WO1999018191A1

WO1999018191A1 - Transgenic animals with knocked-in vec receptor genes and uses thereof

Info

Publication number: WO1999018191A1
Application number: PCT/US1998/020717
Authority: WO
Inventors: Larry Witte; Bronislaw Pytowski; Peter Bohlen; Thomas Sato
Original assignee: Imclone Systems Incorporated
Priority date: 1997-10-02
Filing date: 1998-10-02
Publication date: 1999-04-15
Also published as: JP2001519145A; EP1007652A1; CA2305045A1; AU9598598A

Abstract

The present invention provides a non-human transgenic animal whose cells express a foreign DNA sequence that encodes a functional vascular endothelial cell receptor domain, but do not express a substantially homologous native DNA sequence. This invention also provides methods of using these transgenic animals for identifying therapeutic and other reagents that affect angiogenesis. Additionally provided are methods of using transgenic animals whose cells contain any donor genes as models for testing various reagents.

Description

TRANSGENIC ANIMALS WITH KNOCKED-IN VEC RECEPTOR GENES AND USES THEREOF

BACKGROUND OF THE INVENTION

This invention relates to the development and use of transgenic animals with a "knocked-in" gene. Such animals can be used for testing substances, such as small molecules, antibodies, and other reagents, that have human or veterinary uses.

The knock-in method is a gene replacement technique developed as a tool for developmental studies (Hanks et al., Science, vol. 269, Aug. 4, 1995), and is a variation of targeted gene inactivation (the "knock-out" method). In a conventional transgenic approach, a gene is introduced into the germ line of the animal at an early developmental stage. In the "knock-out" method, an irrelevant DNA sequence is used to inactivate a gene (Mansour et al., Proc. Natl. Acad. Sci. USA 87: 7688-7692 (1990); Shalaby et al., Nature, vol. 376, July 6, 1995). In the knock-in method, a gene is not merely inactivated, as in the sometimes lethal knock-out method, but is simultaneously replaced at that site by another gene by known methods, such as homologous recombination. For example, in the knock-in method, a cDNA from gene 1 is placed under the control of regulatory elements of another gene (gene 2) while gene 2 is simultaneously inactivated. One advantage of the knock-in technique is that a gene can be introduced into a genome at a specific site. Further, knocking out certain genes can be lethal to an animal, whereas the animal can survive if the appropriate homologous gene is replaced using the knock-in method.

The knock-in method has been used as a powerful tool to discover the functions of specific genes in mammalian pathology. A variation of the knock-in method was used, for example, to understand better the mechanism by which CBFB-MYH11 contributes to leukemogenesis by introducing into mice the human fusion oncogene CBFB-MYH11 by knocking-in a portion of the human MYH11 cDNA into the murine Cbfb locus using homologous recombination (Castilla et al., Cell, 87:687-696, November 15, 1996).

The invention described herein is the first disclosure of the knock-in technique being used to substitute a complete gene from one species with the corresponding homolog from another species or with a chimeric construct. Further, since the knock-in technique has until now been used for discovering functions of genes in mammals, the invention described herein is the first disclosure of knock-in techniques being used for development of an animal model for testing therapeutics.

More specifically, this invention also relates to the development of transgenic animal models with knocked-in vascular endothelial cell-specific receptor genes. Until now, there have been no sufficient animal models to test therapeutics that are specific for vascular endothelial cell-specific receptors of a particular species. The invention would allow routine testing with quantitative endpoints for therapeutics that inhibit angiogenesis. In contrast, presently available in vivo model systems, such as the SCID mouse model, are difficult to use in a routine manner. For instance, the SCID mouse model requires laborious surgical manipulations of the animals. Additionally, these models are often subject to complications that are not easily controlled, such as the unwanted vasculahzation of the surgically implanted skin grafts by the mouse vasculature.

The knock-in animal models of the invention can be used to test the role of vascular endothelial cell-specific receptors as well as their function in pathological conditions, such as tumohgenesis. Further, the knock-in animal models of the invention are useful for the identification of therapeutics, such as target molecules, antibodies, and other reagents that react with molecules of a specific species only. An example of a knock-in animal model of the invention is one that would be useful specifically for the identification of inhibitors of pathological angiogenesis. Such inhibitors are likely to be specific to human vascular endothelial cell-specific receptors, but need to be tested in an animal model expressing the human receptor gene, such as a knock-in mouse of the invention.

Since one aspect of the invention is the development of knock-in animal models useful for the identification of inhibitors specific to vascular endothelial cell-specific receptors, the following provides a background of the components of angiogenesis, and therapeutic approaches to inhibiting pathological angiogenesis.

Vascular endothelial cells (VECs) play critical roles in blood vessel formation during human development and in many pathological conditions such as tumorigenesis, ocular diseases, arthritis, and arteriosclerosis. Neovascularization, the de novo formation of new blood vessels, as occurs in embryogenesis, is regulated by multiple cellular processes involving endothelial cells, which includes proliferation, migration, cell-cell interactions, cell-matrix interactions, morphological changes and tissue infiltration. Angiogenesis, the formation of new blood vessels from pre-existing blood vessels, involves the same multiple cellular processes, as well as an enhanced need of tissue infiltration of capillary endothelial cells from pre-existing blood vessels. Both of these events, that is, neovascularization and angiogenesis, will herein be referred to as angiogenesis. Angiogenesis is important in normal physiological processes including embryonic development, follicular growth, and wound healing as well as in pathological conditions involving tumor growth and non-neoplastic diseases involving abnormal neovascularization, including neovascular glaucoma (Folkman, J. and Klagsbrun, M. Science 235:442-447 (1987), retinal neovascularization of diabetes and macular degeneration of aging, as well as chronic inflammatory diseases such as rheumatoid arthritis.

Many of these diseases currently have few, or no, satisfactory therapies. Consequently, intense research is underway to explore the hypothesis that blocking the angiogenic component of these diseases is an effective treatment. Many pharmacological agents with angiogenesis-inhibiting properties have been identified but, clinical trials have not progressed sufficiently. Moreover, most known angiogenesis inhibitors act through poorly understood mechanisms of action. It would therefore be desirable to develop new types of angiogenesis inhibitors which act by interfering with known and specific angiogenesis-related mechanisms.

Angiogenesis can be controlled by interfering with a variety of VEC functions, including mitogenesis, migration, adhesion (Cavada, L, et al. J. Clin. Invest. 98: 886-893 (1996); Brooks, P.C., et al. Cell 79: 1157-1164 (1994)), invasion, and maturation. One approach is to interfere with the action of principal endothelial cell growth factors, especially vascular endothelial growth factor (VEGF).

Vascular endothelial growth factor (VEGF), an endothelial cell-specific mitogen, acts as an angiogenesis inducer by specifically promoting the proliferation of endothelial cells. VEGF is a homodimeric glycoprotein consisting of two 23 kD subunits with structural similarity to PDGF. Several different monomeric isoforms of VEGF exist, resulting from alternative splicing of mRNA. These include two membrane bound forms (VEGF₂₀₆ and VEGF₁₈9) and two soluble forms (VEGF₁₆₅ and VEGF₁₂ι). In all human tissues except placenta, VEGFι₆5 is the most abundant isoform.

VEGF is expressed in embryonic tissues (Breier et al., Development (Camb.) 114:521 (1992)), macrophages, proliferating epidermal keratinocytes during wound healing (Brown et al., J. Exp. Med., 176:1375 (1992)), and may be responsible for tissue edema associated with inflammation (Ferrara et al., Endocr. Rev. 13:18 (1992)). In situ hybridization studies have demonstrated high VEGF expression in a number of human tumor lines including glioblastoma multiforme, hemangioblastoma, central nervous system neoplasms and AIDS-associated Kaposi's sarcoma (Plate, K. et al. (1992) Nature 359: 845-848; Plate, K. et al. (1993) Cancer Res. 53: 5822-5827; Berkman, R. et al. (1993) J. Clin. Invest. 91 : 153-159; Nakamura, S. et al. (1992) AIDS Weekly, 13 (1 )). High levels of VEGF were also observed in hypoxia induced angiogenesis (Shweiki, D. et al. (1992) Nature 359: 843-845). The biological response of VEGF is mediated through its high affinity VEGF receptors, which are selectively expressed on endothelial cells during embryogenesis (Millauer, B., et al., (1993) Cell 72: 835-846) and during tumor formation. Recently, a number of receptor tyrosine kinases (RTKs) that are specifically expressed in endothelial cells have been cloned and characterized. While some RTKs are broadly expressed on diverse cell types, two families have been shown to be primarily restricted to endothelial cells and the early hematopoietic system.

One family includes the VEGF receptors, such as murine FLK-1 , and its human homolog KDR; FLT-1 ; and FLT-4. FLK-1/KDR encode a receptor for VEGF-A, VEGF- B, and VEGF-C. FLT-1 and FLT-4 encode a receptor for VEGF-A and VEGF-C, respectively.

The other family includes TIE-1 and TIE-2. TIE-2 is also known as TEK. The ligand for TIE-2, angiopoietin-1 , has only recently been cloned (Davis, S., et al., Cell 87: 1161-1169 (1996)). The ligand for TIE-1 has not yet been characterized.

The VEGF receptors, in particular FLK-1 /KDR, have been strongly implicated in angiogenesis associated with diverse human pathologies. This realization has led to a major effort to identify inhibitors of tumor angiogenesis with the principal targets being the VEGF molecule and its receptor FLK-1/KDR (Kim, K.J., et al., Nature 362: 841-844 (1993); Strawn, L.M., et al., Cancer Research 56: 3540-3545 (1996)).

VEGF receptors typically are class III receptor-type tyrosine kinases characterized by typically having seven immunoglobulin-like loops in their amino-terminal extracellular receptor ligand-binding domains (Kaipainen et al., J. Exp. Med. 178:2077-2088 (1993)). The other two regions include a transmembrane region and a carboxy- terminal intracellular catalytic domain interrupted by an insertion of hydrophilic interkinase sequences of variable lengths, called the kinase insert domain (Westermark et al., Prog. Growth Factor Res. 1 (4): 253-266 (1989); Terman et al., Oncogene 6:1677-1683 (1991 )). VEGFs elicit their function as proliferation inducers of endothelial cells by binding to and activating their corresponding receptor tyrosine kinases expressed on the surface of endothelial cells.

Critical roles for these two families of receptors in embryonic development have been conclusively shown by studying knock-out mice of each gene: FLK-1 /KDR is critical for endothelial cell differentiation; FLT-1 is important for the organization of primary capillary plexus during the early embryogenesis; TIE-2 was shown to be critical for remodeling of vascular network during angiogenesis in embryos; TIE-1 was identified as a critical molecule for maturation of the vascular network (Sato, T.N., et al., Nature, 376:70-74 (1995)). However, their roles during later embryonic development and pathological conditions could not be studied since knock-outs of these genes resulted in embryonic lethality. A critical role of FLK-1 /KDR in tumor angiogenesis has been clearly shown by retrovirus mediated gene transduction of a dominant negative form of FLK-1 which resulted in prevention of neovascularization in glioma and the consequent prevention of tumor growth.

There are several reasons why KDR is a therapeutic target with highly desirable properties. Of particular importance is that the KDR receptor is expressed almost exclusively on endothelial cells. Further, KDR is strongly up-regulated in activated (proliferating) endothelium as opposed to resting endothelium. In addition, KDR presents a readily accessible target because of its expression on the surface of blood vessel cells. Accordingly, drugs directed against the extracellular domain of KDR can be particularly useful because they act in a highly specific manner, do not need to enter the endothelial cell, and do not have to reach beyond the vasculature to exert their effects on tissues and thus can be effective at lower doses. Additionally, these advantages may contribute to favorable safety profiles of anti- KDR drugs.

These properties of KDR also suggest that it is advantageous to interfere with the VEGF-KDR system at the level of KDR rather than VEGF. KDR is localized on the surface of vascular cells in a restricted manner. The VEGF ligand, on the other hand, is present more widely and at higher concentration deep in the interstitial space of tissues. The VEGF ligand is probably found largely in association with heparan sulfate proteoglycans.

Interfering with the formation of new blood vessels by inhibiting the function of KDR can produce successful new therapies. Further, this approach is advantageous since it offers the possibility of highly specific interference with growing endothelium, as opposed to the generally far less specific treatments now in use. It may be easier to control malignant tumor growth by means of curbing its blood supply with a cytostatic, specific, and potentially non-toxic drug as opposed to directly attacking tumor cells, which is generally done with less specific and frequently cytotoxic drugs. For example, it is advantageous to interfere with an angiogenic receptor that is specifically expressed on the surface of endothelial cells as opposed to another target (e.g. on tumor cells) which is be distributed more widely and at higher concentrations deep in the interstitial space of tissues. More importantly, the availability of effective, non-toxic anti-angiogenesis drugs can provide long-term or lifetime therapies needed to control a variety of disease, such as, but not limited to the metastatic growth of tumors or rheumatoid arthritis.

Studies by the inventors show that neutralizing antibodies to FLK-1 and to KDR are species-specific, and therefore do not cross-react. Some anti-KDR antibodies have been shown to have no effect on the binding of FLK-1 to VEGF, and some anti-FLK- 1 antibodies have also been shown to have no effect on the binding of KDR to VEGF. Accordingly, it would be futile to attempt to test such anti-KDR antibodies, and probably other antagonists, for their anti-angiogenic effect in existing murine tumor models.

Until the present invention, it has been difficult to test human antigen-specific antibodies and other potential inhibitors of human angiogenesis, such as tumor angiogenesis, due to the lack of a sufficient animal model. Typically, with testing of various anti-cancer therapeutics, a human tumor cell line is injected into immunodeficient nude mice and the mice are treated with the anti-cancer therapeutic following a period of tumor growth. Potential anti-angiogenesis approaches such as inhibition of the receptor KDR are unique because the target tissue is the host (murine, for example) vasculature rather than the human tumor cells. Most KDR-specific murine monoclonal antibodies, as described above for example, cannot function in such a model.

Possible approaches to circumvent this problem of testing human-specific antagonists in murine tumor models with murine vasculature include: (1 ) searching for tumor models in non-murine mammals whose FLK-1 receptors exhibit higher homology with KDR; and (2) the chimeric human skin/SCID mouse xenograft model (Brooks, P.C, et al., J. Clin. Invest. 96: 1815-1822 (1995)). However, these approaches are unsatisfactory. There are obvious advantages offered by the use of murine tumor models. A large number of syngeneic and xenogeneic murine models of tumor growth have been developed, and the relevance of FLK-1 in these models has been established (Millauer, B., et al., Cancer Research 56: 1615-1620 (1996)). The substitution of another model (approach (1 )) suffers from major drawbacks, such as: (a) there are no guarantees that anti-KDR antibodies will cross-react with homologous receptors in other species, including primates; (b) establishing consistency and reproducibility would be time-consuming and probably difficult to achieve in non-murine tumor models; (c) satisfactory non-murine tumor models are not common and not readily available.

With regard to approach (2), the chimeric human skin/SCID mouse xenograft model is technically demanding, time consuming and difficult to quantitate (Brooks, P.C, et al., J. Clin. Invest. 96: 1815-1822 (1995)). Additionally, there is the added expense associated with the cost of the SCID mice, their maintenance, and the labor needed to conduct the experiments. Further, only limited information is obtained because the duration of treatment is limited due to the finite time of survival of the human skin- grafts on the mouse. Additionally, there is a problem of increasing skin graft vascularization by the mouse vasculature as skin grafts age, especially when the skin graft contains a tumor.

Accordingly, there is a need for in vivo transgenic animal models which express the human receptor homolog (e.g., KDR) instead of the native receptor (e.g., FLK-1 ) for testing potential therapeutic molecules to treat humans. There is also a need for in vivo transgenic animal models which express homologous genes from other species for testing potential veterinary therapeutic molecules. An object of this invention is to provide in vivo animal models to test the roles of species-specific receptor tyrosine kinases. This strategy would provide a unique approach to understanding the role of each endothelial cell-specific receptor tyrosine kinase during pathological angiogenesis, and facilitate the identification of therapeutic target molecule(s). Another object of the invention is to provide transgenic animal models to test potential therapeutic reagents for their effectiveness and specificity, and especially species-sensitive or species-specific reagents.

SUMMARY OF THE INVENTION

These and other objects have been met by providing a method of testing a substance for use in animals comprising administering the substance to a non- human transgenic animal whose cells express a foreign gene or functional gene fragment from a different species, but do not express a substantially homologous native gene or functional gene fragment, and evaluating any effects of the substance on the animal. In this invention, the foreign and native gene or gene fragments can encode vascular endothelial cell receptor domains.

The present invention also provides a non-human transgenic animal whose cells express a foreign gene from a different species, but do not express a substantially homologous native gene. More specifically, the present invention provides a non- human transgenic animal whose cells express a foreign DNA sequence that encodes a functional extracellular vascular endothelial cell receptor domain, but do not express a substantially homologous native DNA sequence.

DESCRIPTION OF THE FIGURE

Figure 1. The targeting strategy for the knock-in of KDR cDNA into the FLK-1 locus. The outline is a modification of the FLK-1 knock-out targeting strategy of Shalaby et al., (Nature, vol. 376, July 6, 1995). The black rectangle represents the Not I fragment containing the upstream FLK-1 genomic region. The KDR cDNA and the polyadenyiation signal are assembled in an intermediate vector. Approximate lengths of the original genomic FLK-1 DNA and of the restriction fragments relevant for Southern blotting are indicated. Restriction enzymes: B, Bam HI; H, Hind III; N, Nco I; Nt, Not I; P, Pst I, S, Sma I; SI, Sal I; X, Xho I.

Figure 2. The targeting strategy for the knock-in of a KDR/FLK1-1 chimeric cDNA into the FLK-1 locus. The outline is a modification of the FLK-1 knock-out targeting strategy of Shalaby et al., (Nature, vol. 376, July 6, 1995). The KDR/FLK-1 chimeric cDNA, including a transmembrane region (TM), and the polyadenyiation signal are assembled in an intermediate vector. Approximate lengths of the original genomic FLK-1 DNA and of the restriction fragments relevant for Southern blotting are indicated. Restriction enzymes: B, Bam HI; H, Hind III; N, Nco I; Nt, Not I; P, Pst I, S, Sma I; SI, Sal I; X, Xho I.

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a transgenic animal whose cells express a foreign DNA sequence that encodes a functional vascular endothelial cell receptor (VECr) domain. The VECr domain can be a fragment of a VECr, such as the extracellular portion of the receptor, or can be an entire VECr. The foreign DNA sequence is substantially homologous with a native DNA functional VECr domain sequence of the transgenic animal. The cells of the transgenic animal of the invention do not express this homologous native DNA VECr domain sequence.

The VECr foreign DNA sequence can be a fragment or a complete cDNA coding for a VECr of a given species. Further, the VECr foreign DNA sequence can encode a chimeric receptor, wherein the DNA sequence can encode receptors of different species. In one preferred embodiment, the extracellular portion of the receptor is from a species different from the recipient animal.

The transgenic animal of this invention can be produced using the knock-in method. In the preferred embodiment, homologous recombination using a targeting vector containing the foreign DNA sequence results in insertion of the foreign DNA sequence at the site of a homologous native DNA sequence, simultaneously inactivating the native DNA sequence. In this specification, the term "replacement," as well as various forms of the term "replacement," refer to the insertion of a foreign DNA sequence at the site of a homologous native DNA sequence, with simultaneous inactivation of the native DNA sequence. The inactivation of the native DNA sequence occurs upon the disruption of an exon of the native gene when the foreign DNA sequence is inserted into the gene. The foreign DNA sequence is then under the control of the native promoter of the inactivated DNA sequence.

The transgenic animal is preferably a member of a species different than the donor species of the VECr encoded by the foreign DNA. Preferably, both the transgenic animal and the donor are vertebrates, and more preferably, they are mammals. In one embodiment, the transgenic animal is a non-human mammal, such as a mouse, rat, pig, goat, sheep or monkey, and the donor is a human. In a preferred embodiment, the transgenic animal is an animal typically used in biomedical or veterinary research, i.e., a laboratory animal. A laboratory animal can be, but is not limited to being, a mouse, rat, rabbit, dog, pig, cow, horse, goat and sheep.

In a more preferred embodiment of the invention, the donor DNA sequence is human, and the transgenic animal is a mouse. In such a preferred embodiment, the human DNA sequence is preferably under the control of murine tissue-specific regulatory elements, such as a murine endothelial cell specific promoter. In the preferred embodiment, the donor human VECr DNA is constructed without a promoter. This promoterless VECr DNA construct is targeted using a vector of the invention into the mouse genome at a site downstream of the promoter for the mouse VECr.

Receptors of the invention can be any VEC receptor. Examples of VEC receptors include, but are not limited to, the protein tyrosine kinase vascular endothelial growth factor (VEGF) receptors KDR, FLK-1 , FLT-1 , and FLT-4. KDR is the human form of a VEGF receptor having MW 180 kD. FLK-1 is the murine homolog of KDR. FLT-1 is a form of VEGF receptor different from, but related to, the KDR/FLK-1 receptor. Both FLK-1 and KDR encode a receptor for VEGF-A , VEGF-B and VEGF-C. FLT-1 and FLT-4 encode a receptor for VEGF-A and VEGF-C, respectively. VEC receptors of the invention also include the TIE family receptor tyrosine kinases, comprising TIE-1 and TIE-2. TIE-2 encodes a receptor for the angiopoietin-1 ligand.

In the preferred embodiment of this invention, the FLK-1 gene of a mouse is replaced with cDNA of KDR from a human donor, under the control of the murine FLK-1 promoter. Murine recipients produced in this manner express both native FLK-1 and KDR receptors. Cross-breeding these murine recipients produces homozygous KDR KDR mice.

In another embodiment of this invention, the FLK-1 gene of a mouse is replaced with chimeric KDR/FLK-1 cDNA under the control of the murine FLK-1 promoter. Preferably, in the chimera, sequences coding for the extracellular and transmembrane domains of KDR are fused with those for the intracellular domain of FLK-1 , although the transmembrane domains can be from either the KDR or the FLK-1. These clones can be used for the generation of homozygous KDR/FLK-1 mice, so that the intracellular murine FLK-1 domain would be compatible with the murine cell.

This invention also provides a transgenic animal whose cells contain a donor gene from an animal of a different species that has replaced a substantially homologous native gene of the transgenic animal or of an ancestor of the transgenic animal, wherein the cells no longer express the native gene. The transgenic animal is preferably a mouse and the donor gene is preferably human. The donor gene can be any gene of the animal or any synthetic versions or derivatives thereof that are substantially similar to such donor gene.

UTILITY

The invention provides a method of testing a substance that interacts specifically with a protein expressed by the donor VECr DNA sequence comprising administering the substance to the transgenic recipient of the invention and evaluating any effects of the substance on the recipient. For example, since studies indicate that murine VEGF binds to and activates the human KDR receptor, the homozygous KDR KDR mice of the invention are useful as animal models for testing various small molecules, antibodies, and other reagents that affect angiogenesis. Such reagents can either inhibit or increase angiogenesis. Examples of small molecules that can affect angiogenesis include heterocyclic molecules, aromatic molecules, and oligopeptides, among others. Examples of antibodies that can affect angiogenesis are well known in the art.

Further, the invention provides a method of identifying a substance capable of inhibiting abnormal angiogenesis comprising administering the substance to the transgenic KDR animal of the invention and determining whether the substance inhibits abnormal angiogenesis. The invention also provides a method of identifying a substance capable of inhibiting angiogenesis, including tumor angiogenesis, comprising administering the substance to the transgenic KDR animal of the invention and determining whether the substance inhibits angiogenesis. The invention also provides a method of identifying a substance capable of inhibiting tumor growth comprising administering the substance to the transgenic animal of the invention and determining whether the substance inhibits tumor growth. The invention also provides a method of identifying a substance capable of promoting wound healing comprising administering the substance to the transgenic KDR animal of the invention and determining whether the substance promotes wound healing.

The invention also provides transgenic animals of the invention for use in a method of testing any substance for human or veterinary use. The method of testing a substance for human use comprises administering the substance to a transgenic non-human animal whose cells contain a DNA sequence of a human donor. The donor (foreign) DNA encodes a particular gene of interest, and has replaced a substantially homologous native DNA sequence of the animal or of an ancestor of the animal, wherein the cells no longer express the native DNA sequence. The transgenic animals are then evaluated for any effects of the substance on the animal. In a preferred embodiment, the transgenic non-human animal is a mouse. Further, in another preferred embodiment, the donor DNA sequence is under the control of the transgenic animal's tissue-specific regulatory elements. The method of testing a substance for veterinary use comprises administering the substance to a transgenic animal whose cells contain a DNA sequence of a donor that has replaced a substantially homologous native DNA sequence of the animal or of an ancestor of the animal, whereby the cells no longer express the native DNA sequence, and evaluating any effects of the substance on the animal. In a preferred embodiment, the transgenic animal and the donor are members of different species. In another preferred embodiment, the donor DNA sequence is under the control of the transgenic animal's tissue-specific regulatory elements.

DNA ENCODING VEC RECEPTORS

Total RNA is prepared by standard procedures from endothelial receptor-containing tissue. The total RNA is used to direct cDNA synthesis. Standard methods for isolating RNA and synthesizing cDNA are provided in standard manuals of molecular biology such as, for example, in Sambrook et al., "Molecular Cloning," Second Edition, Cold Spring Harbor Laboratory Press (1987) and in Ausubel et al., (Eds), "Current Protocols in Molecular Biology," Greene Associates/Wiley Interscience, New York (1990).

The complete gene or the cDNA of the receptors can be amplified by known methods. For example, the cDNA can be used as a template for amplification by polymerase chain reaction (PCR); see Saiki et al., Science, 239, 487 (1988) or Mullis et al., U.S. patent 4,683, 195. The sequences of the oligonucleotide primers for the PCR amplification are derived from the sequences of mouse and human VEGF receptor respectively. The oligonucleotides are synthesized by methods known in the art. Suitable methods include those described by Caruthers in Science 230, 281-285 (1985).

Additionally, the complete gene can be obtained by standard methods of isolating genomic clones from genomic phage libraries using standard hybridization techniques.

In order to isolate the entire protein-coding regions for the VEC receptors, the upstream PCR oligonucleotide primer is complementary to the sequence at the 5' end, preferably encompassing the ATG start codon and at least 5-10 nucleotides upstream of the start codon. The downstream PCR oligonucleotide primer is complementary to the sequence at the 3' end of the desired DNA sequence. The desired DNA sequence preferably encodes the entire extracellular portion of the VEGF receptor, and optionally encodes all or part of the transmembrane region, and/or all or part of the intracellular region, including the stop codon. A mixture of upstream and downstream oligonucleotides are used in the PCR amplification. The conditions are optimized for each particular primer pair according to standard procedures. The PCR product is analyzed by electrophoresis for cDNA having the correct size, corresponding to the sequence between the primers.

Alternatively, the coding region can be amplified in two or more overlapping fragments. The overlapping fragments are designed to include a restriction site permitting the assembly of the intact cDNA from the fragments.

DNA encoding the VEC receptors of the invention are inserted into a suitable targeting vector and inserted by homologous recombination into a suitable recipient. The DNA inserted into a recipient can encode the entire VEC receptor, or a fragment of the VEC receptor.

The nucleic acid molecules that encode the VEC receptors of the invention, or portions thereof, can be inserted into targeting vectors using standard recombinant DNA techniques. Standard recombinant DNA techniques are described in Sambrook et al., "Molecular Cloning," Second Edition, Cold Spring Harbor Laboratory Press (1987) and by Ausubel et al., (Eds) "Current Protocols in Molecular Biology," Green Publishing Associates/ Wiley-lnterscience, New York (1990). A suitable source of cells containing nucleic acid molecules that express the VEC receptor includes VECs.

Suitable vectors for use in mammalian cells are known. Such vectors include well- known derivatives of SV-40, adenovirus, retrovirus-derived DNA sequences and shuttle vectors derived from combination of functional mammalian vectors, such as those described above, and functional plasmids and phage DNA. EXAMPLES:

The Examples which follow are set forth to aid in understanding the invention but are not intended to, and should not be construed to, limit its scope in any way. The Examples do not include detailed descriptions of conventional methods employed in the construction of vectors and plasmids, the insertion of genes encoding polypeptides into such vectors and plasmids or the introduction of plasmids into hosts. Such methods are well known to those of ordinary skill in the art and are described in numerous publications including Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) "Molecular Cloning: A Laboratory Manual," 2nd edition, Cold Spring Harbor Laboratory Press.

Knockinq-ln a KDR Gene Into a Murine Recipient

The homologous recombination step occurring in murine embryonic stem cells simultaneously disrupts the first exon of the FLK-1 gene ("knock-out") and replaces it with cDNA for KDR ("knock-in"). The resulting heterozygous mice are expected to express both the native FLK-1 and the KDR receptors. Homozygous KDR/KDR mice are obtained in an F1 intercross.

Experimental Design and Methods:

Production of the targeting construct: The following is a description of the methods of obtaining the necessary DNA components for the targeting vectors (FLK-1 genomic fragments and KDR cDNA). Also described are the preparation of both a full-length KDR construct and a novel chimeric cDNA consisting of the extracellular portion of KDR fused to the intracellular portion of FLK-1. FLK-1 genomic and cDNA clones: A 15 kb FLK-1 genomic clone is obtained. A 200 bp cDNA fragment that includes the signal peptide sequence was obtained by PCR from a mouse lung cDNA library using primers complimentary to the published FLK-1 sequence (Matthews, W., et al. Proc. Natl. Acad. Sci. USA 88: 9026-9030 (1997)). This fragment was random primed with 32P-dCTP and used as a probe to screen a 129/SV mouse genomic library (Stratagene). Purified phage DNA from clone #8 was digested with Sal I and an approximately 15 kb fragment was inserted in both orientations into the Sal I site of pBluescript SK II (+) vector (Stratagene) to give the plasmid pmgFLK1.2. This clone was confirmed by restriction analysis to encode the mouse FLK-1 genomic DNA (Shalaby, F., et al., Nature 376: 62-66 (1995)). Full length FLK-1 cDNA was obtained in accordance with Matthews, W., et al., Proc. Natl. Acad. Sci. USA 88: 9026-9030 (1991 ).

KDR cDNA clone: Full length KDR cDNA was obtained by RT-PCR using primers complementary to the published sequence (Terman, B.I., Oncogene 6: 1677-1683 (1991 )). The template for RT-PCR was human fetal kidney mRNA obtained from spontaneously aborted human fetuses (Clontech). Two overlapping fragments encoding 5' and 3' regions of the cDNA were obtained and assembled in the expression vector pcDNA 3 (Invitrogen) using a unique Bam HI site to give the vector KB 113. The cDNA was completely sequenced on both strands.

Vectors for targeting construct: Neomycin (pPGK neo bpA) and thymidine kinase (TKpSL1190) (Sato, T.N., et al., Nature 376: 70-74 (1995)) vectors are used for the construction of the targeting vector.

Preparation of the chimeric KDR/FLK-1 cDNA: FLK-1 and KDR cDNAs share a unique Bam HI site located at the codon for methionine 806 of the KDR sequence. To amplify the sequence coding for the FLK-1 cytoplasmic domain, PCR primers are designed such that the 5' primer is located just upstream of the Bam HI site and the 3' primer just downstream of the termination codon. In addition, the 3' primer is designed to encode a Not I site. Full length FLK-1 cDNA serves as the template for the amplification. The PCR product is cloned into the vector pCR 2.1 (Invitrogen) and sequenced on both strands. The FLK-1 cDNA is digested with Bam HI and Not I and subcloned into the KDR expression vector KB 113 (see above) replacing the sequence coding for KDR cytoplasmic domain. In the resulting chimeric cDNA, the first 20 amino acids of the cytoplasmic domain are derived from the human sequence. However, this region contains only a single difference between the murine and human proteins with glycine (KDR) and glutamic acid (FLK-1 ) at amino acid 794 of the KDR sequence.

Verification of the expression constructs: The expression constructs tested for the ability to mediate the expression of a functional hybrid receptor molecule by transient transfections into COS 7 cells. The full length KDR expression construct KB 113 serves as a positive control. 48 hours after transfection, the expression of KDR or KDR/FLK-1 is tested by 1251-VEGF binding, Fluorescence Activated Cell Sorting (FACS), Western blotting and in a receptor autophosphorylation assay.

For 1251-VEGF binding, VEGF165 is iodinated with 1251. COS 7 cells are plated at near confluency in 24-well plates. All subsequent steps are carried out at 4°O The cells are washed 1X with binding buffer (BB = MCDB-131/15 mM HEPES/0.1% gelatin/1 μg/ml heparin), then incubated for 2 hours in 0.5 ml BB containing 2 ng/ml of 1251-VEGF. Following incubation, the cells are washed 3X with BB, 2X with PBS, dissolved in 1 % Triton X-100 and counted for radioactivity.

For FACS, the cells are removed with 2 mM EDTA in PBS, washed with cold Hanks balanced salt solution supplemented with 1 % BSA (HBSS-BSA) and then resuspended in 100 μl of the same buffer at a concentration of 105 cells per sample. The cells are incubated for 30 minutes with 10 μg of the appropriate anti-KDR or control FLK-1 specific monoclonal antibody. After washing, a 1 :40 dilution of goat anti- mouse or anti-rat IgG conjugated to FITC (TAGO) is added for a final 30 minute incubation on ice. Cells are then analyzed on a Coulter Epics Elite Cytometer. Data is expressed as the measurement of the mean fluorescent intensity of anti-KDR monoclonal antibody binding to cells relative to the control measurement of anti-FLK-1 monoclonal antibody binding.

For Western blot analyses, transfected and control COS 7 cells are lysed in a buffer containing 20 mM Tris-HCI pH 7.4, 1 % N-octylglucoside, 137 mM NaCI, 10% glycerol, 10 mM EDTA, 100 μg/ml Pefabloc (Boehringer Mannheim), 100 μg/ml aprotinin, and 100 μg/ml leupeptin. Following low speed centrifugation the lysates are separated by SDS-PAGE and transferred to nitrocellulose. The KDR and chimeric KDR/FLK-1 receptor proteins are detected with affinity-purified polyclonal rabbit antibodies developed at ImClone against the soluble KDR extracellular domain. The blots are incubated with 1251-labeied Protein A (Amersham) and detected by autoradiography.

For the receptor phosphorylation assay, the control and transfected COS 7 cells are starved for 24 hours in DMEM containing 0.5% CS and then stimulated with 20 ng/ml VEGF for 10 minutes at room temperature. Following ligand stimulation, cells are washed with cold PBS containing 1 mM sodium orthovanadate, lysed in a buffer containing 20 mM Tris-HCI pH 7.4, 1 % N-octylglucoside, 137 mM NaCI, 10% glycerol, 10 mM EDTA, 0.1 mM sodium orthovanadate, 10 mM NaF, 100 mM sodium pyrophosphate, 100 μg/ml Pefabloc (Boehringer Mannheim), 100 μg/ml aprotinin, and 100 μg/ml leupeptin. Following centrifugation at 14,000 x g for 10 minutes, receptors are immunoprecipitated from cleared lysates with Protein A Sepharose beads coupled to rabbit anti-KDR antibodies. The beads are washed, mixed with SDS loading buffer and subjected to Western blot analysis. The phosphoprotein patterns of the stimulated receptors are detected using an anti-phosphotyrosine monoclonal antibody (UBI) and developed by chemiluminescence (ECL; Amersham).

All routine molecular biology procedures such as restriction digests, ligations, PCR and Southern blotting is performed using standard procedures (Sambrook J, Fritsch EF, and Maniatis T, editors. "Molecular Cloning. A Laboratory Manual." 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (1989)). Assembling the target vector: The targeting vectors that direct the homologous recombination of the two KDR receptor forms (full-length KDR and chimeric KDR/FLK- 1 ) into the FLK-1 locus are assembled as follows.

The strategy for disruption of the FLK-1 gene in mouse embryonic stem (ES) cells and the expression of KDR or chimeric receptors under the control of the FLK-1 regulatory elements is outlined in Figure 1. The cDNA in the targeting construct in Figure 1 and in the discussion below is referred to as KDR but represents either KDR or chimeric forms of the receptor. The strategy is broadly based on that used by Shalaby et al., for the targeted disruption of the FLK-1 gene (Nature 376: 62-66 (1995)).

An upstream FLK-1 genomic DNA fragment and KDR cDNA are assembled in a cloning vector designed for this task. The assembly vector consists of the pCR II backbone plasmid (Invitrogen) in which the multiple cloning site (MCS) between the Nsi I and Xba I sites are replaced with a synthetic MCS containing the required restriction sites in the following order: Not I, Bam HI, Nsi I, Sma I, Kpn I, Eco RV and Not I. The correct construction of the assembly vector is verified by digesting with Sma I as this enzyme is not present in the parental pCR II plasmid and by sequencing. A 1.8 kb genomic FLK-1 fragment extending from an upstream Bam HI site to the Pst I site in the first coding exon (Figure 1 ) is cloned into the Bam HI and Nsi I sites of the assembly vector with the simultaneous inactivation of the compatible Pst I and Nsi I sites. The KDR cDNA and bovine growth hormone polyadenyiation signal (pA) are excised from the KB 113 expression plasmid with Kpn I and Pvu II and cloned into the Kpn I and Eco RV sites of the assembly vector (Pvu II and Eco RV are compatible, blunt-cutting enzymes).

A Not I fragment from the assembly vector containing the upstream FLK-1 genomic sequence and the KDR cDNA are inserted into the unique Not I site of pPGK neo bpA and clones are selected in which the orientation of the inserted DNA matches that indicated in Figure 1. A 5.7 kb fragment of genomic FLK-1 DNA extending from a Sma I site downstream from the first coding exon to a Sal I site further downstream is cloned into the Hind III and Sal I sites of pPGK neo bpA by filling the digested Hind III site with the Klenow fragment of DNA Polymerase I. Finally, a blunted 2 kb Hind III fragment containing the thymidine kinase expression cassette is cloned into a blunted unique Sac II site of pPGK neo bpA.

The correct assembly of the final targeting vector is verified by PCR with primers located on adjacent DNA fragments and by restriction digests.

Production of the Mice with a Knocked-ln KDR Gene: The following describes a method of producing lines of ES cells in which at least one allele of the FLK-1 gene is inactivated and in which the expression of the KDR or KDR/FLK-1 chimeric mRNA is under the control of the native FLK-1 regulatory elements.

129/sv ES (Nagy, A., et al., Proc. Natl. Acad. Sci. USA 90: 8424-8428 (1993)) cells are electroporated with the targeting vector using the ECM 600 electroporator (Gentronics) in HEPES-buffered saline at 160 V, 50 μF capacitance and 360 ohms resistance. After electroporation, 2 x104 cells are cultured on a 100-mm dish containing feeder STO fibroblasts (Mansour, S.L., et al., Proc. Natl. Acad. Sci. USA 87:7688-7692 (1990)). At 48 hours post electroporation, the cells are selected with gancyclovir and G418 and individual colonies are isolated.

Genomic DNA from double-selected ES clones is prepared and tested by Southern blotting for homologous recombination. Genomic DNA is digested with either Nco I or Xho I. A probe is generated from a FLK-1 Pst l-Xho I fragment downstream of the targeted locus (see Figure 1 ) and labeled with 32P. This probe is expected to detect a 6.5 kb Nco I fragment in the wild-type locus (Shalaby, F., et al. Nature 376: 62-66 (1995)) and a much larger fragment in the targeted locus resulting from the insertion of the KDR and Neo cDNAs. Similarly, this probe should detect an approximately 3.8 kb Xho I fragment generated by the insertion of a novel Xho I site just downstream of the KDR cDNA. The targeted ES cells are used for the subsequent development of mice that express the chimeric KDR/FLK-1 or the full length KDR receptors under the control of the FLK- 1 promoter. First, heterozygous germ-line chimeric (KDR/FLK-1 )/FLK-1 or full length KDR/FLK-1 mice can be produced using conventional knock-out procedures. Expression of KDR in the endothelium of heterozygous mice is confirmed by immunocytochemistry using KDR-specific polyclonal and monoclonal antibodies and by RT-PCR with KDR specific primers. The mice are then cross-bred to produce homozygous full length KDR/KDR or homozygous chimeric (KDR/FLK-1 )/(KDR/FLK-1 ) mice.

These mice are cross-bred with an immunodeficient mouse strain such as FvAG -/- and the progeny can serve as recipients for the implantation of various murine and human tumor cell lines. The efficacy of administration of monoclonal antibodies that inhibit KDR-VEGF binding or the administration of other KDR-directed agents can then be determined.

Transgenic Animal Models with Knocked-ln VEC Receptors and Their Uses:

The following assays can be used to identify target molecules for therapeutic intervention. In addition, the angiogenesis models described below can be used to test the resulting therapeutic reagents as to their effectiveness and specificity.

a. Tumor Angiogenesis:

The mice with knocked-in VECr genes described above are crossed with RAG-1(-/-) mice and the resulting progeny used for implantation of tumor cell lines. Various human tumor cell lines are injected into these immuno-compromised knock-in mice and the effect of therapeutic antibodies, target molecules, and other human species- specific reagents is evaluated. b. Ocular Neovascularization:

The locally induced expression of each knock-in form of receptor tyrosine kinase is tested in the mice with knocked-in VECr genes during ocular neovascularization induced by various angiogenic factors such as VEGF and FGF. These models would be useful in the study of ocular conditions such as retinopathy.

c. Acute and Chronic Inflammation Models:

The mice with knocked-in VECr genes described above are used to study the effects of therapeutic agents on a variety of induced inflammatory conditions. These knock- in mice, in which an inflammatory condition has been induced, would be of particular value in studying therapies for a variety of acute and/or chronic inflammatory conditions, such as rheumatoid arthritis.

d. Psoriasis Models:

Psoriatic skin is characterized by microvascular hyperpermeability and angioproliferation. The hyperplastic epidermis of psoriatic skin expresses strikingly increased amounts of vascular endothelial growth factor. Accordingly, the mice with knocked-in VECr genes described above are useful to study the effects of therapeutic agents on psoriasis, which is often characterized by an increase in vascular endothelial growth factor.

e. Bullous Disease:

Vascular endothelial growth factor is strongly expressed by epidermal keratinocytes in bullous diseases such as erythema multiforme and bullous pemphigoid. These conditions are characterized by increased microvascular permeability and angiogenesis. The development of erythema as a result of hyperpermeable blood vessels is also a common feature after excess sun exposure. To test various therapeutic compounds that have an effect upon these various conditions the mice with knocked-in VECr genes described above are useful.

f. Wound Healing:

The role of angiogenesis in wound healing is well-known. In particular, an increase in VEGF expression has been reported in wound healing. The mice with knocked-in VECr genes described above are useful in testing the effects of agonists and antagonists on the expressed receptors, as they relate to wound healing. Such effects would further an understanding of the wound healing process, and would allow therapeutic intervention of the process.

g. Arteriovenous Malformations:

Arteriovenous malformations (AVMs) are congenital lesions composed of abnormal vasculature, with no capillary component, and are clinically significant due to their tendency to spontaneously hemorrhage. The endothelial cell- specific receptor tyrosine kinase, TIE, has been shown to be elevated in AVM and surrounding brain vasculature. Additionally, upregulation of VEGF mRNA was observed in the cells of brain parenchyma adjacent to the AVM, and VEGF protein was detected in this tissue as well as in AVM endothelia. Normal brain, in comparison, expressed little or no TIE or VEGF. The significant upregulation of VEGF and TIE in AVMs indicates ongoing angiogenesis, contributing to the slow growth and maintenance of the AVM. Accordingly, the mice with knocked-in VECr genes described above are used to study the effects of therapeutic agents on these congential lesions.

Claims

CLAIMSWe claim:

1. A non-human transgenic animal whose cells express a foreign DNA sequence that encodes a functional vascular endothelial cell receptor domain, but do not express a substantially homologous native DNA sequence.

2. The transgenic animal of claim 1 , wherein the transgenic animal is a laboratory animal.

3. The transgenic animal of claim 2, wherein the laboratory animal is a mouse.

4. The transgenic animal of claim 1 , wherein the foreign vascular endothelial cell receptor DNA sequence is chimeric.

5. The transgenic animal of claim 4, wherein the chimeric DNA sequence is a FLK-1/KDR gene.

6. The transgenic animal of claim 1 , wherein the foreign DNA sequence is mammalian.

7. The transgenic mouse of claim 3, wherein the foreign DNA sequence is human.

8. The transgenic mouse of claim 7, wherein the human foreign DNA sequence is under the control of murine tissue-specific regulatory elements.

9. The transgenic mouse of claim 8, wherein the human foreign DNA sequence is under the control of a murine endothelial cell specific promoter.

10. The transgenic mouse of claim 9, wherein the promoter is a FLK-1 or a TIE-2 promoter.

11. The transgenic mouse of claim 10, wherein the promoter is a FLK-1 promoter.

12. The transgenic mouse of claim 7, wherein the native DNA sequence is the FLK-1 gene.

13. The transgenic mouse of claim 12, wherein the human foreign DNA sequence is a KDR gene.

14. The transgenic mouse of claim 7, wherein the foreign DNA sequence is an extracellular human KDR gene fragment.

15. A non-human transgenic animal whose cells express a foreign gene from a different species, but do not express a substantially homologous native gene.

16. A method of testing a substance that interacts with a protein expressed by the foreign DNA sequence of claim 1 comprising administering the substance to the transgenic animal of claim 1 and evaluating any effects of the substance on the animal.

17. A method of identifying a substance capable of inhibiting angiogenesis comprising administering the substance to the transgenic animal of claim 1 and determining whether the substance inhibits angiogenesis.

18. A method of identifying a substance capable of inhibiting tumor growth comprising administering the substance to the transgenic animal of claim 1 and determining whether the substance inhibits tumor growth.

19. A method of testing a substance for use in animals comprising administering the substance to a non-human transgenic animal whose cells express a foreign gene or functional gene fragment from a different species, but do not express a substantially homologous native gene or functional gene fragment, and evaluating any effects of the substance on the animal.

20. The method of claim 19, wherein the transgenic non-human animal is a mouse.

21. The method of claim 19, wherein the donor DNA sequence is under the control of the transgenic animal's tissue-specific regulatory elements.