Ragl-Cre knock-in allele was previously shown to be specific for lymphoid cells using a reporter assay dependent on Cre-mediated deletion of a /oxP-flanked (floxed) segment of the Lmo2 gene. We have sustained these observations using an additional reporter assay, namely the ROSA-loxSTOP-lacZ (ROSA26-R) allele where βgal expression is activated by deletion of a loxP-pA site. Haematopoietic populations from mice carrying both Ragl-Cre and ROSA26-R alleles were stained with fluorescent antibodies binding to various surface markers and flow cytometry carried out for co-detect ion antibody and βgal-derived fluorescence, βgal signal was found in thymus and spleen (almost all cells) and in BM, a population of βgal+ cells exist but did not co-express with CD34, Seal, Ckit or Terl 19 (or Macl, not shown). This means that the Ragl-Cre is restricted to lymphoid cells as previously shown. The Ragl-Cre gene is expressed in both B and T cells but only T cell tumours have arisen in the Ews-ERG; Ragl-Cre invertor line. The possible inversion of the Ews-ERG gene in B cells was investigated using RT-PCR analysis of expressed Ews- ERG fusion mRNA (Fig 16). RNA was prepared from whole spleen or thymus or from flow sorted B220+ spleen cells (3400 cells) or thyl.2+ spleen cells (6400 cells) and RT-PCR performed. Pax5 and CD3 primers were used for specific detection of transcripts B cell and T cells respectively. Pax5 transcripts were detected in cDNA made from spleen and B220+ sorted cells and CDS in the spleen, thymus and thy 1.2+ sorted cells. No evidence of CD3 expression was found in the sorted B220+ cells, nor of Pax5 in the sorted thy 1.2+ cells, showing that the sorted cells were practically free of contaminating T or B cells respectively. The presence of Ews-ERG fusion RNA was analysed with RT-PCR primers yielding a product spanning the fusion junction and detected with an internal junction probe. Ews-ERG RT-PCR product was detected in the unfractionated spleen and thymus sources, as well as in the purified, sorted B220+ and thyl.2+ cells. Therefore, Cre-mediated inversion of the Ews-ERG gene occurs in both T and B cells.

Ews-ERG induces malignancies of mature T cells The presence of thymomas in all afflicted Ews-ERG; Ragl-Cre mice suggested that the hematological malignancy in these mice comprised T cells. This was confirmed by FACS analysis of T cell surface markers and by determination of T cell or B cell receptor gene rearrangement status. The majority of normal mouse thymic T cells express both the CD4 and CD8 surface markers (double positive or DP cells), as well as the pan-T cell marker Thyl (CD90). The situation with the thymomas of inverter mice varied from mainly CD4+CD8+ DP cells mainly CD8+ SP cells, or mainly DN cells. Thirteen of the cohort of 25 leukaemic invertor mice were analyzed for their thymic T cell profile by FACS. While we found either CD8+ SP or CD4+CD8+ DP thymomas, none showed a CD4+ SP phenotype. Two of the thymomas had a mixed phenotype comprising mainly CD8+ SP cells, with a sub-population of DP cells. The clonality of these T cell tumours was assessed by filter hybridization of genomic thymoma DNA with a T cell receptor β probe (Tcrb) from the Jβ2 region. The probe detects a 5Kb band in non-lymphoid DNA corresponding to the intact Tcrb gene; if V-D-J or D-J joins have occurred, new restriction fragments are created giving rise to 'rearranged' bands on the hybridization autoradiograph. All except one (M9) showed at least one rearranged Jβ2 allele and several have two rearranged alleles (e.g. Ml 8). Variable amounts of residual germline allele were present. The status of the immunoglobulin locus was examined with a probe from the region between JH and Cμ segments. Single Igh allele rearrangements were observed in about half of the cohort including ones in which FACS analysis and Tcrb clonality hybridizations showed the tumours to be clonal T cell malignancies. The Igh rearrangements are therefore likely to be abortive rearrangements. Confirmation that the Tcrb rearrangements observed by hybridizations were due to productive V-D-J joins, was made by sequence analysis of the Tcrb genes in four of the cases. PCR amplification from thymoma DNA was carried out with pools of Vβ primers and a Jβ2 reverse primer (primer sequences from and sequences identified with the IGMT database. Table 4 shows the V-D-J junctional sequences 37

73

indicating the presence of productive T cell receptor β genes. In all cases, N-region diversity is present between the V-D and D-J junctions. In the thymoma DNA from M2, a non-productive join has also occurred, in addition to the productive one, resulting in an out-of-frame joint.

Discussion: Our results show that the Ews-ERG fusion, normally restricted to sarcomas in humans, can initiate T lymphocyte tumours, if conditionally expressed in committed lymphoid cells. Leukaemogenesis in Ews-ERG invertor mice thus concurs with the hypothesis that EWS fusions can cause neoplasia arising in various cell types. Further, our data shows that the cellular context of EFPS-associated chromosomal translocations in humans does not have to be a stem cell or even a multi-potent progenitor, as committed lymphocyte precursors expressing Rag recombinase genes are the leukaemic precursors in our invertor model. The same conclusions can be applied to Frø-associated chromosomal translocations that seem to be largely interchangeable with EWS, given the spectrum of cancers in which these are found and the relatedness of FUS and EWS coded proteins. Thus, EWS or FUS chromosomal translocations probably arise more by virtue of accessibility of the genes to chromosomal translocations than by the precise cellular specificity of the resultant fusion proteins.

Example 5: MLL- AF4 invertor mice develop B-cell tumours.

The recurrent chromosomal translocation t(4;ll)(q21;q23) is the most common abnormality in acute leukemias of patients less than 1 year old, and overall the most common MLL fusion protein, giving rise to B lineage tumors invoking a poor prognosis. New methods to combat these leukemias are needed and a mouse model of MLL-AF4 tumorigenesis is needed. Modeling MLL-AF4 effects in mice has been hampered by embryonic lethality due to expression of MU-AF 4 knock-in alleles, even using /oxP-STOP conditional alleles. The present inventors have obviated this problem using the invertor mouse conditional technology in which MIl-AF 4 fusion cannot take place until Cre expression occurs, giving many options for cell-specific generation of M11-AF4 fusion. The results show that the MII-AF4 invertor mice develop B cell tumors with high penetrance providing an ideal model for studying MLL- AF4 biology.

Results MII-AF4 fusion is lethal for mouse embryos We initially used the knock-in approach to produce an MLL-AF4 mouse model analogous to the method used for the MII-AF9 knock-in (Corral et al., 1996; Dobson et al., 1999). An AF 4 cDNA-polyA cassette was fused into exon 10 of the mouse Mil gene in ES cells. Expression of the MII-AF4 fusion transcript could be demonstrated in these knock-in ES cells. However, injection of two different clones of targeted ES only yielded one low level chimeric mouse in 334 transferred blastocysts. This chimera was runted and died after 14 days. The failure in chimera production was due to embryonic lethality at around E12.5. In an attempt to circumvent this embryonic lethal effect, a conditional knock-in version was generated using a fioxed transcription stop (MII-AF4 /αxP-stop) analogous to that described for Aml-Eto fusion (Higuchi et al., 2002). Like the MII-AF4 fusion knock-in, we were unable to produce chimeric mice with the MII-AF4 loxP-stop ES cells, whereas control ES cells with only a loxP sequence added at the same intronic position resulted in usual numbers of chimera. We concluded that transcriptional read through of the /σxP-STOP allowed the embryonic lethal effect of the MII-AF4 fusion transcript.

Therefore MII-AF4 invertor ES cells were created, injected into blastocysts and germline carriers of the cassette were made. This MII-AF4 knock-in model was not subject to embryonic lethality, allowing inter-breeding with lines of mice expressing Cre recombinase in various specific hematopoietic cell types. Two sets of MII-AF4 invertor ES cells were made in which the AF4 cassette was flanked by either wild type or mutant loxP sites (Albert et al., 1995; Hoess et al., 1982) (MII-AF4 loxP and MIl- AF4 Iox66/71 invertors,). Following expression of Cre recombinase in the ES cells, the invertor cassette knocked-in to Mil intron 10, became inverted and thus was orientated in the same transcriptional direction to Mil and was transcribed into a fusion mRNA of MII-AF4. I U I / U iJ -. UUJ / u

75

MII-AF4 invertor mice develop lymphoid tumors. Targeted ES cells were used to make MII-AF4 invertor mice that were crossed with mice expressing Cre recombinase under control of the lymphocyte specific promoters of Rag-1, Lck, and CD19 (McCormack et al., 2003) to generate cohorts that were observed for a period of 18 months (we were unable to generate MII-AF4 invertor mice crossed with Lmo2-Cre due to embryonic lethality similar to MII-AF4 knock-in, unpubl.). High proportions of these mice developed lymphoid tumors in this period, ranging from -70-90%. Survival curves are shown in Fig. 15 for the six groups of mice studied, demonstrating a similar time course of disease incidence irrespective of the Cre strain of mice used to invert the AF 4 cassette into the MIl-AF 4 fusion mRNA. Prominent findings in the post-mortem examinations were extreme splenomegaly and mesenteric lymphadenopathy due to infiltration by lymphoma. Macroscopic infiltration of lymphoma into liver and kidney was only seen in late stage disease. The thymus was often slightly enlarged and discrimination from mediastinal lymph nodes was not always possible. Lymphomas were not usually found in other sites. Blood smears of tumor bearing mice showed no striking abnormalities compared to wild-type animals. The number of circulating white blood cells was usually only slightly increased (WBC 4.8 ± 2.8 vs. 3.5 ± 2.0), whilst the red blood cell and platelet numbers were slightly reduced (RBC 7.6 ± 1.3 vs. 9.8 ± 0.5; PLT 893.6 ± 387.0 vs. 1095.1 ± 378.2). In differential white blood cell counts, granulocyte and monocyte numbers were moderately increased (lymphocytes 1.5 ± 0.8 vs. 1.7 ± 1.0; monocytes 0.7 ± 0.4 vs. 0.4 ± 0.2; granulocytes 2.7 ± 1.6 vs. 1.4 ± 1.1).

MIl-AF 4 fusion transcript is present in lymphomas Tumor incidence in the MII-AF4 cohorts was dependent on the Cre-expression allele (Fig. 21) but there was only a significant difference in time of onset of tumorigenesis between mice with the invertor cassette flanked by wild type or mutant alleles loxP in the case of the Rag-1-Cre mice. Filter hybridization analysis of genomic DNA was used to establish the occurrence of AF 4 inversion in spleen, thymus, lymph node, liver, and kidney (tail biopsy DNA was used as control to show the germline and knock-in Mil alleles) of mice that succumbed to lymphoma. The inverted segment was readily detected in spleen, thymus and lymph node and the intensity of the inverted allele correlated with the level of tumor infiltration in the respective tissues.

Transcription of the resulting MII-AF4 fusion gene was confirmed by RT-PCR with primers positioned in exon 10 of Mil and exon 5 of AF4.. After inversion of the AF4 cassette, chimeric mRNA is transcribed and spliced to give MII-AF4 fusion mRNA representative of the fusion product found in human leukemia. Due to higher sensitivity of RT-PCR compared to filter hybridization, fusion transcript was also detected in low grade infiltrated tissues such as the liver and kidney. The RT-PCR products were cloned and sequenced to demonstrate the in-frame fusion of Mil and AF4 exons as shown for transcripts in the targeted ES cells- Specific Cre expression in defined hematopoietic populations has been demonstrated previously using a highly sensitive reporter assay for Cre-mediated deletion (McCormack et al., 2003). In brief, Rag-1 Cre is highly expressed in both, B- and T-progenitor subsets, but is not detectable in Macl+ cells. Lck-Cre is exclusively expressed in the T-cell compartment from the DNl stage onwards. There was no evidence of Cre activity in cells sorted for the expression of CD34, Seal, Terll9, Mac-1 or Gr-I. Further, no Cre activity could be detected after selection of cells expressing B cell markers (CD 19, B220, IgM, IgD or IgG). These results with Lck-Cre were confirmed by flow cytometry using the ROSA-YFP lox-STOP deletion (Srinivas et al., 2001) as a fluorescent gene reporter assay (L. Drynan & THR unpubl). CD19- Cre expression is limited to B-cells, with different levels according to the respective differentiation state. Single and double positive T-cells, Terll9+, and Mac-1 + populations were Cre negative. Specificity of Cre activity was additionally assessed by differential PCR to estimate the efficiency of inversion in sorted subsets of spleen cells from young asymptomatic invertor mice. MII-AF4 Iox66/71; Rag-Cre mice showed inversion of about 65% unsorted spleen cells, whereas in the Macl+ fraction inverted alleles were only found at low frequency, attributable to the level of contørαinating lymphocytes in the sorted populations. Inversion was induced almost equally in B220+ and Thyl.2+ cells. By contrast, MII-AF4 Iox66/71; Lck-Cre mice showed inversion exclusively in Thyl .2+ cells. Efficiency of inversion was higher than in MII-AF4 Iox66/71; CDl 9 mice, in which Cre expression was tightly restricted to B220+ lymphocytes.

MIl-AF 4 invertors develop diffuse large B cell lymphoma.

The post-mortem examination of the three types of invertor mice showed these animals all developed lymphoid neoplasms. Histology and irπmiinohistochernistry classified the tumors, by the Bethesda system (Morse et al., 2002), uniformly as diffuse large cell lymphoma of the centroblastic subtype, typically involving spleen, thymus, and mesenteric lymph nodes, although some mice had only splenomegaly, without mesenteric lymph node enlargement. The infiltration of non-lymphatic organs such as liver and kidney was more variable and dependent on the stage of disease at the time of post-mortem examination. Bone marrow was rarely affected. Enlarged mesenteric lymph nodes were full of lymphoid cells that could be identified as B cells by their staining with the anti-B220 B cell marker. The lymphoma cells were characterized by their large vesicular nuclei, clumping chromatin, and 1 to 5 prominent nucleoli. Numerous mitoses and many apoptotic cells were present in tumor areas.

Irnmunohistochemistry was used to demonstrate the B-cell phenotype of the lymphoma cells. These were large B cell centroblastic lymphomas (DLBCL) and typically, were infiltrated by small reactive T cells to some level. Myeloid cells were absent in tumor areas, as judged by the absence of cells expressing the F4/80 myeloid marker in the lymphomas. In a minority of tumors, small areas of histiocytic infiltrates were observed in a pattern compatible with histiocytic subtype of DLBCL. There were no differences between Rag, Lck, or CD19-Cre expressing mice in respect of lymphoma morphology and immunophenotype. The B cell phenotype of the MII-AF4 invertor tumors was confirmed by studying the status of the immunoglobulin heavy chain and T cell receptor genes. Overall, about 60% of the lymphoma infiltrated spleens or lymph nodes analyzed showed clonal Igh rearrangements of at least one allele. The remaining germline band originated from residual normal cells or reactive T cells in spleens, which were only partially replaced by lymphoma cells or reactive T-cells in the mesenteric lymphoma tissue. In addition, two samples displayed one Tcr rearrangement.

MIl-AF 4 invertor B cell lymphomas are transplantable tumors

The B cell tumors arising in the MII-AF4 invertor mice are invasive tumors, extensively occupying sites outside of the primary lymphoid organs, such as perivascular deposits in liver and kidney. Lymphadenopathy was a characteristic feature found in the mice. This tumor profile suggested that these could be serially transplanted into syngeneic mice. Single cell suspensions of lymphoma infiltrated spleens and mesenteric lymph nodes of CD19-Cre were tested in transplantation assays using Rag-1 null recipient mice, which lack lymphocytes (Mombaerts et al., 1992). Single cell suspensions, purified by density gradient centrifugation (comprising a population of B lymphoma cells together with reactive T-cells and low numbers of Mac- 1 -positive cells), were injected into a Rag-1-/- mouse. FACS analysis of the donor cells from the MII-AF4; CD19-Cre lymphoma showed 50% B220 (B cell), 34% Thyl.2 (T cell) and 12% Macl (myeloid cell). A transplanted tumor arose in the recipient Rag-1 -/- mouse in just under 2 months and was exemplified by splenomegaly and extensive infiltration of liver and kidneys. FACS analysis of the tumor cell population showed that there had been expansion of only the lymphomatous B cells and very few T and myeloid cells remained. Secondary transfer of these splenic tumor cells into secondary Rag-1-/- recipients (four mice were used) gave rise to tumors in 17-23 days and the surface marker phenotype was explicitly that of a B cell neoplasm. The correspondence of the primary tumor with titiat of the transplanted tumors in the Rag-1-/- recipients was investigated by analysis of clonal Igh gene rearrangement.. Filter hybridization showed a faint band corresponding to clonal rearranged Igh allele in the DNA from the invertor tumor from lymph node and spleen but this rearranged band became prominent in the first and second Rag-1-/- recipient mouse tumors. No rearrangement of the Tcrb gene was found.

Discussion. The invertor technology has allowed the first generation of B cell specific models for MLL-AP4. This is a valuable addition to the repertoire of MLL-models allowing studies of secondary mutations co-operating with MLL-fusions, lineage specific target genes and an important element to development of new therapies to target MLL- leukemia. This last point is of particular clinical relevance as the prognosis of MLL- AF4 associated leukemia is very poor, especially if it occurs before 1 year of age (Pui et al., 2002) and new drugs are needed to treat this leukemia.

B cell tumors in MII-AF4 invertor mice In human MLL-AF 4 translocations, cytogenetic studies have indicated that the crucial event lies on the der(l l) chromosome. Similar conclusions apply to MLL- AF9 positive leukemia (Huret et al., 2001). In the MII-AF9 knock-in mouse model, in which only expression of the MU-AF9 fusion product occurs, it was sufficient to induce myeloid leukemia (Corral et al., 1996; Dobson et al., 1999). We made an invertor conditional MII-AF4 mouse model in which expression of the MII-AF4 fusion corresponds to the human der(l 1) product. We used three distinct lines of Cre expressing mice to facilitate the inversion of the AF4 cassette and an identical lymphoma occurred in all three situations with high penetrance. A long latency period was observed in all three cases and little difference was found when the invertor cassette was flanked by wild type loxP sites or the mutant loxP sites, except for the MII-AF4; Rag-1-Cre in which the mutant loxP allele increases the rate of tumor formation somewhat. The long latency suggests, however, that expression of the fusion transcript alone is not sufficient for formation of overt lymphoma, but secondary mutations have to occur. This interpretation is supported by the detection of clonal Igh rearrangements in 60% of the cases. As Cre-mediated inversion and expression of the MII-AF4 chimaeric mRNA will occur in a large number of cells, oligo- or polyclonal tumors with diverse immunoglobulin gene rearrangements might be expected to occur.

Cellular origin of M11-AF4 leukemias In infant MLL- AF4 positive leukemia, the tumors typically show the phenotype of the B-lymphocyte lineage, mainly immature pre/pro-B cell leukemias (Pui et al., 1991). The tumors in the MIl-AFA invertor mice were always more mature B cells.. This difference may reflect the timing with which the MII-AF4 fusion is available to induce neoplasia. There is a high percentage of patients with t(4;l l) translocations without clonal immunoglobulin heavy chain and T-cell receptor gene rearrangements suggesting that an immature progenitor cell is the target of the chromosomal translocation (Peham et al., 2002), supported by rare cases of lineage switch in some infants with ALL/t(4,l l) (Ridge et al., 1995). Recent microarray data showed co- expression of many myeloid and monocytes/macrophage-specific genes in B-lineage ALL with MLL aberrations, suggesting that leukemia with MZZ-translocations might originate from a bi-potential progenitor or perhaps MLL- AF4 protein drives the trans- differentiation of an early lymphocyte progenitor (Armstrong et al., 2002). However, characterization of sorted cells from leukemic bone marrow of childhood high-risk ALL/t(4;l l) suggested an origin in a primitive lymphoid-restricted progenitor cell (Hotfilder et al., 2005). Moreover, studies in monozygotic twins with leukemia, carrying an identical MLL-AF4 chromosomal translocation, but different immunoglobulin rearrangements strongly supported the model that the primary MLL- AF4 target cell was a progenitor cell at a differentiation stage before efficient RAG expression (Greaves et al., 2003; Mori et al., 2002). This issue could be addressed using the MII-AF4 invertor model and an inducible Cre driven by a promoter of a gene, such as Lmo2, expressed in bone marrow progenitor cells. Our data show that M11-AF4 fusion can be oncogenic in cells restricted to the lymphoid lineage (i.e. not only in uncommitted progenitors). The resultant tumors are transplantable and can infiltrate primary and secondary host tissue aggressively. AU the lymphomas showed an identical phenotype, even MII-AF4; Lck-Cre mice. Previous data showed that the Lck-Cre expression is tightly restricted to cells in the T cell developmental pathway from the DNl stage onwards (McCormack et al., 2003) which is just a stage before endogenous Lck activity is apparent (Wolfer et al., 2001; Wolfer et al., 2002).

The intriguing possibility is that the expression of Mll-Af4 in a postulated non- committed T cell progenitor could also cause reassignment of the lineage, in this case to the B cell lineage. While the examination of such developmental alteration by the Mll-Af4 fusion is outside the scope of the current establishment of a tumor system, second generation models should be able to address this question using Cre-expressing mice with the capability to cause AF4 inversion at precisely defined times in hematopoiesis.

Generation and characterization oϊ MIl-AF 4 invertor mice The AF4 cDNA segment from nt 1504 to nt 4304 (AccNo. NM_005935) was amplified from a human B-ALL sample RNA and cloned into the Notl site of pC2-neo (Forster et al., 2005). The entire fragment from the clone (AF4-pA-MClNeo-pA) was excised as an Sfϊl fragment, the restriction site ends filled in and the fragment was introduced into the blunted BamHI site of pLOXKI (Forster et al., 2005) so that the AF4-pA-MClNeo-pA sequences were immediately downstream of the short intron sequence and acceptor splice site in pLOXKI and all flanked by loxP sites (to give ρLOXKI-AF4). The invertor cassette of ρLOXKI-AF4 was cloned as a Scel-Clal blunt fragment into the blunt-ended BgIII site of the Mil targeting vector (a 5.5Kb EcoRI fragment (Collins et al., 2000)). Finally the HSV tk gene was added as a blunt end filled Xbal fragment into the blunted CIaI site of the Mil targeting vector. For electroporation into ES cells, the final construct was linearized with an Xhol within the polylinker sequence of the Mil targeting vector. For the MIl-AF 4 Iox66/71 version, a vector designated pLOXKI2 was constructed. The Iox66 sequence was introduced as oligonucleotides into SacIIτNotI digested pMG2 (Forster et al., 2005) followed by the short segment of the Af4 intron as an Xbal-BamHI PCR fragment. The Iox71 site was cloned as oligonucleotides into the EcoRV-Hindlll sites to give pLOXKI2 (analogous to pLOXKI (Forster et al., 2005)). The AF4 cDNA fragment described above, was cloned as a blunt Sfil-Notl into the blunt BamHI site of pLOXKI2 and the SV40pA-MCl-neo-pA fragment was blunt end cloned into the polylinker EcoRV site to yield pLOXKI2-AF4. This invertor cassette was cloned as a Scel-Clal blunt fragment into the blunt-ended BgIII site of the genomic Mil EcoRI fragment targeting vector. The HSV tk gene was added as a blunt Xbal fragment into the blunt CIaI site of the Mil targeting vector and linearized with Xhol for transfection into ES cells.. Mice carrying MIl-AF 4 loxP or MIl-AF 4 Iox66/71 alleles were bred with those expressing Cre recombinase under the control of the Rag-1, Lck, or CD 19 promoter (McCormack et al., 2003). Animals were culled at the first signs of ill health. After post-mortem examination, samples for RNA and DNA extraction were snap frozen and tissues for histological analysis were fixed in 10% buffered formalin. Blood smears were stained with May-Grϋnwald-Giemsa (MGG) stain and an automated differential blood count from EDTA-blood was carried out (Animal Blood Counter, Vet abc). 4 μm sections of paraffin embedded tissues were stained with hematoxylin and eosin.

Tissue sections for immunohistochemistry were dried overnight at 600C and subsequently deparaffinized in xylene, dehydrated in ethanol, and washed in water. Antigen retrieval was accomplished by placing slides in a pressure cooker at full pressure for 3 minutes in 0.01 M Citric Acid, pH 6.0. Sections were stained using a Dako TechMate 500 automated immunostainer with a primary (rat anti-mouse B220 (BD Pharmingen; 1/25 dilution), CD3 (Dako; 1/50 dilution), and F4/80 (Serotec; 1/50 dilution)) antibody incubation time of 45 minutes and a standard biotinylated secondary (goat anti-rat; 1/250 dilution) antibody for 30 minutes and a streptavidin tertiary complex for 30 minutes with a DAB visualisation system. Sections were counterstained with Mayer's haematoxylin, for a blue nuclear contrast to the DAB- derived brown immunopositive end-product.

Filter hybridization of Sphl digested genomic DNA from various tissues was carried out using the Mil 5 'probe (Collins et al., 2000) (Fig. 16A) to detect the inverted allele in the various organs. To estimate the proportion of inverted versus non-inverted MII-AF4 alleles in various cell populations, DNA was prepared (Qiagen Blood Kit) from 10,000 unsorted or FACS sorted cells expressing Mac-1, B220 or Thy 1.2 Purity of the sorted populations was >97.5%. 1/10 of the DNA solution, primer MEF, the fluorescent primer AIR, and a primer set for the single copy gene Lmo2 were used in a PCR reaction with the following conditions: 940C 10 min, 35 cycles each of 940C 30s, 66°C 30s, 72°C 60s and final extension at 72°C 7 min. PCR products were separated by capillary electrophoresis (ABI Prism 3100 Genetic Analyzer), quantified by 03337

83

calculation of the area under the curve (Genescan Analysis Software, Applied Biosystems), and the ratios of inverted allele/Z,mσ2 allele were calculated.

The MII-AF4 fusion transcript resulting from the inversion, post-transcriptional splicing and processing was detected by RT-PCR. Total RNA was extracted from tissues or sorted cells and reverse transcribed. 100 ng of cDNA was amplified in a PCR using nested primers primer sets positioned in Mil exon 10 and AF4 exon 5 (Fig. 16B and C. PCR products were visualized on 2% agarose gels stained with ethidium bromide.

Flow cytometric analysis of cell surface marker expression Single-cell suspensions were prepared from mouse thymus, spleen, lymph node, and bone marrow. Cells were stained with either PE or FITC conjugated antibodies to CD4, CD8, Thyl.2, Gr-I (Ly-6G), Mac-1 (CDl Ib)5 B220 (CD45R), IgM, IgG, CD43, CD24, c-kit, Sca-1, and CD34, or matching isotope control antibodies (BD Pharmingen) and analyzed using a FACSCalibur flow cytometer and Cell Quest software (Becton-Dickinson).

Immunoglobulin and T-cell receptor gene rearrangement assays Genomic DNA from various tissues was digested with HindIII (or EcoRI as in Fig. 6) and filter-hybridized with probes for Igh (Neuberger and Williams, 1986), Tcr Jβ2, and Tcr Cβl (Malissen et al., 1984). A diagrammatic map of the respective regions is shown in Fig. 21 A and B.

Cell transplantation into Rag-/- recipients Single cell suspensions of tumor cells from infiltrated spleens or mesenteric lymph nodes were purified by density gradient centrifugation and injected intravenously at a dose of 5x106 cells into Rag-/- mice (Mombaerts et al., 1992). Recipients were monitored and sacrificed at the first signs of lymphoma. Tissues were tested for the presence of Igh rearrangements by filter hybridization and immunophenotype using FACS analysis as described above. All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention which are apparent to those skilled in molecular biology or related fields are intended to be within the scope of the following claims.